Armament Research
Armament Research
Armament Research
AD NUMBER
AD159528
CLASSIFICATION CHANGES
TO:
FROM:
unclassified
restricted
LIMITATION CHANGES
TO:
Approved for public release, distribution
unlimited
FROM:
Controlling Organization: British Embassy,
3100 Massachusetts Avenue, NW, Washington,
DC 20008.
AUTHORITY
DSTL, DEFE 15/997, 31 Jul 2008; DSTL, DEFE
15/997, 31 Jul 2008
UNCLASSIFIED
AD NUMBER
AD159528
CLASSIFICATION CHANGES
TO
restricted
FROM
confidential
AUTHORITY
RARDE Reclassification list no. 42 dtd 4
Aug 1970
UNCLASSIFIED
AD NUMBER
AD159528
CLASSIFICATION CHANGES
TO
confidential
FROM
secret
AUTHORITY
RARDE Reclassification list no. 36 dtd 19
Jul 1967
0)
CD
1~
Co
0
0
C\J
4 KZ
9.A
g,
QANII_
7
-T
UA
. ...
...
...
-4
10
!
Ai
Tj
4
K. hlustritc
Ministry of Supply
.2
BA1AJ* RESEARCH IJJ) DEVELOPNM
ESTAMSIZIaT
2Pj
!3
A.R.D.B. 111FORT (B)6/58
w. ShepArd (B5)
S. A. Beac.
(B5)
Summary
of the component errors to be used in
This report gives a list
chance of hit calculations, and discusses the magnitude of these errors for
JU.
'tes
U. 3. ( nfidential
CONTENTS
Para
INTRGDUCTION
Variable biases
15
Comments
18
MAGNITUDE OF COW02MT
22
Permanent biases
23
Variable biases
28
Random errors
40
Summary
46
47
.49
53
APPENDICES
App
E
Estimation
B.r >
rIt:
,;
i
__I
__
".
IMPORTAN4T
This document should be returned to the Reports Officer, Armament Research and Development
Establishment, Fort Halstead, Sevenoaks, Kent, when no longer required.
INITIAL DISTIBUTION
Internal
No.
Director
Deputy Director
SA/Director
4-5
8
PS/B
PS/ER
Ps/NAx
PS/P
Si
r34
10
11
1APC (p)
NPC (L)
12
S/1
13
S/L1
15
16
17
18
S/L2
s/pi
S/F2
S/r5
s/S1
19
/S/8
2D
s/xi
21-22
23-24
25
B Division Library
RO & Ed..
United Kinjzdm
Ministry of Sup ly
26
CM
27
28
29-30
31
DK
Chief Scientist
DG of A
DGSR(M)
32
D of A (W)
33-34
DGFV
35
CI-YL
36
37-41
42-W,
Ministry of Defence,
45-46
DRPS
MINI
jr;
esrila
X
SECF& DISCM
War Office
47
D0]9S
48
SA/A0
49
50
CS/AORG
DMD
51
DQM
52
53-54
55-56
57-58
59-CO
DRAC
61
62-63
DRA
D Inf.
61+-65
TIL
iMC of S
School of Infantry (Att.SWW(l))
for retention
77
78-95
96
97-104
70
71-76
3035M60
307-ID9
- ONR (London)
Gemary
-HQ
Stock
M10 -153
SECEr DISORE
SECRET-DISCREFT
INTRODUCTION
As a result of disoussions held during the First and Second
1.
Tripartite Technical Conferences on Tank Armament, held in 1956 and 1957,
Panel C (Assesament)was able to reach a considerable measure of agreement
on the subject of chance of hit with tank and anti-tank guns.
Details of the discussions are given in Annexure 3 of "Report
2.
of Tripartite Conference on Medium Tank Armament 1956" (1)x. In particular,
the following agreement between 1K, US and Canada is tabled:
"IT WAS AGREED that the component errors which enter into
The list
chance of hit calculations should be listed.
should be split into three parts corresponding to:(a) Permanent biases (or fixed biases)
These are errors which affect the position of the
Variable biases
These are errors which affect the position of the
3*
present paper gives a list that has received general UK approval, and, on the
basis of this, the chances of hit to be expected with the 20 pr gun in
It is hoped that this evaluation will prove a
Centurion- are evaluated.
ready guide to the method in which chances of hit under quasi-battle conditions
are calculated, and will also provide a general reference to the magnitude of
certain of the errors involved.
LIST OF ERRORS AD THEIR DEFINITIONS
Component errors in chance of hit calculations
4.
"
1
SRCET-DISRPET-
SEaRET-DISCREET
Table 1
Classification of component errors
lass
Errors in line
Errors in elevation
11ean Jump
Sight I off-set
Errors in range-graduation
Mean throw-off,
Drift
Sight off-set
(sight or
Bias in rangefinding
Shooting-in error
Aligment error
Mechanical errors & biases in sighting system and linkage
Shooting-in error
Aligrnment error
Throw-off, occasion-to-occasion
S Juurpp occasion-to-occasion
Droop, occasion-to-occasion
Ballistic coefficient, effect of
errors
occasion
Gunwear, effect on W
Charge temp, occasion-to-occasion
Cross-wing
Head-w nd
Range-finding error
Error due to discrete range intervals
resulting
from
attempted
correction of
Ithese
Jump, rd-to-rd
Droop, rd-to-rd
Yaw, rd-to-rd
o o
Throw-off, rd-to-rd!
Lateral bend, rd-to-rd
Yaw, rd-to-rd
5.
Gustiness in head-wind
Gustiness in cross-wind
2
SHCRET-DISCRIMT
-DISCRE7
7.
Variable biases
These are errors that affect the position of the WPL, but vary in
8.
magnitude from engagement-to-ngggemeAt (occasion-to-occasion), or from tankto-tank or anti-tank gun to anti-tank gun; they are constant during any
particular engagement.
The most important component error of the variable biases is the
shooting-in error, which results from the combination of a zeroing error with
There
These errors arise in the following manner.
an alignment error.
exists, for most guns, a drill whereby the axes of the sight and gun can be
A typical example of
aligned using a purely mechanical or optical method.
on to a distinct
was
bore-sighted
gun
the
in
which
drill would be one
to the gun
relative
adjusted
then
was
the
sight
and
target
distant
point on a
9.
*this
The axis of the gun with which the sight has been aligned (usually
either the apparent axis or the muzzle axis) and the mean line of departure
of the shot may not coincide; in order therefore that the sight can be
aligned with the mean line of departure it
ing to the target rang.e, is then applied to the gun using the range-drum and a
*of
rounds is fired to determine the MPI; the total of these errors is known as
The combination of this error with the alignment error,
the zeroing error
which gives the total error in the zeroing correction, is known as the
shootina-in error.
The gun and sight are normally aligned each day, or on each
Ii.
occasion that the weapon is moved to a new site; as has been mentioned
previous3y, however, the zeroing correction which is applied subsequent to
each aligment procedure is only determined infrequently. 3 Thus in a particular
Values of
engagement both aligment and shooting-in errors will be present.
*
3SWIMT-DISCRT
SEaRET-DISCREET
12.
The remaining variable biases are relatively simpler to describe.
Occasion-to-occasion variations in jump and droop affect the mean trajectory
in elevation, and variations in throw-off and bend affect it in line;
allowance must be made for the differences between the values of these parameters on the occasion the gun was zeroed, and their values during the
engagement of an enemy target.
Similarly, an error in elevation arises from
variations in ballistic coefficient due to differences in air density, and
from the effects on muzzle velocity of changes, from the conditions of zeroing, of xun wear and charge t=rature;
in line, occasion-to-occasion
variation in trunnion tilt
(and/or errors resulting from an attempted
correction) will cause biases during an engagement, again because of the
differences between the conditions during the engagement and those under
which zeroing was carried out.
Head-wind and cross-wind also cause
differences in the mean trajectory on different occasions.
13.
In any particular engagement an error arises due to ranefinding;
that is to say, the range that is set initially on the range-drum or sight
graticule is usually inaccurate (even if a rangefinder is used).
A small
error also arises because the range drum or sight graticule is often engraved
only in discrete range intervals and ranges thus have to be taken to the
nearest increment.
14.
It must be noted that many of the errors mentioned above are often
Further, it
grouped under the general term "non-retention of zeroing!'.
sheuld be mentioned that it is assumed in this paper that a cleaner round is
always fired some time before an engagement starts since it is not considered
feasible to attempt to take account of the erratic behaviour that is often
observed with such a round.
Random errors or dispersions
15.
These are the errors that affect the dispersion of individual rounds
about the MPI.
16.
The most important error in this category is the ballistic error.
It includes, in addition to variations due to windage and to difference in
projectile weights within the permitted design tolerances, inaccuracies due
to round-to-round variations of
, droop, yaw, ballistic coefficient and
muzzle velocity in elevation, and due to throw-off, lateral bend, and yaw in
line.
In most British tank and anti-tank gun ammunition, propellant lots
are mixed; dispersion in elevation is increased by attendant lot-to-lot
variations due in part to errors in charge adjustment.
Similarly, round-toround variation in charge temperature increases dispersion in elevation.
17.
4
SECRE,T-DISCREE T
SERE-ISCREMT
MAGNITUDE OF O0PONEN
*Gun
20 prAAPMMk 4
Projectile
Muzzle velocity
48O f/se
26)0 ft/sec
Ballistic coefficient
3.695'1
.. 9,
7.636'
8.5724
(13.278')
28.930'
.25*
(14.366')
500 yards
Angle of
Projection for:- 1000 yards
.
2.263i
343
20.740) (69.995'
1 6 -411-9A4756
.1 56
500 yards
1.12%
( 3.798?)
1500
200 yars
yards
Angle of
descent for:-
"A000yards
1500 yards
2000 yards
2.398.075')
3792 V12-793'
5.3
(18.303')
5
7CT-DSCR10T
9.759f (32.937'
18. n
29.7
61.147)
(98.471')
SEMW-DISCREET
Permanent biases
23.
Mean jump and throw-off. These biases are equal to the differences
between the mean jump and throw-off of the ammunition used for zeroing and
those for the ammunition fired in anger.
No measurements have yet been made
of the jump and throw-off of 20 pr APDS Nk 4; and in fact little
is known
about the causes of these errors (4)(5).
It seems reasonable however for
present purposes to assume that the jump and throw-off for APDS Mk 4 are the
same as those that have been measured for APDS Mk 1; similarly it has been
assumed that a HESH shell in 20 pr would have the same jump and throw-off as
the 20 pr HE mk 1 shell (6) which also has a low velocity.
Table 3
Jump an
thow-off for 20 pr
Projectile
Jump
Throw-off
AMS Mk 4
-0.822
HEM
-0.874A (-2.950')
Diff erence
(-2.775')
-. l00$W
(T.0338')
Ro.06.)f (RO.212')
o.0546 (0.175')
0.16lo
(0.550')
It is assumed in this report that the gun is zeroed by firing ADS Mk 4 (see
Appx A).
The biases due to mean jump and throw-off will therefore have
values of zero when A=DS is fired in anger, and values of -0.0546 (-0.175')
and RO.16.6 (RO.550') respectively for HESH.
24.
Sight off-set. The sight in Centurion is 22.5 inches higher, and 26.5
inches to the right of the axis of the gun at the trunnions.
The bias due
to sight-offset. has been calculated accordingly and is given in the following
table; zeroing has been assuped to be carried out at 1000 yards range, and the
effect of the normal practice of adjusting the telescope 1 minute right to
make some allowance for drift (see 3ara 27 below) has been included.
Table .
Bias due to sight off-set
Range
500 yards
00 yards
1500 yards
2000 yards
Bias in elevation
-0.6 V(-2.1
Bias in line
.0.0
+0.212A(+0,7161
+0.318(+l.0741
IO .29(
(L3.500
L1.000Fi
I-0.04V(O. 157
R0.079 ( O.265
25..
Error in ranie-graduation This error is assumed to be negligibly small
for Centurion; the range-drum is engraved using the best possible data from
R and A firings.
Divergencies in an actual gun from range table values, due
to factors such as gun wearp charge temperature differences and so on, are
considered below.
6
SECRET-DISCREET
determination.
27.
Drift.
in the vertical plane of fire; the distance by which it falls away from the
0 tan X mins, where
plane of departure is called the drift and has a value G
20 pr AMS Mk 4
Range
500 yards
1000 yards
1500 yards
2O yards
10 .043A(T. .138)
0
R0.0+,7#6(RO .- 43'
R0.08*(R0.296')
RO.88(R.2.99111
R0.5T?ARl.9)+7'
Variable biases
28.
Shooting-in error
component errors based on the drill normally adopted when shooting-in the 20
29.
also;
0.516(l.741')sd
0.45,(1.536t)sd
0.33*(1.137')sd
0.34
*7
SECERT-DI:REET
(1.164')sd
SECRET-DISCREET
30.
Mechanical errors and biases in sighting system and linkage
errors arise from two main causes:-
Theso
500
1000
1500
10O
yards
yards
yards
yards
20 pr APDS Mk 4
0.04,%(0.1511)
0.044(0.1501
0.04% O.l5l'
o.0:46(0.156'
20 prHESH
sd
sd
sd
sd
0.04%(0.153, )sd
0.05.% O.l81) sd
0.074#60.250 ' sd
0.10)60.346' )sd
31.
Occasion-to-occasion variation in J,ump and throw-off There is
evidence that there is no variation in the mean jump or throw-off .of the 2D pr
gun firing AEDS Mk 1 and HE Mk 1 shell from occasion-to-occasion or seriesto-series (6)(19).
It is assumed here therefore that these errors are
also negligible for 20 pr AE)S Mk 4 and HESH ammunition.
32.
Occasion-to-occasion variation in droop and lateral bend
Major
alterations to droop and lateral bend may occur due to changes in atmospheric
conditions, particularly in a hot gun.
For instance, a shower of rain, or
freshening of the wind, have been found to alter droop and lateral bend. (8)
(2)(21) Values for these errors, which include variations between barrels,
are therefore naturally difficult to obtain. A perusal of the references
noted above however suggests that values of 0.2966 (1.O00')sd for occasion-tooccasion variations in droop, and 0.074A (0.2501)sd for lazteral bend are
probably reasonable for present purposes.
33.
Effect of variation in air density on ballistic coefficient Changes
in air density due to alteration in air pressure, temperature and humidity cause
roughly proportional changes in ballistic coefficient.
Large variations can
naturally occur from season to season and from theatre to theatre of operations.
A range scale designed for 60F will therefore have considerable errors at
long ranges in cold or hot climates with lower velocity ammunition(22). No
data have been found from which an estimate can be made of the change in air
density that may occur from occasion-to-occasion.
For the purposesof the
present paper, however, it has been assumed that changes greater than 20.0 in
air density from occasion-to-occasion (takel equivalent to an sd of Yl about
the mean) are rare.
On this basis the values determined for the error under
consideration are given below:Table 7
Errors due to variation in ballistic coefficient
Range
500 yards
1 oo yards
1500 yards
2000 yards
o___oo
ooo_
___ i o_o_oo___
0.Olyi.43)d
20 pr APDS Mk 4
0.002A6(0.0061 )acl
,sd
o.0 05k(.016
OOl (O037'sd
0.023,4(0.071 sd
8
SECRET-DISCREET
20 pr HESH
. 05 A(O .l
)sd
0.1274(0:429 1 sd
0 271(0.9121 sd
34.
Thus the muzzle velocity of a gum when it is fired in anger will be less
than the velocity at which it was p=eviously zeroed and there will be a
A reasonable method of estimating the
corresponding drop in the MPI.
value of this bias is given in Appendix B; it is shown that the most likely
from a new gun to a fully worn gun. The mout likely biases in KPI
calculated on the above basis are given in Table 8, Gun wear also affects
the accuracy of rounds fired, ballistic dispersion increasing as the gun
wears. Little is knbwn, however, of the comparative size of this error
component for the anmunitions under consideration here, and it has been
thought sufficient in this paper to consider that the values for round-toround dispersion given in paragraph 4.0 below are for a partly worn gun.
Table
Range
500 yard
1500 yards
2oo yards
2 pr APDSMk 4
20 prHEM
-O013A(-0 .044
-o.0 8(-0.278,1)
-0-040A -0-135'j
5%(-0.181'
-0:3*1-1-038t)
-0.4.6(-1.568')
It has not
35.
Ocasion-to-ocasion variation in charge temperature.
been found possible to trace any data giving directly the magnitude of this
error.
An indication of the size of possible changes in charge temperature
may, however, be obtained from certain trials carried out by the Ordnance
Board with 25 pr ammunition (23) and from conclusions reached by a cmmittee
set up to investigate the accuracy of predicted fire from FBA. guns (4).
These suggest that, even if ammunition is not well insulated from atmospheric
changes in temperature or from the sun's radiation, the diurnal variation in
charge temperature is unlikely to exceed 20oF at the vexy most; in a tank it
Little is known about the
is likely that this variation will be much less.
the season-to-season variation in charge temperature. However, if it is
assumed that the tank's gu is zerobd in the sae theatre of operations as
that in which it is used in action, it would seem reasonable to take an sd
of 5P, say, as representing the variation in charge temperature likely to
be encountered from occasion-to-oocasion in tank guns. The angular effects
of the corresponding changes in muzzle velocity are shown below; a temperature
correction of 46ftsec for 10Op has been taken for AIMS and 4ftsec for 10'F
for HESH.
SIETDIjRE
SEE-DISCREE3!
Table 9
Errors due to occasion-to-occasion variation in charge temperature
Range
500 yards
1000 yards
1500 yards
2000 yards
MkA.
20 pr AS
0
o.o16(0.053, sd
C.02%(0.090' ad
0.036/i(o.12o, sad
2D pr HESH
s
0.o0#(0.035' sd
0.01(0.059' sad
0.026)(0.0891 )sd
36.
Effects of trunnion tilt. It is obvious that a tank in battle will
rarely take up a firing position in which it is perfect]i level; in general
therefore a trunnion tilt will be present.
The Centurion
is not equipped
with a trunnion tilt corrector and the full effect of the tilt must be allowed
for.
Investigations into this source of error (25X26)(27)(28) both in the
UK and the US suggest that a value of 50 sd of trunnion tilt is reasonable
for typical firing positions in action.
It must be admitted that the data
on which this value is based are not altogether satisfactory, and it is
suggested that further work might well be undertaken to obtain reliable
values for this important error.
Values for the ammunitions under
consideration are given in the following table:-
Table i0
Errors due to trunnion tilt
Range
500 yards
1000 yards
20 pr APDS Mk 1,
00O944A(0.31')d
0.194(0.653'
20 pr HESH
0.33%)t(1-136' ad
0.73*(2.475'
1500 yards
0.29e(I.005')sd
1.20
2000 yards
0:434 1(l382')1sd
1. 77%i(5.98911sd
(4.071' ad
When zeroing is being carried out more care can be taken in selecting a site
and a value for trunnion tilt 1.00 ad for zeroing in the field is talmn in
Appendix A.
37.
Head-wind and cross-wind.
Recent work (29) has shown that the
standard-deviations of head-wind and cross-wind velocity distributions in
NW Europe each have values of 4.9ft/sec.
It is not possible using the
fire-control system in Centurion readily to allow for wind, and thus the full
effects of the wind mentioned above must be taken into account; values are
given below:-
SECREET-DISCREET
Effets of hand
Wind
Head
croswind
Range
20ip'APDS N c
500 yards
1000 yards
All
1500 yards
201r HEM
0 .i 0 0003 f sd
0.004;6(0.+04' sd
negligible
0-0106 0.037' sd
ZM
0yards
0.2%0084Is
500 yards
1000
yards
1500 Yards
1ross
0 .0
0 .341ad
0
9,
0.0311 0.1121sd
0 .250 0:4049
0399 1 34
. 8
2000 yards
0.04-t0.160
o.5
Id
sd
ad
l1. 93 7 , s
38.
RMSefinding error. Although there is some evidence that the errors
in ranges,estimated either
visually or using a coincidence-type ragefinder
are not normally distributed (13) (30), it is generally accepted that the
divergence is slight and that a rangefinding error can be expressed with
meaning in terms of a standard-deviation in the usual way (26).
The following
3," of range, sd
Rangefinder estimation
=
(metre-base coinoidence-
type Btp)
6f yards.
Values for these errors, expressed as angles at the gun, for the anmunitions
under consideration are given in the following table:Table 12
Rangefindjag errors
Method of
determining
Ii
Range
20 prBEESRM1 4
20 I"ESH
range
Visual
50yards
1D00 yards
0.336A(l.135')sad
~0.7
(2.394' d
range
15DO yards
1.
estimation
2000 yards
1.54(5
Metre-base
500 yards
1000 yards
0.013AO0036' )sd
0.0)
o,62
o.%
2000 yards
0 T, i0. 682' s d
3.724.'
)sd
1'
rg n0 rDICdIMT
Ti
1. 12536(14o2281 )Sd
2.87 6 9.695' sd
5.0 A(16.918' sd
7.82*(26.417'sd
0:041037
! o.)sd.,647' sd
1.oq6fi3.699') sd
SECR-DISCP=ET
It should be noted that there is in fact no rangefinder on the Centurion tank;
figures for errors in range determination using a rangefinder are however
included so that estimates can be made later of the effects of including a
rangefinder in the fire-control system.
39.
Error due to discrete range intervals. A tank comander gives, in his
fire order for opening fire, a range rounded off to the nearest 100 yards. (36).
An error of as much as t 50 yards may therefore be present from this drill3e.
The frequency distribution between these limits can be taken to be rectangular
anZ represented as having a standard deviation of 100/[12 = 28.9 yards.
The
resulting magnitude of the error under consideration for the ammunitions
concerned is shown in the table below:Table 13
Error due to discrete range intervals
Range
20'pr AIDS
100 yards
0.06rp0.2i31'
k 4__20
100.ars06%,0:219'
1500 yards
0.073A(0.239'
2000 Yards
prH
0.277d0.93'
sd740934~l'
ad
sd
0.32
1.087' ad
s07,%10.253,
d
0.377(1.2721 -sd
Random errors
40.
Ballistic error. Recent accuracy firing trials with 20 pr AFDS
Nk 4 anmunition carried out by ADE, indicate that the ballistic error (which
includes the intrinsic dispersion of gun and ammunition, the effects of
round-to-round dispersion in jump, throw-off, droop, lateral bend, yaw,
ballistic coefficient and muzzle velocity) has a value of 0.276A (0.93') sd
in elevation and 0.2)+$p(0.83')sd in line at 1000 yards.
These values have
been taken to apply at all ranges, however, in this paper.
It has also
been assumed that the ballistic error for 20 pr HESH ammunition is 0.296A
(1.000')sd in both elevation and line at all ranges.
41.
Lot-to-lot variation in propellant. AFDS propellant for the 20 pr
QF tank gun is, at present, proved to a pressure rather than, as is usual,
to a velocity.
Study of recent firings by OSR, as yet unpublished, show
that, for 29 lots tested, the measured muzzle velocity had a standard
deviation of 14.8ft/sec.
This value has therefore been taken as representative of the error incurred by lot-to-lot variation in propellant for 20 pr
APDS Mk 4.
I-ERI propellants are proved, in the normal way, to a velocity,
and figures of HESH shell firings indicate that a sd of 3f/sec is reasonable
to take for the variation to be expected from lot-to-lot for the fictitious
2D pr HEH shell.
These errors are given, in Table 14 below, expressed as
angular deviations at the gun:Table 14
Errors due to lot-to-lot variation in propellant
Range
20 pr APDS Mk 4
20 pr HESH
i
!
500 yards
1000 yar,ds
1500 yards
2000 Xars
0.008 (0.027' sd
0.0J.
0.048 ad
0.024A(o.081, sd
0:03,
1 '1sd
0.008A(0.026' ad
0.02(o.o7,) ad
0.03.(0.129' ad
0.05340.l1811,d
2
This error is the difference between the range ordered and the
estimated range. The error between the estimated range and the true range is
discussed in para 38 above.
12
SECRTT-DISCREET
42.
The results of
trials
carried out on Cromwel and Sherman tanks suggest that under normal
operating conditions, the round-to-round variation in charge temperature is
unlikely to exceed 20F for rounds stowed in the turret and 50F if rounds in
It will be asumed that these
the co-driver's position are includd.
variations are unlikely to be exceeded in the Centurion
and a standard
deviation of -," will be taken as typical.
The temperature corrections
given in para 35 are used to determine values given in the table below:Table 15
Errors due to round-to-roun.
Range
500 yards
1000 yards
150yad
variation in
charge temperature
20 pr AMESMk 4
0.0010.0081 )d
0.00
0.00400W19
0.006 0.019' ad
00094 0.013191 d
sd
ad
007A005
rHS
0.009)sd
45.
*Gusts
Range
20 pr AIDS Mk 4
500 yards
Head
1000 yards
150 yards
0.00JA(0.003: )sd
Al
negligible
0.00% 0.017' ad
0.0 i
(0.046'
2QO yards
500 yards
Cross
1000 yards
1500 yards
20 pr HEMI
02'1sd
0.0 l6 .0 4 l1
ad
yr 0,0961 )d
I 0.028A
o.4-i 0.138')sa
10
0.l44%0.4951)sd
0.3
1.052' )sd
0.484 1.65o'a
13
SECRT-DICRYE
EXM-DISCEE
Summary
46.
Values have now been given for Centurion of all the component errors,
as listed in Table l, which enter into chance of hit calculations. For
convenience, the values are tabulated in Tables 18 to 21 inclusive, which are
at the end of this report.
CWTCES OF BIT WITH CENTURION UIRD
QUASI-BATTLE CODITIONS
47.
In this section, the results are given of calculations to detennine
the first round chance of hit under quasi-battle conditions for AFDS Mk 4 and
a fictitious HH shell fired from the 20 pr tank gun in Centurion, at ranges
of 500, 1000, 1500 and 2000 yards.
The values of the component errors given
in Tables 18 - 21 have been used, all of which enter into the calculation
when chances of hit under quasi-battle conditions are required.
It has beenassumed (as is usual) that the total errors in line and elevation are
sufficiently Gaussian in form to allow tables of the Normal error integral to
be used, and that errors in line and elevation are uncorrelated.
48.
Table 17 below gives values for chance of hit on a 7ft 6 ins square
target when visual range estimation is used.
For the sake of comparison,
values for chance of hit when a metre-base rangefinder is used are also given.
Table 17
First round chances of hit against a 7?61 square target with
Centurion under quasi-battle conditions
Method of
Probability of hit
determining
range
Range
20 pr APDS Mk 4
Visual
500 yards
97.7%
range
estimation
21000 yards
1500 yards
71.Q
36.2,
Metre-base
500 yards
:1000 yards
98.7j
rangefinder
2000 yards
1500 yards
I 2000 yards
245,
83.4;,
55.52.
20 ;r HESH
86.y
25.O.,
5
l. Y;
98.7T
65.3'
28.6;'o
14
SECRET-DISCREET
= 2.5-46
SE=-D ISP
15
SECHT-DISCRET
SE= I-DISOCM
Table 18
Summary of errors in Centurion 20 pr QF gun (mils at 14n)
AMlS
EMI.SATION
Error
Permanent
bias
Mean junp
Sight off-set
0
0
-0.637
Shooting-in err0.516sdl
Alignment
Meoh errors in sighting
System
Jump, oco-to-oco
Droop, ooo-to-oc
Co
Gun wear
Charge tamp, oco-to-occ
Rneidn:Visual
Sum of biases
Total ad: Visual
RF
!dispersion
0
-o.637
15 0OX
0
+0.212
Z)Q()X
0
+0.318
RF
Discrete range intervrals
eor
an.Lay,
)f00)X
Sum of biases
~Variable
bias
5 OOx
+0.212
+0.318_
0.516sd
0 .337sd, 0.337sd
0 .045sd' 0 .044sd
0.516sd: 0.516sdl
0.337sd 0.337sdi
0 .045sd 0 .046sd
0.296sdi 0.296ad
0 .296ad: 0.296sd
0.002sdi 0.005ad
o.olsd 0.021sd
-0.013
-0.025
0.00 8a& 0.016sd
0.336sd, 0.709sd
0.01Usd, 0.048sd
o.o65sd- o.o68sd
-0.040
0.0 27sd. .3s
1.10 3sdf 1.554sd
0.108sd. 0,202s
0 .O7lsd 0.075sd1
-0.013
-0.040
-0.-025
-0.055
0,766sd;'0*988sd
o.688i3d, 0.690sd
1.301sd1,70la
0.698sd 0.720sd
Ballistic error
Charge temp, rd-to-rd
rd-to-rd
Backlash in sight/gun---Gustiness-
0.276adi 0.276sd
0.0276sd 0.276s6
0.002sd, 0 .004scl
00119ad: 0.11946.
0 .007sd
0.119sd
*Totalsad
0. 30ld0.30lsd
---
TOTAL BIAS
TOTAL SD:
0 .Ol si
0.119d
-0.650
Visual
1-0,025
0:823sd! 1-0335d
0.751,sd' 0.753sd
RP
16
SECRET-flIWREET
+0.172
1+0.263
1 336sd1 1.728s&.
0.761sd: 0 .781sd,
SB=DISCR=~
T abe f.97
Sumxaxy of errors in Centurion 20 pr QF
AFD@-
LUNE
x
Error
5 00
;Mean throw-off
L1.04.6
,.041
O
Sum of biases
Variable
bias
Ll.087
0
LO.296
150
0
I1W.047
R0.079
O.043
RO.087
Lo=4
m.166
0
10.296
Shooting-in error
0 .455sd 0.455ad
O.455sd
0.455
0.345a
4.X58d
0.345ad
0.34581
0.074sd ;0.07"d
0.094%sd 0.lId
0.OlOsd 0.023,s
0 .074ad
0.298sd
0.033sad
0.07a
0.416sa
O.047sa
0.584d i 0.606s
0.64.9sd
0.8kgsd
0.24.6sd
0.348 d
0.246a
.
I Lateral
bend, oco-to-oo
I
Cross wind
Total ad
'Ballistic error
Random
ILay, rd-to-rd
terror or BaoklAsh in sight/gn
ion
1 000
Alignment
Throw-off, occ-to-ooo,
i Trunnion tilt
dispers-
un (Als at gun)
0.246ad
0.148d
-
Gustiness
_
0.012sd
_
0.028d
0.04lsd
0.058d
.....
0.287sd 1
TOTAL BIAS
Ll.087
0.288ad
10.296
Ii
TTAL S
_...
Total ad
.246Bd
0148sd
o.651sad
17
SC3T-ISI
ItT
U.6td
0.290as
10.004
0.293ad
1 R0.166
I
0 .73lsd
0.898s
SEE-DISCREET.
Table 20
Sumnary of errors in Centurion 20 pr QF gun (mils at gun)
HEH
NLWATION
x
Error
Permanent
bias
500
Mean jump
Sight off-set
Range graduation
1-0.052
-0.637
'-0.052
0
2 00 0 x
-0.052
+0.212
-0.052
'+0.318
....
Su of biases
-0.689
$hooting-in error
:-0.052
+ 0o.160
0.516sd i 0.516sd
Alignment
Mech errors in sighting
system
bias
1500
Bias in rangefinding
Variable
I000X
0.37sd337sd
0.045sd
Jump, occ-to-occ
37sd
0.055sd
-
Droop, ooc-to-occ
Co
,Gun wear
Charge temp, occ-to-ooc
Head wind
Rangefinding: Visual
RF
'Discrete range
0.296sd i 0.296sd
0.013sd 0.o0os
-0.082
intervals
Sum of biases
-0.183
0.337sd
O.103sd
0.296sd
0.127sd
-0.308
0.296sd
0.271sd
-0.465
O17sd
0 .0llsd
5.013s
0.510sd
0.026sd
0.025sd
7.827sd
1.096Sd
0. 241,ad
0.322sd
0.377sa,
00277d
J.183
0.728sd 0.766sd
-0.308
5.072sd
0.924ad
-0.465
7.871sd
1.377sd
0.296sd
1.448sd- 2.967sd
RF
0.516sd
0.337sd
0.074sd
0.005sd 00.010d
0.OOlsd 0.004sd
1.253sd 2.873sd
0.04.sd 0.192sd
-0.082
o.516sd
+0.266
Ballistic error
0.296sdi 0 .296sd
0.296sd
Rn
Propellant, lot-to-lot
0 .O08sd
0.038ad j 0.053sd
error or
0.003sa 0.oo6sd
0.009sd
0.U9sd
0.119sd
sight/gun
Gustiness
Total
ad
TOTAL BIAS
RF
0.l.9sd
-
0.034sd
0.ll9sd
-
0.O01sd 0.O05sd
0.Ol3sd
0.030sd
0.319sd 0.320sd
0.322sd
0.325sd
-0.771
0.023sd
-0.235
-0.li+8
-0.199
l.483sad 2.984sd
5.08-sd
7.878sd
0.978sd
1.415sd.
18
SECRET-DISCREET
Table 2
Sumary of errors in Centurion 20 p QF
-
IError
LINE
I0X
200OXp
RO.163
LO .047
Ro.163
R0079
iRO .104
RO.320
RO.577
Ro.886
I..779
PDO.187
lRo.693
R1.128
03
Shooting-in-error
0.3,5sd
Aligment
Throw-off, ooc-to-ooo
O.,sd
-
1 0.55a
:1,4554
0 345sc
0.455sd
0.345sd
0.074sd
0.337sd
0.074sd
0.733sd
0.074sd
o .074sd
1.206di 1.775sd
Cross wind
0.120sd
0.255sd
0.399sd
0.574sad
0.678sc
0.966ad
.1.395scl
1.952sd
0.296sd
0.148sd
0,296sd
0.1)+85
0.296sd
0.248a5
0.296ad
0.148sd
0.2145sd
0.312sd
0.i89s
0.703sd
Total sd
0.361ad
0.455sc
0.5950sd
0.777ad
TOTAL BIAS
.779
Total sd
Random
error or
15 0)OX
RO.163
LO .296
Sun of biases
Variable
bias
1O
Ro.163
LI.046
-ff
Penaanet
bias
I mean
Sight thoff-Set
SDrift
un (ils at gun)
Ballistio error
Lay, rd-to-rd
diapers-
Baoklash in sightgun
ion
Gustiness
....
0.768ad
TOTAL SD
19
RO .187
1.068sd
Ro .693
1.515sd
RJ.128
2.10lsd
SECE-DIS
Appendix
A
Dertion
Notation
OS
TI.
VAL
Shooting in error
A-lign
error
OH)
OBL
rfaL -
( J -
0"D
TLB -
'7'00 -
rC
O'HW -
OTT -
0QR
VGVW
oB
Ballistic error
OIL -
IT
O1
OI
OIGH -
(rG0 -
refers to the
Shooting-in error
2.
A desirable drlI
for zeroing the QF 20 pr gun in Oenturion is desoribed below* A target with a distinct aiming mark at a known range is used;
it is assumed here that, in accordance with current PWb.doctrine, this range
is 1000 yards.
20
SECR12-DISCEETI
The gun
The gun and sight are first aligned (T & A'd) as follows.
3.
is layed by boresighting through the firing-pin hole to muzzle cross-wires
onto the aiming mark; the telescope gratioule is then adjusted, using the
T & A knobs, with the range-drun set at the T & A mark, until it also is
The total misalignment
judged to be coincident with the aim g mark.
L) will therefore
between gun and sight at 1000 yards (the aligment error,
be obtained by compounding the errors due to inaccuracies in the boresighting
0 jZ, with the errors in
datum points, O*D, and in laying the boresightm,
laying the telescope graticule' TL, and these due to backlash in the sightgun
system, O , and due to the discrete intervals on the T & A knob adjustments3,'R
O*CL. Thus:E62
aa
EL$
07.2
2.
..
2.+.a..
. .
.i
The appropriate
The next step is to determine the zeroing correction.
4.
elevation is applied to the gun using the range-drum, and a cleaner round fired.
The sight is laid on the aiming mark on the target and an APDS round fired.
Two more rounds are then fired, relaying between each, and the position of the
MPI on the target estimated from the points of strike of the three rounds.
The error in estimating the position of the MPI is obtained by compounding the
is the standard
relevant variable biases and random dispersions; thus if
elevation:in
deviation of this error (the zeroing errQr)
++
04
J
(<9
+l/
+ t
....
(i)
The sight is relayed on the aiming mark again, and then adjusted using
5.
the T & A knobs onto the estimated position of the )dPI of the three shots. The
standard deviation, 0*, of the error in carrying out this part of the drill is
given by:( 2
21
2
22
Mi,
02
< A=+<(71,+
=
" +
It +
+ 0 331ra
Ther
it
.
3.33
+ 1.33<
a +
+.....2
+ 0.33 C2+ 0. 33
+ 0 .33
1 000000*
(iv)
report.,
x .1
(1.14'); hence
distribution (17)A
21
SCR-DIFTRWT
CM
SECME-DISCRM
Appx A
Similarly, in line,
V2
+ 133CS4
333
2cr 2
++
+
.33 aB2 +
0.33 v
+i
()
Alignment error
7.
An expression has already been given for the value of the alignment
error to be adopted for the shooting-in procedure.
On a subsequent occasion,
when gun and sight are T & A'd., the error in elevation must be taken as slightly
greater as it is umlikely that the range to the target will be exactly the
same as that to the target used when shooting-in.
A small error, OR-'
is
therefore included to cover this; an s.d in range difference of 20 yards at 1000
yards (equivalent to 0.050)6 at the gun) has been taken as a reasonable estimate
of this error.
Thus, in elevation,
.(a+)
. . . . .
. .......
. .
.
..
=.2 2
ITO
ab
CE
++
....
ali=ment errors
8.
Values for the shooting-in and alignment errors for Centurion have
been calculated using the expressions given above.
The magnitudes of the
component errors have already been given, in the majority of instances, in the
body of the text or in this Appendix. As has already been implied, however,
certain of the errors have values for the conditions of zeroing which are
different from those which apply in action. In particular, because it is
usual for an especial effort to be made to stand the tank on a level piece of
ground when zeroing, and also because zeroing;is only carried out on a day in
which the wind is considered moderate, the following values have been taken
for the errors due to trunnion tilt
and wind when zeroing (with ADS) at
1000 yards:Error due to trunnion tilt
= 0.039(0.132')sd
of 1.0(27J
: negligible
(0.068')sd
: 0.0206
Negligible
0.028
(0.094')sd
(29))
22
SWRET-3ISCMT
SECM-DISIEET
Appx B
Ap-pendix B
Estimation of bias due to Mn wear
There is an approximately linear drop in
i.
life of a gun; this is shown diagramatically in
is the velocity drop over the gun life L rounds
,v locity .
--
. .
.I
V1
i
II
.md.
fired
2.
It can be assumed that whenever a new barrel is mounted in a tank
it will be zeroed before the tank is in action.
An opportuity may not
~arise
for the gun to be zeroed again during the course of its life; if,
~however,
there is a chance of zeroing subsequently, this will naturally be
taken.
It seems unlikely under present circumstances that the gun will be
~zeroed
more than once subsequent to its initial zeroing.
3.
Assume that this second zeroing takes place when 1 rounds have been
Then the average drop in MV since first zeroing is given by: -
~fired.
vi =" lv/L
~Similarly,
~of
* v
(L._l)vL
~~~L;
~or
lv, + (L.-l)va
v = (l2 + (L-_)
3)v/la 2
23
SECI T-DlISCX
1) rounds
SEC R-DISCREET
4.
For present purposes, it will be assumed that the gun is not used
after the third quarter, and is either only zeroed once at the beginning of
its life or is, in addition, zeroed at the c mencement of the 2nd or 3rd
quarters of life; each of these occurrences is assumed to be equally probable.
Usin2 an approach similar to that outlined above it is found that i
19v/72
SEORET DISCREET
SECR-DISRENr
REFERENCES
(i)
DRAC
War Office
(2)
Frankford Arsenal
Report NR R138D
(3)
AORG
Memo
D13
(4)
FVRDE
Report BR IWD
(5)
ML
(6)
OBProc Q 6898
(7)
AORG
Report 83
(8)
AORG
Report 84
(9)
AORG
Report 125
(10)
AORG
Report 240
AORG
Report 225
(13)
FVDE
Report
(3J)
Range
RW90
PMDE
Report RW 99
(15)
FYRDE
Report RW57
(16)
OB Proc 31007
(17)
AORG
Report 13/51
(18)
FVRDE
Report AR172
(19)
(20)
*
Gunnery: Centurion 2
Centurion 1
FVRDE
Report RW101
FVRDE
Report RW119
R
(21)
FVRDE
(Contd)
Report RW85
(22)
Ordnance Board
Chance of Hit
Committee
(23)
OB Proc 25308
(24)
DRA
War Office
(25)
AORG
Report 517
(26) BRL
Memo Report 619
(27) BL
Memo Report 749
(28)
TAR
Report 38
(29)
ARDE
Memo (B)59/57
(30)
Nature
166.- P
(31)
(32)
906-907
Ordnance Board
Chance of Hit
Committee
TAR
Report 7
(33)
TAR
Report 8
(31+)
(35)
D & PS
(36)
War Office
Code No 8745
(37)
FVIWE
Report RW81
(38)
AORG
Memo E17
(39)
AFG, Maryland
26
SECTMT-DISCREET
For Indexing
MCRETISCFEET
Armament Research & Developmant Establishment
A.R.D.E. Report (B)6/58
531.567:
531.56.088:
Feb. 1958
chance
the QF
shell.
battle
6P. .412.6
SECRET/DISCREET
Armament Research & Development
Establishment
A.R.D.E. Report (B)6/58
531,567;
531.56.088:
623.412,6
Feb.1958
SECRET/D ISCREET
SECRET/D ISC EET
*of
531.567:
531.56.088:
62 412.6
cond it Ions.
R.W.Shephard, S.A.Beach.
chance
the QF
shell.
battle
Feb 1958
531.567:
531.56.088:
623o412.6
Feb. 1958
SECRET/D ISCREE T
U. i. C"oni,idential
Dritish Restricted
dstll
[dstl I
AD#: AD159528
Date of Search: 31 July 2008
Record Summary: DEFE 15/997
Title: Classification of errors involved in determination of chances of hit and evaluation of
chance of hit with Centurion tank under quasi-battle conditions
Availability Open Document, Open Description, Normal Closure before FOI Act: 30 years
Former reference (Department) Report No (B) 6/58
Held by The National Archives, Kew
This document is now available at the National Archives, Kew, Surrey, United
Kingdom.
DTIC has checked the National Archives Catalogue website
(http://www.nationalarchives.gov.uk) and found the document is available and
releasable to the public.
Access to UK public records is governed by statute, namely the Public
Records Act, 1958, and the Public Records Act, 1967.
The document has been released under the 30 year rule.
(The vast majority of records selected for permanent preservation are made
available to the public when they are 30 years old. This is commonly referred
to as the 30 year rule and was established by the Public Records Act of
1967).