Report Imperia Team

Team 12 Technical Report to the 2024 EuRoC
Academic Supervisor: Dr. Aaron Knoll

Project: Imperial College London Rocketry
(ICLR)
Academic year: 2023-2024
Authors: ICLR Engineers
Date: September 2024
Imperial College London

South Kensington Campus
London SW7 2AZ
U.K.
Revision List
Revision Date Description

1.0 16/09/24 Initial Report
2.0 16/09/24 Minor Formatting Errors
3.0 27/09/24 Propulsion flight qualification test, recovery mechanical separation test
and parachute reefing test reports added. Updated Checklists.
i
Contents
Revision List i
List of Figures vi
List of Tables ix
1 Introduction 1
2 Project Goals and Mission Objectives 1

2.1 Mission Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
3 Nimbus 24 Overview 2
4 Propulsion 2
4.1 Plumbing and Tanks layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
4.2 Propellant & Pressurant Tanks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.3 Electronic Tank Pressure Regulator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.3.1 Controller Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.3.2 Regulator Safety and Abort Logic . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.3.3 Integrated Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.4 Bi-propellant Engine (THANOS-R) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.4.1 Injector Head . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.4.2 Thrust Chamber Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.4.3 Thrust Chamber Cooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.4.4 Ignition Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.5 Valve Actuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.6 Ground Support Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.6.1 Remote pressurised fluid filling system . . . . . . . . . . . . . . . . . . . . . . . 13
4.6.2 Manual Fuel Loading System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.7 Rocket Filling Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.7.1 Manual Fuel Filling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.7.2 Remote Nitrogen Filling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.7.3 Remote Oxidiser Filling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5 Airframe 17
5.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5.2 Skeletal Airframe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5.2.1 Stringers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
5.2.2 Engine Truss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5.2.3 Bulkheads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.2.4 Couplers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.2.5 Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3 Monocoque . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.3.1 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.3.2 Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
6 Recovery 22
6.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
6.2 Clamp Band . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
6.3 Reefed Parachutes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
6.4 Recovery Bulkheads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
ii
6.5 Shock Load Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
7 Avionics 26
7.1 Avionics Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
7.1.1 Flight System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
7.1.2 Ground Support System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
7.1.3 Mission Control System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
7.1.4 COTS Avionics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
7.2 Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
7.2.1 Ricardo Hardware Template . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
7.2.2 Flight Controller - Pickle Rick (2024) . . . . . . . . . . . . . . . . . . . . . . . . 28
7.2.3 Engine Control Unit - Stark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
7.2.4 Servo Actuation Board - Chad (2024) . . . . . . . . . . . . . . . . . . . . . . . . 30
7.2.5 E-Reg Board - Greg (2024) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
7.2.6 High Powered Actuation Board - Flint & Steel (2024) . . . . . . . . . . . . . . . 31
7.2.7 Power Distribution Unit - Lightning McQueen (2024) . . . . . . . . . . . . . . . 31
7.2.8 Sensor Board - Kermit (2024) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
7.2.9 CAN Repeater (2024) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
7.2.10 Canard Board - Geddan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
7.2.11 Miniaturised Altimeter Board - Artemeter . . . . . . . . . . . . . . . . . . . . . 33
7.2.12 Power Switchover Board - Witcher (2024) . . . . . . . . . . . . . . . . . . . . . 33
7.2.13 Antenna Bay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
7.3 Electrical Quick disconnects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
7.3.1 RBUS Electrical Quick Disconnect (EQD) . . . . . . . . . . . . . . . . . . . . . 34
7.3.2 Adapter Power Quick Disconnect (AQD) . . . . . . . . . . . . . . . . . . . . . . 34
7.4 Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
7.4.1 Apogee Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
7.5 Ground Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
7.5.1 Ricardo-Backend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
7.5.2 Ricardo-CommandServer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
7.5.3 Grafana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
7.6 Arming and Ignition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
8 Active Control Systems - Canards 37

8.1 Canard Roll Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
8.2 Profile and Planform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
8.3 Actuation Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
8.4 Control System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
9 Payload - FAWKES 39
9.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
9.2 Payload Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
9.3 Electronics System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
9.4 CubeSat Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
9.4.1 Guided Recovery In Flight For Improved Navigation (GRIFFIN) . . . . . . . . . 40
9.4.2 External Outreach - Northern Ireland Advanced Composites Centre (NIACE) . . 41
9.5 CubeSat Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
9.5.1 Deployment mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
10 Trajectory Analysis 44
10.1 Nominal Thrust Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
10.2 Nimbus 24 Trajectory Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
iii
10.3 Monte Carlo Simulations - Landing Zone . . . . . . . . . . . . . . . . . . . . . . . . . . 45
10.4 Nimbus 24 Stability Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
10.5 Nimbus 24 Nominal Trajectory Summary . . . . . . . . . . . . . . . . . . . . . . . . . . 46
10.6 Deployable Payload Descent Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . 47
11 Mission Concept of Operations 48
12 Conclusions and Outlook 50
References 51
A System Data 52
B Additional Flight Simulations 54

B.1 Canard-less Flight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
C Project Tests 55
C.1 SRAD Avionics - Apogee Detection and Flight Telemetry Test . . . . . . . . . . . . . . . 55
C.2 COTS Avionics - CATS Vega Nichrome Recovery Actuation Test . . . . . . . . . . . . . . 57
C.3 SRAD Avionics - Radio Range Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
C.4 SRAD Avionics - High Altitude Flight Test with Actuation Events . . . . . . . . . . . . . 62
C.5 SRAD Avionics - High Temperature Test . . . . . . . . . . . . . . . . . . . . . . . . . . 64
C.6 SRAD Avionics - Low Temperature Test . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
C.7 Propellant Loading of Nimbus’ Propulsion System . . . . . . . . . . . . . . . . . . . . . 69
C.8 Hot Fire Testing of THANOS-R Engine and NIMBUS 24 flight propulsion system . . . . 72
C.9 Propellant Unloading of Nimbus’ Propulsion System . . . . . . . . . . . . . . . . . . . . 77
C.10 Hydrostatic Testing of THANOS R Combustion Chamber . . . . . . . . . . . . . . . . . 81
C.11 Hydrostatic Testing of SRAD Pressure Vessel (Fuel Flight Tank) . . . . . . . . . . . . . 83
C.12 Hydrostatic Testing of SRAD Pressure Vessel (Oxidiser Flight Tank) . . . . . . . . . . . 85
C.13 Hydrostatic Testing of COTS Pressure Vessel (Nitrogen Flight Tank) . . . . . . . . . . . 87
C.14 Ground test of the deployment mechanism of the payload . . . . . . . . . . . . . . . . 89
C.15 Ground test of the actuation mechanism of the guided recovery system . . . . . . . . . 91
C.16 Inflation test of folding configuration of parafoil . . . . . . . . . . . . . . . . . . . . . . 93
C.17 Engine Truss Connector Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
C.18 Parachute reefing deployment testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
C.19 Recovery mechanical separation testing . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
D Flight System Overview 103
E Ground Support System Overview 105
F Material Tables 106
G Hazard Analysis 109
H Risk Assessment 110
I Checklists 113
I.1 Checklist A - Paddock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
I.4 Checklist B - Final Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
I.5 Checklist B2A/5A - Payload Assembly and Integration . . . . . . . . . . . . . . . . . . . 118
I.6 Checklist B2B - Recovery Assembly and Integration . . . . . . . . . . . . . . . . . . . . 119
iv
I.7 Checklist B2C/5B - Electronics System Checks . . . . . . . . . . . . . . . . . . . . . . . 120
I.8 Checklist C - Pyro Tent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
I.9 Checklist D - Launch Rail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
I.10 Checklist E - Prelaunch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
I.11 Checklist F - Launch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
J Engineering Drawings 132
K Data Task Request Handler Example JSON Config 149
L Data Task Request Handler UI 150
M Geddan Power Source Logic 151
N SRAD Tank Failure Calculations 153

N.1 Hoop Stress Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
N.2 Bolt Shear Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
N.3 Hole Tear Out Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
N.4 Hole Bearing Surface Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
O Modelling the dynamics of post parachute deployment behaviour 155
P Reefing characteristics equations, calculation and procedure 156
Q Airframe Finite Elements Simulation Reports 157

Q.1 Body Tubes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Q.2 Stringers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Q.3 Payload Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Q.4 Guided Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Q.5 Recovery Bulkheads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
R Payload Addendum 180
v
List of Figures
1 Subsystem breakdown of Nimbus 24. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2 The full rocket propulsion system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3 P&ID for the in-flight feed system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4 Cross-section views showing the mechanical construction of the oxidiser tank. The fuel
tank uses identical construction, with a shorter tube. . . . . . . . . . . . . . . . . . . . 5
5 Tank pressure and E-Reg valve demand angle data from a cold flow performed with the
flight feed system. A slight overshoot occurs at startup which helps give a boost in thrust
off the pad. This oscillation damps out quickly within 1.5 seconds . . . . . . . . . . . . 7
6 (left) Exhaust plume of the THANOS-R engine during the flight qualification hot fire
(right) close-up of the assembled engine . . . . . . . . . . . . . . . . . . . . . . . . . . 8
7 Cross Section of the THANOS-R Engine showing all the major components . . . . . . . 8
8 (left) closeup of the chamber side of the injector, showing the film cooling holes. (right)
Assembled injector next to thrust chamber, showing pintle insert and all connections. . 9
9 (left) Plot of chamber wall stresses and chamber wall yield strength along the cham-
ber (middle) Plot of gas side wall, coolant side wall and coolant temperature along the
chamber (right) Plot of wall safety factor based on the local yield strength at temperature 11
10 Modified thermal simulation results with lower OF ratio and 85% with 23% water and
2% PDMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
11 Servo Actuated Ball Valve design, utilizing a 4 bar linkage to transfer torque. First image
with lid and potentiometer and second image without lid . . . . . . . . . . . . . . . . 12
12 (left) The 2x pressurised fluid quick disconnect lines running down either side of the
engine (right) the 3x quick disconnects with the flame deflector which pulls off the 2x
pressurised hoses during launch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
13 GSS filling system Plate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
14 P&ID of the GSS filling system. Further details are given in the sections below . . . . . 14
15 Plot of the nitrogen COPV pressure during filling and propellant tank pressurisation. . . 16
16 Plot of the oxidiser tank pressure and temperature during propellant loading, chill down
and pressurisation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
17 The breakdown of Monocoque and Skeletal Sections within the rocket . . . . . . . . . . 17
18 Repeating pattern of the stringers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
19 Stringer Artefacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
20 Engine Truss Generative design study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
21 Bulkheads in Skeletal Airframe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
22 Couplers on Nimbus 24 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
23 Body Tube Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
24 The Clamp Band system in Nimbus 24 . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
25 Recovery system lines diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
26 The Toroidal Reefed Parachute in both reefed and disreefed states. . . . . . . . . . . . 24
27 The recovery Bulkheads in Nimbus 24 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
28 Loads on the airframe due to parachute deployment in the reefed state . . . . . . . . . 26
29 Nimbus Electronics Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
30 Flight Controller, Pickle Rick . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
31 Engine Control Unit, Stark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
32 Actuator Board, Chad . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
33 E-reg add-on board, Greg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
34 High Powered Actuation Board, Flint and Steel . . . . . . . . . . . . . . . . . . . . . . 31
35 Power distribution unit, Lightning McQueen . . . . . . . . . . . . . . . . . . . . . . . . 31
36 Sensor Board, Kermit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
37 CAN Repeater . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
vi
38 Canard Control Board, Geddan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
39 Altimeter, Artemeter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
40 Power OR’ing board, Witcher . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
41 Antenna Bay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
42 Nimbus electrical quick disconnect (EQD) . . . . . . . . . . . . . . . . . . . . . . . . . 34
43 Mission control command interface during flight qualification hotfire . . . . . . . . . . 37
44 Canard Module Render . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
45 Payload Structure Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
46 Guided Recovery Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
47 Payload Parafoil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
48 Left to right - (i) Undeployed state of the deployer; (ii) Deployer with arm unlocked (iii)
Deployed state of the deployer with the payload. (iv) Render of the entire assembly. . . 43
49 Flight Qualification Hotfire Thrust Curve . . . . . . . . . . . . . . . . . . . . . . . . . . 44
50 Nimbus 24 Predicted Trajectory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
51 Nimbus 24 Velocity and Acceleration Profiles . . . . . . . . . . . . . . . . . . . . . . . . 45
52 Apogee and landing points for ascent sims . . . . . . . . . . . . . . . . . . . . . . . . . 46
53 Apogee and landing points for descent sims . . . . . . . . . . . . . . . . . . . . . . . . 46
54 Nimbus 24 In-Flight Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
55 Nimbus 24 In-Flight Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
56 Concept of Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
57 Nimbus 24 Nominal Stability (with Canards) . . . . . . . . . . . . . . . . . . . . . . . . 54
58 Nimbus 24 Stability without Canards . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
59 The apogee detection algorithm plotting parabolic trajectories during flight until apogee
is detected . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
60 Assembled avionics bay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
61 Flight controller radio station during test. . . . . . . . . . . . . . . . . . . . . . . . . . 60
62 Booster trajectory recorded from Pickle Rick telemetry downlink . . . . . . . . . . . . . 63
63 Avionics Bay of Sirius . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
64 Sensor readings of flight controller under high temperature conditions . . . . . . . . . 65
65 Pickle Rick Flight controller with dry ice placed upon . . . . . . . . . . . . . . . . . . . 67
66 Sensor Readings for Flight Controller under extreme low temperature conditions . . . . 68
67 Ground Support System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
68 Propellant Loading Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
69 Propellant Loading Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
70 Thrust and Pressure Graph from Qualification Hotfire . . . . . . . . . . . . . . . . . . . 74
71 From left to right, photos from the final qualification hot fire . . . . . . . . . . . . . . . 75
72 Photos from the hydrostatic test. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
73 Photos from the hydrostatic test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
76 Deployer Unlocked via actuating servos . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
77 Payload Falling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
78 Payload Fully Deployed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
79 Parafoil banked port side . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
80 Parafoil in neutral position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
81 Parafoil banked starboard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
82 Parachute Deployment Stages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
83 Photos from the truss testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
84 Graphs from the truss testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
87 Layout of Electronics Subsystems and wiring . . . . . . . . . . . . . . . . . . . . . . . . 103
88 Layout of Electronics Subsystems and wiring . . . . . . . . . . . . . . . . . . . . . . . . 104
vii
89 System Diagram of the ground support system for Nimbus . . . . . . . . . . . . . . . . 105
90 Checklist Legend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
91 Screenshot of the data task request handler user interface . . . . . . . . . . . . . . . . 150
92 Flowchart of Geddan desired deployment power behaviour. . . . . . . . . . . . . . . . . 151
93 Final hardware implementation of Geddan deployment power logic. . . . . . . . . . . . 152
94 Rocket body tube . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
95 Buckling mode 1 of Body Tube . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
96 Buckling mode 2 of Body Tube . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
97 Buckling mode 3 of Body TUbe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
98 Displacement of Body Tube under shock load . . . . . . . . . . . . . . . . . . . . . . . 159
99 Stress of Body Tube under shock load . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
100 Upper propulsion stringer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
101 Buckling Mode 1 of upper propulsion stringer . . . . . . . . . . . . . . . . . . . . . . . 162
104 Stress under shock loading of upper propulsion stringer . . . . . . . . . . . . . . . . . . 163
105 Displacement under shock loading of upper propulsion stringer . . . . . . . . . . . . . 164
106 Lower propulsion stringer: Stringer . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
107 Buckling Mode 1 of lower propulsion stringer . . . . . . . . . . . . . . . . . . . . . . . 165
110 Stress resulting from shock load for lower propulsion stringer . . . . . . . . . . . . . . 166
111 Displacement resulting from shock load for lower propulsion stringer . . . . . . . . . . 166
112 Payload Stringer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
113 Buckling Mode 1 of payload stringer . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
116 Stress from shock loading of payload stringer . . . . . . . . . . . . . . . . . . . . . . . 170
117 Displacement from shock loading of payload stringer . . . . . . . . . . . . . . . . . . . 170
118 Maximum stress of vertical load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
119 Maximum stress of horizontal load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
120 Buckle mode 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
121 Buckle mode 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
122 Buckle mode 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
123 GRIFFIN Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
124 Maximum stress of vertical load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
125 Maximum displacement of vertical load . . . . . . . . . . . . . . . . . . . . . . . . . . 177
126 The lower recovery bulkhead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
127 The stress distributions in MPa on the lower recovery bulkhead . . . . . . . . . . . . . . 179
128 The displacement distributions in mm on the lower recovery bulkhead . . . . . . . . . . 179
129 Hermes in detail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
130 Hermes engineering drawing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
131 Fold I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
132 Fold II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
133 Fold III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
134 Fold IV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
135 Fold V . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
136 Fold VI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
137 Folding order to minimise risk of entanglement of lines . . . . . . . . . . . . . . . . . . 184
138 Actuation Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
139 Guided Recovery Build . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
viii
List of Tables
1 Nimbus 24 length breakdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2 Nimbus 24 mass breakdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
3 Propulsion System Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
4 Design specification of the two SRAD propellant tanks . . . . . . . . . . . . . . . . . . 5
5 Design specification of THANOS-R Engine at nominal operating point . . . . . . . . . . 8
6 A table showing the stringer lengths and safety factors . . . . . . . . . . . . . . . . . . 19
7 Design specifications of the Flight Controller, Pickle Rick . . . . . . . . . . . . . . . . . 28
8 Design specifications of the Engine Control Unit . . . . . . . . . . . . . . . . . . . . . . 29
9 Design specifications of the servo control board . . . . . . . . . . . . . . . . . . . . . . 30
10 Design specifications of the E-Reg board . . . . . . . . . . . . . . . . . . . . . . . . . . 30
11 Design specifications of the High Powered actuation board . . . . . . . . . . . . . . . . 31
12 Design specifications of the Power Distribution unit . . . . . . . . . . . . . . . . . . . . 31
13 Design specifications of Kermit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
14 Design specifications of the CAN Repeater . . . . . . . . . . . . . . . . . . . . . . . . . 32
15 Design specifications of the Canard board . . . . . . . . . . . . . . . . . . . . . . . . . 32
16 Design specifications of the Miniaturised Altimeter board . . . . . . . . . . . . . . . . . 33
17 Design specifications of the Power OR’ing board . . . . . . . . . . . . . . . . . . . . . . 33
18 Nimbus 24 Rail Departure States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
19 Nimbus 24 Burn Out States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
20 Nimbus 24 Apogee and Impact States . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
21 Nimbus 24 Ascent Maximum State Values . . . . . . . . . . . . . . . . . . . . . . . . . . 47
22 General Specifications of Nimbus 24 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
23 Nimbus 24 Mass Breakdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
24 Nimbus 24 Length Breakdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
25 Nimbus 24 Recovery Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
26 Nimbus 24 THANOS-R Engine Specifications . . . . . . . . . . . . . . . . . . . . . . . . 53
27 Nimbus 24 Battery Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
28 Nimbus 24 Avionics Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
29 Ground test of the deployment mechanism of the payload . . . . . . . . . . . . . . . . 89
30 Ground test of the actuation mechanism of the guided recovery system . . . . . . . . . 91
31 Materials and Manufacturing Methods for Components . . . . . . . . . . . . . . . . . . 106
32 Geddan power switchover truth table . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
33 Key Results - Buckling of a Body Tube . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
34 Key Results - Shock Load Applied to Body tube . . . . . . . . . . . . . . . . . . . . . . . 159
35 Mesh Convergence Study of Shock Load on Body Tube . . . . . . . . . . . . . . . . . . 160
36 Key Results - Buckling of upper propulsion stringer . . . . . . . . . . . . . . . . . . . . 162
37 Mesh convergence study of Buckling of upper propulsion stringer . . . . . . . . . . . . 162
38 Key Results - Shock loading of upper propulsion stringer . . . . . . . . . . . . . . . . . 163
39 Mesh Convergence study of Shock loading on upper propulsion stringer . . . . . . . . . 163
40 Key Results - Buckling of lower propulsion stringer . . . . . . . . . . . . . . . . . . . . 165
41 Mesh Convergence study of Buckling of lower propulsion stringer . . . . . . . . . . . . 165
42 Key Results - Shock loading of lower propulsion stringer . . . . . . . . . . . . . . . . . 166
43 Mesh Convergence study of Shock Loading of lower propulsion stringer . . . . . . . . . 166
44 Key Results - Buckling of Payload stringer . . . . . . . . . . . . . . . . . . . . . . . . . 168
45 Mesh convergence study of buckling simulation of payload stringer . . . . . . . . . . . 168
46 Key Results - Shock loading of payload stringer . . . . . . . . . . . . . . . . . . . . . . 170
47 Mesh Convergence study of shock loading of payload stringer . . . . . . . . . . . . . . 170
48 Key Results - vertical acceleration load . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
49 Key Results - Horizontal acceleration load . . . . . . . . . . . . . . . . . . . . . . . . . 173
ix
50 Key Results - Vertical buckling load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
51 Key Results - Static Shock Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
52 Parafoil Further Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
x
1 Introduction
Imperial College London Rocketry (ICLR), founded in 2018, is a student-led rocketry team from Im-
perial College London. The team consists of approximately 100 students from various engineering
departments. The primary goal for the team is to educate and engineering students in both the prac-
tical and theoretical aspects of rocketry and spacecraft engineering. However, from a more technical
standpoint, the team’s long-term ambition is to reach space with a fully student-researched and devel-
oped vehicle. This will be achieved through incremental developments of smaller rockets to mark key
technical milestones towards this goal.
The team is split into five key engineering sub-teams: Payload, Flight Dynamics, Airframe and Recovery,
Electronics, and Propulsion. Each sub-team is managed by a Team Lead who oversees the overall
direction of the team, and one or more Technical Leads who ensure the technical goals of each sub-
team are met. To further ease project governance, an Administrative and Supportive Team is also in
place.
The team has competed at EuRoC for the past 3 years, setting increasingly challenging goals, push-
ing ourselves to become more technically proficient. EuRoC ’21 saw our first launch of a large-scale
commercial off-the-shelf (COTS) solid rocket, Constant Impulse. The following year at EuRoC ’22, the
team attempted to launch their first student researched and developed (SRAD) hybrid rocket, Sporadic
Impulse, which unfortunately was not launched after encountering an electronics voltage instability.
For EuRoC ’23, the team decided to further upgrade and improve the design and manufacturing of its
flight vehicles, culminating in Nimbus, our first ever liquid bi-propellant rocket. Nimbus was powered
by THANOS A, our first liquid engine and the first of it’s kind to be successfully fired in the UK amateur
rocketry community. Unfortunately, Nimbus suffered a tank failure during a launch attempt due to
incorrect assembly which ended the vehicle’s launch campaign.
This year for EuRoC ’24, the team has re-engineered our newest vehicle, Nimbus 24, from the same
foundation as its predecessor, Nimbus, with several upgrades to improve the vehicle. Some noticeable
upgrades include the regenerative liquid engine, radax couplers and asymmetric fluid lines to help pre-
vent propulsion assembly errors. Nimbus 24 will be competing in the 3000m SRAD Liquids Category.
2 Project Goals and Mission Objectives

In line with ICLR’s goals, the aim is to provide members with the opportunity to learn about the fun-
damentals of rocketry through developing components for a rocket, gaining hands-on experience by
manufacturing & testing the components and researching and developing novel new technologies in
the field of student rocketry. All projects reflect our core aim and attempt to provide the aforemen-
tioned opportunities. In order to encourage the hands-on development approach and give members
the challenge and experience of in-house development, COTS components are only used where safety
or reliability demands it.
2.1 Mission Objectives

Nimbus 24 is the second bi-propellant rocket developed by Imperial College London Rocketry, building
on the foundation laid by Nimbus in 2023. This iteration addresses the criticalities and shortcomings
identified from the previous rocket, ensuring a smoother and more reliable experience from assembly
to launch. Each subsystem has specific technical objectives aimed at enhancing the team’s capabilities
and driving continuous technical improvement. The design incorporates lessons learned to improve
performance and demonstrate the feasibility of novel subsystems.
• Achieve a 3000 m apogee using an SRAD airframe, propulsion and avionics system
• Demonstrate a working throttling system for a liquid bi-propellant engine
1
• Demonstrate an SRAD mechanical clamp band recovery system
• Demonstrate a deployable 3U CubeSat payload with a guided parachute and integrate external
experiments on board
• Demonstrate roll rate control using a canard-based control system
3 Nimbus 24 Overview
Nimbus 24 is a 4 m tall, 190 mm diameter, 50 kg liquid bi-propellant rocket. The rocket has been de-
signed with constant reference to the requirements mapped out by EuRoC, alongside those introduced
by the team at the start of the project. Design drivers this year consisted of both technical and opera-
tional requirements. These technical drivers included the SRAD regeneratively cooled liquid propulsion
system, the deployable payload mechanism and the inclusion of movable canards for active roll control.
Whilst the latter comprised of ease of assembly and safe operation. A high level subsystem overview
of Nimbus 24 is displayed in Figure 1. Running from the nose to the tail, the rocket consists of: a nose
cone; clamp band separation mechanism; recovery and antenna bay; canard and camera bay; payload
and payload deployer; avionics; nitrogen, fuel and oxidiser tanks; plumbing; regeneratively cooled
combustion chamber; and the fin can.
Figure 1: Subsystem breakdown of Nimbus 24.
Tables 1 and 2 summarise the masses and lengths of each major section and sub-system of Nimbus 24.
Table 2: Nimbus 24 mass breakdown
Table 1: Nimbus 24 length breakdown

Section Mass (kg)
Section Length (mm) Payload 3.0
Electronics 4.0
Propulsion Section 2461
Recovery 1.8
Payload Bay 662
Airframe 18.3
Recovery Tube 800
Propulsion 24.2
Nosecone 340
Propellant 11.5
TOTAL 4263
TOTAL Dry 51.3
TOTAL Wet 62.8
4 Propulsion
Nimbus 24 is a bi-propellant pressure fed propulsion system with an electronically regulated nitrogen
pressurant system. The philosophy of the propulsion system is to engineer it to be 1. Safe 2. Testable 3.
Reusable 4. High Performance 5. Throttleable. The system uses nitrous oxide as its oxidiser and ethanol
as its fuel with an electronically regulated nitrogen pressurant system. 300 bar nitrogen pressurant is
stored in a COTS COPV tank, with both fuel and oxidizer stored in SRAD tanks. The propellants are
delivered into a regenerative cooled rocket engine developed this year. The high level propulsion system
specs are shown in Table 3.
2
Table 3: Propulsion System Specifications
Peak Thrust [N] 4,000

Total Impulse [Ns] 21,500
Chamber Pressure [bar] 20
Tank Pressure [bar] 40
Propellant Mass [kg] 11.5
Specific Impulse [s] 170
4.1 Plumbing and Tanks layout

The feed system consists of three sections: upper, mid, and lower sections, with the high-pressure
nitrogen tank placed above the upper feed system; the fuel tank placed between the upper and mid
feed systems, and the oxidizer tank positioned above the lower feed system section, as shown in Figure
2. The P&ID of the rocket propulsion system is shown in Figure 3. The upper feed system consists of an
electronic pressure regulator (E-Reg), which regulates the nitrogen pressure down from 300 bar to 40
bar at static conditions. To ensure that both tanks are supplied with the same pressure, a single output
from the pressure regulator is split into two leading to the two tanks. Each line incorporates an inline
check valve to ensure the ullage gas from both tanks does not mix. Both tanks incorporate a dip tube
to allow for 10% ullage volume.
The fuel and oxidizer tanks are fitted with safety pressure relief valves (SPRVs) calibrated to relieve
pressure at 60 bar and a COTS burst disc from Wehberg Safety [1] with a nominal burst pressure of
70 bar ± 10% and a rupture diameter of 8mm. This ensures the safe discharge of pressurized gas
from either tank in the event of any over-pressure event. Each tank has a normally open solenoid valve
attached for fail-safe depressurization. Solenoid valves for nitrogen reside in the upper feed assembly,
whereas the solenoids for the fuel and oxidizer tanks reside within the mid-feed assembly. The solenoid
valve attached to the lower end cap of the fuel tank uses a dip tube for venting high-pressure gases.
The oxidizer tank has the solenoid attached to the upper end cap. This system improves safety by only
venting high-pressure gases without displacing fluid out of the tanks, in the rocket assembly, or to the
ground beneath. All vented gases are directed out of the sides of the rocket, and each is labeled on the
panel. This aids in the visual confirmation of any venting event that occurs. An oxidizer vent ball valve
is also present in the mid-feed system, actuated by a servo motor to adjust the vent rate during filling.
For ease of transportation, the propulsion system can split into two parts at the mid feed assembly.
The lower feed system consists of the main 3/8" fuel valve and a 1/2" oxidizer valve, which controls
the flow of propellants to the engine. The lower feed system also contains the attachment points to
the ground support system (GSS), which includes a needle valve for fuel and two quick disconnects for
oxidizer and high-pressure nitrogen.
All the pipes are made out of 20 SWG stainless steel, with high pressure stainless steel Swagelok
components used for sealing. Using Swagelok simplifies the assembly of the propulsion system making
it safer and reusable.
3
Figure 2: The full rocket propulsion system.
Figure 3: P&ID for the in-flight feed system.
4.2 Propellant & Pressurant Tanks

The high-pressure nitrogen is stored in a COTS composite overwrapped pressure vessel (COPV) (as
shown in Figure 2), which is rated to a pressure of 300 bar with a test pressure of 450 bar and a burst
pressure of 1000 bar. This tank has been hydrostatically proof tested after purchase to a pressure of 400
bar. The tank is connected to the plumbing with M18x1.5 threaded port adapted to a 1/4" Swagelok
tube adapter.
4
The fuel and oxidizer tanks are both SRAD components. Each tank is made of two in-house machined
aluminum end caps attached to either side of an aluminum tube. As shown in Figure 4, radial bolts are
used for pressure containment, and O-rings are used for sealing the tank. This allows for disassembly
of the tanks for inspection or cleaning. The flat outer face of the end caps allows for multiple fittings to
be attached to the tank, giving the benefit of components such as burst discs being threaded into the
tank directly. For simple assembly and reusability, tank end caps incorporate BSPP fittings with dowty
washers adapted to Swagelok fittings. A 10 mm gap is left between the outer wall of the tanks and the
composite skin of the rocket to pass pipes and cables across the tank. The SRAD tanks are designed to
yield at a pressure above 120 bar. The tanks have a maximum expected operating pressure (MEOP) of
60 bar, which has been hydrostatically proof tested to 90 bar, 1.5 times the MEOP.
Table 4: Design specification of the two SRAD propellant tanks
Propellant Oxidiser Fuel

Dry Mass (kg) 6.1 4.6
Capacity (L) 11 5.5
Prop Mass Capacity (kg) 7.5 4.0
Length (mm) 639 370
Diameter (mm) 170 170
Maximum Expected Operating Pressure (bar) 60 60
Minimum Burst Pressure (bar) 120 120
Figure 4: Cross-section views showing the mechanical construction of the oxidiser tank. The fuel tank uses
identical construction, with a shorter tube.
To ensure the tank would be able to handle the required pressure, four separate failure modes were
analysed. These are hoop stress failure of the tank tube, shear failure of the radial bolts, hole tear-out
of the radial bolts and bearing surface yielding of the bolt holes. The calculations for all the tanks are
shown in more detail in Appendix N.
4.3 Electronic Tank Pressure Regulator

To keep the propellant tanks at the correct pressure during the engine burn, an SRAD electronic pres-
sure regulator, or E-Reg, has been created to regulate high pressure nitrogen in the COPV down to the
40 bar propellant tank pressure. This is used instead of a mechanical regulator which has been used
5
previously, as a properly sized regulator system for the flow rates required would be very heavy and
bulky, or would exhibit a large ’droop’ in the outlet pressure during engine startup, meaning the tanks
would have to be initially pressurised to much higher pressures as a result.
The E-Reg system consists of a servo actuated high pressure ball valve controlled by a closed loop
control system with a pressure transducer in the fuel tank used as feedback. The controller performs
initial pressurisation of the tanks up to their flight pressures prior to launch, and then during the engine
burn it maintains the tank pressure against the set point as the nitrogen COPV is depleted.
4.3.1 Controller Design

The E-Reg utilises a proportional controller with feed forward and gain scheduling. The overall equation
for this controller output is shown by Equation 1. This was chosen as it is very simple and has a
predictable output, without issues such as windup that could come with adding an integrator. The
feed forward and gain scheduling are used to adjust the controller based on the current inlet pressure
from the nitrogen COPV as well as the tank ullage volumes, which are significant non-linearities in the
system which would be hard to compensate for with a proportional controller alone.
θservo = Kp · e + θF F (1)
At the start of the burn, the COPV pressure is very high (between 250 and 300 bar) and the tank
ullage volume is very small. This makes the tank pressure very sensitive to the E-Reg valve position and
requires much smaller angles to be used, so the Kp gain and feed forward angle are at their smallest.
Then, as the burn progresses, the COPV pressure drops rapidly, requiring the feed forward angle to
increase. The ullage volume also increases, requiring larger changes in angle for the same change in
tank pressure and so Kp is increased as well. Equation 2 and Equation 3 show the relationships used
to determine Kp and the feedforward angle, θF F respectively based on the nitrogen COPV pressure,
PN 2 at any given time. The constants in the equations were determined empirically. The value of these
parameters are also limited to a specific range as shown, so in the event of bad data from the pressure
transducer on the nitrogen tank, the control system will still operate within acceptable bounds. A
similar measure is a servo angle limiter for the first 0.5s of the burn, which prevents the E-Reg valve
fully opening at startup and creating a large spike in tank pressure.
β
Kp = Kp,0 + , 2 < Kp < 3 (2)
PN 2
α
θF F = θ0 + , 55◦ < θF F < 85◦ (3)
PN 2
4.3.2 Regulator Safety and Abort Logic

Since the regulator is a critical component of the rocket with high consequences in the event of a failure,
all the possible failure modes were considered and mitigated.
The most obvious system failure would be the regulator valve becoming stuck open due to either a
mechanical failure of the actuator linkage or a loss of power to the servo motor or control board.
During static operations on the pad, this would result in the tanks increasing in pressure until the burst
discs fail, which would then fully vent the system. Since each burst disc has an opening diameter of
8mm and the E-Reg piping internal diameter is only 4.3mm, the tanks will vent far faster than they
can be filled. If the regulator valve becomes stuck open during flight, the tank pressure will not change
significantly compared to normal operation, as the propellants will still be flowing into the engine. In
this case, having the regulator remain open would prevent an early engine shut down and a potential
crash of the rocket back near the launch pad.
6
While a stuck regulator valve is the worst case scenario, other scenarios involving failures of the pres-
sure transducer providing feedback could also occur. The regulator control board has a dedicated PT
attached to the fuel tank, but can also read the pressure of two other PTs in the fuel and oxidiser tank
across the CAN bus. Due to the risk of the regulator closing during flight causing an early engine shut-
down, a ’half abort’ and a ’full abort’ structure has been implemented. The ’half abort’ is triggered if
any tank PT exceeds 55 bar or if any two become disconnected, and it stops any closed loop control
of the regulator, and moves it to its lowest feed forward angle. A ’full abort’ is triggered if any tank PT
exceeds 65 bar or if all three become disconnected, and in this case the E-Reg valve is closed fully. The
full abort should never end up being triggered during a launch attempt unless there is an issue with
either main engine valve during ignition which prevents one of the propellants from flowing.
4.3.3 Integrated Testing

25 cold flow tests with the E-Reg were performed on a separate test setup prior to use in the rocket
system to tune the different empirical constants used in the controller. After that, 6 cold flows and
4 hot fire tests were performed with the full rocket system, including 2 hot fires with the final flight
hardware. The tank pressure plots from one of these tests is shown in Figure 5.
Figure 5: Tank pressure and E-Reg valve demand angle data from a cold flow performed with the flight feed
system. A slight overshoot occurs at startup which helps give a boost in thrust off the pad. This oscillation damps
out quickly within 1.5 seconds
4.4 Bi-propellant Engine (THANOS-R)

THANOS-R (Terrifyingly Hot Alcohol Nitrous Oxide System - Regenerative) is a regeneratively cooled
liquid bi-propellant engine using a pintle injector and an injector-mounted C-Class Klima rocket motor
for ignition. The engine consists of two main parts: The injector and the thrust chamber. The injector
consists of a main body that connects the plumbing for propellants and instrumentation and an inter-
changeable pintle insert. 16 successful hot fires have been performed on a prototype engine, with a
flight version successfully hot fired and tested.
7
Figure 6: (left) Exhaust plume of the THANOS-R engine during the flight qualification hot fire (right) close-up
of the assembled engine
Table 5: Design specification of THANOS-R Engine at nominal operating point
Mass (kg) 3.0

Nominal Thrust (kN) 4.0
Nominal Chamber Pressure (bar) 20
Nominal OF Ratio 2.0
4.4.1 Injector Head
Figure 7: Cross Section of the THANOS-R Engine showing all the major components
The main purpose of the injector head assembly (Figure 7) is to take the fuel and oxidiser and in-
ject it into the combustion chamber in the correct ratio, as well as mix and atomise it to allow rapid
combustion. The secondary purpose is to provide film cooling to the combustion chamber walls.
The injector consists of an aluminium main body and an interchangeable stainless steel pintle insert. On
the back of the main body are 5 threaded ports for Swagelok fittings. The central 1/2" Swagelok fitting
is the oxidiser inlet. The side 3/8" BSPP port is for the igniter cartridge, and the remaining ports are
8
for the fuel injector pressure transducer, chamber pressure transducer and heated fuel thermocouple
port. The fuel comes into the injector manifold through the bottom of the injector via the outlets of the
regenerative cooling channels of the combustion chamber.
The pintle insert fits into the back of the main body and is held in place with the 1/2" Swagelok fitting.
There is an o-ring groove to prevent fuel from leaking out the top of the injector and a sealing washer
to prevent nitrous leaking between the pintle insert and the 1/2" Swagelok fitting. This also means that
there are 2 seals between the fuel and oxidiser with a vent in the middle, which reduces the likelihood
of any mixing before injection.
The oxidiser injection is achieved with 48 x 1.5 mm diameter holes drilled radially into the pintle. The
number of holes was determined by empirically scaling our previously hot-fired rocket engine injectors’
according to the new oxidiser mass flow rate of 1.55 kg/s at a pressure drop of 4.1 bar, 24% of chamber
pressure. This reduces the likelihood of pressure oscillations due to coupling between the injector and
chamber pressures.
The fuel is injected through an annular orifice around the outside of the pintle, which impinges perpen-
dicularly with the radial oxidiser injectors, ensuring good mixing and atomisation. The annulus gap was
sized using the mass flow vs pressure drop equation for an orifice, using an experimentally determined
discharge coefficient of 0.82. The annulus injects fuel at 0.78kg/s with 5.8 bar of ∆P . Proper injector
mixing generally requires the momentum ratios of both propellants to be closely matched, which is
challenging with this engine due to the high OF ratio of 2.5, meaning the oxidiser will generally have
much greater momentum. High concentricity is maintained with a 10-degree taper on the pintle insert,
which fully constrains it in multiple axes and ensures the annulus gap is well maintained. This orifice
sizing gives a momentum ratio between propellants of around 2, which is still on the high end but has
worked well so far in the static tests.
Finally, a thin film of coolant is injected into the combustion chamber through 44 x 0.4 mm diameter
holes placed near the edge of the combustion chamber face of the injector body as shown in Figure 8.
This leads to a discharge of 0.17 kg/s of fuel or 20% of the total fuel mass flux that goes into combustion.
This results in a total fuel mass flow rate 0.95 kg/s.
Figure 8: (left) closeup of the chamber side of the injector, showing the film cooling holes. (right) Assembled
injector next to thrust chamber, showing pintle insert and all connections.
4.4.2 Thrust Chamber Geometry

The thrust chamber of THANOS-R guides the well-mixed and atomised fuel and oxidizer through its
combustion process and accelerates the products through a supersonic nozzle.
The length of the combustion chamber before the nozzle gives it an L* of 0.53m. The length came
from the maximum print height of the printers. This is smaller than most nitrous oxide alcohol rocket
engines in literature, but it hasn’t been shown to degrade the specific impulse by a significant amount.
The nozzle of the thrust chamber is made using Rao’s parabolic estimation of the ideal nozzle geometry,
at a 80% nozzle half angle with starting angles 21.5° then ending at 14°. Under sea level conditions,
9
the nozzle is designed to be expanded to 0.8 bar and is thus over-expanded. However, this only reduces
the specific impulse of the rocket engine at sea level by one second, so it is deemed negligible.
4.4.3 Thrust Chamber Cooling

The thrust chamber also has to be cooled as it is in direct contact with the hot combustion products
contained within. The thrust chamber is regeneratively cooled with only its fuel.
The thrust chamber is additively manufactured in AlSi10Mg by our sponsor Alloyed LTD for its high
thermal conductivity and low weight. A cross-section view of the engine is shown in Figure 7. Thermal
calculations for the engine were computed using an in-house thermostructural solver and validated in
RPA (Rocket Propulsion Analysis).
The fuel first enters the engine assembly through an inlet at the top, which, through a tube printed
integral with the chamber, guides the fuel into a plenum near the nozzle exit to distribute the fuel
radially around the thrust chamber. Once distributed, the fuel enters the 40 radially symmetric cooling
channels until it exits into the injector body as warm fuel.
The cooling channels were designed using both an in-house built thermo-structural solver and then
validated with RPA (Rocket Propulsion Analysis). To size the chamber cooling, CEA and isentropic
relations are used to find the thermodynamic properties of the chamber gasses at each axial position n
along the thrust chamber. Then, the heat flux across the wall can be found using the Bartz equation:
" #
0.026 µ0.2 0.8 D 0.1 A 0.9
0 cp 0 pc t t
hg = σ (4)
Dt0.2 P r00.6 c∗ R A
Where the subscript ()0 means at stagnation and σ is this correction factor also given by Bartz:
1 −0.68 γ − 1 2 −0.12

1 Tw γ−1 2
σ= 1+ M + 1+ M (5)
2 Tc 0 2 2 2
Where Tw can be assumed to be 600 K for most of the hot gas side of our aluminium chamber.
Assuming 1D heat transfer from the hot combustion gasses to the wall, then to the coolant on the other
side using the Dittus-Boelter equation, the total heat flux at each axial position n can be transformed
into one equation:
1
q˙n = 1 tw n 1
(6)
hg n + k + hc n
The hot gas wall temperature Twg can then be calculated. As rectangular cooling channels are used,
the equation for the max transverse σt and longitudinal σl at the hot wall can be found with:
2
(pco n − pg n )

w Eaq̇tw
σt = +
2 tw 2(1 − ν)k
σl = Ea(Twg − Twc )
Using the worst case where pg is negligible at engine startup and considering changing material prop-
erties as temperature of the material changes, the thermostructural safety factor at all points within
the thrust chamber can be evaluated. This is the graph for THANOS-R using pure Ethanol as the fuel:
10
Figure 9: (left) Plot of chamber wall stresses and chamber wall yield strength along the chamber (middle) Plot
of gas side wall, coolant side wall and coolant temperature along the chamber (right) Plot of wall safety factor
based on the local yield strength at temperature
To be even more conservative thermally. As the Bartz equation isn’t the most accurate, polydimethyl-
siloxane (PDMS) is employed as a fuel additive. PDMS deposits a refractory silicone dioxide lining with
an incredibly low thermal conductivity of 1W/mK as opposed to AlSi10Mg’s 100-200W/mK. In addi-
tion, by adding water to the fuel, the specific impulse of the engine is lowered. This further increases
the safety margins by lowering the combustion chamber temperature. The engine will also have to
run at a lower O/F ratio, meaning more fuel will be pushed through the cooling channels, increasing
Reynolds number and thus heat removed from the chamber walls. An updated simulation is shown
here:
Figure 10: Modified thermal simulation results with lower OF ratio and 85% with 23% water and 2% PDMS
The thermal effect of the PDMS is modelled as an additional thin insulating layer in Equation 6 with a
thickness of at least 50 microns as measured from our hot fire testing. This leads to a final fuel mixture
of 85% Ethanol, 13% Water and 2% PDMS, a very safe and reliable mixture that has worked for many
hot fires. Increased efficiency can be achieved by removing water and changing the O/F ratio, but the
lowering of risk was more prioritised in “Nimbus 24”.
4.4.4 Ignition Method

The engine is ignited using a Klima C6-P solid rocket mounted in a cartridge in the injector. The motor
is ignited using an e-match which passes through the nozzle. Having the igniter hard mounted in
the injector significantly reduces the chance of a hard start as it can’t be ejected from the engine as
the propellant flow rate is ramped up. A solid motor also gives a choked flame, meaning it operates
11
independent of chamber conditions and can’t be snuffed out. The solid motor fires directly into the
well-mixed pintle impingement fan to guarantee a successful ignition. The igniter is fired 500 ms
before the main valve open command is sent to ensure it is fully ignited before the valves open. The
thrust then ramps up from below 200N to 4 kN in less than 100ms.
4.5 Valve Actuation

The rocket utilises 4 servo actuated ball valves and 3 normally open solenoid valves. The ground station
includes 4 more servo ball valves. The solenoid valves are used without modification, but all the ball
valves require additional hardware to actuate as they are COTS manual valves.
The valve shaft of every servo ball valve is attached to a 4-bar linkage. This 4-bar linkage is actuated
using servos as shown in Figure 11. The ratio of lever lengths is chosen to output a high torque and
angular precision at the extremes of the movement, which helps overcome the “breakaway torque”
present in ball valves. The valve lever consists of grub screws that clamp to the valve shaft. This
reduces the backlash effect, which mainly occurs due to manufacturing precision.
35kgcm servos with full metal gearboxes are used, and the linkage construction is also fully metal and
uses rod end bearings to prevent wear and account for minor misalignment. For additional data, rotary
potentiometers are attached to the valve shafts to measure the true valve angle, which can be used to
characterise backlash and debug actuation failure.
The rocket consists of 1/4", 3/8" and 1/2" servo ball valves. The 3/8" valve controls the fuel flow to the
engine whilst the 1/2" valve controls the oxidiser flow to the engine. Remote venting of the oxidiser
tank is conducted through the 1/4" valve, to help cool down N2 O and increase its density. All the
ground support system valves are 1/4" to minimise cost where high mass flows are not required.
Figure 11: Servo Actuated Ball Valve design, utilizing a 4 bar linkage to transfer torque. First image with lid and
potentiometer and second image without lid
4.6 Ground Support Systems

The propulsion ground support system (GSS) is designed to safely handle the high pressure nitrous
oxide and nitrogen remotely with the manual opening of cylinders only pressurizing their attached
hoses. The quick disconnects (QDs) for pressurised fluid filling can be found in the lower feed assembly
(as shown in figure 12). A flame deflector protects the launch rail and weighs the rocket during remote
filling operations. The flame deflector also aids in disconnecting the QDs attached to the rocket. On
the rocket, the needle valve for fuel filling is also in the lower feed assembly for ease of access.
12
Figure 12: (left) The 2x pressurised fluid quick disconnect lines running down either side of the engine (right)
the 3x quick disconnects with the flame deflector which pulls off the 2x pressurised hoses during launch
4.6.1 Remote pressurised fluid filling system

The remote pressurized fluids filling system consists of a nitrogen filling line at the top and an N2 O fill
line at the bottom, as shown in Figure 13. The P&ID for this system is shown in Figure 14. Each filling
system has two servo-actuated ball valves, one to control the flow to the rocket and the other to vent
the flexible hose connecting the filling system and the rocket. The filling systems also feature a pressure
transducer, manual gauge, and needle valve for manual venting. To remove a potential failure point
from the operations, an actuated quick disconnect system has been removed, and the hose is passively
pulled off at launch by holding down the quick-release collar to the launch pad with the steel flame
deflector. A backup steel cable tied to the launch rail is also added as a redundancy in the unlikely event
of the flame deflector detaching from the rail. Entirely different quick-release connectors are used for
the nitrogen and oxidiser filling hoses to prevent swapping, and the internal attachment points are
designed so that they cannot be assembled the wrong way around.
Figure 13: GSS filling system Plate
13
Figure 14: P&ID of the GSS filling system. Further details are given in the sections below
4.6.2 Manual Fuel Loading System

The fuel is loaded manually using a commercially obtained low-pressure spray bottle, which features
a built-in manual pump, an outlet hose for the liquid, and an over-pressure valve to vent the tank. A
plastic hose uses pneumatic push fittings adapted to Swagelok to connect the spray bottle to a needle
valve present in the lower feed assembly of the rocket. The hose incorporates an in-line on-off plastic
valve to quickly restrict the flow of fuel. A 1/4" dip tube is attached to the fuel tank, which restricts
the volume of fuel to the desired amount. The dip tube is extended down to the lower feed assembly,
where fuel can be collected once it reaches the dip tube. This feature helps with visual confirmation
of the fuel tank being filled to the correct amount. This pipe is capped and also acts as a vent during
filling.
4.7 Rocket Filling Procedure

The rocket filling procedure consists of filling the fuel tanks of the rocket manually, setting up the
ground system for the remote filling of the pressurant and oxidiser, and connecting the filling station to
the quick disconnect of the rocket. The filling procedure starts when the rocket is vertical on the launch
pad in the order mentioned above, and it ensures that all steps of the filling process, which are more
than several bars above atmospheric, are done remotely. This is to ensure the safety of all members and
to increase the speed of launch operations, as fuel filling can be performed in parallel to other actions
and at any time during a launch window. The filling procedure of NIMBUS 24 is done with great safety
given Ethanol is non-toxic, and safety procedures and all pressurised operations are done remotely, but
it is also quicker as more members can be on the launch pad without PPE, performing parallel tasks.
4.7.1 Manual Fuel Filling

The filling of the Ethanol fuel mixture into the fuel tank consists of a low-pressure filling station, which
is manually operated. The ethanol is measured and filled in containers from the assembly tent by the
fuel filling team and transported to the launchpad. Once the rocket is vertical on the launchpad, a
side panel is removed, allowing access to the fuel-filling connection. Loading the fuel in the rocket
requires opening the fuel vent valves to allow air to escape, connecting the fuel port to the fuel fill hose
of the manual pump, and opening the needle valve to allow for filling. The pump is then filled from
14
the containers into the manual pump and is measured to be 6L of Ethanol. The weight of the pump
is measured and recorded, and the scale is zeroed to that value. After the connection is secured, the
manual pump is pressurised to 1-3 bar, and the valve on the fuel fill hose is opened to begin the flow
of Ethanol into the tank. The pump is measured until 6L or 4.7 kg of the fuel mixture is filled into the
tank. In the case of overflow, the outlet of the vent valve should be watched to see if any ethanol spills;
in the case it does, it is gathered with cleaning paper and is left to evaporate. Once filling is complete,
the fuel fill hose valve is closed, and any excess Ethanol is returned to the containers. The manual
pump is then vented, and the hose is disconnected from the rocket. The fuel vent valve is closed.
In the case of abandoning launch, the ethanol can be drained by connecting the fuel fill hose, opening
the fuel needle valve and allowing the fuel to flow back into the container through gravity alone.
4.7.2 Remote Nitrogen Filling

The high-pressure nitrogen is loaded into the rocket completely remotely, flowing the nitrogen through
the GSS through the quick disconnect fittings on the rocket.
On the launch pad, a high-pressure nitrogen cylinder is transported to and connected to the GSS high-
pressure nitrogen line, which is then tightly secured. The GSS nitrogen hose is attached to the quick
disconnect nitrogen fitting on the rocket. All valves including the manual needle valve, the Nitrogen
vent and fill valves are in a closed position. Once everyone except the PFS has evacuated the launch
pad, the nitrogen cylinder is opened slowly, and the isolated nitrogen line before the remote fill valve
is pressurised. As this is a high-pressure section, only the PFS is authorised to be near the GSS. The
pressure reading on the manual pressure gauge is checked to ensure it is above 250 bar. Once the N2 O
Cylinder is also opened and checked all personnel evacuates the launch pad.
The filling procedure begins once everything on the checklist has been checked and is done by the
electronics mission operator and the propulsion mission operator. Commands are given to open the fill
valve on the GSS while reading pressure on the interface for pressure on the GSS and the Pressurant
tank. When the pressure transducer on the rocket has reached a steady-state value, the nitrogen tank
on the rocket is filled, and oxidiser filling is then commenced. Just prior to launch, both propellant
tanks are pressurised by the electronic regulator, and the COPV is then topped up from the cylinder
one last time to between 230 and 250 bar. The filling valve is then closed and the hose is vented via a
remote vent valve on the GSS. Once the hose is vented, the vent valve is closed and the pressure gauge
monitored to ensure that the check valve in the rocket is not leaking. The pressure transducer on the
tank is also monitored to ensure no leaks elsewhere. This process is shown by the plot in Figure 15.
4.7.3 Remote Oxidiser Filling

The Oxidiser is filled remotely in the same manner as the nitrogen pressurant, connecting the nitrous
filling line from the GSS to the N20 quick disconnect fitting on the rocket. During the launch pad
operations, the PFS connects the N2 O hose to the designated N2 O fitting on the GSS. The Nitrous
Oxide bottle is then opened prior to pad evacuation.
When the remote filling operations are performed the fill valve is opened and the tank vent valve is
opened and adjusted to maintain a lower pressure in the flight tank compared to the “fill tank” (i.e.
the N2 O Cylinder). This creates a ∆P which keeps the fill rate steady. This process also cools down the
oxidiser to increase its density. To monitor the fill level, the rocket sits on a load cell which is bolted
onto the T-slot profile of the launch rail. A flame deflector sits above the load cell to protect both it and
the oxidiser and pressurant filling umbilical during lift-off.
To fill the tank, the vent valve is first cracked open to the ’slow vent’ position, and then the fill valve
is opened. The pressure in the flight tank rises to between 30-60 bar depending on the temperature.
After 2kg of propellant has been filled, the vent valve is opened further to the ’fast vent’ position, which
increases the fill rate by venting more gaseous nitrous oxide. When the tank is filled, a white vapour
15
Figure 15: Plot of the nitrogen COPV pressure during filling and propellant tank pressurisation.
plume is visible from the vent valve as liquid flows through the dip tube at the top of the tank. The
dip tube is sized so that this leaves a 10% ullage volume at the top of the tank. The vent valve is then
left open to allow more nitrous oxide to boil off, which cools down the remaining liquid, increasing its
density. This is done until the nitrous vapour pressure reaches 32 bar. The fill valve is then reopened
to top off the tank, without decreasing the density by much. This process is illustrated in Figure 16.
Once filling is completed, the hose is vented to atmospheric pressure to allow the quick disconnect to
be pulled off easily when the rocket launches. The rocket can hold at this point until launch go ahead
has been given. As the oxidiser warms up, a short blip of the vent valve can be done every few minutes
to keep it cool. When launch is imminent, the tanks are then pressurised from the nitrogen COPV.
Figure 16: Plot of the oxidiser tank pressure and temperature during propellant loading, chill down and pres-
surisation.
16
5 Airframe
5.1 Overview
The airframe is the major supporting structure of the rocket, whose primary function is to enable inte-
gration of the various individual subsystems, whilst also providing an aerodynamic shell. For the design
of Nimbus 24’s airframe, two different approaches have been taken for individual sections: monocoque
and skeletal. A monocoque construction involves the usage of a lightweight, one-piece, composite tube,
and the internal components such as bulkheads are mounted directly to the body tube. This structure is
chosen for minimising airframe mass and maximising RF transparency; however, it creates potential is-
sues in terms of integration and assembly. A skeletal airframe on the other hand, consists of aluminium
stringers passing through the entire section, where bulkheads, stiffening rings and various other com-
ponents are then mounted, the sections are then sealed by composite panels. However, this comes at
the penalty of increased mass, given the substantial weight of the aluminium stringers.
On Nimbus 24, the skeletal airframe is used from the engine connector to the top of the payload bay.
The skeletal airframe’s design is primarily driven by the thrust load. It is also designed to maximise on
the rail bending stiffness. The benefits of a skeletal airframe include:
• easy access to the propulsion and payload sections,
• increased mounting points for wires and boards,
• cheap and simple to manufacture.
The skeletal airframe greatly simplifies the assembly of the lower section of the rocket. However, the
upper section of the rocket requires RF transparency as it houses the antenna bay. For this reason, a
glass fibre monocoque is chosen for this section. Use of a monocoque reduces accessibility however, its
use is justified here as this section only accommodates the main parachute and antenna bay.
The breakdown of skeletal and monocoque sections through the rocket is shown in Figure 17.
Figure 17: The breakdown of Monocoque and Skeletal Sections within the rocket
5.2 Skeletal Airframe

The aluminium skeletal airframe is used in regions where RF transparency is not required, and for sec-
tions where frequent access and maintenance is essential. Each skeletal section of the rocket is made up
of three stringer sections mounted between two couplers with bulkheads mounted in between. Three-
fold symmetry has been applied within this rocket to align with the fins, that have been selected for
optimal stability characteristics. Furthermore, a threefold symmetry provides more space in between
the stringers, this therefore improves access to the rocket, whilst also allowing for the integration of
a deployable payload structure. Couplers have been incorporated at the end of stringer segments to
mount the composite panels, as well as provide a rigid, well-toleranced interface between sections.
Brackets and bulkheads are mounted on the stringers to allow mounting electronics, and supporting
tank structures.
For the skeletal airframe, Aluminium 6082-T6 has been chosen throughout given its high strength-
to-weight ratio and affordable cost. The majority of components of the skeletal airframe have been
17
Figure 18: Repeating pattern of the stringers
designed to be manufactured using only a water jet cutter and a lathe, this has been done to minimise
costs. However, exceptions have been made to some components, this is primarily in the radax-style
couplers and the engine truss, where 5 axis milling and SLS printing have been used respectively.
5.2.1 Stringers
Stringers are long slender structures, which are thus prone to compressive failure in buckling. As a
result, their design was first informed using the Euler-Buckling theorem as shown in Equation 7
π 2 EI
Pcrit = (7)
L2ef f
where Pcrit is the critical buckling load, E is Young’s Modulus, I is the second moment of inertia, and
Lef f is the effective length that is applied based on boundary conditions. The second moment of inertia
is a function of the geometry of the stringer as depicted by
bh3
I= (8)
12
where b is the width of the stringer, and h is the thickness of the stringer.
The major compressive load received by the stringers is the thrust from the Thanos-R engine, thus the
stringer cross-section was designed to resist a compressive load of 5 kN distributed across the three
stringers with a safety factor.
The first design consideration involved fixing the thickness of the stringers to 6 mm, this was necessary
in order to accommodate for locking washers at bolted interfaces and also to allow for sufficient space
for integration with other sub-systems. This limits the design variables to only the width of the stringer,
hence simplifying design. Based on the Euler buckling calculations on the longest stringer, the width
must be approximately 59 mm, to accommodate for a minimum safety factor of 1.5. However, this
provides minimal space between each stringer of only 100 mm, which is insufficient for hardware
requirements. As a result, during analyses, to create more mass optimised and efficient structures,
the stringer buckling segments were designed and analysed taking into consideration the presence of
bulkheads along the stringer, which delays earlier buckling modes and increases critical buckling load.
Based on these considerations, the longest length that the stringers needed to be designed for was
effectively 713 mm i.e. the length of the Payload Deployer. This requires a critical width of 26 mm.
The stringer design had to accommodate for mounting points to attach bulkheads, integration com-
ponents for electronic hardware, tanks, etc. To accommodate for these, holes needed to be drilled
into the stringers. To achieve this, a repeated pattern was cut into the top of the stringers. This was
chosen to create mass savings, and also provide uniform mounting points, so that during integration
components can be shifted along the stringers. Additional holes were added to accommodate for any
additional lengths. The repeated pattern is shown in Figure 18. The width and thickness of the stringer
are unchanged by the mass saving and mounting holes, this therefore, ensures that the second moment
of area, is not greatly affected and the buckling load capabilities remain high. To verify this, finite
element simulations were carried out on the stringers, in their mounted configuration, including all
bulkheads, the final stringer lengths and safety factors are shown in Table 6
18
Section Stringer Length / mm Safety Factor
Lower Propulsion 1082 2.9048
Upper Propulsion 1030 2.6978
Payload 714 2.7683
Table 6: A table showing the stringer lengths and safety factors
Additionally, to improve integration various special features have been included. This is primarily to
prevent slipping of the stringers as well as to ensure alignment. As a result, quarter circle elements
have been added at the interface between the sides of the stringer and the couplers. Furthermore,
it prevents the two edges of stringers from touching each other which can affect the alignment of
stringers. A similar feature has been added to the interface between the stringer and the engine truss,
the quarter circles have been replaced with a larger width bottom section to provide a more secure
interface.
(a) Truss Interface Artefact (b) Coupler Interface Artefact
Figure 19: Stringer Artefacts
5.2.2 Engine Truss

Given the fact that the engine injector plate contains only vertical mounting points and that the engine
is of a smaller diameter than the rocket, a component is thus required to serve as the interface between
the two. To manufacture this part using manual methods would require multiple complex components,
as well as considerable fasteners to mount it together, and this thus creates considerable complexity in
the design. Furthermore, such a component would incur large weight gains, and drastically limit the
internal space provided for propulsion plumbing as well as the igniter.
As a result of the increased complexity of said component, it was decided to use generative design as
an approach to create optimised weight saving components. The process of generative design involves
defining loads and boundary conditions, as well as geometries to preserve or to serve as obstacles
throughout. Through combinations of these the software generates the optimum structure, to minimise
mass, as well as adhere to minimum safety factors for both buckling and yield of the material. The
geometry used for generative design is shown in Figure 20a.
Following initial designs, the optimised structure that was created showed excessively high safety fac-
tors. This was primarily due to the fact that there was no limit on thickness for individual elements,
furthermore, during the process of finite element analyses, stress concentrations developed through
non-smooth transitions led to additional material being added, to create buckling safety factors of 25
or more. As a consequence, to provide more space for plumbing as well as decrease the mass of the
structure, a starting geometry was defined; this provides an initial “guess” to thus inform the design.
An upside down stringer was used in this case to connect the 3 mounting holes to the engine injector
19
(a) The geometries used to define the generative design study
(b) Engine Connector
Figure 20: Engine Truss Generative design study
to the bottom of the stringers. The thickness of the stringer was varied until a reasonable safety factor
in buckling was demonstrated.
Following the generative design process, additional fillets were added to eliminate any lingering stress
concentrations. The final design is shown in Figure 20b, and it demonstrates a safety factor of 2.6 in
buckling and 42 in yield.
To verify the results from finite element analysis, and test the structure in a real-life application, the
engine truss was SLS printed out of aluminium, in an identical state to the flight version. A test rig was
developed, and a compressive test was carried out using a mechanical testing machine to verify if the
flight structure could withstand the necessary 5 kN, and also to test a duplicate until destruction. It
ultimately failed at 16 kN, thus demonstrating the effectiveness of 3D printing as well as the viability
for using such a structure in a rocket design. The full version of the report is in Appendix D.
5.2.3 Bulkheads
In the skeletal airframe of the rocket, bulkheads have been used to provide mounting points for com-
ponents as well as serving to increase the rigidity of the stringer segments. For bulkhead design, unlike
the stringer design which focused on repeatability, and reusability, the bulkhead designs have been
highly optimised for each of their respective functions. Bulkheads are present in the skeletal airframe
section to support the weight of the nitrogen tanks, and also to support the avionics and payload de-
ployer system. Each of these bulkheads has been designed to be manufactured using water jet cutting
where possible, with post machining achievable on a manual mill, this was to minimise costs. With
the exception of the upper payload bulkhead, which has increased complexity due to the requirement
of 30 degree offset panels in the payload section. For the nitrogen bulkheads, a 30 degree taper with
foam padding has been included to align and support the weight of the tanks; whilst also allowing for
space if the tank moves vertically due to thermal expansion and contraction. Within each bulkhead,
considerable cut outs are present for weight savings, and holes have been cut to attach raised inserts
for avionics. These are shown in Figure 21.
5.2.4 Couplers
Couplers ensure that the joints between each section are rigid. This is critical between stringer joints,
to limit the loss in strength in the overall rocket airframe; and also to create a tight joint between
the stringer and monocoque sections. Furthermore, loose coupler joints in previous rockets caused
by Aluminium composite interfaces have caused significant tip deflection when the rocket is mounted
on the launch rail. For the coupler designs, the key design points to focus on are alignment, ease
of integration and creating a stiff interface in the axial direction. As a result, the latest couplers have
been designed similar to a radax coupler, wherein the bolts are mounted at 35 degrees to the horizontal
20
(a) Nitrogen Tank Bulkhead (b) Upper Payload Bulkhead
Figure 21: Bulkheads in Skeletal Airframe
to create a secure axial joint, and also align the coupler. Furthermore, the bolt mounting holes thus
point outwards from the main rocket, as a result, the couplers can be fastened from the outside. This
greatly simplifies the integration process, since there is enough space for a fastener and appropriate
tool. However, the usage of angled holes in the coupler greatly increases complexity of manufacture,
since a 5-axis CNC would be required. As a result, this component was manufactured by one of our
sponsors, to thus ensure high quality tolerances. The couplers are shown in Figure 22
(a) Stringer Stringer Coupler (b) Stringer Body Tube Coupler
Figure 22: Couplers on Nimbus 24
Additionally, various features have been added to the coupler to improve its rigidity. Firstly, a set of
flat faces have been included around the circumference to allow the component to be mounted inside
of a lathe chuck and thus simplifies manufacture. Secondly, a series of overhanging edges have been
included that allow the top and bottom couplers to slot into one another prior to fastening, this again
improves integration. Thirdly, the couplers have been designed to ensure that the angled faces are the
first to meet, to therefore ensure a well toleranced rigid joint betweeen all parts.
To mount the composite components to the couplers, radial holes have been used for the panels, and
in the body tube couplers, a 30 mm long smooth section allows the body tubes to be epoxied to the
couplers. To create the best joint, the coupler wall has been offset by 0.2 mm from the composite wall
to allow for epoxy to fill it completely.
5.2.5 Integration
For the remaining mounting points, a series of structures have been manufactured especially for each
case. These components are 3D printed to cater for the unusual geometries necessary to mount avionics,
quick disconnects etc at angles from the stringers. These components use brass threaded inserts in order
to allow them to be removed and replaced frequently. The repeated slots on the stringers serve as ideal
mounting points, given the fact that they allow the avionics to be moved up and down the rocket,
post manufacture. 3D printed components have been deemed appropriate given the minimal weight
21
of avionics components, and have been simulated and tested to resist the inertial loads of their weight
on lift off.
5.3 Monocoque
A glass fibre monocoque is used in sections where the skeletal airframe cannot be used. This is the case
where RF transparency is needed and easy access to the interior of the rocket is not required. The main
considerations for the monocoque section are that the structure withstands the applied loads and the
tube is manufactured to a high standard. The latter point is particularly important given the mechanical
properties of the monocoque are sensitive to the quality of the composite lay-up. For composite parts,
the presence of air bubbles, dry fibres, or frayed fibres can create weak points. As a result, within this
section there is a considerable focus on the manufacturing processes.
5.3.1 Materials
For composites in Nimbus 24, glass fibre reinforced plastic (GFRP) and carbon fibre reinforced plas-
tic (CFRP) have been used. Glass fibre has been utilised for RF transparency, minimal cost, and also
its simplicity in machining, given the fact that glass fibre can be processed with a high quality HVAC
system, and also causes minimal damage to electronic equipment, contrary to carbon fibre which has
conductive dust. As a consequence the skeletal airframe panels of the rocket, which require consider-
able precision have been manufactured with glass fibre. The recovery tube has been manufactured of
glass fibre given the fact that the antenna bay is mounted inside, and thus requires RF transparency.
Because this composite structure would be under the shock loads of the rocket, the thickness of the
tube is designed to be considerably larger, and care has been taken to mount the recovery bulkhead on
thickened ply packs, that therefore greatly increase the thickness at holes. The fin can and nose cone
however, have been manufactured from carbon fibre, this is because of the decreased density of carbon
fibre compared to glass fibre, and thus higher specific strength, and stiffness. This therefore allows for
thinner walls at equivalent strength, thus decreasing the weight of these components.
The glass fibres utilised in Nimbus 24 is 200 gsm 2x2 Twill E-glass fibres, the carbon fibre utilised is 180
gsm 2x2 Twill Carbon fibre, and the consolidating epoxy is quick set epoxy, to therefore accelerate the
lay up process, and eliminate opportunities for sagging or air bubbles to form. 2x2 twill composites have
been employed as compared to uni directional composites, so that tubes can be manufactured from a
continuous section of cloth, to increase strength, and also to provide the strength in the hoop and axial
directions. Furthermore, twill composites are considerably easier to work with than uni directional
composites, which therefore reduces the likelihood of frayed material, that can reduce the integrity of
composites.
5.3.2 Simulations
Given the fact that the quality of composite manufacture has a significant effect on the structural be-
haviour, as a result the critical, structural composite components were designed to withstand a higher
safety factor, than the rest of the rocket. The main critical composite that had to be simulated is the
recovery tube under the compressive thrust load case, and also the tensile load case created from the
shock load. From the figures below, it can be seen that the tube has a safely factor of over 2, thus it is
sufficient to be used in the rocket. The full simulation report can be found in Appendix P.
6 Recovery
6.1 Overview
Nimbus 24 uses a dual stage recovery mechanism consisting of separation and A clamp-band system is
used to separate the nose cone from the monocoque section allowing the pilot chute to inflate. Conse-
22
(a) First mode of buckling for the tube from compression (b) Displacement contour plots for the tube from shock loads
Figure 23: Body Tube Simulation Results
quently, the force from this pilot chute is used to deploy the reefed main parachute.
6.2 Clamp Band

The clamp band system consists of a series of Aluminium C-clamps attached to two spring steel bands.
The bands are connected together with a loop of Dyneema which is tensioned using a bolt and hinged
joint. Nichrome wire encased in a 3D printed pla housing is actuated to melt the Dyneema as detailed
in Section C.19. The ability for spring steel to store elastic energy makes it ideal for a fast and reliable
mechanical separation.
(b) Exploded Diagram of Clamp Band

(a) Clamp Band in locked configuration
Figure 24: The Clamp Band system in Nimbus 24
On Nimbus 24, the clamp band, also features a 3D printed aero cover, that protects the nichrome coil
from adverse weather conditions during ascent. The cover is attached via string and therefore opens
up but remains attached to the rocket.
Nimbus 24 uses a discontinuous reefing mechanism. A pilot chute is inflated during separation which
pulls the main reefed parachute out of the recovery tube. In its reefed state, the parachute has a 1ft
skirt diameter producing a reefing drag ratio of 0.0124 acting as a traditional drogue. The unreefed
parachute is toroidal in shape with a Cd of 2.2 and a 3m diameter.
6.3 Reefed Parachutes

The decision to use reefing was made to greatly simplify the recovery hardware required. As indicated in
Figure 25, the reduction in components greatly decreases chances of failure such as lines entanglement.
The toroidal parachute is a lightweight, high drag coefficient geometry with a packing diameter of
100mm which loosely fits into the recovery tube. Contrary to previous year, the main chute is brightly
coloured orange to increase visibility upon deployment.
23
Figure 25: Recovery system lines diagram
The dimensions of the parachute were determines based on the constraints detailed by the mission
profile. To satisfy the terminal decent velocity constraint, the COTS parachute was decided to have
a canopy diameter, DC , of 120 in (10ft), giving a nominal diameter, D0 , of 4.29m. To figure out the
diameter needed for the reefed state of the parachute, the maximum allowable descent speed under
drogue of 46 m/s was used. The reefed drag area, and hence the reefed diameter, were calculated using
experimental data presented by Knacke, for similar parachutes, and by extrapolating the relationships
presented in the manual. The steps of the calculations are presented in appendix P, using which the
reefed diameter was found to be 1 ft.
The reefing mechanism is actuated using a similar nichrome-Dyneema system which will be tested in
a runway tow test as detailed in Section C.18.
(a) Reefed State (b) Disreefed State
Figure 26: The Toroidal Reefed Parachute in both reefed and disreefed states.
Other recovery hardware includes nylon shock cords which are rated to 6.6kN giving a safety factor of
3. Light-weight locking carabiners rated to 24kN are used to secure shock cords to 8kN eyebolts.
6.4 Recovery Bulkheads

To manage the shock loads on parachute deployment, bulkheads are required to transfer the load from
a central eyebolt to the main rocket assembly. The loads are transferred into the recovery tube. The
recovery bulkhead has been designed with a focus on simplified manufacture, supporting loads and
24
also minimising weight. To simplify manufacture, the bulkhead has been designed to be manufactured
via water jet cutting. Large cut outs have been added to make it lighter. Furthermore, 6 bolt holes with
steel inserts are used to mount the bulkhead to the tube walls, to distribute the load evenly, and also
reduce the size of the bolt holes, thus reducing the damage to the composite material. Further, the steel
inserts are necessary to allow for the bulkhead to be removed and replaced frequently. The Recovery
bulkhead has a safety factor of 5.
For the upper recovery section, instead of using a conventional bulkhead, which proves challenging for
two reasons. Firstly, the design of a nose cone should be as smooth as possible to improve aerodynamic
performance and therefore bolts cannot pass through the outside of the nose cone. Secondly, to mount
the bulkhead to an epoxied inner ring proves complex given the necessity to position and mount the
bulkhead without being able to visually align it. As a result, in Nimbus 24, the nose cone tip mount is
sufficiently long to allow an M8 eyebolt to screw into the bottom, and the aluminium tip to screw into
the top. This greatly simplifies integration.
(b) A cross section of the nose cone tip and its

(a) The Recovery Bulkhead integration component
Figure 27: The recovery Bulkheads in Nimbus 24
6.5 Shock Load Calculations

The shock loads on the rocket when the parachutes deploy are substantial and hence require accurate
modelling. They were simulated using a custom SIMULINK model which was developed in-house. The
SIMULINK model is described in more detail in Appendix O. The model was used to obtain the load
on the main lower body due to the deployment of the parachute, in its reefed state, which is the most
significant axial load that the deployment would exert onto the rocket. The graph below shows the
loading on the rocket body in the aforementioned load case. Figure 28 shows that the maximum shock
load exerted on the rocket would be 2200N. Since the dis-reefing event in no way exerts an axial load
on the cords, due to the radial nature of the loading, the reefed to dis-reefed transition was redundant
to be simulated to check for the axial loads on the recovery bulkhead, as the load, for all intents and
purposes is almost zero.
25
Figure 28: Loads on the airframe due to parachute deployment in the reefed state
7 Avionics
Nimbus 24 incorporates the team’s third iteration of Electronics, learning upon failures lessons learnt ac-
crued during testing. The Ricardo avionics ecosystem provides a unified hardware and software frame-
work which scales from simple, single, solid motor rockets, to larger rockets with complex propulsion
systems. The on-going development of this ecosystem allows for reliable implementations, and faster
development times for fully featured avionics systems on ICLR rockets. During the previous year these
avionics have proved successful on all 18 hotfires conducted and also many side projects, proving the
versatile and modular capability to adapt to various implementation.
To achieve this, the following three main tenets were key to the development of the Ricardo ecosystem:
A distributed architecture breaks the avionics system into discrete nodes, where each node maintains
a single responsibility. For example, the flight controller would request a valve to move to a certain
position and it would be up to the valve control board to ensure the valve moves to the required position
either by simply moving a servo or potentially changing the set-point of a more complex control system.
This separation of responsibility allows the avionics system to be expanded and upgraded easily. To
allow the Ricardo ecosystem to adapt easily between rockets of differing complexity, configurability
of both the electronics system layout and the events occurring during the flight is crucial. This con-
figuration should also be changed during run-time rather than compile-time to reduce the complexity
when re-configuring an avionics system for a different rocket system. Transparent networking en-
ables communication with any node on the network over any physical layer. This is important as many
different physical layers are used within a rocket. Ensuring all nodes are accessible from anywhere
in the network also helps to enforce single responsibility within the system, reducing the complexity
of firmware: for example, nodes such as the flight controller only need to focus on the control of the
rocket during flight, and operations like filling the rocket can be implemented separately.
The progression of the Ricardo avionics ecosystem this year mainly focused on improving upon the
previous systems implemented within Sporadic Impulse, a hybrid rocket built for EuRoC 2022, and
Nimbus, ICLR’s first liquid bi-propellant rocket built to compete in EuRoC 2023. A number of areas for
improvement were identified following rigorous post-launch attempt analyses.
Firstly, an emphasis was placed on ensuring interfaces to the avionics system were consistent and
simple, to improve user-operability. Secondly, more robust power conditioning was identified as an
essential feature to include on all new boards. Thirdly, the difficulty of integrating some subsystems has
been recognised, thus leading to a redesign of those systems to aid integration. Finally, more extensive
26
and representative testing of the full avionics system is beneficial to ensure the hardware and software
works as expected, as well as ensuring the team members are experienced with the operational side of
the avionics system.
7.1 Avionics Overview

An overview of the electronics for Nimbus 24 is presented in Figure 29, and can be broken into three
main sections: the flight system, the ground support system, and the mission control system.
Figure 29: Nimbus Electronics Overview
7.1.1 Flight System

The flight system, displayed in Appendix D, consists of the components on board Nimbus 24 powering
and controlling the propulsion and recovery systems and relaying vehicle telemetry to mission control.
The flight system uses a centralised bus, called Ricardo Bus (RBUS), to provide logic power (3.3 V),
deployment power (14.8 V nominal) to all actuators, and communication between all on-board compo-
nents through a CAN 2.0 bus. CAN provides a noise-tolerant and multi-master physical layer, supporting
the goal of an extensible avionics system. This is thanks to CAN’s hardware arbitration, meaning that
regardless of the number of nodes present on a bus, the packets deemed most important will always
be prioritised. The deployment power rail is centrally controlled, thus the ensuring the vehicle is in a
safe state by switching the rail off.
The flight system is powered by 2 packs of dual 2s 3000 mAh Li-Ion batteries connected in series, each
pack capable of providing over 9 hours of system run-time. Each battery pack is connected to RBUS
through an independent power distribution unit (PDU), creating a redundant power system.
7.1.2 Ground Support System

The Ground Support System (GSS) (Appendix E) provides Nimbus 24 with ground power (Unlimited
power) and A CAN bus connection. The ground support system also provides control for the ground-
based valves for remote oxidizer and pressurant filling. The GSS connects to mission control via a high
bandwidth, commercially-available point-to-point WiFi link, enabling high-rate vehicle telemetry to be
transferred between mission control and Nimbus 24 whilst on the rail. This is the primary data and
command link to the rocket during ground operations. A ground station overview is seen in Figure 89
27
7.1.3 Mission Control System
The mission control system enables launch operators to effectively command, control and monitor the
rocket’s subsystems during ground operations such as remote propellant loading. Mission control is also
responsible for communication while the rocket is in flight, providing the ability to track the rocket and
locate it for recovery. This is supported through the development of station-specific user interfaces as
detailed in Section 7.5.3.
7.1.4 COTS Avionics

The CATS Vega board is used as the COTS redundant flight computer for recovery system actuation as
well as official tracking. It is powered separately from the main power bus by a standalone 3s 2500 mAh
Li-Ion battery. A 3s charging circuit, connected to the COTS system battery is additionally linked to
ground power, allowing on-the-rail charging of the battery.
7.2 Hardware
The hardware developed for Nimbus 24 builds upon previous years with the modular Ricardo-Template
(Section 7.2.1), providing a standardised schematic and layout for common components such as the
ESP32 micro controller, power management and communication circuits. This standardisation provides
a stable foundation to build hardware on top of, allowing development focus to be placed into the
board’s differentiating features. Additionally, the use of similar, well tested components across all
Ricardo hardware ensures reliability while allowing the team to keep ample back-up stock of critical
components.
7.2.1 Ricardo Hardware Template

The Espressif dual-core ESP32-S3 microcontroller has been used in the Ricardo ecosystem for the past
two years for its fast performance and cost efficiency. A CAN transceiver is included to enable any board
to access the rocket-wide CAN 2.0 bus.
The hardware template uses a buck-boost converter for logic power rectification, meaning that every
board design can support logic power inputs from 1.8-5.5V, increasing tolerance to any power distri-
bution faults. The layout of the converter and supporting passives is also provided in the template, as
loop stability and EMI output of DC-DC converters are highly influenced by component positioning and
copper plane shape. The output of this converter is connected to a low dropout regulator (LDO) for
post-rectification, resulting in a stable, low ripple 3.3 V local power supply. As the LDO may require
a minimum 500 mV dropout voltage, the buck-boost converter is set to output 4.0 V, allowing enough
headroom for the LDO to regulate down to 3.3 V.
7.2.2 Flight Controller - Pickle Rick (2024)
Table 7: Design specifications of the

Flight Controller, Pickle Rick
Dimensions (mm) 100 x 37

Sensors 6 axis IMU, Hi-g IMU
3 axis Mag, Baro, GNSS
Pyro Channels 4
Servo Channels 4
Input Voltage (V) 2.8 - 22 Figure 30: Flight Controller, Pickle Rick
Pickle Rick is the primary flight controller of Nimbus 24. The flight controller manages the propulsion
and recovery systems during flight, computes the rocket’s predicted apogee, records the trajectory and
28
handles communications with mission control.
The flight controller integrates a 6-axis accelerometer and gyroscope, 3-axis magnetometer, a baro-
metric pressure sensor, and a GNSS chip capable of receiving 4 different constellations, all for in flight
telemetry and state estimation. During various flights of previous versions of Pickle Rick, it was found
that the microSD card would either eject or crack during hard landings. Therefore, a lockable and shock
resistant microSD card slot is included for logging of sensors and deployment events during flight. A
buzzer is also included to provide auditory feedback of the flight computer’s state.
Pickle Rick features a LoRA module operating in the 868 MHz band, providing a radio link between
mission control and the rocket during flight. This module has a peak power output of +20 dBm, which
can be configured lower to comply with local regulations. Coupling this with LoRA’s excellent receiver
sensitivity and signal to noise ratio capabilities enables the flight controller to return telemetry and
tracking data at distances as high as 65km, even when using omni-directional antennas as detailed in
Section C.3. The flight controller logic power is provided by a similar architecture to the one described
in Section 7.2.1, however with a different buck-boost converter allowing for input voltages up to 22V
and therefore allowing the board to be powered by lithium batteries ranging from 1-5S. Pickle also
includes pyro and servo channels (detailed in Table 7) which can also be powered from the same
battery as logic power. The wide input range supported by the controller, as well as the inclusion of
pyro and servo channels allows it to act as the sole flight and deployment controller on smaller rockets.
7.2.3 Engine Control Unit - Stark
Table 8: Design specifications of the Engine Control Unit

Pyro Channels 1
Servo Channels 2
ADC Channels 6
Sensor Options 5V, 12V, 4-20ma
Input Voltage (V) 2.8 - 22
Figure 31: Engine Control Unit, Stark
Stark is the engine controller used to control the engine, Thanos R. Previous iterations of the engine
controller have required 4 separate boards for servo and pyro actuation, as well as sensor polling for
O/F control. These boards performed individual tasks while communicating over the CAN bus, leading
to increased hardware complexity and integration challenges. Stark provides full engine control and
fast data logging through a single PCB by integrating the functionality of the 4 boards. Consequently,
removing the engine controller’s reliance on the CAN bus increases reliability and reduces load on the
network.
Two servo channels powered at regulated 6 V control the oxidiser and fuel main valves. These valves
have independently calibrated potentiometers which relay accurate positions which can be compared
with the valves’ demand angles. This provides a useful debugging tool which can be used to check
for nominal operation. To protect the PWM channels on the ESP-32 micro controller, a TVS protection
device has been added.
Engine e-match ignition is achieved through a single N-channel MOSFET with a 12V regulated input. An
e-fuse, set to current limit at 4.98 A, protects the board in case of a short circuit while also providing
a continuous current monitor reading. Continuity across the pyro channel is detected by the micro
controller when deployment voltage is switched on and is indicated by a bright LED.
Stark has 6 ADC channels providing 24 bit resolution, primarily used for engine chamber and injector
pressure sensing. The fast polling of this data is used to implement reliable throttle and OF-ratio control.
29
Additional features include a locking micro-SD cardholder and 6 pin GPIO breakout for providing extra
functionality.
7.2.4 Servo Actuation Board - Chad (2024)
Table 9: Design specifications of the servo control board
Servo Channels 2
Potentiometer Channels 2
Servo Voltage (V) 6/7.4
Continuous Output Current (A) 12
GPIO Breakouts 6
Figure 32: Actuator Board, Chad
Each servo actuation board can control two servos, allowing only one board to be used if there are
multiple servos in close proximity, thus easing integration. The board is also used to read each valve’s
potentiometer, which provides direct feedback of the valve’s absolute position. This allows to charac-
terise the valve’s backlash, detect valve failures and automate system checks.
The servos used are commercial RC servos, hence they require voltage regulation down to 6-7.4V to
avoid damaging their integrated controllers, which Chad supplies through a buck converter. The posi-
tion of such servos is set through a PWM signal, which is provided by the microcontroller.
7.2.5 E-Reg Board - Greg (2024)
Table 10: Design specifications of the E-Reg board
Dimensions (mm) 30x24

Sensor Channels 1
Sensor Options 5V, 12V, 4-20mA
Figure 33: E-reg add-on board, Greg
Greg is a sensor breakout board for Chad. A single sensor channel can be powered using regulated
5V or 12V supply. The signal is read by an ADC on the Chad micro controller, which is used for high
data rate E-Reg pressure transducer feedback. The identical GPIO pin configuration and sizing of both
boards means that they can be integrated as a single unit using pin headers and standoffs.
30
7.2.6 High Powered Actuation Board - Flint & Steel (2024)
Table 11: Design specifications of the High Powered ac-

tuation board
Pyro Channels 4
Total Continuous Output Current (A) 20
Total Peak Output Current(A) 30
Figure 34: High Powered Actuation Board, Flint
and Steel
Flint & Steel consists of 4 actuation channels which are used for various applications within Nimbus 24.
These involve solenoid valves, used in the propulsion section, as well as Nichrome wire cutters, used in
main separation and main parachute release.
N-channel MOSFETs with good thermal dissipation properties allow each channel to operate individ-
ually for long periods of time without overheating. Flint & Steel also incorporates a current shunt
to monitor the output current, and a high side switch allowing power to connected actuators to be
switched off entirely. This allows safety limits to be introduced which protect the rest of the rocket
from a bus overload due to a malfunctioning actuator. A continuity check is also available on each
channel to ensure that devices are connected correctly.
7.2.7 Power Distribution Unit - Lightning McQueen (2024)
Table 12: Design specifications of the Power Distribution

unit
Input Voltage (V) 4-25.2
Maximum Deployment Current (A) 20
Maximum Logic Current (A) 10
Figure 35: Power distribution unit, Lightning McQueen
Lightning McQueen is the Power Distribution Unit(PDU) responsible for regulating input voltage from
the Ground Support System (GSS) or the onboard batteries to 3.3 V, which is required for the ESP32
microcontrollers. It also enables remote control over switching the logic and deployment power rails
on and off, ensuring effective power management for the rocket’s systems. Additionally, Lightning
McQueen features active power OR’ing, allowing automatic switchover between power sources in case
of failure while facilitating load sharing. To protect electrical components and onboard batteries, the
PDU also incorporates current limiting on the logic and deployment power rails during short circuit
events.
31
7.2.8 Sensor Board - Kermit (2024)
Table 13: Design specifications of Kermit

ADC Channels 4
Sensor Options 5V, 12V, 4-20mA
Thermocouple Channels 2
Nominal Data Rate (SPS) 500
Maximum Data Rate (SPS) 64000
Figure 36: Sensor Board, Kermit
Kermit is a sensor data acquisition board which features four analog to digital converter (ADC) channels
sampled with a 24 bit resolution. The ADC channels are highly configurable, supporting both differen-
tial and single-ended readings, current sensing (4-20mA) and voltage sensing for 5V and 12V sensors.
This allows for any of the channels to be used to read a wide range of sensors. Additionally, power
rails are supplied from boost converters which are post-rectified by linear regulators, thus preventing
switching noise from propagating into sensor readings. Two thermocouple channels are also available,
allowing temperature readings from any type of thermocouple.
7.2.9 CAN Repeater (2024)
Table 14: Design specifications of the CAN Repeater
CAN Yes
Figure 37: CAN Repeater
The CAN repeater allows multiple independent CAN buses to connect together, while also retaining
hardware arbitration. A correctly set up CAN bus requires both ends to be terminated to prevent signal
reflections, and this termination can not be removed during a bus’ runtime. The CAN repeater provides
termination at one end of the bus, and also retransmits any CAN packets across the electrical junction
so that after QD disconnection both buses can continue to operate nominally. The CAN repeater also
provides an option for galvanic isolation, eliminating ground loops.
7.2.10 Canard Board - Geddan
Table 15: Design specifications of the Canard board

Servo Channels 3
Servo Voltage (V) 7.4
Continuous Output Current (A) 12
Peak Output Current (A) 14
Figure 38: Canard Control Board, Geddan
32
The canard control board, Geddan, is designed to control the rocket’s canard module. This module is
nominally powered independently with a single 3s Li-Ion but, for reliability, can switch over to RBUS
deployment power if the primary battery voltage is lost. This switchover logic is implemented at the
hardware level by an ideal diode. A more detailed overview of the board’s power system is provided in
Appendix M. The battery voltage is regulated down by a buck converter, with key statistics outlined in
Table 15. The board also features a 6-DOF accelerometer and gyroscope for roll rate feedback, as well
as a 3-axis magnetometer. The board features an SD card for data logging of measured roll rates and
their corresponding actuator control outputs.
7.2.11 Miniaturised Altimeter Board - Artemeter
Table 16: Design specifications of the Miniaturised Al-

timeter board
Dimensions (mm) 18x54.5

Input Voltage (V) 1.8 - 5.5
Sensors 6 axis IMU, Hi-g IMU, 3
axis Mag, Barometer Figure 39: Altimeter, Artemeter
Storage 4Gbit NAND Flash
Artemeter is a smaller altimeter board than other COTS alternatives so has increased flexibility when
installing into a rocket. The data read from the sensors is passed into a Kalman filter and then saved
onto a NAND flash chip through a translation layer. Its use in Nimbus 2024 will be for testing and
supplementary data acquisition. The altimeter will be powered by an independent 1s1p LiIon battery
and will function independently from the rest of the electronics.
7.2.12 Power Switchover Board - Witcher (2024)
Table 17: Design specifications of the Power OR’ing

board

Maximum Input Voltage (V) 28
Maximum Continuous Current (A) 20
Figure 40: Power OR’ing board, Witcher
The power input OR’ing board, Witcher, is designed for switchover between the ground power and the
onboard batteries. Ground power is favoured over the batteries, allowing the avionics system to have
an indefinite lifespan whilst on the rail. The batteries will only be used for flight and recovery, meaning
they require less capacity and are therefore lighter. Additionally, Witcher allows for both power sources
to be switched off, hence providing the rocket avionics with remote reset capabilities. The main design
specifications of Witcher can be found in Table 17.
33
7.2.13 Antenna Bay
Figure 41: Antenna Bay
The propulsion sections of Nimbus 24 consist of a metal skeletal structure which attenuate radio fre-
quency (RF) signals significantly. Whilst in flight, live telemetry is achieved through LoRa communica-
tion where omni-directional antennas are used on the rocket to facilitate the transmission of data. The
3D Printed PLA antenna bay as seen in Figure 41 protrudes outside of the metal structure where it is
surrounded by a glass fibre body tube. Both glass fibre and PLA are RF transparent materials.
7.3 Electrical Quick disconnects

The avionics system of Nimbus 24 requires two electrical quick disconnects (QDs): one for the main
rocket bus, RBUS QD, and one for the ground power supply to the rocket, called the Adapter Power
QD. The connectors used on both QDs are directional, meaning they have to be angled downwards, to
ensure the QDs do not damage any hardware as they are pulled off at launch.
Figure 42: Nimbus electrical quick disconnect (EQD)
7.3.1 RBUS Electrical Quick Disconnect (EQD)

Whilst on the launch rail the RBUS of the rocket and the GSS need to be connected for CAN commu-
nication, voltage rail sensing and power bus reset control. In total this requires six contacts. For this
application, the current throughput of each contact is negligible. A USB-C connector has 12 contacts
and a retention force of 11 N, which was deemed suitable for this application.
7.3.2 Adapter Power Quick Disconnect (AQD)

Whilst on the launch rail the rocket is powered from the GSS through an 18 V rail with a current
limit of 20 A. A D-Sub 3 pin connector has a maximum current rating of 20 A per contact, providing a
significant margin. Combined with a retention force of approximately 20 N, this connector was chosen
for this application.
34
7.4 Software
The Ricardo avionics ecosystem contains a multi-language software stack. Embedded targets based
around the ESP32-S3 feature software are written in C++17 on top of the FreeRTOS layer, a real-
time operating system. The usage of modern C++ allows for highly modular and reusable code, as
well as the ability to take advantage of highly optimized C++ libraries such as Eigen, which provides
efficient linear algebra subroutines. Middlewares, for example interface programs running on the GSS
Raspberry Pi, are written in Python for cross platform support and better widespread familiarity within
the team.
Three libraries constitute the base libraries, and are used extensively throughout all embedded soft-
ware. LibRNP (Ricardo Network Protocol) provides a packet based network protocol, and LibRRC
(Ricardo Rocket Components) abstracts the rocket to two base classes, actuators and sensors, allowing
succinct description of the rocket flight system. A further companion library for LibRNP called pylibr-
RNP provides helper classes to serialize and deserialize RNP packets in python based applications. To
improve standarisation of embedded software, the core framework running on every PCB is included
in Libriccore. This library provides a base ’system’ object, which can be derived to implement board-
specific features - allowing team members to focus on the unique functionality of every board. The
system also implicitly runs some fundamental modules: a finite state machine to break down the op-
erational runtime of the program into discrete states, a LibRNP network manager enabling networking
on every node, a global command handler where commands can easily be registered and defined, a
core logging framework providing a unified method for system message logging, and a system status
objects that tracks error states of the application in a single bit field providing an efficient method for
tracking of multiple error flags.
7.4.1 Apogee Detection

To robustly detect apogee, an algorithm based around successive quadratic polynomial fitting is used
on altitude data from the flight controller Kalman estimator, sampled at 50Hz, across a shifting win-
dow of 100 data points. The subsequent time series is summed and the resulting system is solved using
householder QR decomposition. The resultant quadratic fit coefficients allow for a complete character-
isation of the vehicle’s trajectory under an arbitrary constant acceleration. As the vehicle approaches
apogee, the vertical acceleration of the vehicle tends towards gravitational acceleration, and hence the
quadratic fit to the vehicle trajectory will converge. The maxima of the altitude-time quadratic fit as
well as the time when the rocket should reach this maxima can easily be calculated analytically from
the quadratic coefficients. Finally, by simply comparing whether the current time is greater than the
predicted apogee time, as well as verifying the current altitude is lower than the apogee, apogee can
be detected in a computationally efficient, and noise-resilient way. To further improve the algorithm’s
robustness, a minimum altitude of 100m is required for the algorithm to be able to trigger as well as a
Mach lockout, which prevents apogee prematurely being detected due to compression shocks forming
on the rocket above transonic speeds. This process is further detailed in Section C.1.
7.5 Ground Software

Multiple software running together in a stack provide the frontends and backends which are used
by operators to interact with the rocket and ground support equipment. The stack is comprised of a
combination of both in-house software, such as Ricardo-Backend and Ricardo-CommandServer, and
existing third-party solutions, namely Grafana.
The individual services in the stack are capable of running as standalone, native applications, since com-
munication between the services occurs over Internet Protocol (IP) sockets. Nevertheless, it was decided
to use a containerised architecture for the various components. This provides a number of benefits, no-
tably ease of deployment and management. The use of Infrastructure as Code (IaC) methodologies
35
means that a ground system server, in most cases a Raspberry Pi running Debian, can be provisioned
automatically in a number of minutes, as opposed to requiring an extended, manual configuration.
7.5.1 Ricardo-Backend
Ricardo-Backend provides a translation layer between the RNP network, connected to the flight and
ground support systems, and the IP network used for the mission control network. This translation
allows for the use of typical IP clients, such as laptops and tablets, to receive telemetry and send com-
mands to the various systems onboard the rocket and the on the ground.
One of the key components of Ricardo-Backend is its Socket.IO server which provides a two-way
publish-subscribe queue. This server provides multiple channels, including a telemetry channel which
provides a telemetry stream as it is decoded from the RNP network, and a packet channel which can
be used to send encoded packets to nodes on the RNP network. A lower-level WebSocket fallback is
also available for use with clients that do not support the full Socket.IO implementation.
Telemetry is provided by nodes on the RNP network by request. This is managed by the Data Task
Request Handler (DTRH) which generates telemetry requests and deserialises data returned from the
nodes on the rocket and ground systems. Request tasks are defined within a JSON configuration file
(see Appendix K) which contains the expected packet structure and the structures of bit fields to support
their decoding into individual variables. A browser-based front end, as shown in Appendix L, allows
for live modification of the configuration, including creating new tasks, modifying existing tasks, and
starting and stopping tasks as required. Finally, logging functionality saves received telemetry to disk
for use during post-flight analysis.
7.5.2 Ricardo-CommandServer
Ricardo-CommandServer is used to streamline the process of sending a command to a node on the RNP
network. A series of HTTP endpoints are provided, corresponding to common commands. These can be
as simple as sending a single, preset command to a single node, a command which accepts arguments
via a JSON-encoded payload, or even more complex behaviour where multiple commands are sent to
multiple nodes in preprogrammed sequence.
Ricardo-CommandServer provides the important capability of enabling any client capable of generating
HTTP requests to send commands to nodes. Individual clients are not required to generate their own
encoded RNP packets and, therefore, do not require specific knowledge of the RNP network configu-
ration.
7.5.3 Grafana
Grafana is used as the primary method for operators to interact with the rocket and ground systems,
both for visualising telemetry and for sending commands. The large number of available plugins provide
key features, such as integration with the WebSockets provided by Ricardo-Backend to receive telemetry
and the ability to add buttons to dashboards to send commands to Ricardo-CommandServer.
Well-designed user interfaces are key for providing operators with information in a timely and quickly
intelligible fashion. This improves reliability and safety when conducting operations. Grafana dash-
boards are easily customised, enabling rapid iteration during development based on the feedback of
operators. Examples of the display of data with Grafana, as well as control buttons can be seen in
Figure 43.
7.6 Arming and Ignition

The arming and ignition of Thanos-R within Nimbus 24 must be carried out in the correct order and at
the correct time to ensure nominal operation. To ensure this process remains safe, multiple levels of
36
Figure 43: Mission control command interface during flight qualification hotfire
arming are required.

The power system operates in four functional states:
• Off - Both logic and deployment rails are turned off.
• Powered - Only the logic rail is powered.
• Ready - PDU awaits command to power the deployment rail.
• Deployment Live - The deployment rail is energized.
In the Powered functional state, the logic rail is switched, on which allows on-board microcontrollers
to function normally. However, actuation power is off, meaning the rocket is in a safe state. Once the
remove-before-flight pins are removed from the arming interface, the PDU transitions into the Ready
state, where the deployment rail is still inactive. After all personnel have moved to a safe distance
from the rocket, a remote command will be sent from mission control to energize the deployment rail.
Although the deployment rail is live, each flight component must also be armed by a discrete command
before the component can be actuated.
When propellant filling is complete, the flight computer is transitioned into the launch state. It then
performs a flight check, where it attempts to arm each rocket component. If any component returns
an off-nominal status, the flight check fails and ignition cannot proceed. All rocket ground operations,
such as switching the deployment line on, arming components, and the vehicle launch command all
require multiple, deliberate actions to be performed.
8 Active Control Systems - Canards

This year, the team decided to incorporate an active control system into Nimbus 24’s design, a system
typically necessary for higher altitude launch vehicles to ensure they follow their predicted trajectory.
Hence, Nimbus 24 will be used as a technology demonstrator for active orientation control and as this
will be a flight demonstration, it was ensured that the control system would be simple, reliable and
non-critical for flight to ensure its safety. As a result, of the three rotational dynamics: pitch, yaw and
37
roll, only roll control was selected for Nimbus 24 to prevent the system from drastically impacting the
rocket’s translational dynamics.
8.1 Canard Roll Control

The team hopes to employ active roll control through the inclusion of three actuating canards which
all deflect by the same angle of attack in order to impose a rolling moment on the vehicle with minimal
impact on pitch and yaw. A rendered image of the canard module is presented in Figure 44.
Figure 44: Canard Module Render
With the rocket having a potentially asymmetric body due to aerodynamic shrouds and launch lugs, the
vehicle will have a natural tendency to roll. Moreover, in the worst case of a 1 degree fin cant caused by
misalignment during the manufacturing stage, simulations show the vehicle can reach a maximum roll
rate of 500◦/sec. This inherent roll might be beneficial for spin (gyroscopic) stabilisation however, can
be problematic at apogee when the payload is deployed and the rocket separates for recovery. Hence
roll control will be implemented on the vehicle beyond 1500 m, at which point the canards deflect to
de-spin and reduce the roll rate of the rocket until apogee to minimise the risk of entanglement of any
chute lines during recovery separation and deployment of the payload parafoil.
Being one of ICLR’s first flight attempts at an active control system, the canards are intentionally un-
dersized as the system’s overall objective is to observe a non-negligible impact on Nimbus 24’s roll
orientation without compromising the stability of the vehicle and ensuring it is not a flight critical
component. In doing so, the team can also gain experience in handling active control systems in prepa-
ration for moving to more complex systems such as canard pitch control and thrust vector control in
the future.
8.2 Profile and Planform

The NACA 0012 aerofoil was selected for the canard profile due to its high lift-to-drag ratio and low
drag coefficient. Whilst the aerofoil’s symmetry allows it to be more easily manufactured and its 12%
thickness to chord ratio provides sufficient space for the attachment of a shaft for rotating the canards.
The canard’s trapezoidal planform was again chosen for its ease of manufacture with its overall size
being informed by static stability constraints. This is because a too large a canard would ’pull’ the
centre of pressure much further up the rocket’s longitudinal axis and closer to its centre of gravity,
thus reducing its static margin for a given fin size. Subsequently, the fin planform size would have to
increase to lower the centre of pressure or ballast added to the nosecone to move the centre of gravity
closer to the nose to increase the static margin. However, both these options result in either increased
38
mass or aerodynamic drag, reducing the maximum achievable apogee of the rocket. So the final canard
planform ensured a balance between shifting the centre of pressure and minimising aerodynamic drag.
Through Python simulations, the aerodynamic performance of the final canard was computed by de-
termining the 3D lift curve slope of the canard based on planform parameters and the aerofoil’s 2D
aerodynamic characteristics where the maximum allowable deflection of the canard before stalling was
14 degrees. Using this, the maximum lift produced by each canard at maximum deflection, velocity and
air density was approximately 200 N. The change in trajectory and stability caused by flying with and
without canards can be seen in greater detail when comparing the nominal and off-nominal simulated
flight data detailed in Section 10 and Appendix B.
8.3 Actuation Mechanism

The canard module is actuated using a mechanically linked bevel gear system driven by a single servo.
Each canard is mounted to their respective shaft by bonding an aluminium insert to the the back surface
of the canard. Following this, each shaft is constrained in place by two support blocks along its length
with a single bevel gear mounted with a grub screw on the opposite end of the shaft to the canard.
The three canard bevel gears are then driven by a single bevel gear connected to a 35 kgcm torque
servo. Moreover, in the event of a power loss to the servo the servo and as a result the canards become
free spinning. However as the canard’s pivot point on the shaft is located at its centre of gravity with
the centre of pressure lying further downstream, if the canards become free spinning, they will have a
stabilising moment aligning them with the rocket’s angle of attack.
By using a bevel gear system with a single servo to actuate the canards, all three canards are mechan-
ically linked together resulting in all three canards always rotating by the same angle. Consequently,
the canards are only able to impact the roll dynamics of the rocket, and therefore reduces the chance
of any unwanted pitch or yaw changes caused by the canards.
8.4 Control System

The control system for the canards activates once the rocket has surpassed 1500 m AGL and aims to de-
spin the vehicle once it reaches apogee. Since the canards are actuated during the coasting phase the
rocket is decelerating which results in reduced control authority and lifting capability of each canard.
Hence, it is unlikely that an exact non-zero roll rate can be achieved, although it is expected that the
roll rate of the vehicle will experience a visible decrease as it nears apogee.
A simple P controller is implemented to compute the desired angle of attack required to minimise the
error between the roll rate set-point of 0 ◦/sec and the measured current roll rate of the rocket. The
team opted for the inclusion of only a proportional gain as the system is inherently stable due to the
main fins’ damping effect, with the proportional controller aiming to reduce the roll moment’s transient
period. Using this simpler controller on a stable dynamics system also helped to minimise the canards’
risk to Nimbus 24’s nominal flight. Adding to this, the control law’s simplicity allowed the team to
easily tune the controller and avoid issues associated with more complex controllers such as integral
wind-up.
9 Payload - FAWKES
9.1 Overview
EuRoC’s objectives for the ‘payload’ are scored primarily upon two objectives. The first is in the form of
‘creative scientific experiments and technology demonstrations’, the second is in the form of outreach
and using the payload as an ‘educative’ tool for students in the local area, through collaboration.
39
The 3U CubeSat payload, FAWKES (Flight Autonomous Winched Kite Experimental System), adheres
to these principles with the following features:
– An educational-focused 0.5U designed in collaboration with the Northern Ireland Advanced Com-
posites Centre (NIACE)
– To be a second flight test of the guided recovery system, where we can characterise vital systems
and in-house dynamic systems (no active feedback control in this iteration)
9.2 Payload Structure

The structure uses aluminium 6012 T6 in various forms to make the stringers, mounting bulkheads and
top & bottom lids, and is made to be RF transparent via GFRP which is pinned to the middle bulkhead.
An effort was made to simplify the platform where experiments are mounted, which was made using
a sheet of aluminum that was water jet cut and post-machined. The boards and batteries are mounted
through Tough PLA prints which are mounted to the stringers. The specific manufacturing methods
and materials can be found in the global bill of materials (Table 31).
Figure 45: Payload Structure Overview
9.3 Electronics System

The electronics system consists of a Pickle Rick Flight controller that conducts the actuation of the
parafoil lines. Pickle Rick will also provide live telemetry such as GPS, and sensor estimation data.
There is also a CATS Vega, which is being used as the EuRoC Mandatory gps tracker. Both systems are
powered by individual 2S Li-ion batteries which are toggled by an XT30 switch.
9.4 CubeSat Experiments

9.4.1 Guided Recovery In Flight For Improved Navigation (GRIFFIN)
A novel feature of the 3U payload for Nimbus 2024 is a fully autonomous guided recovery system.
This system is the second iteration of a system we have flown before, with some adjustments made in
response to the data acquired in the previous flight. The system is designed to alter the trajectory of the
payload via the reefing of two parafoil lines, causing the parafoil to bank when one line is pulled while
the other is extended. From this robust flight test, valuable position and velocity data will be captured,
enabling future flights to employ more advanced trajectory alterations including plans to implement
targeted tracking of a location based on initial coordinates of the payload at deployment, as well as
mid-flight stabilisation when hit by a gust of wind.
40
To note - the actuation of the lines will be pre-programmed and will not rely sensor input for the lengths
of lines retracted. The pre-programmed actuation sequence can be found in Table R.
The actuation of the break lines occurs via a single central servo, indicated by "Gear(Servo)" in Fig-
ure 46. This central gear transmits torque into two neighbouring Gears, "Gear(Shaft)", which each
rotate opposing ways to induce a bank angle on the parafoil by simultaneously extending one line
while retracting the other. The overall efficiency of the system is 79.6 % based on a standard pressure
angle of 20◦ , with a peak force output of 160 N per line. For the actual reefing operation, kevlar line
is tied using an uno knot (commonly used in fishing) around two titanium shafts that are constrained
by two bearings that sit within housings (bearing housing on bottom and the canopy on top). A hole
in the shaft allows a knot to be tied once line is threaded through it, preventing the line from coming
loose. Then the lines pass through M5 eyebolts (rated for 1200 N WLL each) so that when the parafoil
deploys the eyebolt takes majority of the initial 500 N shock load per line. Then the line passes through
two countersunk holes in the top bulkhead to connect to the control lines for the parafoil. Once the line
passes through the holes in the upper bulkhead, a final knot is tied to limit the length of spool that will
be retracted/released to a set amount, providing a physical fail-safe mechanism to prevent the system
from banking too far and risking an unstable descent. Finally, sheaths are mounted around the shafts
to help prevent the line from expanding out whilst the shafts rotate.
Figure 46: Guided Recovery Model
The system has been successfully ground tested in a wind tunnel, as seen in the tests in Table R. Risk
mitigation for this, and the overall payload system as a whole, can be found in the global hazard analysis
table in Appendix G.
9.4.2 External Outreach - Northern Ireland Advanced Composites Centre (NIACE)

Project Hermes ties in all the experiences the NIACE team has had in their time with all levels of
engineers, from hobbyists to high-level professionals. After working with many companies around the
world solving problems and in Research and Development, a need was identified to educate and inform
people on all aspects of the space sector. This is done by creating a low-cost small pocket sat with high
capability and adaptability, which enables students and/or upcoming engineers who may not have
education in various aspects of engineering, to be able to learn through actions, rather than routine
theory. Hermes aims to make space more accessible through different routes. These being:
– Using off the shelf parts to create a fully working, integrated payload system that is capable of
being adaptable, low cost and easy to “play with”;
– Capability of uploading test code while in orbit to allow others to read sensor data, take images,
41
and debug code prior to launching their own Hermes.
Although in its verification and feasibility stages, the hope is that with tests conducted over time,
reliability of the system can be proved, where details can be found in Appendix R.
The aim of Hermes’s first launch is to test all internal systems as listed in Appendix R, where more
in-depth descriptions can also be found. The onboard battery has enough power to last approximately
6 hours without solar intervention. There are 4/5 solar panels placed on the exterior of the frame, each
with dimensions of 50mm x 50mm x 50mm. This configuration allows the use of 8 panels within a 1U
space. The configuration is powered by a NiMH battery in a metallic casing.
Hermes has a pull-before-flight pin that disengages the battery from all other units. This hard separation
serves as both a safety feature and a means to preserve energy on board. Hermes V1 has a range of
approximately 20km without antenna intervention techniques and adopts a transceiver and onboard
storage, allowing Hermes to store data in the absence of a signal.
9.5 CubeSat Recovery

To control the descent portion of flight of the cubesat, a ram-air parachute (parafoil) was designed and
manufactured, pictured in Figure 47a. This parafoil development aimed to refine flight characteris-
tics, which will validate an in-house six degree of freedom (6-DoF) flight simulation and its predicted
aerodynamic coefficients. This validation is crucial for scaling the parafoil design for larger systems
in future applications. Additionally, the parafoil is paired with the GRIFFIN module to evaluate the
guided recovery of the payload.
The parafoil was designed with the following constraints in mind:
– Stability of flight: Ensuring reliable and controlled descent;
– Dimensions: Fitting within the constraints of our rocket’s deployment system;
– Descent velocity: Complying with EuRoC guidelines for safe descent rates.
To support the design, a 6-DoF model of the parafoil was developed based on its dynamics, drawing
on methodologies from relevant literature where the system state space was analysed. This model
facilitated the sizing of the parafoil and allowed for accurate trajectory predictions, which are detailed
in the subsequent section. For precise specifications, refer to Table 52.
(b) Parafoil inflated at 9 ms−1

(a) Parafoil
Figure 47: Payload Parafoil
The parafoil features a span of 1.27 meters, a mean chord of 0.37 meters, and 24 bridle lines connecting
the parafoil cells to two brake lines. The chosen airfoil section, Clark-Y 11.7%, was selected for its
proven reliability and effectiveness in ram-air parachute design. Following, the design of the planform
geometry, the parafoil’s aerodynamic characteristics and performance, used for simulating its descent
42
trajectory, were determined using both analytical methods and wind tunnel testing. These were then
incorporated into the custom six-degree-of-freedom trajectory simulator to determine the descent flight
and in particular the terminal velocity and drift radius, as discussed in Section 10.
For manufacturing, ripstop nylon was used due to its durability and favorable strength-to-weight ratio.
The individual cells were stitched with nylon thread and reinforced to enhance the overall reliability
and safety of the system. Kevlar bridle and brake lines were chosen for their high strength and were
sized to handle anticipated shock loads during deployment. Line lengths were initially determined
using Surfplan, a kite-making software, and were subsequently fine-tuned during a wind tunnel test
campaign to ensure proper sizing and even load distribution across all cells.
The parafoil is folded using a method outlined in the Appendix R. This method has been flight-tested
and wind tunnel-tested, can unravel, inflate the parafoil, and perform nominally in-flight without the
lines becoming entangled. This method includes folding the break lines such that they are separated
when stored in the payload deployer box, whilst ensuring that there is little to no interference between
the two sides in deployment. The parafoil is then placed into the cup to keep it compressed.
9.5.1 Deployment mechanism

Deployable payloads enable us to incorporate scientific experiments and prototypes from institutions
who aim to flight-test their hardware.
The mechanism is inspired by door locks, where a rotational force via a servo releases a ‘lock arm’
which gives the box the freedom to rotate. The box rotates via the pre-tension in two spring hinges,
which enables the payload deployer box to rotate and the payload to be ejected. There are two lock
systems on either side of the deployer box, with one acting as a fail-safe in the case the other lock arm
fails before or during flight. The tension which enables the lock system to actuate is via 3mm bungee
cord. A white box encases the payload’s parafoil as the payload falls out of the deployer to minimize the
chances of it getting caught with the side of the deployer box. The deployment of the entire payload
system, with the parafoil included, has been successfully ground tested Section C.14.
The sides of the box itself has been made from folded aluminium components which itself is pinned
to the top plate of the deployer box via four angle brackets. The bottom plate of the deployer opens
through a simple hinge mechanism, where the door opens under the weight of the deployer.
Figure 48: Left to right - (i) Undeployed state of the deployer; (ii) Deployer with arm unlocked (iii) Deployed
state of the deployer with the payload. (iv) Render of the entire assembly.
The FEA simulations on the deployer box for the required load cases needed on this mission can be
found in the Section Q.3.
43
10 Trajectory Analysis
The flight simulations for Nimbus 24 were completed using OpenRocket and RocketPy v1.4.2, both of
which employ six degree of freedom ascent simulations and three degree of freedom descent simula-
tions. The final trajectory results presented were computed using RocketPy, whilst OpenRocket and
Autodesk Fusion 360, were used for both high and low level component modelling, allowing for a
detailed characterisation of the rocket’s mass, centre of gravity and inertia tensor. Simulations were
also computed using a Python-based in-house trajectory simulator with Nimbus 24’s flight being cru-
cial to the simulation environments validation. Moreover, as the deployable payload descends under a
parafoil, RocketPy could not be used and so the team designed a separate custom six degree of freedom
flight simulator to predict and analyse the payload’s trajectory.
10.1 Nominal Thrust Curve

The nominal thrust curve forn Nimbus 24 is based on the profile achieved during the full duration flight
qualification hotfire on 14/09/2024. This final thrust profile is presented in Figure 49. During this test,
the propulsion system reached a total impulse of 21,200 Ns with average and peak thrusts of 3.0 kN
and 4.1 kN respectively. Additionally, the nominal burn time of the system is 7 seconds.
Figure 49: Flight Qualification Hotfire Thrust Curve
10.2 Nimbus 24 Trajectory Simulations

Nimbus’ trajectory was simulated using a 12 m launch rail with an inclination of 85 degrees and heading
of 133◦ with the launch sites latitude and longitude being 39.23229◦ and -8.17203◦ (corresponding to
those of the EuRoC launch site at Campo Militar de Santa Margarida). Additionally, the atmospheric
and wind models were taken from a GFS forecast model for the specified launch coordinates with a
surface wind speed of 8 m/s.
As seen in Figures 50 and 51, Nimbus 24 is expected to reach an apogee of 3441 m AGL with a full
duration burn, experiencing a maximum velocity and acceleration of 271 m/s (Mach 0.789) and 56.3
m/s2 (5.74 g) respectively. The time to apogee is estimated to be 27.8 s whilst engine burnout occurs
at approximately 7 s into the 187 s flight at an altitude of 1067 m. To ensure the rocket is stable
at launch, EuRoC imposes a minimum off-the-rail departure velocity hence careful consideration was
required when designing Nimbus to meet these requirements. From the simulations performed, Nimbus
24’s predicted off-the-rail velocity is 31.167 m/s with a rail departure time of 0.902 s and rail departure
thrust-to-weight ratio of 6.231. During the rocket’s descent, a reefed chute is used to slow the rocket
before ground impact with its 13 inch drogue configuration used for fast descent and its disreefed 10
ft main configuration used for slow descent and touchdown. The parachute’s drogue configuration is
44
deployed at apogee with a deployment velocity of approximately 20 m/s. Then at 450 m the system
is disreefed and the main chute inflates at a velocity of 39 m/s and slows the rocket to a ground impact
velocity of 4.41 m/s.
Figure 50: Nimbus 24 Predicted Trajectory
Figure 51: Nimbus 24 Velocity and Acceleration Profiles
10.3 Monte Carlo Simulations - Landing Zone

Monte Carlo simulations were performed to examine the variability in the rocket’s flight performance
and in particular its landing zone due to uncertainties arising from a combination of: environmental
conditions, estimation of inertial properties, and manufacturing tolerances of aerosurfaces.
Uncertainties were quantified for the rocket’s inertial and geometric properties. The ascent and descent
simulations were ran separately, this rationale was adopted to model the payload deployment which
would change the inertial properties of the rocket during flight. The descent simulations were all ini-
tialised using the final state vector of the ’nominal’ ascent flight phase. Ascent and descent simulations
were iterated over under different wind speeds, launch elevations and headings to estimate the range
of landing zones for the rocket as illustrated in Figures 52 and 53.
45
Figure 53: Apogee and landing points for descent
Figure 52: Apogee and landing points for ascent sims
sims
Using both Figures 52 and 53 the maximum drift radius and touchdown location for Nimbus 24 is
approximately 2.7 km under worst case wind speeds of 8.9 m/s. Nonetheless, the rocket’s descent
trajectory remains within the maximum drift limits stipulated in the EuRoC requirements.
10.4 Nimbus 24 Stability Analysis

Nimbus has an initial static margin of 2.379 cal which increases to 2.425 cal at rail departure meet-
ing EuRoC’s minimum stability requirement of 1.5 cal. The static margin then continues to rise to a
maximum of 2.989 cal during flight.
Figure 54: Nimbus 24 In-Flight Stability
10.5 Nimbus 24 Nominal Trajectory Summary

Tables 18, 19, 20 and 21 summarise Nimbus 24’s rail departure, burn-out, apogee and impact states:
Table 18: Nimbus 24 Rail Departure States Table 19: Nimbus 24 Burn Out States
Time 0.902 s Time 7.00 s

Velocity 31.167 m/s Altitude 1067 m (AGL)
Static Margin 2.425 cal Velocity 264 m/s
Thrust-Weight Ratio 6.231 Mach Number 0.766
46
Table 20: Nimbus 24 Apogee and Impact States
Apogee 3441 m (AGL)

Apogee Time 27.8 s
Time of Impact 187 s
Velocity at Impact 4.406 m/s
Maximum Drift 2174 m
Table 21: Nimbus 24 Ascent Maximum State Values
Maximum Speed 271 m/s at 6.28 s

Maximum Mach Number 0.789 Mach at 6.29 s
Maximum Acceleration 56.3 m/s2 at 1.00 s
Maximum Gs 5.74 g at 0.54 s
10.6 Deployable Payload Descent Simulations

With RocketPy unable to accurately model the dynamics of a parafoil, an in-house six degree of free-
dom simulator was designed to simulate payload parafoil’s descent trajectory. The two descent flight
variables of particular interest to simulate were the terminal velocity and maximum drift distance.
These variable were computed using aerodynamic coefficients derived from two methods. The first
were normal and tangential force coefficients from wind tunnel testing and the second were lift, drag
and pitching moment coefficients from analytical methods.
Focusing on the simulations using the analytical aerodynamic coefficients, the terminal velocity reached
by the payload before descent is approximately 9.845 m/s. Moreover, the maximum drift radius of the
payload under this descent velocity and assuming it is falling from 3 km is predicted to be 4.7 km.
In contrast, the terminal velocity achieved when using the aerodynamic coefficients derived from the
wind tunnel tests was approximately 8.85 m/s.
The difference between the two terminal velocities is most probably due to simplifications made in
the analytical aerodynamic analysis and assumptions from the wind tunnel testing. Despite this, both
terminal velocities are within EuRoCs requirements for a deployable payload.
Figure 55: Nimbus 24 In-Flight Stability
47
11 Mission Concept of Operations
1. Manual Filling: The rocket is loaded onto the launch rail, electrical quick disconnects are con-
nected, the power rails are energised, the boards and actuators are checked for nominal opera-
tion whilst horizontal, meanwhile fuel loading equipment is prepped for filling. The launch rail is
erected and filling lines for methanol, nitrogen and nitrous oxide are connected. Methanol filling
is performed first and is carried out manually. Once this has been completed, the launch area is
cleared of unnecessary equipment, the nitrous oxide and nitrogen bottles are opened, the launch
pad crew moves to a safe distance from the rocket and the deployment rail is remotely switched
on in preparation for remote filling.
2. Remote Filling: The ground support system and rocket actuators are armed. Nitrous filling
commences and is performed first, followed by nitrogen. Once the propellants and pressurant
have been filled, the nitrous oxide and nitrogen hoses are vented for automatic disconnection on
lift-off. The propellant tanks are pressurised and the final flight system checks are performed. If
no anomalies are identified, the avionics are transitioned to the launch state and the countdown
for ignition can commence.
3. Lift-Off: At zero, the ignition command is sent, igniting the engine. The rocket begins accelerat-
ing, detaching the electrical, nitrogen and nitrous oxide quick disconnects under the thrust of the
rocket. Once an acceleration of at least 2g and 50m AGL is reached, the avionics system detects
liftoff.
4. Powered Flight: This stage begins once the rocket is clear of the rail and liftoff has been detected.
The engine continues to burn for 5.5 seconds up to approximately 900 m (the simulated burn-out
altitude) reaching a peak Mach number of 0.877. The flight computer is constantly recording data
from its sensors and the rocket’s telemetry, which is relayed back to mission control. During this
flight phase, the canards are used to reach and maintain a specified roll rate for spin stabilisation.
5. Coasting Flight: This phase begins once the engine burns out. The engine shuts down by clos-
ing the oxidiser valve fully; venting and burning the remaining methanol through the engine,
safely disposing of the fuel. Upon depletion of the methanol, remaining nitrogen exits through
the engine. Nitrous oxide and fuel tanks are also vented through the designated normally open
solenoid vent valves. The canards continue to spin-stabilise the system whilst coasting until 2000
m at which they are used to de-spin the rocket in preparation for separation and payload deploy-
ment.
6. Drogue Chute Deployment and Rapid Descent: The process of recovery begins at the apogee
mark, during which the parachute in a reefed configuration is deployed by activating the mechan-
ical separation system with a pilot chute attached to the parachute bag to ensure deployment.
This is initiated by the SRAD avionics system with the COTS flight computer acting as a redun-
dancy. The drogue slows the rocket down to an average speed of approximately 42 m/s. Shortly
after drogue deployment, the payload with an independent recovery system is released which
descends under a guided parafoil, following a pre-planned spiral trajectory.
7. Main Chute Deployment and Slow Descent: At an altitude of 450 m, the parachute is disreefed
to its main chute configuration. The actuation of the disreefing mechanism is triggered by the
SRAD avionics system with the COTS flight computer acting as a redundancy.
8. Touchdown and Recovery: This phase begins when the rocket and payload have touched the
ground. The ground impact velocity of the rocket is approximately 4.5 m/s and is slow enough
to ensure that none of the rocket components structurally fail on impact. Live GPS data will be
continuously transmitted from the rocket and payload by both the SRAD and COTS flight comput-
ers on two separate frequency bands ensuring successful recovery of the rocket and payload. An
inbuilt speaker on the SRAD flight computers will also be emitting audio for easier identification
of the exact touchdown location. This phase ends with the recovery of the rocket and payload
48
and marks the end of the mission.
Figure 56: Concept of Operations
49
12 Conclusions and Outlook
The report presented a comprehensive overview of ICLR’s 2024 EuRoC entry. It highlighted the main
design features of Nimbus 24 and how the team implemented improvements and changes from previous
learning experiences.
The liquid bi-propellant engine, THANOS-R, has been successfully test fired 2 times as of the initial
submission date of this report; producing up to 21,000 Ns of impulse, a peak of 4.1 kN of thrust, and
an ISP of 200 s. The thrust at liftoff and throttle-up time have exceeded requirements, and are within
75% of our target impulse of 20000 Ns. The teams’ filling procedures have been perfected and reliably
demonstrated 20+ times, including tests for filling, pressurisation, nominal engine runs as well as abort
modes and propellant off-loading. Further full wet dress rehearsals are set to be performed throughout
September to simulate launch day procedures.
The skeletal airframe design has been tested extensively through simulations and mechanical structural
tests. This has helped develop a mass-optimised airframe with sufficiently high safety factors under
max thrust and shock loads. This ensures that critical load bearing structures will not buckle nor yield
during flight. Use of light weight, high strength composite materials in panels has further reduced our
dry mass while providing an aerodynamic skin to the rocket. Broadening our manufacturing techniques
to CNC milling has allowed us to create intricate coupler and bulkhead designs necessary for increasing
bending stiffness at lower mass expense.
Nimbus 24’s SRAD avionics design has kept reliability at its core with rigorous testing through 40+
coldflows, 20 static fires as well as flight and ground tests. The use of well structured, modular avionics
developed over the past 3 years has enabled easier rocket integration as well as providing a platform
for faster debugging and safe operation.
As a whole, the knowledge and lessons learnt through the development, manufacture and testing our
previous rockets: Constant Impulse; Sporadic Impulse and Nimbus have allowed the technical advance-
ments and reliability of this year’s rocket to improve exponentially. But a thanks must be extended to
all of those who have supported us, without whom Nimbus 24 wouldn’t have been achievable. Firstly,
ICLR would like to thank all of their sponsors and supervisors for remaining faithful, believing in the
project and continuing to provide support and advice. ICLR would like to thank in particular Dr. Aaron
Knoll who has invaluably supported and guided the team continuously since its inception, taking us
through the difficulties of navigating and managing an ever-evolving team. We would like to extend
a large thanks to the Aeronautics Technicians Jordan Farrar, Mark Thornton, Ian Pardew and Roland
Hutchins for their contributions and assistance with CNC milling processes and manufacture of the
airframe; an important thank you to Ela Sapinska-Elise, Ayesha Khan, Dr. Errikos Levis for assisting
us to seamlessly integrate and grow the finances of ICLR over the past year with the Aeronautics and
Mechanical Engineering Departments; thank you to Dr Kevin Gouder, Will McArdle, Ricardo Huerta
Cruz and Paul Howard for providing access to the 10x5 Wind Tunnel; and also the incredible support
of Gary Senior, Mark and Harry at Mech Eng Stores, the and all of our sponsors.
50
References
[1] Safety W. Hybrid Assembly®;. Available from: https://www.wehberg-safety.de/bursting_
discs_hybrid_assembly.html.
[2] Inc AASM. Aluminum 6061-T6;. Accessed: 1st April 2023]. Available from: https://asm.
matweb.com/search/SpecificMaterial.asp?bassnum=ma6061t6.
[3] Supplies TR. Bolts-Commercial grade;.

Available from: https://tridentracing.co.uk/
product/bolts-commercial-grade/#:~:text=12.9%20grade%20Socket%20Cap%20heads,
75%2C000%20(517MPa)%20shear%20strength.
51
A System Data
Table 22: General Specifications of Nimbus 24
Length 4263 mm
Diameter 190.0 mm
Wet Mass 62.8 kg
Dry Mass 51.3 kg
Stability Off-the-Rail 2.425 cal
Predicted Off-the-Rail Velocity 31.16 m/s
Predicted Apogee 3441 (AGL) m
Predicted Max Acceleration 56.3 m/s2 (5.74 g)
Predicted Max Velocity 271 m/s (Mach 0.789)
Flight Time 187 s
Table 23: Nimbus 24 Mass Breakdown
Section Mass (kg)

Payload 3.0
Electronics 4.0
Airframe 18.3
Recovery 1.8
Propulsion 24.2
Propellant 11.5
Dry Mass 51.3
Wet Mass 62.8
Table 24: Nimbus 24 Length Breakdown
Section Length (mm)

Propulsion Section 2461
Payload Bay 662
Recovery Tube 800
Nosecone 340
Total 4263
Table 25: Nimbus 24 Recovery Specifications
Reefed Drogue-Configuration Diameter 13 in

Reefed Drogue-Configuration CD A 0.3936
Reefed Drogue-Configuration Deployment Altitude Apogee
Reefed Drogue-Configuration Deployment Velocity 22 m/s
Reefed Main-Configuration Diameter 10 ft
Reefed Main-Configuration CD A 29.128
Reefed Main-Configuration Deployment Altitude 450 m
Reefed Main-Configuration Deployment Velocity 39 m/s
Reefed Main-Configuration Terminal Velocity 4.6 m/s
Ground-impact Velocity 4.6 m/s
Max Drift Radius from Launch Pad 2700 m
52
Table 26: Nimbus 24 THANOS-R Engine Specifications
Total Impulse (Ns) 21500

Specific Impulse (s) 200
Maximum Thrust (N) 4000
Burn Time (s) 6
Oxidiser Type Nitrous Oxide
Fuel Type Ethanol
Oxidiser Mass (g) 7000
Fuel Mass (g) 4000
OF Ratio 2.5
Chamber Pressure (bar) 20
Tank Pressure (bar) 40
Table 27: Nimbus 24 Battery Specifications
Name Cell Configuration Capacity (mAh) Nominal Voltage (V)

Flight 4s 1p 3000 14.8
Canard 2s 1p 2500 7.4
Camera 3s 1p 2500 11.1
Payload 2s 1p 2500 7.4
Table 28: Nimbus 24 Avionics Specification
Battery Voltage 14.8 V

Battery Type Li-ion
Battery Capacity 3000 mAh
Endurance 9 hours
RF protocol LoRa
Maximum RF Transmission power 100 mW
Tested range 65 km
Estimated maximum range 100 km
RF frequency 868 MHz
Rocket antenna type Patch
Rocket antenna gain -2.4 dBi
GCS antenna type Rubber Ducky
GCS antenna gain +2.7 dBi
Maximum bit rate 37.5 kbps
Kalman filter refresh rate 200 Hz
53
B Additional Flight Simulations
Presented below are the trajectory simulations for several off-nominal flight conditions computed using
RocketPy for Nimbus 24
B.1 Canard-less Flight

In the event of not flying the physical canard on the rocket, the main aerodynamic parameter affected
is the centre of pressure which moves further down towards the tail of the rocket as a lifting surface
further towards the nose is removed. With the centre of pressure moving further towards the tail and
it already being below the centre of gravity, the stability of rocket increases. The stability margin off
the rail increases to 3.19 from 2.425 when the canards are removed before flight. The stability margin
plots for ascent with and without canards is displayed in Figures
Figure 57: Nimbus 24 Nominal Stability (with Canards)
Figure 58: Nimbus 24 Stability without Canards
54
C Project Tests
C.1 SRAD Avionics - Apogee Detection and Flight Telemetry Test
Testing: SRAD Avionics Recovery Actuation

Test Name: SRAD Avionics Apogee Detection Version: 1 Date: 14/04/2024
Responsible Engineers: Shiven Chandarana, Kiran De Silva, Andrei Paduraru
Test Outcome: Successful Location: Midlands Rocketry Club: United Kingdom
Aim: Test the apogee detection algorithm and verify the actuation of recovery by the SRAD flight computer.
Parts List:
• Pickle Rick Flight Controller
• I class Solid Rocket
55
• Pickle Rick Ground Station

• Ricardo Backend
Procedure: Pickle Rick was flight tested within a test rocket with the main aim of testing apogee detection and verifying the intent of deployment of
parachutes. Pickle Rick was armed and the rocket was ignited. During the flight the different flight phases was seen with the triggering of certain events.
This was seen after the flight via the flight logs which were saved to the SD Card.
Results & Conclusions: Before the flight, live telemetry was established which streamed data to the Grafana graphing software. A high data rate of
all sensors including GPS was communicated validating the radio implementation and user interfaces. During the flight, the flight computer performed
nominally successfully recognising apogee and initiating its recovery state. It recognised lift off and both boost and coast phases of the upward trajectory
as well as identifying correct drogue and main deployement events.
Figure 60: Assembled avionics bay
Figure 59: The apogee detection algorithm plotting parabolic trajectories during
flight until apogee is detected
56
Completion Signed For by the Responsible Engineers:
Shiven Chandarana (11/09/24) Kiran De Silva (11/09/24) Andrei Paduraru (11/09/24)

C.2 COTS Avionics - CATS Vega Nichrome Recovery Actuation Test
Testing: CATS Vega

Test Name: CATS Vega Pyro testing Version: 1 Date: 16/08/23
Responsible Engineers: Shiven Chandarana, Andrei Paduraru, Mohammad Kapadia
Test Outcome: Successful Location: London, United Kingdom
Aim: To ensure the CATS Vega is a suitable secondary flight computer on nichrome based separation systems
Parts List:
• CATS Vega
• CATS ground station
• 2x 2.4GHz LoRA antenna
57
• 21 cm, 0.2mm diameter nichrome

• 2 mm Dyneema
Procedure: The nichrome wire was set up in the flight configuration where it wraps the Dyneema with a specific number of coils. It was then placed into
the CATS vega channel 1 with a WAGO connector. The battery was connected to the board and the board was powered on using the switch.
Using the inbuilt testing mode in the CATS Vega ground station, pyro channels can be fired remotely and allows for manual non-flight based triggers. After
moving to a safe distance the pyro event was triggered for 5 seconds.
Results & Conclusions: The system performed nominally allowing for the nichrome wire to heat up to red-hot without breaking and to successfully cut
the dyneema rope. The result provided evidence to approve the use of the CATS Vega on Nimbus 24’s separation system and parachute disreefing.
Shiven Chandarana (08/08/23) Soham More (08/08/23) Andrei Paduraru (08/08/23)

58
C.3 SRAD Avionics - Radio Range Test
Testing: Avionics System

Test Name: Telemetry and tracking system Version: 1 Date: 22/06/2024
range test
Responsible Engineers: Kiran de Silva, Andrei Paduraru, Artem Sheykin
Test Outcome: Success Location: Various hills in South England
Aim:
To test the maximum range of the telemetry and tracking subsystem with omni-directional antennas at maximum output power (100mW).
Parts List:
• Antenova SRF2I019-100 -2.4 dBi patch antenna
59

• Rubber Ducky +2.7 dBi antenna
• 2x SMA GNSS antenna (any)
• Ricardo-Backend
Radio settings:
• Spreading Factor - 7
• Bandwidth - 200 kHz
• Power - +20 dBm
• Frequency Band - 868 MHz
Procedure for testing radio range:
1. Connect both LoRa and GNSS antennas to the Pickle Rick ground station
2. Connect both LoRa and GNSS antennas to the Pickle Rick flight controller
3. Start telemetry request on both the ground station and flight controller, and verify radio link is functional at short range.
4. Move one or both radio nodes to an elevated position, such as a hill, so that the radio horizon is above the test range
5. Orient the antennas so that there are no major obstructions between them.
6. Repeat the procedure, and move one or both nodes to different locations to test at higher distances.
Results and Conclusions
Telemetry from both flight controller and ground station was successfully received by both stations via radio at distances up to 65 km every 0.5 s. The
operating signal to noise ratio (SNR) at 65 km suggests that it is possible to achieve radio communications at higher distances. However no suitable locations
in the vicinity of the 65km test locations were found, and as such the system was not tested any further.
60
Figure 61: Flight controller radio station during test.
It is important to note that this maximum range was achieved with an output power of 20 dBm (100mW), which is only legal for unlicensed use in a narrow
band of 869.4-869.65 MHz. As such, if this frequency band is congested, the output power needs to be reduced to 14 dBm (25mW) to operate in the rest
of the 868 band. Since the received power varies with R2 , this means that the achievable range is reduced to 32.5 km, which is still adequate for tracking
Nimbus throughout its flight.
61
Andrei Paduraru (22/06/2024) Kiran De Silva (22/06/2024) Artem Sheykin (22/06/2024

C.4 SRAD Avionics - High Altitude Flight Test with Actuation Events
Testing: SRAD Flight Controller

Test Name: Flight Phase Detection and High Version: 1 Date: 11/08/24
Powered Actuation
Responsible Engineers: Shiven Chandarana, Kiran De Silva, Soham More
Test Outcome: Successful Location: Friends of Amateur Rocketry Site, Mojave Desert, USA
Aim: Test the telemetry downlink at large distances, test the flight phase (boost, coast, apogee) detection algorithm, verify the triggering of high powered
actuation channels by the SRAD flight computer.
Parts List:
• I class Solid Rocket
62

• Ricardo Backend
Procedure: The flight controller is placed in both stages of a two-stage rocket with a predicted apogee of 40+ km. The recovery actuators of both stages
and the igniter for the second stage are connected to their respective flight controllers. The electrical system is energised, the flight controller is armed
and an actuation signal is sent when the conditions specified in the flight configuration file are met. During the flight, various events are expected to be
triggered, which will be logged to the on-board SD card for post-flight analysis or may even be observed from the ground (e.g. second stage ignition,
parachute deployment).
Results & Conclusions: Before take-off, a radio link was established which streamed data at the expected data rate and successfully transmitted com-
mands to the rocket from the ground station. During the flight, the flight controller performed nominally, successfully detecting the coast phase of the first
stage and initiating ignition of the second stage, whilst maintaining a stable telemetry downlink.
Figure 62: Booster trajectory recorded from Pickle
Rick telemetry downlink Figure 63: Avionics Bay of Sirius
63
Shiven Chandarana (11/09/24) Kiran De Silva (11/09/24) Soham More (11/09/24)

C.5 SRAD Avionics - High Temperature Test
Testing: Ricardo Avionics High Temperature Test

Test Name: High Temperature test of Ricardo Version: 1 Date: 09/09/23
Avionics
Responsible Engineers: Shiven Chandarana, Kiran de Silva, Andrei Paduraru
Aim: The avionics are expected to perform in the high temperature environment in Portugal aswell as the increased heat environment inside the fully
assembled rocket. To ensure this is the case the avionics can be subjected to higher temperatures to ensure full working capability during these expected
conditions.
Parts List:
64
• Computer
• Ricardo-Backend
• Thermocouple attached to a multimeter
Procedure: The oven was preheated to approximately 70 degrees Celsius. The Ricardo avionics flight controller was put inside the oven, upon a heatproof
soldering mat. The sensors on the flight controller were continuously monitored during the test for any inconsistencies that may be due to temperature.
Results & Conclusions: The flight controller - Pickle Rick performed nominally both measuring data and relaying it to the connected computer as seen
in Figure 64. The flight controller also maintained command response ensuring command-ability during high temperature conditions. Due to the inherent
Faraday cage properties of the oven, GPS and LoRa were not tested but both before and just after (still approximately 50 degrees Celsius) antennas were
connected and both systems worked nominally.
Figure 64: Sensor readings of flight controller under high temperature conditions
65
Shiven Chandarana (09/09/23) Kiran De Silva (09/09/23) Andrei Paduraru (09/09/23

C.6 SRAD Avionics - Low Temperature Test
Testing: Ricardo Avionics Low Temperature Test

Test Name: Extremely Low Temperature test of Version: 1 Date: 31/08/23
Ricardo Avionics
Responsible Engineers: Shiven Chandarana,Andrei Paduraru, Artem Sheykin
Aim: To demonstrate full controllability of the flight controller and the high powered actuation board under extreme low temperatures. Inside the rocket
the tanks and plumbing will be subject to low temperatures and the avionics system is expected to operate in these conditions.
Parts List:
66
• Flint and Steel High Powered Actuation Board

• Computer
• Ricardo-Backend
• Dry Ice
• Nichrome
• Dyneema
Procedure: Dry Ice was put into an insulated foam box. A string of Dyneema was tensioned with a spring and a nichrome wire was encircled. The
nichrome wire was screwed into the high powered actuation board. Both the high powered actuation board and flight controller were placed into the solid
dry ice mass. The temperature will be left to reach equilibrium. Command-ability and data retrieval will be tested many times during this test. The final
test being to command the ignition of a pyro channel for 5 seconds.
Results & Conclusions: The flight controller performed nominally, measuring data and relaying it to the connected computer as seen in Figure 66. The
flight controller and high powered actuation board also maintained command responsiveness. During the test the ambient temperature reached equilibrium
at approximately -60◦C. At this temperature the pyro was fired successfully and the Dyneema separated releasing the attached spring.
In an effort to stress test the Ricardo Avionics System solid dry ice was placed on top of the flight controller to visualise the absolute minimum limits as seen
in Figure 65. The controllers’ sensor readings instantly jumped from -60 Degrees Celsius to -90 Degrees Celsius and continued to operate nominally. During
the test it was not possible to get much colder without more apparatus therefore was deemed a huge success of the Ricardo Avionics System. Although
conditions re-enacted are somewhat of an exaggeration of the low temperatures the opportunity to test under this condition was presented and therefore
was tested successfully. Nonetheless this test gives the approval of the use of the Ricardo Avionics system at temperatures up to -90 Degrees Celcius. It is
worth mentioning the fact that alike the high temperature test (Section C.5) Lora and GPS could not be tested due to the interference in the box as well as
the box being made from metal walls. After the test was concluded, both GPS fix and LoRa radio were tested successfully after.
67
Figure 65: Pickle Rick Flight controller with dry ice placed upon
Figure 66: Sensor Readings for Flight Controller under extreme low temperature conditions

68
Shiven Chandarana (31/08/23) Artem Sheykin (31/08/23) Andrei Paduraru (31/08/23

C.7 Propellant Loading of Nimbus’ Propulsion System
Testing: Fuel, Oxidiser, and Nitrogen Flight Tank Propellant Loading using Ground Support Systems
Test Name: Fuel, Oxidiser, and Nitrogen Filling Version: 1 Date: 14/09/24
of SRAD Pressure Vessel
Responsible Engineers: Martin Leung, Hardik Modi, Andrei Paduraru
Test Outcome: Successful Location: Imperial College London, Silwood Campus
Aim:
Demonstrate methanol, nitrous oxide, and nitrogen loading onto the Nimbus flight system using flight electronics and ground support equipment.
Procedure:
1. Assemble feed system, run tank, vent system, filling system, and fill tank on the test site.
2. Prepare the ethanol filling equipment with necessary PPE.
69
3. Fill the ethanol transfer tank with 5 Liters of ethanol and attach plumbing to the ethanol fill port on the rocket.
4. Open the ethanol vent and fill valve, pressurise the transfer tank and begin loading the ethanol onto the rocket.
5. Monitor the ethanol level in the tank and watch for ethanol dripping from the vent port, when the ethanol stops dripping, close the vent valve and
fill valve.
6. remove ethanol filling equipment and safely store.
7. Ethanol is now loaded
8. Attach the nitrogen fill line to the nitrogen fill port on the base of the rocket.
9. Attach the oxidiser fill line to the oxidiser fill port on the base of the rocket.
10. Open the nitrogen bottle.
11. Open the oxidiser bottle and evacuate the test site.
12. From mission control, open the nitrogen fill valve until a desired pressure of above 250 bar is reached.
13. Close the nitrogen fill valve.
14. Nitrogen is now loaded
15. From mission control, open the oxidiser fill valve remotely and wait until the mass in the run tank stops increasing and the pressure has equalised.
16. Open the fill valve and vent valve slightly until the tank mass reaches 7 kg or a white plume vents from the rocket.
17. Close the oxidiser fill valve and vent valve.
18. Vent the nitrous filling line
19. Oxidiser is now loaded
20. Upon launch the nitrogen and oxidiser filling line will be disconnected by the force of the rocket launch, however since this is a static test the filling line was
left attached.
Results & Conclusions:
The propulsion team have conducted multiple filling tests of oxidiser flight tanks over the teams test campaigns and are now very experienced in the pro-
cedures, having performed over 50 loading and unloading operations of fuel systems The loading and unloading of the NIMBUS 24 flight system has been
performed more than 5 times during the official test following the exact procedure outlined. This year, the new remote loading and unloading of oxidiser
and pressurant gas introduced a variation in the process followed with previous team operations and has been additionally practiced to ensure coherent
and safe operation. Complete filling tests have gone smoothly every time with no issues, and members responsible for the filling an procedures have been
70
trained and are experienced to perform each task and diagnose problems should they appear.
Figure 67: Ground Support System Figure 68: Propellant Loading Setup Figure 69: Propellant Loading Setup
Martin Leung (14/09/24) Hardik Modi (14/09/24) Andrei Paduraru (14/09/24

71
C.8 Hot Fire Testing of THANOS-R Engine and NIMBUS 24 flight propulsion system
Testing: THANOS R Engine

Test Name: Hot Fire Testing of THANOS-R Version: 1 Date: 14/09/24
Engine and NIMBUS 24 flight propulsion system
for Flight Duration
Responsible Engineers: Elliot Brookes-Gayton, Martin England, Jason Aretaios, Martin Leung, Hardik Modi, Andrei Paduraru
Aim:
The goal of the Hot Fire Testing of the NIMBUS 24 propulsion system, is to characterise the exact performance of the flight propulsion system, demonstrate
the reliable ignition and firing of the THANOS-R engine, while also practicing procedures such as fuel loading, oxidiser loading, tank pressurisation, and
tank venting after firing. The purpose of the test is to also confirm that the THANOS-R engine will achieve its target performance in the full flight config-
uration and meets the off-the-rail startup requirements, including the : peak thrust of 4 kN and a sustained burn of over 3 kN during the 7 second flight
duration, producing a total Impulse of over 20,000 Ns to reach the target altitude.
72
Procedure:
1. Prepare the test site for test operations
2. Assemble sub systems including, thrust structure, engine assembly, engine feed system, vertical test stand, rocket feed system, flight electronics on
the test pad
3. Prepare filling systems including Ethanol filling system, GSS filling system, and gas cylinders on the test site
4. Assemble together the flight system and tighten all feed system and engine components
5. Perform electrical systems checks on all hardware including valve actuation and data acquisition
6. Perform leak checks using inert pressurant
7. Perform exact fuel mixture and fuel tank filling with the Ethanol mixture, filling with 4.7 kg of fuel
8. Perform connection of the oxidiser and pressurant gas to the GSS and the rocke and opening of the high pressure gas bottles as prescribed by the
Filling Test Procedure
9. Raise Red flag status and evacuate test area tp begin remote filling of oxidiser and pressurant as prescribed by the Filling Test Procedure
10. Fill until the run tank mass reaches 7 kg
11. Open tank pressurisation valve to pressurise both tanks with nitrogen to 55 bar.
12. Arm the ignition, confirm all stations and operators are GO/NO GO with audible statements, and commence engine firing procedure
13. Begin countdown and send command for engine ignition
14. Hot Fire until depletion of propellants
15. Safe the system venting all tanks ensuring no operator approaches until fully depressurised
16. Approach to close oxidiser and pressurant bottles
17. Vent the lines
18. Perform post test inspections and review data
19. Perform trajectory simulations with full flight configuration hot fire data
73
The propulsion team has conducted 2 successful final hot fires of the THANOS-R engine and flight configuration system of NIMBUS 24, achieving the
desired performance an demonstrating the reliability and safety of the system. In total the team has performed over 17 + hot fires in this years testing
campaign, bringing great confidence in the performance of the propulsion and electronics systems that it has developed and in its testing procedures,
proving to be safe reliable and high performing. The team has also conducted numerous controlled vents of the fuel and oxidiser tanks, full cold flow
tests of the engine, and oxidiser loading and unloading procedures including over operation of over ten full hot fires of the previous THANOS and Nimbus
system. The amount of experience that have been gained by the team diagnosing and repairing issues in the field and testing each component thoroughly
brings us much confidence in the reliability of the NIMBUS 24 flight propulsion system, in the team operation of these tests, and full trust in its performance
for the completion meeting all of its requirements.
74
Figure 70: Thrust and Pressure Graph from Qualification Hotfire

75
Figure 71: From left to right, photos from the final qualification hot fire
Elliot Brookes-Gayton (14/09/24) Martin England (14/09/24) Jason Aretaios (14/09/24)

76
Martin Leung (14/09/24) Hardik Modi (14/09/24) Andrei Paduraru (14/09/24)

C.9 Propellant Unloading of Nimbus’ Propulsion System
Testing: Fuel, Oxidiser, and Nitrogen Flight Tank Propellant Unloading using Ground Support Systems
Test Name: Fuel, Oxidiser, and Nitrogen Version: 1 Date: 14/09/24
Unloading of SRAD Pressure Vessel
Responsible Engineers: Martin Leung, Hardik Modi, Andrei Paduraru
Aim:
Demonstrate fuel, oxidiser, and nitrogen unloading procedures for the Nimbus’ flight propulsion system using the ground support systems and flight elec-
tronics.
Procedure:
In the event of test fire or launch abort the propellant will need to be safely unloaded. Primary concern is venting the nitrous oxide from the SRAD pressure
vessel, unloading the ethanol, and venting the nitrogen. There are two cases depending on the severity of the situation and the response time required.
77
Both of the abort cases have been tested by the team to ensure safe opperations.
CASE 1:
1. Assess the rocket/test site for fires or any other hazards that could increase risk of explosion when venting the oxidiser. If no hazards are present,
unloading is clear to proceed.
2. Disarm ignition/launch procedure
3. Ensure filling valves and main engine valves are closed.
4. Open the nitrous oxide vent valve and observe white plume exiting the vent. Mass from the oxidiser tank should begin to decrease.
5. Wait until the pressure in the oxidiser tank has returned to atmospheric, oxidiser mass to zero, and no observable venting plume.
6. Nitrous has been unloaded.
7. Open the pressurisation valve and vent the nitrogen through the nitrous vent valve.
8. Wait until the pressure has dropped to atmospheric in the nitrogen tank.
9. Open the ethanol vent valve to vent the remaining nitrogen in the ethanol tanks ullage volume.
10. Nitrogen has been unloaded and the Rocket is De-Pressurised.
11. At this stage the large nitrous and nitrogen filling bottle is still open and needs to be closed manually, nitrous remains in the filling hose. The
Pressurised Fluids Safety officer will approach wearing full nitrous safety gear and close the nitrous and nitrogen bottles manually.
12. Once the nitrous bottle is closed the line will be vented manually using the needle valve.
13. The launch pad is now De-Pressurised and safe to approach.
14. Safety officer now disconnects the ignition leads.
15. To remove the remaining ethanol, the filling procedure will be preformed in reverse.
16. Attach an empty ethanol transfer tank to the ethanol fill port and crack the top of the tank to allow air to vent out.
17. Open the ethanol fill valve and vent valve and ethanol will drain out under gravity into the ethanol transfer tank.
18. Once all the ethanol is drained from the tank, place a small container is under the nozzle and open the main fuel valve to drain the remaining ethanol
in the plumbing.
19. Ethanol has been unloaded.
78
20. The rocket has been unloaded.

21. Disconnect the ethanol transfer tank and store the ethanol appropriately.
22. Leave all the ethanol filling valves open until transport to air out as much of the ethanol vapours as possible and wipe up any spills.
23. If launch/testing has been concluded for the day, the team can begin to disassemble the engine to remove the fuel and ignition motors. If another
launch/test is to be attempted, ignition motors can be left inside the engine.
CASE 2:
1. Assess the rocket/test site for fires or any other hazards that could increase risk of explosion when venting the oxidiser. If no hazards are present,
unloading is clear to proceed.
2. Disarm ignition/launch procedure
3. Ensure all filling valves and main engine valves are closed.
4. Open the nitrous oxide vent valve and observe white plume exiting the vent. Mass from the oxidiser tank should begin to decrease.
5. In the case of need for rapid unloading the main Oxidiser valve is oppened and the nitrous oxide is vented through the main feed system, through
the engine, unloading it rapidly. This happens occurs with out the main fuel valve remaining closed, to prevent any fuel and oxidiser mixture.
6. Nitrous has been unloaded.
7. Open the pressurisation valve and vent the nitrogen through the nitrous vent valve.
8. Wait until the pressure has dropped to atmospheric in the nitrogen tank.
9. Open the ethanol vent valve to vent the remaining nitrogen in the ethanol tanks ullage volume.
10. Nitrogen has been unloaded and the Rocket is De-Pressurised.
11. At this stage the large nitrous and nitrogen filling bottle is still open and needs to be closed manually, nitrous remains in the filling hose. The
Pressurised Fluids Safety officer will approach wearing full nitrous safety gear and close the nitrous and nitrogen bottles manually.
12. Once the nitrous bottle is closed the line will be vented manually using the needle valve.
13. The launch pad is now De-Pressurised and safe to approach.
14. Safety officer now disconnects the ignition leads.
15. To remove the remaining ethanol, the filling procedure will be preformed in reverse.
16. Attach an empty ethanol transfer tank to the ethanol fill port and crack the top of the tank to allow air to vent out.
79
17. Open the ethanol fill valve and vent valve and ethanol will drain out under gravity into the ethanol transfer tank.
18. Once all the ethanol is drained from the tank, place a small container is under the nozzle and open the main fuel valve to drain the remaining ethanol
in the plumbing.
19. Ethanol has been unloaded.
20. The rocket has been unloaded.
21. Disconnect the ethanol transfer tank and store the ethanol appropriately.
22. Leave all the ethanol filling valves open until transport to air out as much of the ethanol vapours as possible and wipe up any spills.
23. If launch/testing has been concluded for the day, the team can begin to disassemble the engine to remove the fuel and ignition motors. If another
launch/test is to be attempted, ignition motors can be left inside the engine.
The propulsion team have conducted more than 5 propellant unloading procedures with the Nimbus 24 propulsion system both with the rapid unloading
sequence and the slower unloading sequence and the unloading sequence. In total the team has performed over 30 nitrous unloading procedures in its
testing campaign using retired flights systems. Prior to the competition further wet dress rehearsals will be performed to practice this procedure again
ensuring the response is rapid and smooth. The team is well practiced and confident in performing propellant unloading in any situation that requires so.
80
Martin Leung (17/09/23) Hardik Modi (17/09/23) Andrei Paduraru (17/09/23)

C.10 Hydrostatic Testing of THANOS R Combustion Chamber
Testing: THANOS R Combustion Chamber

Test Name: Hydrostatic Testing of THANOS R Version: 1 Date: 31/08/24
Combustion Chamber)
Responsible Engineers: Martin Leung, Shiven Chandarana, Hardik Modi
Aim:
Demonstrate that the THANOS R engine combustion chamber can withstand a hydrostatic pressure 1.5 times Maximum Expected Operating Pressure
(MEOP) for 30 minutes.
Procedure:
1. Increase the pressure using the test pump up to 30 bar.
2. Shift the toggle to high pressure mode and continue operating handles until desired pressure is achieved (37.5 bar).
81
3. Stop increasing pressure and watch the pressure gauge for any leaks over a period of 30 minutes.
4. Relieve pressure by opening fine adjustment valve and then main pressure release valve.
Test performed as expected, no pressure loss was observed over the 30 minute test period.
Figure 72: Photos from the hydrostatic test.

82
Martin Leung (01/09/24) Shiven Chandarana (01/09/24) Hardik Modi (01/09/24)

C.11 Hydrostatic Testing of SRAD Pressure Vessel (Fuel Flight Tank)
Testing: SRAD Fuel Pressure Vessel

Test Name: Hydrostatic Testing of SRAD Version: 1 Date: 14/06/24
Pressure Vessel (Fuel Flight Tank)
Responsible Engineers: Martin Leung, Hardik Modi, Martin England
Aim:
Demonstrate that the SRAD Fuel pressure vessel can withstand a hydrostatic pressure 1.5 times Maximum Expected Operating Pressure (MEOP) for 60
minutes (twice the maximum expected operating period).
Procedure:
2. Shift the toggle to high pressure mode and continue operating handles until desired pressure is achieved (90 bar).
83
Figure 73: Photos from the hydrostatic test
84
Martin Leung (17/07/24) Hardik Modi (17/07/24) Martin England (17/07/24)

C.12 Hydrostatic Testing of SRAD Pressure Vessel (Oxidiser Flight Tank)
Testing: SRAD Oxidiser Pressure Vessel

Test Name: Hydrostatic Testing of SRAD Version: 1 Date: 14/06/24
Pressure Vessel (Oxidiser Flight Tank)
Responsible Engineers:Martin Leungm Hardik Modi, Martin England
Aim:
Demonstrate that the SRAD oxidiser pressure vessel can withstand a hydrostatic pressure 1.5 times Maximum Expected Operating Pressure (MEOP) for 60
minutes (twice the maximum expected operating period).
Procedure:
85
86
Martin LEung (17/07/24) Hardik Modi (17/07/24) Martin England (17/07/24)

C.13 Hydrostatic Testing of COTS Pressure Vessel (Nitrogen Flight Tank)
Testing: COTS Pressure Vessel

Test Name: Hydrostatic Testing of COTS Version: 1 Date: 14/06/24
Pressure Vessel (Nitrogen Flight Tank)
Responsible Engineers:Martin Leung, Hardik Modi, Elliot Brookes Gayton
Aim:
Demonstrate that the COTS Nitrogen pressure vessel can withstand a hydrostatic pressure of higher than the expected operating conditions. Although a
COTS tank,this was done to ensure thorough testing of all system components.
Procedure:
87
88
Martin Leung (17/07/24) Hardik Modi (17/07/24) Martin England (17/07/24)

C.14 Ground test of the deployment mechanism of the payload
Table 29: Ground test of the deployment mechanism of the payload
Testing: Actuation of the deployment mechanism under simulated flight conditions

Test Name: Deployer Test Version: 1 Date: 11/09/24
Responsible Engineers: Arfred Garcia, Ishan Dubey, Lara Alves
Test Outcome: Successful Location: Imperial College London, South Kensington Campus
Aim:
Demonstrate successful deployment of the payload with entire assembly inside, parafoil included
Procedure:
1. Following the method outlined in ??, fold the parachute to ensure it can fit within the constraints of the deployer.
2. Put the payload into the deployer box alongside the deployer
89
3. Lock into the rocket body

4. Arm the servos
5. Actuate both servos simultaneously
The payload deploys out of the deployer as expected. Numerous iterations have been made to ensure the box which encapsulates the parafoil does not
become stuck when the payload is falling.
90
Figure 76: Deployer Unlocked Figure 78: Payload Fully De-

Figure 77: Payload Falling
via actuating servos ployed
Arfred Garcia (11/09/24) Ishan Dubey (11/09/24) Lara Alves (11/09/24)
C.15 Ground test of the actuation mechanism of the guided recovery system
Table 30: Ground test of the actuation mechanism of the guided recovery system
Testing: Actuation of the recovery mechanism under simulated flight conditions

Test Name: Recovery Actuation Sequence Version: 1 Date: 11/09/24
91
Responsible Engineers: Arfred Garcia, Jayden Jackson, Ishan Dubey

Aim:
Demonstrate the successful actuation mechanism of our payload’s guided recovery system under different flight conditions
Procedure:
1. Assemble the payload guided recovery
2. Setup parafoil in the same folded configuration to be used in flight
3. Mount into the wind tunnel under a range of terminal velocity speeds
4. Turn on the wind tunnel and activate the actuation sequence
5. Visual checks for stability and try to adjust lines
The guided recovery system was successfully actuated at a wind speed of 9 m/s and demonstrated the movement of the lines, which enables us to bank
our payload from side to side in an arc of approximately 30◦ in both directions. Due to the tight turnaround as the wind tunnel was needed for use by
other members in our faculty, more in-depth dynamics were not able to be captured. However, the torque transmission was sufficient enough to move our
lines and bank the payload accordingly, hence increasing our confidence of the system as a whole.
Figure 79: Parafoil banked port side Figure 80: Parafoil in neutral position Figure 81: Parafoil banked starboard

92
Arfred Garcia (11/09/24) Jayden Jackson (11/09/24) Ishan Dubey (11/09/24)

C.16 Inflation test of folding configuration of parafoil
Testing: Under the folding configuration we have planned, how the parafoil successfully unravels in flight and undergoes stable flight
Test Name: Folded Parafoil Test Version: 1 Date: 11/09/24
Responsible Engineers: Arfred Garcia, Ishan Dubey, Rosalind Aves
Aim:
Testing: successful deployment of the parafoil with it folded in the correct configuration
Procedure:
1. Fold the parafoil as outlined previously (??) in a configuration which is akin to the real life case.
2. Release the parafoil and conduct visual checks for inflation of all the panels and stability
93
The parafoil has proved successful in inflation and stability when released under the configuration specified for flight, hence increasing the confidence of
our team on the parafoil’s design.
(a) Folded (b) Partially Deployed (c) Fully Deployed
Figure 82: Parachute Deployment Stages

Arfred Garcia (11/09/24) Ishan Dubey (11/09/24) Rosalind Aves (11/09/24)

94
C.17 Engine Truss Connector Testing
Testing: Engine connector in compression loading to qualify for flight

Test Name: Engine truss compression test Version: 1 Date: 30/08/24
Responsible Engineers: Ahmed Gilani, Joshua Bobin, Charlie Aveline, Pablo Duhamel
Aim:
To test the engine connector truss structure, the structure that connects the engine to the airframe of the rocket, under compressive loads using a 250kN
mechanical testing Instron machine. This is to verify that the part can withstand the maximum nominal thrust force from the engine during flight (5kN).
Procedure:
Two tests were conducted, one on the flight truss to proof test it and one on a spare truss to test it to failure.
1. Assemble the truss with a custom-design and waterjet-cut jig in order to mimic loading conditions during flight (see Figure 83a).
95
2. Turn on Instron machine and set height limits.

3. Balance force and displacement readings on Instron console.
4. Use Bluehill software to take machine off standby.
5. Create a new “method”, i.e. test procedure (compression). Also create and select folders for exporting data.
6. Run the Instron test procedure. Force and displacement data will display on the software window.
7. Perform the test to 5kN on a test truss structure, and to 5kN on the flight component and record load and displacement data.
8. Finally, test the test truss structure to failure, recording the load and displacement data.
FEA simulations conducted on ABAQUS predicted that at failure, 0.496mm of displacement would be experienced at approximately 30kN of applied
compressive force. In comparison, the laboratory testing showed that initial fracture of a truss member occurred at 14.98 kN, with a displacement of
1.287 mm. The testing also showed the overall stiffness of the flight-ready component to be 16.98 kN/mm. The stiffness for the test component was
calculated to be 16.38 kN/mm. The percentage difference between the two is only 3.5%, meaning that the results for the test component can be
extrapolated to the flight-ready component. It can be seen in Figure 84b that there is an initial upward curve before the test enters the linear elastic
region. As such, the stiffness was calculated using points on the graph with a displacement of 0.5mm onwards, to ensure that the calculated stiffness is
the most accurate and has the highest R2 values.
(a) Render of testing jig (b) 80x magnification of surface (c) Broken horizontal member
96
(d) Crack in horizontal member of test component (e) Full set up of compression test
Figure 83: Photos from the truss testing.

(a) Force-displacement curve for test component to
yield.
97
(b) Force-displacement curve for test and flight-ready

components.
(c) Force-displacement curve for test component to yield.
Figure 84: Graphs from the truss testing.

Conclusion:
This test has demonstrated clearly the flight-readiness of this component. The test structure should be able to withstand the required nominal maximum
thrust load of 5kN. As the component first yielded at 14.98kN, it can be said that there is a safety factor of approximately 3. Moreover, it was found that
the test structure had voids internally, which led to the formation of micro-cracks in loading. It was also noted that the test component had visible surface
cracks. However, the flight component does not have any, making crack propagation and thus failure in a similar manner is less likely under similar loads.
98
Ahmed Gilani (01/09/24) Pablo Duhamel (01/09/24) Martin England (01/09/24)

C.18 Parachute reefing deployment testing
Testing: Transition of the parachute from its reefed state to fully deployed state in flight conditions.
Test Name: Reefing Deployment Test Version: 1 Date: 27/09/24
Responsible Engineers: Pablo Duhamel, Soham More, Mohammad Kapadia, Andrei Paduraru
Test Outcome: Successful Location: Lasham Airfield, Alton
Aim:
The aim of this test is to validate Nimbus24’s reefing deployment by which the main rocket’s parachute transitions from the its reefed state to its fully
opened state.
Procedure:
1. Mount the main parachute carabiner to the test vehicle tow bar.
2. Connect a load cell in series with the shock cords.
99
3. Connect the reefing actuation cables to the reefing dyneema.

4. Twill the nichrome wire around the dyneema chord within the reefing case.
5. Connect the actuation cables to the Flint and Steel board located in the test vehicle.
6. Accelerate to 30mph. Once the parachute is inflated and stable, activate the dyneema cutter.
Two runs were performed during the test day. The first run’s objective was to showcase that the reefing system would not undo itself at high velocities.
The nichrome wire cutter was connected and live but not activated. Upon inflation, the reefed parachute maintained its reefed position for the duration of
the run, proving that the reefing method is effective. Following the test, the electronics test confirmed that they had continuity on the nichrome wire
during the entirety of the test. This shows that our dyneema wire cutting system is robust and can sustain loads at high velocity.
The second run’s objective was to demonstrate the disreefing system is flight ready and that the shock cords can sustain the shock loads. The parachute
was mounted in the same configuration as previously only this time the wire cutter was activated after reaching a steady-state on the parachute. The
Dyneema cord holding the parachute in its reefed state was cut 6 seconds after the pyro command was sent which correlates well with our bench tests.
Once disreefed, the parachute fully inflated into its final configuration. The figures below show the parachute in its reefed and disreefed states.
(a) Reefed chute fully inflated, t=0s (b) Disreefed chute fully inflate, t+6s
100
Mohammad Kapadia (27/09/24) Pablo Duhamel (27/09/24) Soham More (27/09/24)

C.19 Recovery mechanical separation testing
Testing: Mechanical separation of the body tube from the nose cone for recovery.
Test Name: Recovery separation Version: 1 Date: 26/09/24
Responsible Engineers: Pablo Duhamel, Soham More
Aim:
To bench test the recovery separation system and qualify it for flight using solely flight hardware.
Procedure:
1. Place the nose cone recovery coupler onto the body tube recovery coupler, aligning the nose cone spring pin into the body tube coupler.
2. Loop the nichrome wire around the dyneema cord.
101
3. Close the nichrome case shut.

4. Assemble the clamp around both couplers and tighten securely.
5. Connect the wire cutter actuation lines to the SRAD and COTS boards.
6. Configure the boards to fire the nichrome for 5 seconds.
7. Initiate test by deploying both boards at the same time.
The nose cone, body tube, clamp band and dyneema cutter were assembled in their flight configuration. Following commands being sent to the pyro
channels on the SRAD and COTS boards, the nichrome wire heated up to cut the clamp band dyneema. The clamp band opened up immediately allowing
the nosecone to separate from the rest of the rocket.
(a) Clamp band assembly at t=0s when the command to (b) Clamp band assembly at t=+3s when the nichrome (c) Clamp band assembly at t=+6s when the dyneema
cut the dyneema is sent cutter is heated and cutting through the dyneema is fully cut and the clamp band opened
102
Pablo Duhamel (26/09/24) Shiven Chandarana (26/09/24) Soham More (26/09/24)

D Flight System Overview
103
Figure 87: Layout of Electronics Subsystems and wiring

104
Figure 88: Layout of Electronics Subsystems and wiring

E Ground Support System Overview
105
Figure 89: System Diagram of the ground support system for Nimbus
F Material Tables
Table 31: Materials and Manufacturing Methods for Components
Component Material Manufacturing Justification

Method
Airframe
Stringers Aluminium 6082 Waterjet Cut 2D geometry,
T6 high strength
Couplers Aluminium 6082 CNC machined High precision,
T6 high strength
Tank Bulkheads Aluminium 6082 CNC machined High precision,
T6 Lightweight
Panels Glass Fibre Wet Layup Low cost, Easier
Manufacturing
Bodytube Glass Fibre Wet Layup Low cost, RF
Transparency
Nosecone Carbon Fibre Wet Layup Stiffer, Aesthetic
Fin Can Carbon Fibre Wet Layup Stiffer, Aesthetic
Electronics
PCB Boards FR4, IC’s PCB Manufacturing Excellent elec-
trical insulating
properties
Wires Silicon insulated COTS Low Cost, High
Copper range of tempera-
ture resistance
Connectors Injection COTS Low Cost, Lock-
Moulded Ny- ing
lon
Board Mounts Onyx Additively Manufac- Low Cost, High
tured Strength
Antenna Bay PLA Additively Manufac- Low Cost, RF
tured Transparency
Pressure Transducers Metal Cas- COTS High reliability
ing/PCB internal off the shelf
Servos Metal COTS High reliability
Gears/Metal off the shelf
Casing
Flight Dynamics
Canard Fins Carbon Fibre Wet Layup Strength, Aes-
thetic
Canard Mounting Aluminium 6082 CNC Machined Strength, Holding
mounting plates T6 canard shafts
Stringer Mounting Aluminium 6082 Additively Manufac- Low Cost, intri-
Blocks T6 tured cate geometry
Canard Shafts Aluminium 6082 Lathe/manual milling Circular geometry
T6
Payload
Continued on next page
106
Method
Stringers Aluminium 6012 Manual Mill Affordable, High
Strength
Mounting Bulkheads Aluminium 6012 Waterjet Cut Affordable, High
Strength
GFRP Mounts Aluminium 6012 Additively Manufac- Complex Geome-
tured try, High Strength
Panels Glass Fibre Wet Lay-Up Lightweight,
Transparent
Shafts Grade 2 Titanium Lathe High strength to
weight ratio, high
shear modulus
Gears Stainless Steel COTS High Torque,
High Strength
Bearings Stainless Steel / COTS Smooth Rotation,
Bronze Torque Transmis-
sion
Box Aluminium 6012 Waterjet Cuts, some Weight Saving,
folded High Strength
Lock System Aluminium 6012 Additively Manufac- Complex Geome-
tured try, High Strength
Hinges Stainless Steel COTS -
Propulsion
Regen cooled engine AlSi10Mg Additively Manufac- Complex internal
chamber tured geometry
Regen cooled engine in- Aluminium 6082 Lathe / Manual Milling Circular geome-
jector head T6 try, lightweight
Regen cooled engine in- Stainless Steel Lathe / Manual Milling Circular geome-
jector insert 316L try, high thermal
strength
Nitrogen Tank Carbon Over- COTS High Safety Fac-
wrapped Pressure tor off the shelf
Vessel
Fuel/Oxidiser Tank Aluminium 6082 Lathe Low weight, In-
walls T6 house able
Fuel/Oxidiser Tank Aluminium 6082 Manual Milling/Lathe Low cost, In-
endcaps T6 house manufac-
turable
Swagelok Plumbing Stainless Steel Assembly High Safety Fac-
tor off the shelf
Ball Valves Stainless Steel COTS Operational Reli-
ability
Solenoid Valves Aluminium/Nylon COTS Normally open
Capability
Recovery
Parachute Ripstop Nylon Purchased Standard sizing
Recovery Bulkhead Aluminium 6082 Waterjet Cut 2D geometry
T6
Continued on next page
107
Method
Wire Cutter Nichrome Purchased Good thermal
properties
Clamp Band Coupler Aluminium 6082 CNC Machined Tight Tube toler-
T6 ance
Clamp Band Clamps Aluminium 6082 CNC Machined Tight Internal Tol-
T6 erances
Clamp Band Band Spring Steel Purchased High Yield
Strength
General
Zip-ties Nylon COTS Cable keeping
Screws Stainless Steel COTS High Strength
108
G Hazard Analysis
Material/Component Potential Hazard Control Measures Risk Level

C/D-Class Klima Ig- Unintended combus- Sealed in fireproof box prior to use. Ignition sources stored Low
nition Motors tion separately. Only EuRoC certified personnel permitted to han-
dle ignitors, under supervision of lead pyrotechnic techni-
cian.
E-matches Accidental ignition Connected to tested firing systems. Activation requires multi- Low
causing fire or burns ple deliberate actions. Handled by certified personnel, under
supervision.
Li-Ion Batteries Fire from over- Charged with proper chargers, balance charged. Careful Low
charge, over- handling to avoid puncture, properly soldered and insulated
discharge, or punc- wiring. Stored in Li-ion safe bags
ture
Nitrous Oxide Dizziness, uncon- Stored in certified vessels, used in ventilated areas. Plumb- Low
sciousness, cold ing and vessels cleaned and tested. Gloves used to prevent
burns, explosions frostbite.
Ethanol Fire, Incapacitation Stored in original sealed container away from ignition Low
from inhalation/con- sources. Protective gear worn during handling. Pumped us-
gestion. ing low-pressure spray bottle, handled outdoors for ventila-
tion.
High Pressure Nitro- Explosion from Loaded last. Components rated for 300 bar. Gas bottles Low
gen overpressure, as- opened slowly, used outdoors for ventilation.
phyxiation
109
H Risk Assessment
Mitigated
Failure Mode Mission Phase Team's Comments and Justification
Probability Severity Risk
Airframe
Buckling of stringers results in a critical failure of the structural integretiy of the rocket.
Buckling of stringers Ascent 1 3 3 This is unlikely to occur as the stringers are designed against the maximum load, with
the resulting simulations indicating a safety factor of at least 1.5.
The engine truss was designed with large buckling and yield strength safety factors.
Two identical trusses tested, one truss to failure to validate the design and ultimate
Engine truss structure failure Ascent/Apogee 1 3 3
load (15Kn). The flight truss has been proof tested to the design load to ensure there
are not any manufacturing defects in the flight truss.
A strong aluminium internal frame increases stiffness and reduces the need for
Bending of the airframe due to weaker
Pre flight 1 2 2 couplers. All joints use machined interfaces to ensure good tolerances. A stiff CFRP skin
joints and couplers near the top
also reduces deflections due to bending.
Bolts within sections are tightened and checked before launch day. Checklist ensure
Airframe bending when being carried Pre-flight 1 2 2 bolts between sections are all tightened correctly. Minimum of 4 people required to
carry the rocket at all times to ensure load is supported evenly.
Bolts appropriately sized to make sure they do not shear. Components along the
Shearing of bolts that screw into body
Mid flight - Chute critcial load path simulated with a resulting safety factor of 1.98. Screw holes in the
tube and/or body tubes and bulkhead 1 3 3
deployment composite tube placed sufficiently far away from the edge so as to not cause any
due to shock loads
damage to the tubes mid flight.
Recovery
Disconnection of nichrome recovery Molex Nanofit connectors are used to connect the nichrome to the electronics cables
Ascent 1 3 3
actuators which are positivley locking hence are unlikely to be experienced in flight.
Nichrome wire melts before cutting Ground tests show that each nichrome heating element can be reused multiple times,
Ascent/Apogee 1 3 3
dyneema due to over-heating therefore the probability of a section of nichrome snapping is very low.
The parachute is made of rip-stop nylon, which is specifically chosen to prevent ripping
Parachute ripping Descent 1 3 3
under the expected shock loads.
The parachute transitions from its reefed state into its fully opened state by using the
same nichrome acutation method used in the main seperation method and previous
Reefed parachute not opening Descent 1 2 2 recovery systems. The full reefing system has been ground tested and pull tested from
a car. Failure to un-reef results in the rocket desending faster than expected, but this
should not result in a critical severtiy risk.
The dyneema is tied with a reef knot which a strong knot commonly used for this
Dyneema knot comes loose mid flight Ascent 1 2 2 loading condition. Tests have been performed with the clamp band showing that the
knot is strong and unlikely to come undone in flight.
Electronics
Connectors and Cable harness tested multiple times. Proper connection verifed at
Quick Disconnect Detaches
Pre-Launch 2 1 2 launch pad physically and electrically during pad operations. Backup communication
Prematurely
link allows for independent control of the rocket if quick disconnect detaches.
Two independant and redundant power management boards and battery packs on the
rocket with automatic switchover reduce the likelyhood. Independent battery system
Deployment power failure Pre-Launch - Flight 1 2 2
for COTS recovery system ensures power failures in primary avionics sytem do not
propagate.
Two independent and redundant power management boards and battery packs on the
Logic power failure Pre-Launch - Flight 1 3 3 rocket. Logic power isolated from deployment power mitigating risk of brownout
during high current output of acutators.
Battery level monitored through power management boards. Quick disconnect and
Batteries depleted Flight 1 3 3
umbilical present for on-the-pad charging. Over 6hr battery life on each battery pack.
Independent SRAD and COTS telemetry links. Hard wired telemetry link present for
ground operations. Range testing of radio systems. During flight, the flight controller
Telemetry Link Failure Pre-Launch - Flight 1 1 1
pre-programmed with all deployment events so no live commands are required for
succesful flight.
Firmware thoroughly tested during real world tests with hardware. Independent COTS
flight computers used to ensure recovery and multi-step igntion implemented to
Firmware Crash Pre-Launch - Flight 1 3 3
prevent un-intentional vehicle ignition. Constant monitoring of all flight boards during
ground operations.
COTS battery system charged using main power umbilical however still fully
independent of main power system. A good compromise between complete
independence of the COTS system and practicality was sought. For this reason, even
COTS recovery power failure Flight 1 2 2 though the COTS recovery system has its own independent battery pack, it is charged
through the main power umbilical in order to ensure infinite battery life on the pad.
Primary avionics system powered speratley ensuring power failures do not propagate
between systems.
The power system has been sized and tested to support multiple igniters firing on a
single battery pack. Current limited output channels ensure the current draw from a
dead short is limited. Dual battery packs give failover redundancy. The battery
Deployment Power Brownout Launch - Flight 1 2 2 chemistry used, as well as the bulk capacitance present on each board was designed to
be able to handle the current draw of the deployment events. Primary and secondary
avionics systems powered seperately ensuring brownouts do not propagate between
systems.
Adequate space between electronics boards designed and appropriate thermal control
measures such as heatsinks used. Simulated hot enviorment testing. Thermal imagery
Electronics heating due to high
Pre-Launch - Flight 2 1 2 used to investiage hot spots on boards. While previous launches during hot, sunny
temeperatures
days have not affected the electronics due to its relatively low power usage, several
mitigations are in place.
110
Electronics configured before-hand with flight configuration verified with hardware-in-
Configuration of electronics with the-loop flight simulations as well as wet dress rehersals. Version control system (git)
Pre-launch - Flight 1 2 2
erroneous settings used to track any changes. Final configuration only verified on launch day. Checklists
are used.
Pre-Launch - Flight - Appropriate insulation present for the voltages used. Connectors designed such that
Electrical fire 1 3 3
Recovery dead shorts through erroneous orientation is not possible.
Pre-Launch - Flight - Battery voltage monitoring. Over current and over voltage protection implemented on
Li-Ion battery fire 1 3 3
Recovery power distribution units. Batteries balanced charged in LiPo Safe bag.
Propulsion
Redundant power systems and prior tests ensure that a valve failure is unlikely.
GSS N2O Filling Valve Actuation Failure Pre-flight 1 2 2 Consequences of filling valve failure are low as vehicle can be vented and approached
to facilitate repairs.
GSS N2 Filling Valve Actuation Failure Pre-flight 1 2 2 Consequences of filling valve failure are low as vehicle can be vented and approached
to facilitate repairs.
Redundant power systems and prior tests ensure that a valve failure is unlikely. Should
the vent valve fail, vehicle can be vented through normally open solenoid valves and
approached to facilitate repairs. Should both the vent valve and solenoid valve fail, a
Rocket N2O Vent Valve Actuation
Pre-flight 1 2 2 calibrated SPRV (Safety Pressure Relief Valve) and burst disc ensures the tank does not
Failure
fail due to an overpressure event. Consequences of vent valve failure are low as N2O
can be offloaded via main engine valve (through a separate board) with ignition and
fuel valve disarmed and approached to facilitate repairs.
COTS solenoid valve and Prior tests ensure that a valve failure is unlikely. Calibrated
SPRV (Safety Pressure Relief Valve) and burst disc ensures the tank does not fail due to
Rocket Fuel Solenoid Vent Valve an overpressure event. Consequences of vent valve failure are low as valve is normally
Pre-flight 1 2 2
Actuation Failure open, resulting in a safe system state in the event of a valve failure. Vent valve only
vents from above the maximum liquid level in the tank, meaning no liquid is lost during
a valve failure.
COTS solenoid valve and Prior tests ensure that a valve failure is unlikely.
Consequences of vent valve failure is low as valve is normally open, resulting in a safe
Rocket N2 Solenoid Vent Valve
Pre-flight 1 2 2 system state in the event of a valve failure. Nitrogen COTS COPV can also be vented via
Actuation Failure
the electronic regulator through fuel and oxidiser tank venting systems, to ensure
rocket is safe to approach.
Consequences of valve failure are low as vehicle can be vented and approached to
facilitate repairs. If valve fails during ignition, ethanol will be dumped and burnt off
Main N2O Valve Actuation Failure Pre-flight, Ascent 1 2 2
immediately presenting low risk. During flight, valve is intended to remain open to full
vent all fluids from the rocket. During flight, tank vents are automatically opened to
ensure rocket is safe to approach.
Consequences of valve failure are low as vehicle can be vented and approached to
Main Fuel Valve Actuation Failure Pre-flight, Ascent 1 2 2 facilitate repairs. If valve fails during ignition, nitrous will be dumped from the rocket.
During flight, valve is intended to remain open to full vent all fluids from the rocket.
During flight, tank vents are automatically opened to ensure rocket is safe to approach.
Engine designed with appropriate safety factors, and tested over 16 times at all
expected operating points, as well as some tests both above target maximum thrust
Engine Combustion Chamber Explosion Ascent 1 3 3
(112%) and for extended durations (12s) . Engine hydrostatic tested to 1.5x the MEOP.
Due to the nature of this failure, there is no way to reduce the severity.
The regenerative cooling channels can fail if the heat transfer to the coolant is
insufficient. This can lead to a structural failure of the channel wall where coolant can
leak into the comubstion chamber. This will reduce the efficiency of the engine as
Engine Combustion Coolant Channel
Ascent 1 2 2 more fuel is dumped than nominal, but should not compromise anything else on the
Failure
rocket, hence is not a critical failure. Test data from instrumented engine indicates
critical parameters within expected design region (e.g fuel injector flow temperature)
indicating this failure mode is unlikely.
Ignition method tested successfully numerous times over the past 3 engine
development cycles. COTS motors with well fitting COTS ematches proven to be
Ignition Failure Pre-flight 1 1 1
reliable. Consequences of ignition failure are low as vehicle can be vented and
approached to facilitate repairs and swapping ignition cartridges.
Tank designed with appropriate safety factors, and tested numerous times integrated
in flight system. Tank fitted with a calibrated SPRV and burst disc to vent overpressure.
Bust disc sized to 4 times the area of the filling pipe. Furthermore, it is designed to fail
SRAD N2O Tank Explosion Pre-flight, Ascent 1 3 3
in a specific manner, wherein radial bolts fail in shear-out, allowing end-cap to pop out
and tank to de-pressurise. Due to the nature of this failure, there is no way to reduce
the severity.
Tank designed with appropriate safety factors, and tested numerous times integrated
in flight system. Tank fitted with a calibrated SPRV and burst disc to vent
overpressure.Bust disc sized to 4 times the area of the filling pipe. Furthermore, it is
SRAD Fuel Tank Explosion Pre-flight, Ascent 1 3 3
designed to fail in a specific manner, wherein radial bolts fail in shear-out, allowing end-
cap to pop out and tank to de-pressurise. Due to the nature of this failure, there is no
way to reduce the severity.
Fluid system components will be hydrostatically tested and full propulsion system leak
tested with nitrogen. Procedure has been developed using industry standards. If leak
Fluid system leak Pre-flight, Ascent 1 2 2 occurs during pre-flight phase, vehicle can be vented and approached to facilitate
repairs. During flight, nothing can be done to fix leakage, however consequences are
low as the system only needs to hold pressure for a few seconds after liftoff.
GSS Filling Hose Vent Valve Failure Pre-flight 1 1 1 Consequences of valve failure are low as vehicle can be vented and approached to
facilitate repairs.
QD collar is held in place allowing rocket to disconnect it as it launches. Redundant
cable and bracket system used to ensure collar is well secured to the launch pad.
Ground tests have been performed to characterize the expected pull out force and
Quick Disconnect Failure Ascent 1 3 3
ensure the collar works as intended so failure is very unlikely. Consequences of failure
would damage filling lines on both the rocket and ground support system, likely
requring extensive repairs.
Consequences of valve failure are low as vehicle can be vented through normally open
solenoid valves and approached to facilitate repairs. If valve opens unintentionally or
Electronical Regulator Failure Pre-flight, Ascent 1 2 2
by too much, high pressure nitrogen can enter the low pressure plumbing parts, where
the SPRV and burst disc will fail first before anything else. Bust disc sized to 4 times the
area of the filling pipe.
Payload
The payload and deployer has been designed considering the expected loading
Strucutral failure of deployer box /
Ascent, Descent 1 3 3 conditions with a safety factor of at least 1.5. While this severity is critical to the
payload
payload, it is not expected to have any significant impact on the launch vehicle.
Two lock arms are used within the system to lock the payload deployer in the stowed
position, providing redundancy and ensuring that failure of a single arm does not cause
Lock arm failure Pre-flight, Ascent 1 3 3 premature deployment. The locking mechanism has been designed to operate
nominaly with a single arm being sufficent to retain the payload in all loading
conditions.
In the case of a steering spool failure, resulting in the loss of directional control, the
payload still has a nominal, simulated flight trajectory where the parafoil will perform
Parafoil steering spool failure Descent 1 2 2
well within the technical limits. The steering mechanism has been tested under
expected loading conditions (wind tunnel) with no anomallies so failure is unlikely.
The parafoil is made of rip-stop nylon, which is specifically chosen to mitigate the risk
Parafoil rips Descent 1 3 3 of ripping the parachute material. Parafoil has been load tested under expected
descent conditions sucssefully.
Arbor, reef, uno knots become undone Knots are unlikely to become undone in flight as shown with previous tests on similar
Descent 1 3 3
mid flight payload as well as ground tests showing they should have sufficient strength.
Parafoil eyebolts sized for expected shockload during deployment and tested under
Parafoil eyebolt failure Pre-flight, Ascent 1 3 3
expected loading conditions with same parafoil.
Flight Batteries tested with multimeter before installation on launch day.
Communication with SRAD flight controller and COTS tracker verifed before launch in
Avionics failure Pre-flight, Ascent 1 2 2
pre-launch checklists. Parafoil deployment is passive and is automatically deployed
during payload deployment.
The scientific experiment acts as an independent body to the rest of the electronics
Science experiment failure Pre-flight, Ascent 1 2 2 system therefore any failure within this system is self contained and has no effect on
the trajectory of the payload
Steering lines are tied through the center of the spool, so it is not possible for the
Steering Line unspooling Pre-flight, Ascent 1 3 3
steering lines to completly unspool.
Protective parachute box encapsulates the parafoil ensuring the parafoil does not get
The parafoil gets caught in deployer
Pre-flight, Ascent 1 3 3 caught when being deployed. This has been tested and verified through a ground
when deployed
testing campaign to ensure succesful deployment in all expected conditions.
I Checklists
In the following subsections are the team’s main checklists from final assembly in the assembly tent on
launch day up until launch of Nimbus 24. Figure 90 details a legend used for the checklists.
Figure 90: Checklist Legend
113
I.1 Checklist A - Paddock
Team 12 - NIMBUS24 Imperial College Rocketry London
A - Paddock
Location: Paddock
Summary: Upon completion, the rocket will be ready to transport to
the assembly.
Lead:
Members: ICLR
Time To
Step Responsible Task Comment Check
Launch
A1 Pablo Airframe Checks
Stringers Visual inspection, fasteners
Engine Truss Visual inspection, fastener

All panels excluding lower
Panels On
propulsion 1
Check spare fasteners
A2 Pablo Recovery Checks
Parachute
Connect reefing line
Connect reefing nichrome
Check reefing continuity
Wrap reefing case in nichrome
Fold parachute and place in recovery tube
Clamp Band
Check hinge and clamps are secured to
spring steel band
Check dyneema is tied correctly to the
clamp band
A3 Arfred Payload Checks
Parafoil
Check lines are untangled
Check all knots are secure
Check parafoil is folded correctly
Payload
114
Check payload battery voltages If low, charge
Check Pickle and CATS are powered
Guided Recovery System
Check parafoil control lines are wound on

the spool
Check the guided recovery is actuating
correctly
HERMES 0.5U
Check 0.5U battery voltage If low, charge
Check 0.5U is powered on
Payload
Check payload panels are mounted
securely
Deployer
Check bungee is taut and knots are tight If not cut and tie new cord
Check dinosaur locks and unlocks
correctly
Load the payload into the deployer
A4 Usmaan Canard Checks
If canards are flying active,

Canard servo mounted ensure
servo is mounted to the top
If canards are locked on

Canard locking plate EuRoC request, ensure servo
is mounted to the top
Check canard bevel gears are greased
Check canards are aligned
Check canards spin
A5 Shiven Electronics Checks
Check all boards are powered

Check all actuators are mapped and
Servos, Solenoids
functioning
Check all sensors are mapped correctly
Check boards have SD cards and
Kermit, Pickle, Stark
formatted ExFAT
Check ignition sequence config for Pickle including E--Reg
115
Zero points are correct and

Check payload deployer and canard
deployer
servos function
servo mirroring
Check recovery sequence config for
Pickle
Check all rocket and payload battery
Charge if necessary
voltages
A6 Martin E Propulsion Checks
Assemble the lower and upper

propulsion sections
Assemble the engine on to the truss
Check all valves are assembled correctly

Follow from tank to filling ball
Check fuel plumbing
valve
Check oxidiser plumbing Follow from tank to QD
Check pressurant plumbing Follow from tank to QD
Check all Swagelok fittings are tightened
Check QDs are functioning correctly
Entire propulsion system should be ready

for transport
as a single system
A7 Shiven Camera Checks
MS-DOS FAT 32 format SD cards
Check SD cards are in the cameras
Check cameras are recording
Rocket is ready for transport to the assembly tent
Approved by:
116
I.4 Checklist B - Final Assembly

B - Final Assembly
Location: Assy Tent
Summary: Upon completion, the rocket will be ready to go to pyro tent.
Lead: Pablo
Members: Arfred, Daniel, Shiven
Step Time To Launch Responsible Task Comment Check
B-1 Pablo Begin Initial Set Up
B-1-1 Arrange tables
B-1-2 Set up the rocket stands
B-1-3 Bring rocket sections inside tent
B-2A Arfred Payload Assembly Refer To Payload Assy Checklist
Refer To Recovery Assy Checklist
B-2B Daniel Recovery Assembly
Request Signed Checklist
B-2C Shiven Electronics Assembly Refer to Electronics Assy Checklist
Hold, only proceed when recovery assembly is complete
B-3 Pablo Recovery - Payload Integration
B-3-1 Soham Connect recovery actuation lines
B-3-2 Rec Assy Align recovery section to payload section
B-3-3 Rec Assy Fasten couplers
B-4 Pablo Payload - Upper Propulsion Integration
B-4-1 Soham Connect RBUS Red line
B-4-2 Soham Connect RF lines Purple lines
B-4-3 Soham Connect COTS line Yellow and black line
B-4-4 Soham Connect recovery actuation lines Green lines
B-4-5 Rec Assy Align payload section to upper propulsion
B-4-6 Rec Assy Fasten couplers
Hold, only proceed when payload assembly is complete
Refer To Payload Assy Checklist
B-5A Arfred Payload Integration
Refer to Electronics Assy Checklist
B-5B Shiven Electronics Systems Check
B-6 Pablo Panel Installation
B-6-1 Pay Assy Assemble payload panels
B-6-2 Pay Assy Assemble canard panels
B-6-3 Rec Assy Assemble upper propulsion panels
B-6-4 Rec Assy Assemble lower propulsion panels
B-6-5 Rec Assy Check lower propulsion panels 1 isn't assembled
B-6-6 Rec Assy Assemble fin can
Final assembly complete, report to C-Pyro Tent Lead
Approved by:
117
I.5 Checklist B2A/5A - Payload Assembly and Integration

B-2A/5A - Payload Assembly & Integration
Location: Assy Tent
Summary: Upon completion, the payload will be fully integrated into the rocket.
Lead: Arfred
Members: Rosalind, Jayden
2A-1 Arfred Remove payload panel
2A-2 Jayden Begin payload electronics assembly
2A-2-1 Jayden Check CATS Vega battery voltage
2A-2-2 Jayden Check Pickle battery voltage
2A-2-3 Jayden Check 0.5U is mounted correctly
2A-2-4 Jayden Power 0.5U
2A-2-5 Jayden Arm CATS power Plug XT30
2A-2-6 Jayden Arm Pickle Power Plug XT30
2A-2-7 Jayden Check CATS telemetry
2A-2-8 Jayden Check Pickle telemetry
2A-2 Rosalind Begin parafoil preparation
2A-2-1 Rosalind Check parafoil lines are untangled
2A-2-2 Rosalind Check parafoil control lines are equal length
2A-2-3 Rosalind Check parafoil knots are correct and tight
2A-2-4 Rosalind Fold Parafoil
2A-3 Arfred Install payload panel
Assembly completed, report to B-Final Assembly lead
Hold for payload integration
5A-1 Arfred Payload Installation
5A-1-1 Rosalind Ensure the locking servos are in their zeroed position
5A-1-2 Rosalind Manually unlatch the dinosaurs and open the deployer
5A-1-3 Jayden Open the bottom deployer hinge and hold in deployed state
5A-1-4 Rosalind Insert folded parafoil into chute box
5A-1-5 Rosalind Loaded the payload into the deployer Ensure no lines are caught on the deployer
5A-1-6 Rosalind Close deployer hinge and lower deployer into the rocket
5A-1-7 Rosalind Ensure the dinosaurs are correctly latched Gently pull on the deployer
Integration completed, report to B-Final Assembly lead
Approved by:
118
I.6 Checklist B2B - Recovery Assembly and Integration

B-2B - Recovery Assembly & Integration
Location: Assy Tent
Summary: Upon completion, the recovery system will be fully integrated.
Lead: Daniel
Members: Shaeel, Alex
2B-1 Daniel Parachute Installation
2B-1-1 Shaeel, Alex Check shock cords are connected to body tube eye bolt
2B-1-2 Shaeel, Alex Check shock cords are connected to nose cone eye bolt
2B-1-3 Shaeel, Alex Check nichrome case assembly
2B-1-4 Shaeel, Alex Connect reefing actuation lines
2B-1-5 Shaeel, Alex Check pilot chute cords are attached to the main parachute
2B-1-6 Shaeel, Alex Place parachute into recovery tube
2B-1-7 Shaeel, Alex Place schock cords in the recovery tube
Hold, report parachute installation is complete to B-Final Assy Lead
2B-2 Daniel Clamb Band Installation
2B-2-1 Shaeel, Alex Check recovery springs are in place
2B-2-2 Shaeel, Alex Check clamp band actuation lines are accessible
2B-2-3 Shaeel, Alex Align nosecone coupler to body tube coupler
2B-2-4 Shaeel, Alex Clamp the band around the couplers tighten on loop side using Dynema cord
2B-2-5 Shaeel, Alex Connect actuation line to nichrome housing
2B-2-6 Shaeel, Alex Wrap the Dynnema in nichrome wire
2B-2-7 Shaeel, Alex Close the nichrome housing
Integration completed, report to B-Final Assembly lead
Approved by:
119
I.7 Checklist B2C/5B - Electronics System Checks

B-2C/5B - Electronics Systems Checks
Location: Assy Tent
Summary: Upon completion, electronics systems will be fully checked.
Lead: Shiven
Members: Soham, Andrei, Artem, Kiran
2C-1A Shiven Battery Checks
2C-1A-1 Check batteries are installed and connected Locations: canard module, nitrogen tank
2C-1A-1 Check batteries voltage
Locations: canard module, payload bay,
2C-1B Shiven Check SD cards are installed .
nitrogren tank, lower prop feed
Initial check complete, report to B-Final Assy Lead
Hold
5B-1 Shiven Begin electronics system check
5B-1-1 Connect PDU0 key
5B-1-2 Insert PDU0 DEPLOY OVERRIDE
5B-1-3 Verify deploy voltage 16.8v > 16v > 15v
5B-1-4 Verify logic power 4v > 3.3v > 3.0v
5B-1-5 Verify valves are in nominal position All valves closed
5B-1-6 Verify all boards logic power
5B-1-7 Remove DEPLOY OVERRIDE
5B-1-8 Remove PDU0 key
Electronics checks completed, report to B-Final Assembly lead
Approved by:
120
I.8 Checklist C - Pyro Tent

C - Pyro Tent
Location: Pyro Tent Summary: Upon completion, the rocket will have passed final
Lead: Shiven flight integration and ready to move to the pad.
Members: Martin E
C-1 Shiven Pre-pyro tent checks
C-1-1 Shiven B-Final Assembly sign-off
C-1-2 Shiven Rocket LRR sign-off
C-1-3 Shiven EuRoC go for pyro sign-off
C-2 Shiven Transport rocket to pyro tent
C-2-1 Move rocket stands to pyro tent
C-2-2 Move rocket to pyro tent
C-2-2 Move lower propulsion panel to pyro tent
C-3 Shiven Ignitor Assembly
C-3-1 Martin E Check the igniton tube assembly
C-3-2 Martin E Feed the e-match cable through the engine
C-3-3 Martin E Attach the e-match cable to the outside of the engine with masking tape
C-3-4 Martin E Thread the ignition tube onto the engine and tighten
C-3-5 Martin E Assemble lower propulsion panel
C-4 Shiven Prepare launch rail toolkit
C-5 Shiven Move rocket to final flight qualification
C-5-1 Collect all members of the launch rail team
C-5-2 Collect launch rail toolkit
C-5-3 Lift the rocket with 4 members of the launch team
C-5-4 Walk the rocket to the range safety tent
Pyro tent assembly complete, move to D-Launch Rail
Approved by:
121
I.9 Checklist D - Launch Rail
D - Launch Rail
Location: Pad
Summary: The rocket is at the pad, upon completion, all personnel will be ready to
evacuate the pad.
Lead: Shiven
Members: Soham, Hardik, Martin L, Ethanol Filling Team
D-1 Shiven Before Moving Rocket To Launch Rail
Has the team received clearance to move to the

D-1-1
launch rail?
Has the launch rail tool kit been collected and
D-1-2
checked?
D-1-3 Is MC set up and ready to go?
D-1-4 Has MC/LC established radio communication? Ensure the team has two fully charged radios
D-1-5 Has the launch rail been prepped for arrival?
D-1-6 Has the team got the relevant safety equipment
D-1-7 Has the launch rail team taken a bio break?
D-2 Shiven Transport the Rocket to The Launch Rail
4 member of the launch team will carry the

D-2-1
rocket
D-2-2 1 member shall guide those carrying the rocket
D-3 Shiven Load Rocket onto the Launch Rail
D-4A Shiven Attach Ethanol Drain Tube and Cup
D-4B Shiven Insert PDU0 PDU1 Pull Before Flight Pins
D-5 Shiven Attach PDU0 and PDU1 Keys
D-6 Shiven Logic Power On
Visual check of the OLEDs for nominal operation

D-6-1
of logic rail
D-6-2 Connect EQD and GPQD Make sure the GSS is turned on
Check comms and power rail monitoring by

D-6-3
Andrei P
D-7 Shiven Systems Check
Verify recovery isolated and engine ematch is

D-7-1 Disconnect recovery xt60s
disconnected
D-7-2 Ping test from mission control
D-7-3 Remove PDU0 RBF pin
122
D-7-4 PDU0 transition to LIVE state
Logic Power 3.3v,

D-7-5 Bus voltage checks
Deployment Power above 14.8v
Listen for nominal valve operation. In the case of

GSS valves,
D-7-8 Valve cycle routine a visual check can also be performed. Verify that
all valves return to a closed state after checkout. If
any valves fail test, see section Valve Debug (H)
D-7-9 Close OX-FILL-N Close the GSS needle valves
D-7-10 Close N2-FILL-N Close the GSS needle valves
ARM, OPEN and CLOSE and DISARM OX-VENT-

D-7-11
GSS
ARM, OPEN and CLOSE and DISARM N2-VENT-
D-7-12
GSS
D-7-13 ARM, OPEN and CLOSE and DISARM OX-FILL-N
D-7-14 ARM, OPEN and CLOSE and DISARM N2-FILL-N
D-7-15 ARM SV-PRESS
D-7-16 ARM SV-OX
D-7-17 ARM SV-FUEL
D-7-18 ARM, OPEN and CLOSE and DISARM OX-FILL
D-7-19 ARM, OPEN and CLOSE and DISARM OX-HOSE
D-7-20 ARM, OPEN and CLOSE and DISARM OX-VENT
D-7-21 ARM, OPEN and CLOSE and DISARM N2-PRESS
D-7-22 ARM, OPEN and CLOSE and DISARM FUEL-VENT
D-7-23 ARM, OPEN and CLOSE and DISARM OX-MAIN
D-7-24 ARM, OPEN and CLOSE and DISARM FUEL-MAIN
D-7-25 Test fire engine ignitor with dummy resistor load Read ignition voltage with a multimeter
D-7-26 Test fire recovery with dummy resistor load Read ignition voltage with a multimeter
Test Continuity on Recovery Channels with

D-7-27
multimeter
D-7-28 Canard operation check
D-7-29 ARM, OPEN, and DISARM FUEL-VENT
D-7-30 Disarm all actuators
D-7-31 Insert PDU0 RBF pin to safe deployment power
D-7-32 Transition all boards to pre-flight/idle
123
Final Electronics Prep Before Rocket Goes
D-8 Shiven
Vertical
D-8-1 Plug in camera board power
D-8-2 De-isolate recovery channels
D-8-3 Replace electronics panel
D-9 Shiven Rocket goes vertical
Announce that the rocket is going vertical and

D-9-1
remove obstruction
D-9-2 Take up slack in the rail winch
D-9-3 Lift rocket until it is at the desired launch angle
D-9-3 Tighten up guide lines and lock in place
D-10 Shiven Attach fluid and electronic quick disconnects
Remove kapton tape and attach nitrous oxide

D-10-1
quick disconnect at the base of the rocket
Remove kapton tape and attach nitrogen quick
D-10-2
disconnect at the base of the rocket
Remove the outward facing lower feed section
D-10-3
panel by undoing the 4x M6X12 TX screws
D-10-4 Attach fuel filling hose to the fuel fill/drain port
D-11 Shiven Begin Ethanol Filling
Check ethanol plumbing line is assembled

D-11-1
correctly and that the equipment is clean
D-11-2 Ensure filling team is wearing appropriate PPE
Fill the ethanol transfer container with 5.3 L of

D-11-3 ethanol and place on a weighing scale.
Zero out the scale
With the filling valve closed, pressurise the

D-11-4 ethanol transfer tank using the pump on the
spray bottle
D-11-5 Place the ethanol tank on the scale and zero it
Begin filling the rocket with ethanol by slowly

opening the fill valve until the reading on the
D-11-6
scale is -2.5 kg
or ethanol starts to drip from the vent line
When filling is complete, close the push fit filling

D-11-7
valve and depressurise the transfer container
Close the needle valve on the rocket and reopen
D-11-8
the push fit filling valve.
D-11-9 Remove the ethanol drain hose and cup
Remove the methanol filling adapter and replace

Disconnect the filling line and store remaining
D-11-10 the 1/4" plug.
ethanol safely
Hand tighten
D-11-11 Move and store the ethanol filling equipment
124
Assemble the outward facing lower feed section
D-11-12
panel by replacing the 4x M6X12 TX screws
D-12 Shiven Prepare Remote Propellant Filling
Check hose fittings are tightened and bottle

adpater is tight on the bottle.
D-12-1 Check nitrous plumbing is assembled correctly
Ensure the needle valve is fully closed (turn
clockwise to close)
Verfiy that Valves OX-MAIN, OX-SV,OX-VENT,

OX-VENT-GSS, OX-FILL-GSS
D-12-2
are both in the CLOSED state (can visually
confirm)
Check that the filling team is wearing the

D-12-3 Nitrous safe gloves and a face shield
correctly PPE
D-12-4 Open the nitrous bottle
Check that the pressure of the bottle is in a

D-12-5 Between 40 - 65 bar
suitable range using the gauge on the GSS
D-12-6 Open the nitrogen cylinder
Check that the pressure of the N2 bottle is in a

D-12-7 Above 250 bar
suitable range using the gauge on the GSS
Launch rail checklist complete, move to E-Prelaunch
Approved By
125
I.10 Checklist E - Prelaunch
E - Prelauch
Location: Pad/MC
Summary: Upon completion the rocket is ready for go/no-go and launch.
Lead: Andrei
Members: Shiven, Martin E
Time To
Step Responsible Task Comment Check
Launch
E-1 Andrei Before Moving to Prelaunch
E-1-1 Have all items on previous checklists been completed?
E-1-2 Is the team prepared to launch in the next 15 - 60 minutes?
Does the weather look good for the upcoming launch

E-1-3
window?
E-1-4 Are all systems running nominally?
E-1-5 Are all nessecary members of the team in mission control?
Has the launch site been cleared of debris and non critcal
E-1-6
equipment?
E-2 Shiven Hold Until Ready To Begin Filling
E-3 Shiven All Unecessary Personel Evacuate
E-4 Shiven Attach Ignitors
E-4-1 Check that no current is going through the ignition leads
E-4-2 Connect e-matches to the igniton leads
E-5 Shiven Pull PDU0 PDU1 RBF Pins Out
E-6 Shiven All Remaining Personel Evacuate
E-7 Andrei PDU Deployment Live Command
E-8 Andrei Valve Arm Command
E-9 Andrei ARM and CLOSE FUEL-VENT
E-10 Andrei N2 COPV Filling
E-10-1 Check N2 HOSE VENT valve is CLOSED
E-10-2 Check OX HOSE VENT valve is CLOSED
126
E-10-3 Check EREG is CLOSED and DISARMED
E-10-4 ARM and SLOW FILL with N2 FILL valve
E-10-5 Wait for N2 PRESSURE to reach slow rate of increase
E-10-6 Move to FAST FILL with N2 FILL valve
Wait until either N2 PRESSURE has reached approximately

E-10-7
160 bar
E-10-8 CLOSE N2 FILL valve and DISARM
Check N2 PRESSURE is stable and OX TANK PRESSURE and

E-10-9
FUEL PRESSURE are at atmospheric
E-11 Andrei OX Tank Filling
E-11-1 ARM and CLOSE (energise) OX SOLENOID VENT valve
E-11-2 ARM and OPEN OX VENT valve to SLOW VENT position
E-11-3 Check that OX TANK PRESSURE is still at atmospheric
E-11-4 ARM and OPEN OX FILL valve
E-11-5 Zero ROCKET MASS load cell
Check that OX TANK PRESSURE is increasing to between

E-11-6
30-50 bar
If OX TANK PRESSURE is above 40 bar, open OX VENT valve

E-11-7
to FAST VENT position
Watch for plume from vent outlet. CLOSE OX FILL valve

E-11-8 when
this occurs
Continue blipping OX VENT valve to FAST VENT from SLOW

E-11-9 VENT until OX TANK PRESSURE reaches 25 - 30 bar.
Do not fully close the vent valve
E-11-10 OPEN OX FILL valve until plume is visible from vent again
E-11-11 CLOSE OX FILL valve and DISARM
E-11-12 CLOSE OX VENT valve and DISARM
Verify that OX TANK PRESSURE is below 38 bar

E-11-13
(if it is, repeat steps E-11-9 to E-11-13)
127
E-11-14 ARM and OPEN OX HOSE VENT valve
E-11-15 Wait until OX HOSE PRESSURE drops to atmospheric
E-11-16 CLOSE OX HOSE VENT and DISARM
E-12 Martin E Propulsion Systems Check (Go/No Go #1)
E-12-1 Are all pressures and temperatures nominal?
E-12-2 Has sufficient propellant been loaded?
Do the propellants seem stable in case of a delayed launch

E-12-3
window?
E-13 Andrei Hold Until Turn For Launch
E-14 Andrei Pressurise Propellants
E-14-1 ARM and CLOSE (energise) FUEL SOLENOID valve
E-14-2 Check that FUEL TANK PRESSURE is still at atmospheric
E-14-3 ARM EREG valve
E-14-4 Run the PRESSURISE command for the EREG
Monitor FUEL TANK PRESSURE to check that they are

E-14-5
increasing steadily
When FUEL TANK PRESSURE reaches the OX TANK

PRESSURE, check that both pressure values keep
E-14-6
increasing
to 40 bar
Check that the EREG closes once the tank pressures reach
E-14-7
40 bar. If it doesn't, close it manually.
Check that both tank pressures are stable and not

E-14-8 increasing
or decreasing
E-15 Andrei N2 COPV Repress
E-15-1 ARM and SLOW FILL with N2 FILL valve
E-15-2 Wait for N2 PRESSURE to reach slow rate of increase
E-15-3 Move to FAST FILL with N2 FILL valve
128
Wait until either N2 PRESSURE has reached approximately
E-15-4
250 bar
E-15-5 CLOSE N2 FILL valve and DISARM
Check N2 PRESSURE is stable and OX TANK PRESSURE and

E-15-6
FUEL PRESSURE still at 40 bar
E-15-7 ARM and OPEN N2 HOSE VENT valve
E-15-8 Wait until N2 HOSE PRESSURE drops to atmospheric
E-15-9 CLOSE N2 HOSE VENT and DISARM
E-16 Andrei Avionics Systems Check
E-16-1 Rocket: Zero Baro
E-16-2 Rocket: Set home
E-16-3 Rocket: Verify LoRa telemetry
E-16-4 Payload: Set home
E-16-5 Payload: Arm valve
E-16-6 Payload: Verify LoRa telemetry
E-17 Martin E Propulsion Systems Check (Go/No Go #2)
E-17-1 Are all pressures and temperatures nominal?
E-17-2 Do the propellants seem stable?
Prelaunch checklist complete, move to F - Launch
Approved by:
129
I.11 Checklist F - Launch

F - Launch
Location: MC
Summary: Upon completion the rocket has been launched
Lead: Andrei
Members:
F-1 Andrei Start Subsystems Go/No Go
F1-1 Recovery go/no go Verify continuity on recovery ignitors
F1-2 Avionics go/no go
F1-3 Propulsion go/no go
F-2 Andrei Flight go/no go
F-3 Andrei Arm Rocket for Launch ARM launch button
F-4 Andrei Launch Control go/no go
F-5 Andrei Inform RSO vehicle ready
F-6 HOLD Andrei Countdown Starts
PRESS Launch Button to initiate
F-7 Andrei Launch
launch autosequence
Rocket has successfully launched
Approved by:
130
131
J Engineering Drawings
132
133
1 2 3 4 5 6 7 8
.5°
°±0 0.5°
A 0.0 80.0°± A
12
40.0°±
0.5°
14
6x
B B
12 x
Ø6.
A A
2
C C
Through hole
134
.05
8±0
5.4
This taper is annotated .05
D 3 ±0 D
in the A-A section view
7.6
A-A (1:1.5)
60.0°±0
.1°
E E
TOLERANCES THIRD ANGLE PROJECTION
MATERIAL: TITLE:
ANGULAR ± 1°
X = ± 0.5
X.X = ± 0.1 SURFACE FINISH Aluminum 6061
X.XX = ± 0.02 MACHINED
FACES Ra 6.3
Fuel Tank Bulkhead v15
ALL DIMENSIONS ARE IN
NAME DATE MILLIMETRES
F DRAWN Hardik Modi 24/07/2023

DO NOT SCALE DRAWING DWG NO. F
CHECKED
Fuel Tank Bulkhead
APPROVED A3 SCALE 1:1.5 SHEET 1/1 REVISION:
1 2 3 4 5 6 7 8
1 2 3 4 5 6 7 8
.5°
±0
° °
A 0.0 80.0°±0.5 A
12 40.0°±0.5
°
14
6x
B B
12 x
Ø6.
2
A A
C Through hole C
135
.05
8 ±0
5.4
This taper is annotated .05
3 ±0
in the A-A section view 7.6
D D
A-A (1:1.5) 35°±

0.1°
E E
MATERIAL: TITLE:
ANGULAR ± 1°
X = ± 0.5
X.X = ± 0.1 SURFACE FINISH Aluminum 6061
FACES Ra 6.3
Nitrogen Tank Bulkhead v12

CHECKED
Nitrogen Tank Bulkhead
APPROVED A3 SCALE 1:1.25 SHEET 1/1 REVISION:
1 2 3 4 5 6 7 8
1 2 3 4 5 6 7 8
0.5x37°
A (use a deburr A
A (5:1) to make this)
B B
2
66)
(36.
52
(Through Hole)
C
A C
17
136
(Ø33)
0.0
+0.1
Ø12.0
D D
35.5 14 M20x1.5 6g
(major diameter is
Ø18 19.968)
E E
MATERIAL: TITLE:
ANGULAR ± 1°
X = ± 0.5
X.X = ± 0.1 SURFACE FINISH Aluminium 6082
FACES Ra 6.3
Nosecone with Aluminium Tip v29

CHECKED Martin Lueng 16/07/2023

Nosecone with Aluminium Tip
APPROVED Jayden Ting 17/07/2023
A3 SCALE 1:1 SHEET 1/2 REVISION:
1 2 3 4 5 6 7 8
137
138
139
1 2 3 4 5 6 7 8
49
47
A A
L
AL
4
Ø1 RU R1
TH
3
xØ
4
Dont require
8.1
Ø10
B B
Ø6
Ø20
C C
4.1 5.1 2x M3x0.5 6H
140
2.5 7.0
This is space for the screw, geometry not important, 17.0

do what is convenient to allow this space 10.38
22
R.5
4
R.
D D
7.5
.5
E E
R1
MATERIAL: TITLE:
ANGULAR ± 1°
X = ± 0.5
X.X = ± 0.1 SURFACE FINISH Aluminimum 6082T6
FACES Ra 6.3
prototype canard shaft v24
F DRAWN Martin Leung 7/8/2023

CHECKED
prototype canard shaft
APPROVED A3 SCALE 3:1 SHEET 1/1 REVISION:
1 2 3 4 5 6 7 8
1 2 3 4 5 6 7 8
A A
93
13
75
B B
Ø23
Ø29
A A
C
(18) C
141
15
G 1/2-14 A x 15/15
D 19mm Tapping Drill A-A (1:1) D
BREAK ALL EDGES
18.9 - 19
83.5
88.9
E E
MATERIAL: TITLE:
ANGULAR ± 1°
X = ± 0.5
X.X = ± 0.1 SURFACE FINISH ----
FACES Ra 6.3
Stronger Ignition Cartrige v6
F DRAWN Martin Leung 8/4/2023

CHECKED
Stronger Ignition Cartrige
1 2 3 4 5 6 7 8
142
143
144
B-B (2:1)
23.4
R6
D (4:1)
8
9.5
11
42.4
60
44.4
4
Ø11
15° OFFSET
48.3
51.4
65 12 HOLES
145
30° SPACING Ø1.5 3.5

A
B B
1°
0.
°±
10
A D A-A (2:1)
Dept. Technical reference Created by Approved by
17.45 (17.43)
ICLR Philip Nome 31/01/2023

17.47
16.7 h9 (16.657)
(19.947)
Document type Document status

16.7
19.98
STAINLESS STEEL 316 UNDER REVISION

Title DWG No.
20 f8
BREAK ALL SHARP PINTLE THA-002

CORNERS Rev. Date of issue Sheet
2 1/1
1 2 3 4 5 6 7 8
19
+0.0 1 n6
Tolerances for Hole Basis, Ø1 0 -0.0
-.008
A Simmons et al 2020, Table 20.1 Ø6-0 h6 A
Ø5 +.019
-.01
n6
TOP VIEW
B B
+.012
H7
Ø3 0
4
C C
146
+0.1
Ø3.0 -0.1
+0.1
30.0 -0.1
D D
+0.1
18.0 -0.1
E E
ISO VIEW FRONT VIEW RIGHT VIEW

DRAWN DATE
IF IN DOUBT ASK Jayden Jackson 25/08/2024 DRAWN CHECKED
MATERIAL SCALE UOS

DO NOT SCALE Titanium 3:1
FINISH/TREAT
Smooth
BS 8888 MASS (g)
F 12.83 F
TITLE UN-DIMENSIONED CAD SURFACES MACHINE FROM CAD
Guided Recovery Shaft BREAK ALL SHARP EDGES WITH MIN RAD OR CHF UOS PROJECT SHEET
IMPERIAL COLLEGE LONDON ROCKETRY, EXHIBITION RD, SOUTH KENSINGTON, LONDON SW7 2AZ ALL THREADS RH UOS Payload - ICLR 1 of 1
1 2 3 4 5 6 7 8
1 2 3 4 5 6 7 8
A A
B B
C C
147
D D
E E
DRAWN DATE
IF IN DOUBT ASK JAYDEN JACKSON 30/08/2024 DRAWN CHECKED
MATERIAL SCALE UOS

DO NOT SCALE TITANIUM AND STEEL 2:1
FINISH/TREAT
BS 8888 MASS (g)

F F
TITLE UN-DIMENSIONED CAD SURFACES MACHINE FROM CAD
ASSEMBLY DRAWING GEARS BREAK ALL SHARP EDGES WITH MIN RAD OR CHF UOS PROJECT SHEET
IMPERIAL COLLEGE LONDON ROCKETRY, EXHIBITION RD, SOUTH KENSINGTON, LONDON SW7 2AZ ALL THREADS RH UOS PAYLOAD - ICLR 1 OF 1
1 2 3 4 5 6 7 8
1 2 3 4 5 6 7 8
A A
19.0
4.5
4.5
75.0
84.0
93.0
121.0
B B
144.0
144.0
213.0
213.0
C C
148
D This way up D
4.5
4.5
1. All are M3 thru, clearance and CSK such that bolt head is flush with
the surface
E E
2. All the holes are the same distance from the edge of the stringer
MATERIAL: TITLE:
ANGULAR ± 1°
X = ± 0.5
X.X = ± 0.1 SURFACE FINISH Aluminum
FACES Ra 6.3
Stringer A
F DRAWN Arfred Garcia 26/08/2024

CHECKED
PLD003
1 2 3 4 5 6 7 8
K Data Task Request Handler Example JSON Config
1 {
2 "task_name": "nim_oxventsrv_telem",
3 "autostart": false,
4 "poll_delta": 1000,
5 "running": false,
6 "logger": true,
7 "receiveOnly": false,
8 "request_config": {
9 "source": 1,
10 "destination": 9,
11 "destination_service": 10,
12 "command_id": 1,
13 "command_arg": 0
14 },
15 "packet_descriptor": {
16 "state": "uint16_t",
17 "value": "int32_t"
18 },
19 "bitfield_decoders": [
20 {
21 "variable_name": "component_status",
149
22 "bitfield": "state",
23 "flags": [
24 {
25 "id": 0,
26 "description": "NOMINAL"
27 },
28 {
29 "id": 1,
30 "description": "DISARMED"
31 },
32 {
33 "id": 6,
34 "description": "ERROR"
35 }
36 ]
37 }
38 ],
39 "rxCounter": 0,
40 "txCounter": 0,
41 "connected": true,
42 "lastReceivedPacket": "",
43 "rxBytes": 0,
44 "txBytes": 0
45 }
L Data Task Request Handler UI
Figure 91: Screenshot of the data task request handler user interface
150
M Geddan Power Source Logic
The Geddan deployment power rail requires special consideration to the rest of the Ricardo avionics.
This is because the board is primarily powered by an external battery, hence with no additional con-
sideration, the board would not respect the bus-wide rule that RBUS deployment line being off means
that actuators are not powered. The desired behaviour is detailed in Figure 92.
Figure 92: Flowchart of Geddan desired deployment power behaviour.
The LTC4412 ideal diode controller’s output can be controlled by the state of its’ control (CTL) pin,
with a LOW (ZERO) CTL pin turning the output on, and a HIGH (ONE) CTL pin turning the output off.
Hence, to implement the logic shown in Figure 92, the truth table shown in Table 32 is required. In
some cases, the output of the logic gates is not important due to the power source not being present,
hence in these cases the output has been marked with a ’-’.
Table 32: Geddan power switchover truth table
Battery Power RBUS Power Battery CTL RBUS CTL

0 0 - -
0 1 - 0
1 0 1 -
1 1 0 1
151
Implementing this in hardware requires a NAND gate connected to the Battery CTL pin, and an AND
gate connected to the RBUS CTL pin. Potential dividers are used to shift the deployment level voltages
to below 3.3V, allowing the logic gates to compare them. This can be seen in the final implementation
in Figure 93.
Figure 93: Final hardware implementation of Geddan deployment power logic.
As a final precaution against a premature enabling of actuators, the EN pin of the buck converter is
connected to the ESP32, meaning that the servos will only be powered when the canard component
has been armed.
152
N SRAD Tank Failure Calculations
N.1 Hoop Stress Failure
The cylindrical tank walls are made from a 170mm aluminium tube. Aluminium 6061-T6 was chosen
as it has a relatively high yield strength of 276 MPa [2]. It is also readily available and so easy to source
in the size required.
Due to the thin walls of the tank relative to its diameter, the thin wall pressure vessel formula for
hoop stress and axial stress can be used to calculate the von mises stress in the tank wall.
PR
σθ =
t
PR
σz =
2t
q
σv = σθ2 + σz2 − σθ σz
With an internal pressure of 120 bar (2x MEOP) and an diameter of 170mm, the minimum thickness
required is 3.1mm. The original tank tube wall thickness is 5mm, but this is machined down to reduce
mass. During machining, the tube was clocked in to within 0.5mm run out and 1.2mm was removed
from the low spot giving a minimum wall thickness of 3.3mm. This gives an overall safety factor against
yield in the tank wall of 2.14.
N.2 Bolt Shear Failure

To constrain the end caps at either end of the tube, 18x 6mm Shoulder Bolts are used, made from
grade 12.9 steel. Grade 12.9 steel has a shear strength of 517 MPa [3], and with a 6mm solid ground
diameter, this gives a shear strength per bolt against yield of 14.6 kN.
The axial force on each end cap is given by the internal pressure multiplied by the internal cross section
of the end cap. This force is divided by the number of bolts to find the shear force in each bolt.
2
πP Dtank
Fax =
4
Fax
Fbolt =
Nbolts
For a tank internal diameter of 160mm, at 120 bar (2x MEOP), the total axial force on each endcap
is 241 kN. Dividing by 18 gives a shear force in each bolt of 13.4 kN, which is just below the shear
strength against yield of 14.6 kN, giving an overall safety factor of 2.18.
N.3 Hole Tear Out Failure

Bolt tear out occurs when there is too little material on the side of the tube that the radial bolts apply
force against when under pressure. It can be modelled by looking at the shear area available for a
rectangular section of material after the bolt, which represents the worst case of stress distribution.
The hole tear-out stress in the aluminium can be calculated using the following formula.
Fbolt
σtear−out =
2txmin
153
where xmin is the minimum distance from the edge of the bolt hole to the edge of the tank. In this
case, xmin is 14.3mm. With a tank wall thickness of 5mm around the bolt holes and a force of each
bolt of 13.4kN at 120 bar (2x MEOP), the maximum bolt tear-out shear stress is 187 MPa. Aluminium
6061-T6 has a shear strength of 207 MPa [2], giving an overall safety factor on the MEOP of 2.20 for
bolt tear-out.
N.4 Hole Bearing Surface Failure

Bearing stress is the stress applied by the bolt to the contact areas with the bolt hole. If these are too
large, the bolt hole can plastically deform, potentially leading to more concentrated stresses in other
areas and eventual failure. The bearing stress can be defined as the force on the bolt over the contact
interface area of the bolt with the material.
Fbolt
σbearing =
Ainterf ace
To provide the maximum possible bearing surface, shoulder bolts were used with precision reamed H7
holes made in the tank endcaps to match. This means the bearing area can be found by multiplying
half of the bolt circumference by the wall thickness of material. This calculation is performed for both
the tank wall and the tank endcap which the bolt goes into. For a 13.4kN load per bolt at 120 bar (2x
MEOP), the bearing stresses on the bolt hole of the tank wall is 284 MPa. Aluminium 6061-T6 has a
bearing yield strength of 386 MPa, and an ultimate bearing strength of 607 MPa [2]. Therefore, there
is a minimum safety factor of 2.72 against bearing yield and significantly more margin against total
failure.
154
O Modelling the dynamics of post parachute deployment behaviour
It was found that third party apps to calculate the shock loads were not accurate and hence a custom
calculator was developed in Simulink. A two degree of freedom system was used to model the parachute
and rocket body system to obtain the initial shock loads experienced by the parachute cords and the
recovery bulkhead located in the rocket fuselage. The system was modelled as a multiple mass, spring,
and damper system where the parachute cord acts as a spring and the drag forces act on both the
parachute and body, and are treated as the inherent damping present in the system, being proportional
to velocity squared. The equations of motion for the parachute and the rocket body were modelled to
be
1
mp ẍ + ρẋ2 Ap CDp − mp g − k(y − x − L) = 0 (9)
2
1
mb ÿ + ρẏ 2 Ab CDb − mb g + k(y − x − L) = 0
2
where,
(
Es As
L , if y − x > L.
k= (10)
0, otherwise.
and where mp and mb are the masses of the parachute and body respectively, ρ is the density of the
air, x and y are the displacements of the parachute and body respectively, Ap is the nominal area of the
parachute, Ab is the area of the base of the nosecone, CDp and CDb are the respective drag coefficients
of the parachute and rocket body, g is the gravitational acceleration, k is the spring constant of the
shock cord, L is the length of the shock cord. Es is the Young’s modulus of the shock cord’s material
and As is the cross-sectional area of the shock cord.
The area, Ap , of the parachute was modelled using two models. First one being
AP = A, for t>0 (11)
where A is a constant number which is the area of the parachute being considered. The second model
described the area increasing linearly which resulted in a more physical model where
Ap t
(
tinf lation , for t <= tinf lation
A= (12)
Ap , for t > tinf lation .
and where tinf lation is the time it takes for the parachute to inflate fully.
The more realistic increasing area model replicates the actual process better and hence yields shock
loads much lower than the upper bound. However, it is worth noting that not even the linear increase of
the area is accurate as the actual relationship between the radius of the canopy of a parachute and time
is very difficult to predict analytically and requires computationally heavy simulations and/or empirical
relations which themselves are not accurate due to the scarcity in research conducted on the topic of
parachute inflation in the subsonic regime.
155
P Reefing characteristics equations, calculation and procedure
To determine the reefed diameter of the parachute, a set of relationships outlined by Knacke was used,
the results of which were then used for subsequent post-parachute deployment analyses performed.
From the selection of parachutes that were tested in the manual, the hemispherical parachute with a
spill hole best represented the selected parachute for the mission, and hence the values were chosen
from the experimental data available for the aforementioned parachute.
The diameter ratios, DC /D0 and DP /D0 , were given as 0.66 and 0.71 respectively. The the wetted
surface areas, SC , SP and S0 were then calculated using Sx = πDx2 /4. Using, the surface areas, and
the drag coefficient, CD0 , the drag area CD0 S0 was calculated, which was set to be equal to (CD S)P ,
as outlined by the Knacke. The drag area required when the chute is reefed, (CD S)R , was calculated
be equating the weight of the rocket in the recovery phase of the mission and solving for the drag area,
as shown below,
1 2
W = ρVdesc (CD S)R
2
2W (13)
(CD S)R = 2
ρVdesc
The reefing ratio was then determined, by finding the ratio of the drag area in the reefed state and the
nominal state using. The reefing line ratio, τ = D0 /DR , was then found using the experimental results
collected and summarised by Kancke, which then allowed the reefed diameter, DR , to be found.
156
Q Airframe Finite Elements Simulation Reports
Q.1 Body Tubes
Erica Keung
Date of Simulation: 10/06/2024
Overview
Figure 94: Rocket body tube
Simulation Description
The rocket body tubes made with bi-axial layered carbon fibre sheets is an important part of the rocket’s
structure as it helps to maintain its structural integrity while the rocket is moving. In order to make
sure the panels can withstand the maximum stress when the rocket is accelerating and can handle the
shock load, simulations are done in both buckling and tension mode to determine if its safety factor is
larger than 2.
Software
The software used for this simulation is Abaqus 2023, the licence used is the teaching license.
Set-Up
The body tubes are made of 8 layers of carbon fibre stacked on top of each other with a 90 degree angle,
thus the material setting of the simulation follows to ensure the most accurate result. The UTS of the
carbon fibre sheet is 2500 and the Poisson’s ratio is 0.47. Because the carbon fibre tube is very thin, a
hollow tube is used to simulate it. The tube is partitioned at distances 25mm,550mm,560mm,740mm
and 765 mm from the start in order to simulate the boundary conditions acting on the tube. A load of
3000N acts on the base of the tube until the first partition for the buckling load,a load of 4500 acts on
the base of the tube also until the first partition for the shock load. The encastre boundary condition
acts on the top of the tube until the partition and a displacement boundary condition acts on the middle
partition, where it allows the tube to move in a direction parallel to itself. 18328 nodes are used to
form the tetrahedral shaped mesh.
157
Simulations
Buckling load
Key Results
Table 33: Key Results - Buckling of a Body Tube
Applied Load kN Buckling mode Eigenvalue Safety Factor

5 1 50.862 50.862
5 2 50.862 50.862
5 3 50.867 50.867
Figures
Figure 95: Buckling mode 1 of Body Tube
Figure 96: Buckling mode 2 of Body Tube
158
Figure 97: Buckling mode 3 of Body TUbe
Discussion and Conclusion
From the above figures and tables, it can be concluded that the safety factor of the carbon fibre tube for
buckling is around 51 for all three modes, which is above the 2 requested, thus the tube is very suitable
to act as the body tube for nimbus.
Shock load
Key Results
Table 34: Key Results - Shock Load Applied to Body tube
Applied Load kN Max stress (MPa) Max displacement (mm) Stress Safety Factor
4.5 4.891 0.126 617
Figures
Figure 98: Displacement of Body Tube under shock load
159
Figure 99: Stress of Body Tube under shock load
Table 35: Mesh Convergence Study of Shock Load on Body Tube
Number of Elements Max Stress (MPa)

638 4.814
2080 4.893
4680 4.904
18204 4.891
From the above figures and tables, it can be concluded that the simulation has converged as the max
stress stays very similar as mesh size decreases for the tube, and that the safety factor of the carbon
fibre tube for forces from acceleration is definitely larger than the 2 requested, thus the tube is very
suitable to act as the body tube for nimbus.
160
Q.2 Stringers
Erica Keung
Overview
Stringers are one of the most important structural parts of the rocket, as they hold the rocket together,
allow the rest of the rocket to be assembled to them and withstand the acceleration and parachute
deploying shock stress so it is vital that the stringers would not break throughout the acceleration
process and parachute deployment. The simulations were done to test the structural design of the 3
stringers on the rocket.
The stringers are made of aluminium and have a Young’s modulus of 70 GPa, a Poisson’s ratio of 0.3
and the entirety of the stringer is a homogeneous section. Because there are 3 identical stringers for
each section of the rocket, the forces each stringer has to withstand are divided by 3. Therefore, the
stringers are simulated with a shock load of 1500N during the release of the parachutes. For the thrust
load, because the maximum thrust load of the rocket engines is 5000 N, the stringers will be simulated
with a force of 1667 N.
Software
The software used for all of the simulations is Abaqus 2023 edition, the license used is the teaching
license.
Upper Propulsion Stringer
Figure 100: Upper propulsion stringer
Set-Up
This is a stringer which is 1030mm in length. It is connected to the coupler at either end of the section
and fuel and nitrogen tank bulkheads along the length of the section. The 1667N acceleration load
and the 1500N Shock load are both applied to the hole at the bottom of the stringer. The holes that
the loads are applied to and the middle holes will have a encastre boundary condition, as the stringer
is connected to the tank, while all other M4 holes will have an encastre boundary constraint. The part
is meshed with a tetrahedral mesh with a mesh size of 1.75 mm for the buckling simulation and 1.675
mm for the shock load simulation to ensure accuracy.
161
Simulation 1 - Buckling
]
Table 36: Key Results - Buckling of upper propulsion stringer
Applied Load kN Buckle mode (mm) Eigen Value (MPa) Buckle Safety Factor
1.667 1 2.6978 2.6978
1.667 2 2.7232 2.7232
1.667 3 2.8404 2.8404
Figures
]
Table 37: Mesh convergence study of Buckling of upper propulsion stringer
Number of Elements Eigenvalue of mode 1
4170 2.8357
5717 2.7565
14621 2.7078
35649 2.6087
110364 2.6952
131570 2.6978
Figure 101: Buckling Mode 1 of upper propulsion stringer
162
Simulation 2-Shock Load
Table 38: Key Results - Shock loading of upper propulsion stringer
Applied Load kN Displacement (mm) Maximum Stress (MPa) Stress Safety Factor
1.5 0.3685 64.57 3.72
Figures
Table 39: Mesh Convergence study of Shock loading on upper propulsion stringer
Number of Elements Max stress (MPa)

4170 57.18
5717 54.53
14621 58.43
35649 61.58
110364 63.7
131570 65.46
134600 64.57
Figure 104: Stress under shock loading of upper propulsion stringer
163
Figure 105: Displacement under shock loading of upper propulsion stringer
From table Table 37 and Table 39, it is shown that both the buckling and shock load simulations con-
verged. For Simulation 1, the Eigenvalue of mode 1 of the simulation decreases and then converges at
110364 elements. For simulation 2, the Max stress increases and then converges at 134600 elements.
Thus both simulations are valid.
For both simulations, the safety factor is larger than the 1.5 standard. The safety factor of the buckling
simulation increases as its mode increases so it can be reasonably assumed that the buckling modes
larger than 3 will not be of concern. This shows that the structural design of this stringer is valid and
it can be used as a part of the rocket.
Lower Propulsion Stringer

Setup
Figure 106: Lower propulsion stringer: Stringer
This is a stringer that is 1082mm in length. It is connected to the rocket with 4 M4 circular holes on
both ends. The 1667N buckling loads and 1500N shock loads are all transmitted through the bottom
2 holes and they along with the 2 holes on the upper side of the stringer will have encastre boundary
conditions as they are meant to be fixed . The stinger is also attached to the oxidiser tank endcaps on
the 7th and 27th slot of the stringer counting from the base, so displacement and rotational boundary
conditions are also applied to the slots. The stringer is meshed with a tetrahedral mesh as it will result
in a more accurate simulation and the mesh sizes for the buckling and shock loads simulations are both
1.75 mm.
164
Table 40: Key Results - Buckling of lower propulsion stringer
Applied Load kN Buckle mode (mm) Eigenvalue (MPa) Buckle Safety Factor
1.667 1 2.9048 2.9048
1.667 2 5.8416 5.8416
1.667 3 11.305 11.305
Figures
Table 41: Mesh Convergence study of Buckling of lower propulsion stringer

7459 3.1232
12635 3.0521
38527 2.9553
104438 2.9065
125554 2.9048
Figure 107: Buckling Mode 1 of lower propulsion stringer
165
Table 42: Key Results - Shock loading of lower propulsion stringer
1.5 0.3662 64.66 3.71
Figures
Table 43: Mesh Convergence study of Shock Loading of lower propulsion stringer

12635 57.93
28134 62.26
104438 63.95
125554 64.66
Figure 110: Stress resulting from shock load for lower propulsion stringer
Figure 111: Displacement resulting from shock load for lower propulsion stringer
166
For the Shock stress simulation, from table Table 43, the max stress of each simulation increases with
the number of elements of the mesh but the teaching license would not allow more elements to be sim-
ulated. However, judging from the increase of max stress, it is reasonable to assume that the max stress
will converge a bit after 125554 elements, thus the simulation is valid. From Figure 110, Figure 111
and Table 42, it can be seen that the stress safety factor is way above the 1.5 needed, and the max
displacement is acceptable, so the structural design of the stringer is valid in terms of shock load.
For the buckling simulation, from Table 41, it can be seen that the eigenvalue of mode 1 of the simulation
decreases as the number of elements increases and converges to around 2.9048 at 104438 elements,
so the simulation is valid. The high safety factor can be contributed to the stringers are connected to
the nitrous tank endcaps which means the tank also takes a substantial proportion of the load, hence
alleviating a lot of the stringer length from carrying the load.
The unconventional geometry of the tank and the use of the limited license restricted a full stringer-
tank simulation from being run which would be a logical next step to accurately predict the buckling
characteristics of the tank.
167
Payload Stringer
Setup
Figure 112: Payload Stringer
This stringer is a stringer of 714 mm in length, it supports the rocket at the payload section. It is also
connected to the rocket couplers with 4 M4 holes on both ends of the rocket. The second pair of holes
counting from the bottom of the stringer is connected to one of the payload bulkheads and the third pair
of holes counting from the bottom of the stringer is connected to the other payload bulkhead. Because
the thrust loads and shock loads act on the stringers through the bottom pair of holes, the constraints
of the first, second and third pair of holes would be a displacement and rotational constraint, allowing
the stringer to move in the direction parallel to the stringer. The uppermost pair of holes are attached
to the other coupler of the payload section and has an encastre constraint. Although there are other
pairs of holes on the stringer, they are not connected to major skeletal structural parts of the rocket and
would not be considered. A tetrahedral mesh is used to obtain the most accurate result and the mesh
size used is
Table 44: Key Results - Buckling of Payload stringer
Applied Load kN Buckle mode (mm) Eigenvalue (MPa) Buckle Safety Factor
1.667 1 2.7683 2.7683
1.667 2 5.6557 5.6557
1.667 3 11.001 11.001
Figures
Table 45: Mesh convergence study of buckling simulation of payload stringer

2797 3.1455
5326 3.1109
11090 2.9423
17735 2.8320
81380 2.7697
103985 2.7683
168
Figure 113: Buckling Mode 1 of payload stringer
169
Table 46: Key Results - Shock loading of payload stringer
1.5 0.2059 65.07 3.69
Figures
Table 47: Mesh Convergence study of shock loading of payload stringer

11090 56.17
17735 61.63
81380 65.59
103985 64.78
136468 65.07
Figure 116: Stress from shock loading of payload stringer
Figure 117: Displacement from shock loading of payload stringer
From table Table 45 and Table 47, it can be seen that both the shock load simulation and the buckling
simulation reaches convergence. For the buckling simulation, the eigenvalue of mode 1 decreases as
mesh size decreases then converges at around 81380 elements. For the shock simulation, the max stress
increases as mesh size decreases then converges at 103985 elements. Thus both of the simulations are
valid.
From Table 44 and Figure 113, Figure 114, Figure 115 it can be seen that the buckling safety factor for
the stringers in all three modes is larger than the 1.5 needed. Again, due to the eigenvalue increasing
as the buckling mode increases, it can be assumed that the eigenvalues for further modes would be
170
higher and would be of no concern. From Table 46, Figure 116 and Figure 117, it can be seen that the
stress safety factor is much larger than 1.5, and the max displacement is under an acceptable range.
Thus the stinger’s structural design is valid and can be used in the rocket.
171
Q.3 Payload Simulations
Deployer Box
Ishan Dubey and Arfred Garcia
Overview
One of the objectives of the launch of the rocket is to carry the necessary payload to the location and
deploy it accurately. Thus a payload box that can withstand the max acceleration forces of the rocket
in both vertical and horizontal directional stress mode and buckling mode is extremely important.
This simulation aims to prove the structural competence of the payload box, especially its truss
structure.
Software
The software used for this simulation is the Abaqus 2024 version, and the license used is the teaching
license, which limits the number of nodes to 250000 nodes.
Set-Up
The payload box is constructed entirely out of aluminium, thus it has a homogeneous section
assignment. The Young’s Modulus is 70GPa and the Poisson’s ratio would be 0.3. Although in reality,
the box would be an assembly, since all of the parts would be assembled together securely through
fasteners and for ease of simulation, the assembly is simulated as a part instead. The part is simulated
with two forces, with the vertical force being simulated in normal stress and buckling mode, and the
horizontal stress being simulated in only stress mode. The vertical force applied is 255N, which is 4kg
times the max acceleration of 6.5g. For the vertical force, it is applied through the rod mound on both
sides of the bottom of the deployer box. While for the horizontal force, it is applied through the
mound on the side of the box, with the load of 60N, representing 1.5g multiplied by a mass of 4kg.
The top four holes where the hinge connects to the rest of the rocket have encastre constraints and
the side is also constrained to simulate the lock arms. The mesh used is tetrahedral mesh and the
mesh size is 5 for the most accurate results.
Simulations
Simulation 1 - Vertical acceleration load
Key Results
Table 48: Key Results - vertical acceleration load
Applied Load N Maximum Stress (MPa) Stress Safety Factor

270 118.9 2.02
172
Figure 118: Maximum stress of vertical load
Simulation 2 - Horizontal acceleration load
Key Results
Table 49: Key Results - Horizontal acceleration load
Applied Load N Maximum Stress (MPa) Stress Safety Factor

60 138.1 1.74
Figure 119: Maximum stress of horizontal load
173
Simulation 3 - Vertical buckling load
Key Results
Table 50: Key Results - Vertical buckling load
Applied Load N Buckling mode Eigenvalue Buckle Safety Factor

270 1 43.745 43.745
270 2 44.342 44.342
270 3 53.261 53.261
Figure 120: Buckle mode 1
174
From all of the figures and tables above, it can be seen that all of the factor for all of the load cases on
the payload deployer box is at least larger than 2, which is way above the 1.5 originally needed.
Moreover, it can be seen from Table 50 that the Eigenvalues for buckle modes 1, 2 and 3 are very
significant. From the results of the simulations, it can be confidently said that the part would be able
to withstand the max acceleration forces applied to it during takeoff and can be used as a part in
Nimbus 24.
175
Q.4 Guided Recovery
Jayden Jackson
Overview
Figure 123: GRIFFIN Module
Software
The software used for this simulation is the Abaqus 2024 version, and the license used is the teaching
license, which limits the number of nodes to 250000 nodes.
Set-Up
A static, vertical load of 1000 N (value obtained from MATLAB SIMULINK Model also used for main
parachute and drogue simulations) is split into two 500 N loads applied at each of the M5 eyebolts.
Both the bottom plate and bulkhead are made from the same material (Aluminium 6012) so a
homogeneous section is used. Material properties are mechanical/elastic, with Young’s Modulus of 70
GPa and Poisson ratio of 0.33. Tie constraints between the inner surfaces of the top plate and
bulkhead are used to model threads, whilst encastre constraints with boundary condition of clamped
are applied to the holes on the sides of the bulkhead to model the bulkhead being fixed inside the 3U
payload. A tetrahedral mesh of size 2.5 with quadratic mesh elements are used for higher accuracy
than standard linear elements at the cost of slightly longer computational time.
Simulations
Simulation 1 - Static Shock Load
Key Results
Table 51: Key Results - Static Shock Load
Applied Load N Displacement (mm) Maximum Stress (MPa) Stress Safety Factor
1000 0.0055 73.58 4.26
176
Figure 124: Maximum stress of vertical load
Figure 125: Maximum displacement of vertical load
Based on the results above, it can be concluded that the guided recovery module is safe to fly on
Nimbus 2024 based on the safety factor of 4.26.
177
Q.5 Recovery Bulkheads
Jayden Ting
Overview
The recovery bulkheads serve as the connection between the rocket airframe and the shock cords, as a
consequence they connect the sections of the rocket around a separation joint together. The major
load that the components have to undergo is the considerable tensile load exerted from the shock
cord creating by the sudden acceleration created during parachute deployment. The simulations
verify the structural design of the components. The recovery bulkheads are manufactured from
Aluminium 6082 T6, and thus have a Young’s modulus of 70 GPa and a Poisson ratio of 0.3. The
loading on the components is transferred via an M8 washer that has been modelled using a 3 mm
thick steel ring with outer diameter of 17 mm and inner diameter of 8 mm. A rough contact was
modelled between the components. The load applied was 4500 N. The boundary conditions were 3
pinned holes for the upper recovery bulkhead and circumferentially fixed for the lower recovery
bulkhead to model the bonded contact.
Software
The software used for all of the simulations is Abaqus 2022 edition, the license used is the teaching
license.
Lower Recovery Bulkhead

Set Up
The lower recovery bulkhead is connected via an M8 eyebolt to the D3, M1 and R1 shock cords.
Through these a shock load of 4500 N acts through the bulkhead. The bulkhead is mounted to the
recovery tube using structural epoxy. The load has been applied to the steel washer as highlighted
above.
Figure 126: The lower recovery bulkhead
Results
The static stress simulation results are shown below.
178
Figure 127: The stress distributions in MPa on the lower recovery bulkhead
Figure 128: The displacement distributions in mm on the lower recovery bulkhead
Based on the above figures, the maximum stress on the structure occurs in the 4 arms of the bulkhead
and it is of the magnitude of 125.9 MPa, Thus presenting a 1.98 safety factor, which is highly suitable
for the rocket. Furthermore, the stress concentrations are well distributed thus indicating that in the
event of an imperfect load the bulkhead would not fail. Further, it is seen in the following figure that
displacements are negligible.
179
R Payload Addendum
The aim of Hermes’s first launch is to test all internal systems including:
– LoRa communications system with ground station.
– 9DOF sensors including Accelerometer, Magnetometer, and Gyroscope.
– Barometer, Temperature sensor, and Altitude sensor.
– Onboard 2-megapixel camera with 56.8° Field of View (FOV).
– Integrated data storage (64GB).
– Ground station and associated software for communication.
Figure 129: Hermes in detail
180
4 3 2 1
F F
9.00
E E
46.00
D D
73.50
11.25
C C
B 73.50 B
UNLESS OTHERWISE SPECIFIED: FINISH: DEBURR AND
DO NOT SCALE DRAWING REVISION
DIMENSIONS ARE IN MILLIMETERS BREAK SHARP
SURFACE FINISH: EDGES
TOLERANCES:
LINEAR:
ANGULAR:
NAME SIGNATURE DATE TITLE:
DRAWN
CHK'D
APPV'D
A MFG A
MATERIAL: DWG NO.
V5 Assembly
Q.A
A4
WEIGHT: SCALE:1:2 SHEET 1 OF 1
4 3 2 1
Figure 130: Hermes engineering drawing
181
Figure 131: Fold I
182
Figure 132: Fold II

Figure 133: Fold III
183
Figure 134: Fold IV

Figure 135: Fold V
Figure 136: Fold VI
Figure 137: Folding order to minimise risk of entanglement of lines
184
Table 52: Parafoil Further Details
Parameter Value
Bridle Lengths 21.26 meters
Adjusted Bridle Length 24.44 meters
Centre Bridled 18.0%
Tip Bridled 18.0%
Wingspan 1.421 meters
Maximum Chord 0.549 meters
Aspect Ratio 3.312
Flat Kite Area 0.610 square meters
Adjusted Kite Area 0.448 square meters
Active Lift Area 0.562 square meters
Leading Edge Length 1.459 meters
Trailing Edge Length 1.497 meters
Coning AoA (Centre) 3.000 degrees
Coning AoA (Tip) 3.000 degrees
Note No profile at wingtips
185
Figure 138: Actuation Sequence
Pre - Drogue
Update altitude
NO
Is current
altitude at
drogue?
YES
Post - Drogue
Update z axis Start recording time

velocity since drogue
NO
Is z axis velocity less T + 10 seconds since
than 10 m/s ? drogue?
NO YES
YES
Rotate clockwise (100% Rotate clockwise (100%
duty cycle) duty cycle)
Continue for 2 seconds Continue for 2 seconds
0% duty cycle (i.e idle) 0% duty cycle (i.e idle)
Rotate anti-clockwise Rotate anti-clockwise

(100% duty cycle) (100% duty cycle)
Rotate clockwise (100% Rotate clockwise (100%

duty cycle) duty cycle)
Landing
Figure 139: Guided Recovery Build
Extra considerations on guided recovery - An unpredictable element of this experiment is the

influence of wind gusts on the behaviour of the parafoil and, by extension, the trajectory of the
payload. Since the dynamic model developed by Payload only considers a fixed wind velocity vector
of 10 m/s, the system is currently unable to react to any influence of wind gusts. In future launches,
the goal is to develop a guided recovery system that can self-correct a plotted trajectory when
encountering wind gusts by banking to counteract the acceleration induced by the gust. Flight data
such as GPS, velocity, and acceleration vectors captured in this launch will be invaluable in developing
such a recovery system.
187

Report Imperia Team

Uploaded by

Copyright:

Available Formats

Report Imperia Team

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Report Imperia Team

Uploaded by

Copyright:

Available Formats

Team 12 Technical Report to the 2024 EuRoC

Academic Supervisor: Dr. Aaron Knoll

Imperial College London

Revision Date Description

2 Project Goals and Mission Objectives 1

8 Active Control Systems - Canards 37

11 Mission Concept of Operations 48

12 Conclusions and Outlook 50

B Additional Flight Simulations 54

D Flight System Overview 103

E Ground Support System Overview 105

F Material Tables 106

G Hazard Analysis 109

H Risk Assessment 110

J Engineering Drawings 132

K Data Task Request Handler Example JSON Config 149

L Data Task Request Handler UI 150

M Geddan Power Source Logic 151

N SRAD Tank Failure Calculations 153

O Modelling the dynamics of post parachute deployment behaviour 155

P Reefing characteristics equations, calculation and procedure 156

Q Airframe Finite Elements Simulation Reports 157

R Payload Addendum 180

2 Project Goals and Mission Objectives

2.1 Mission Objectives

Figure 1: Subsystem breakdown of Nimbus 24.

Table 2: Nimbus 24 mass breakdown

Table 1: Nimbus 24 length breakdown

Peak Thrust [N] 4,000

4.1 Plumbing and Tanks layout

Figure 3: P&ID for the in-flight feed system.

4.2 Propellant & Pressurant Tanks

Table 4: Design specification of the two SRAD propellant tanks

Propellant Oxidiser Fuel

4.3 Electronic Tank Pressure Regulator

4.3.1 Controller Design

4.3.2 Regulator Safety and Abort Logic

4.3.3 Integrated Testing

4.4 Bi-propellant Engine (THANOS-R)

Table 5: Design specification of THANOS-R Engine at nominal operating point

Mass (kg) 3.0

4.4.1 Injector Head

4.4.2 Thrust Chamber Geometry

4.4.3 Thrust Chamber Cooling

4.4.4 Ignition Method

4.5 Valve Actuation

4.6 Ground Support Systems

4.6.1 Remote pressurised fluid filling system

Figure 13: GSS filling system Plate

4.6.2 Manual Fuel Loading System

4.7 Rocket Filling Procedure

4.7.1 Manual Fuel Filling

4.7.2 Remote Nitrogen Filling

4.7.3 Remote Oxidiser Filling

5.2 Skeletal Airframe

Table 6: A table showing the stringer lengths and safety factors

(a) Truss Interface Artefact (b) Coupler Interface Artefact

Figure 19: Stringer Artefacts

5.2.2 Engine Truss

Figure 20: Engine Truss Generative design study