Technical Note Rosemount Twisted Square Thermowell

00840-0100-4952, Rev AA March 2018

Rosemount™ Twisted Square™ Thermowell

Vortex-induced vibration solution
1.0 Introduction
Thermowells subjected to fluid flow experience dynamic stresses imposed by oscillating vortex
pressures. These dynamic stresses can result in vortex-induced vibration (VIV), which is the leading
cause of thermowell stress failures. Emerson™ now offers an innovative solution using patented
Rosemount Twisted Square thermowell technology to provide a robust stem design that dampens VIV,
effectively eliminating the concern of
dynamic stress failure.
This new design was inspired by the
difficulties of designing conventional
thermowells to avoid issues with VIV.
Most often, stress failures occur on
thermowells that have not been
configured following recommendations
based on calculations per ASME PTC 19.3
TW to ensure the thermowell will
withstand the fluid forces and process
pressures. The wake frequency limit
generally calls for the most drastic design
changes, especially for long thermowells
or high-velocity processes. It is critical
that the frequency limit is verified to
ensure that the natural frequency of the
thermowell is safely away from the Strouhal frequency (vortex-shedding frequency). As these
frequencies converge, the thermowell can lock-in to resonant conditions, and the dynamic stresses
caused by VIV are greatly magnified.
The conventional solution to avoid these lock-in regions is to shorten the thermowell and/or increase the
outer diameter to shift the thermowells natural frequency away from resonance. These changes can
result in decreased accuracy and/or slower response time for the measurement. The dimensional
adjustments recommended by the calculations are often acceptable to maintain an appropriate
measurement point, but in some cases, they may compromise temperature measurement accuracy,
making it difficult to implement. These situations are becoming more frequent, as average fluid
velocities increase due to more plants using smaller pipe sizes to save cost on projects.
A complete overview of thermowell calculations per ASME PTC 19.3 TW is available for download from
Emerson.com.
Rosemount Twisted Square Thermowell Technical Note
March 2018 00840-0100-4592, Rev AA

2.0 Rosemount Twisted Square - VIV solution

When facing wake frequency issues, an alternative solution to adjusting the thermowell dimensions is to
damp the dynamic stresses from the vortices, and effectively suppress VIV excitation. This thermowell
design functions by desynchronizing vortices in its wake, so they are not uniformly defined or alternating
at a consistent phase along the length of the thermowell. See Figure 1.

Figure 1. Design Comparison

Conventional Thermowell Rosemount Twisted Square Thermowell

Conventional thermowells have a circular cross The Rosemount Twisted Square’s sharp edges and helical
section, which allows organized vortices to shed profile provide continuously changing separation points for
in sheets along the axial length of the stem. The vortex shedding along the axial length of the stem. The
uniform pressure from the vortices apply vortices are forced out of phase, preventing them from
alternating forces on the thermowell. These becoming synchronized along the span, significantly
forces are greatly magnified by VIV and can lead damping dynamic stress levels and eliminating concern of
to thermowell fatigue failure. VIV.

2.1 Advantages of the Rosemount Twisted Square thermowell

 Reduces dynamic stress by over 90 percent, eliminating concern of VIV
 Simplifies thermowell calculations to reduce design time
 Uses same thermowell design for multiple processes and tag locations
 Reduces inventory since one thermowell profile fits all
 Greatly reduces the risk of thermowell failures during changing process conditions (start-up, shut
down, events)
 Supports increased fluid velocities for more throughput or smaller pipes
 Allows for longer insertion lengths for more specific sensor placement
 Allows for smaller thermowell diameters for better temperature response time
 Avoids the need for complex and expensive pipe modifications

3.0 Understanding VIV and its effects

Vortices are shed around bluff bodied objects subject to flow, which interact with the structure,
imposing oscillating pressures on the surface. These pressures can cause an elastic structure to vibrate,
and these vibrations can be greatly magnified if the vortex frequency nears the natural frequency of the
structure. This VIV effect is of great importance to many fields of engineering because of the possible

2 Rosemount Twisted Square Thermowell

Technical Note Rosemount Twisted Square Thermowell
00840-0100-4952, Rev AA March 2018

destructive effects. For example, VIV can cause destructive vibrations on chimney stacks or bridges
subject to wind, platform risers subject to ocean currents, heat exchanger tubes, thermowells, and
sample probes subject to fluid flow in pipes.
The character of these oscillatory patterns around a circular cylinder is dependent on the Reynolds
Number (Re), which is a dimensionless parameter defined by the free stream velocity U, cylinder
diameter B, and kinematic viscosity v of the fluid.

Flow regimes dependency on Re for a circular cylinder was summarized by Lienhard in1966, and shown
in Figure 2. The location of boundary layer separation about the diameter is also dependent on Re and
correlates with the flow regimes.

Figure 2. Regimes of Fluid Flow across Circular Cylinders (Lienhard, 1966)

Re < 5 Regime of unseparated flow.

5 to 15 ≤ Re < 40 A fixed pair of FÖPPL vortices in the wake.

40 ≤ Re < 90 and 90 ≤ Re < 150

Two regimes in which vortex street is laminar
Periodicity governed in low Re range by wake
instability.
Periodicity governed in high Re range by vortex
shedding.

150 ≤ Re < 300 Transition range to turbulence in vortex.

5
300 ≤ Re ~
< 3 x 10 Vortex street is fully turbulent.

3 x 105 ~< Re 3.5 x 106

Laminar boundary layer has undergone turbulent transition.
The wake is narrower and disorganized.
No vortex street is apparent.

6
3.5 x 10 ≤ Re
Re-establishment of the turbulent vortex street that was
5
evident in 300 ≤ Re ~< 3 x 10 . This time the boundary layer
is turbulent and the wake is thinner.

Vortices are not formed at low Re Numbers (Figure 2), but as the Re increases, the pressure is not
sufficient to send the fluid completely around the cylinder. The fluid separates from the side of the
cylinder, and vortices are formed by fluid folding over itself due to the shear-layer velocity difference
between the free flow and the surface of the cylinder (Blevins, 2001). The vortex on one side of the
cylinder will grow and draw the shear layer from the opposite side, causing the first vortex to shed

Rosemount Twisted Square Thermowell 3

Rosemount Twisted Square Thermowell Technical Note
March 2018 00840-0100-4592, Rev AA

(Gerrard, 1964). This pattern continues, and regularly alternating vortices form an ordered wake pattern,
downstream of the object, which is known as a Von Karman vortex street.

The frequency of the vortex oscillations is known as the Strouhal frequency ( ), which increases linearly
with the fluid velocity and is also a function of Re. One cycle of the Strouhal frequency consists of two
vortices, one from each side of the cylinder (Figure 4). The vortices separate in sheets along the length of
a circular cylinder, causing surface pressures that result in two dynamic forces (Figure 3).
 Oscillating-lift force: Each vortex creates transverse surface pressure on the cylinder perpendicular to
the flow and relative to the side of the cylinder from which it was shed. The lift force fluctuates with at
frequency .
 Oscillating-drag force: Each vortex creates surface pressure on the cylinder, in-line with the fluid flow,
fluctuating at frequency 2 .

Figure 3. Dynamic Forces on a Thermowell Caused by Vortex Pressure

Oscillating Drag (In-line) Oscillating Lift (Transverse)

Point of max stress Point of max stress

from bending moment from bending moment

Oscillating drag force

Fd = Pd A d Oscillating lift force
FLOW F1= P1A1
Oscillating drag pressure
Oscillating lift pressure
Pd = 1 ρCd U ²
2 FLOW P = 1 ρC U²
1 1
2

Cd and Cl are constant drag and lift coefficients for a circular cylinder. Cd = 0.1 and Cl = 1.0.

Figure 4. Thermowell Vibration

In-line Resonance Transverse Resonance Thermowell Natural Frequency = fcn
fs = 0.5 fcn fs = fcn

Strouhal Frequency = fs
1 cycle of f s 1 cycle of f s

Thermowell
FLOW Tip

Z
Y (in-line)
W
FLO X (transverse)

In-line
Transverse

Fluid Velocity

(Above) relationship of the thermowell natural frequency and the Strouhal frequency. (Below) The chart shows vibration
amplitudes calculated per ASME PTC 19.3 TW-2016. The dotted lines are on the resonant frequencies, and the gray area show the
lock-in regions. ASME PTC 19.3 TW-2016 has keep-out regions to avoid lock-in conditions where liquid must have fs < 0.4fcn to
avoid in-line and transverse resonant conditions, and gas must have fs < 0.8fcn with special considerations to pass through 0.4fcn <
fs < 0.6fcn.

A thermowell responds elastically to the fluctuating forces. Bending stress from VIV will increase
dramatically at resonant conditions as the thermowell natural frequency (fnc) converges with the
4 Rosemount Twisted Square Thermowell
Technical Note Rosemount Twisted Square Thermowell
00840-0100-4952, Rev AA March 2018

frequencies of the oscillating lift ( ) or oscillating drag force (2 ). The frequencies do not need to overlap
directly, as the coupled movement of the cylinder (at fnc) will influence the vortex behavior, causing the
vortex frequency to shift and synchronize with the vibration frequency. This effect is known as lock-in.
While in lock-in conditions, the vibrations of the cylinder will increase the strength of the vortices and
correlation of the wake along the thermowell axis. A conventional thermowell can lock-in to the
transverse or in-line direction, though transverse forces are of a higher magnitude due to the higher
force coefficient and higher velocity.

4.0 Damping VIV

A history of design failures and challenges in many applications has prompted much investigation into
damping and suppressing VIV. There is the strategy to avoid resonant or lock-in conditions, which is
traditionally done for thermowells with calculations per ASME PTC 19.3 TW. Other methods include:
 Add damping to the base of the structure
 Streamline the design
 Actively disrupt the flow around the structure by external means
 Add some device or modify the structure’s geometry for passive damping, which has proven to be
cost effective for cylinders and is applicable for many industrial structures.
Passive-damping devices and geometries function by disrupting the formation of an organized vortex
street. Summaries of different passive designs are described by (Blevins, 2001) and (Raghavan A. Kumar,
2008). Helical designs are among the most effective in damping VIV and applicable to thermowells,
providing an ability to withstand harsh process conditions without dependence on orientation. Scruton
was the first to investigate the use of helical damping of VIV (C. Scruton, 1957) by using strakes on scale
models to simulate wind on towers. Other investigations describe recommended designs by optimizing
the number, geometry, and pitch of the strake for certain applications (Htsuo Ishizaki, 1984) (Raed K.
Lubbad, 2011) (J. F. Wilson, 1989).
Helical designs do not eliminate the formation of vortices but rely on flow separation to be tripped over a
protrusion or edge along the cylinder axial length. The helical shape prevents the vortex shedding from
being correlated along the axial length as explained earlier. An ANSYS CFX simulation (Figure 5 on page 6
shows core vortex pressure for fluid around a circular cylinder compared with the helical profile of the
Rosemount Twisted Square. The image illustrates how the vortices shed from a circular cylinder in
sheets, uniformly alternating from one side to the other, while the Rosemount Twisted Square prevents
vortices from correlating in a synchronized phase by forcing separation at different points along the axial
length. This suppresses harmful VIV at resonant frequencies and damps the dynamic forces acting on the
cylinder, as the vortex pressures are not evenly distributed (Figure 6 on page 6). There are also
suggestions that helical designs have potential three-dimensionality in the separating flow which can
destroy regular vortex shedding (Peter Bearman, 2004).
Numerous investigations of the characterization of flow behavior around straight square cylinders show
how the Strouhal Number and force coefficients change with different incidence angles or angle of
attack (McClean, 2014) (R.F. Huang, 2010) (Igarashi, 1984). This dependence on the angle is due to:
 Differences in shear layers emanating from opposite sides of the square
 Separation from the leading edges and flow reattachment (for some angles) to the side
 Changes in wake width and vortex formation length
Fluctuating surface pressure was found to be one of the most sensitive characteristics in relation to the
angle of incidence. Similarly, opposite side faces have trends that do not coincide at most angles
(Igarashi, 1984). These behaviors may be the cause of desynchronized vortex shedding on the
Rosemount Twisted Square and could contribute to the damping of VIV.

Rosemount Twisted Square Thermowell 5

Rosemount Twisted Square Thermowell Technical Note
March 2018 00840-0100-4592, Rev AA

Figure 5. ANSYS CFX Simulation

ANSYS CFX simulation showing core vortex pressure around a conventional cylinder and a Rosemount Twisted Square profile. The
vortex pressure has correlation along the length of the conventional cylinder causing higher forces compared to the varied
pressure on the Rosemount Twisted Square. The Rosemount Twisted Square forces separation causing desynchronized vortices
and vortex pressure acting on the thermowell.

Figure 6. Damped Dynamic Force

Conventional Twisted Square

Oscillating Lift Force

FLOW FLOW Oscillating Lift Pressure

How dynamic force is damped on the Rosemount Twisted Square, as the vortex pressures are not uniformly distributed along the
stem length.

5.0 Rosemount Twisted Square design

Emerson has been investigating VIV suppression on thermowells for over a decade. There has been
extensive research, development and testing on a variety of passive thermowell designs (Figure 7 on
page 7) in attempts to minimize the effects of VIV. The testing compared tip displacement of various
stem profiles subjected to flow by measuring VIV amplitude with accelerometers in thermowell tips. This
lead to development of the Rosemount Twisted Square design using smooth, flat surfaces for a robust,
helical profile with four sharp edges at an optimized pitch angle. The design causes desynchronized
vortex separation along the length of the thermowell, preventing formation of an organized, two-dimen-
sional, vortex street. The effectiveness of the Rosemount Twisted Square has been proven through
Emerson and third-party testing to reduce dynamic stress by more than 90 percent, eliminating the
need for dynamic stress calculations.

6 Rosemount Twisted Square Thermowell

Technical Note Rosemount Twisted Square Thermowell
00840-0100-4952, Rev AA March 2018

Figure 7. Stem Profiles

Examples of some stem profiles tested during Rosemount Twisted Square Research and Development.

5.1 Emerson testing

Test methods
Throughout the years, Emerson performed most of the Rosemount Twisted Square testing using flow
equipment built for calibrating Emerson’s vortex meters. The most recent round of testing was
performed on the final Rosemount Twisted Square design, manufactured using standard production
processes.
The purpose of the test was to compare dynamic tip displacement of a conventional straight thermowell
against a Rosemount Twisted Square thermowell subjected to flow. The test setup (Figure 8 on page 8)
included two 316 stainless steel Van Stone thermowells each having a stem diameter of 3/4 inch (19mm)
with unsupported lengths of 19 inches (480mm). The thermowells were mounted in a 6-in. pipe with
about 5.5 inches (140 mm) of the stem immersed into the flow. The thermowells were designed to
resonate transversely in water at about 20 ft/s (6 m/s), and flow was adjusted from 5 to 28 ft/s (1.5 to 8.5
m/s) in 1-ft/s increments. Reynolds numbers were from 3.0 x 104 to 1.7 x 105. Accelerometers were
secured in the thermowell tips to determine the displacement amplitude from both inline and transverse
vibration. Root mean square (RMS) accelerometer data was collected with a signal analyzer using a
Hanning window for frequencies between 5 to 205 Hz.

Rosemount Twisted Square Thermowell 7

Rosemount Twisted Square Thermowell Technical Note
March 2018 00840-0100-4592, Rev AA

Figure 8. Test Setup Schematic

Signal conditioners
for transverse and
Van Stone Thermowell inline axis
19 inch 316 SST

Flange standoff and

extension assembly

Signal Analyzer

Six inch pipe section

Water Water
FLOW
Pump Reservoir

Accelerometer

View port

Experimental results
Spectrum plots (Figure 10 on page 9) of thermowell tip acceleration show that the peak acceleration
points are close to the thermowell natural frequencies predicted by the calculations. Tip displacement
data was plotted (Figure 9) by integrating the maximum RMS acceleration at each velocity using the
corresponding frequency for each data point. Maximum values of tip displacement and dynamic bending
stress for the inline and transverse directions are shown in Table 1.

Figure 9. Tip Displacement Versus Velocity

.75" Conventional Thermowell vs. Twisted Square

Single amplitude displacement; RMS 10 avg spectrum data
Linear Scale Log Scale
Fluid Velocity [m/s] Fluid Velocity [m/s]
1.5 3.0 4.6 6.1 7.6 9.1 1.5 3.0 4.6 6.1 7.6 9.1
0.4 10.2 1

0.35 8.9

0.3 7.6 0.1

Tip Displacement [inch]

Tip Displacement [inch]

Tip Displacement [mm]

0.25 6.4

0.2 5.1 0.01

0.15 3.8

0.1 2.5 0.001

0.05 1.3

0
0 0.0001
5 10 15 20 25 30 5 10 15 20 25 30
Fluid Velocity [ft/s] Fluid Velocity [ft/s]

Conventional Inline Conventional Transverse Conventional Inline Conventional Transverse

Twisted Square Inline Twisted Square Transverse Twisted Square Inline Twisted Square Transverse

Tip displacement vs. velocity for the conventional and Rosemount Twisted Square thermowells in the in-line and transverse
directions. The linear scale shown on the left illustrates the dramatic reduction of tip vibration for the Rosemount Twisted Square.
The log scale on the right shows more detail of the Rosemount Twisted Square tip vibration.

Note
The conventional thermowell design in this test does not pass ASME PTC 19.3-TW calculations for any velocity over
7.4 ft/s (2.2 m/s). The calculated lock-in velocities are between 7.4 and 11.2 ft/s (2.2 and 3.4 m/s) for inline
resonance and over 14.8 ft/s (4.5 m/s) for transverse.

8 Rosemount Twisted Square Thermowell

Technical Note Rosemount Twisted Square Thermowell
00840-0100-4952, Rev AA March 2018

Table 1. Peak Tip Displacement Points for the Inline and Transverse Direction

Inline Transverse
Rosemount
Conventional Conventional Rosemount
Twisted
thermowell thermowell Twisted Square
Square
Maximum RMS
.013 in.(0.34
displacement of .092 in.(2.34 mm) .358 in.(9.10 mm)
mm)
accelerometer
Velocity at max
8 ft/s (2.44 m/s) 19 ft/s (5.79 m/s) 21 ft/s (6.40 m/s)
displacement No peak
Vibration frequency
56 Hz 54.5 Hz 69.5 Hz
at max displacement
RMS bending stress
at root of 12.4 ksi (85.4 MPa) 48.1 ksi (331.7 MPa) 1.7 ksi (11.9 MPa)
thermowell

Figure 10. Spectrum Plots for Accelerometer

Conventional Thermowell - Transverse Twisted Square Thermowell - Transverse

Spectrum Plot Spectrum Plot
120 120

Velocity (ft/s) Velocity (ft/s)

100 100
5 6 7 5 6 7
8 9 10 8 9 10
80 11 12 13 80 11 12 13
RMS Acceleration [g's]

RMS Acceleration [g's]

14 15 16 14 15 16
17 18 19 17 18 19
60 60
20 21 22 20 21 22
23 24 25 23 24 25
40 26 27 28 40 26 27 28

20 20

0 0
35 45 55 65 75 85 35 45 55 65 75 85
Frequency [Hz] Frequency [Hz]

The results show that the Rosemount Twisted Square thermowell greatly damps the effects of VIV. The
Rosemount Twisted Square machined from 3/4-in. diameter bar stock vibrates at a higher frequency than
the 3/4-in. diameter conventional thermowell, which is expected given its reduced mass. The peak
displacement values of the Rosemount Twisted Square are also at a higher velocity than the circular
thermowell. ANSYS was used to determine the maximum bending stress at the root of the thermowells
based on the tip displacement. Even with peak displacement values at higher velocities, the Rosemount
Twisted Square shows a greater than 96 percent reduction of stress in the transverse direction. The data
shows no significant peak vibration for the Rosemount Twisted Square at inline conditions, which
indicates complete VIV suppression with a 99 percent reduction of stress compared to the conventional
thermowell at inline lock-in conditions.
The max RMS dynamic stress at the root of the conventional thermowell was 48.1 ksi (331.7 MPa) and
1.7 ksi (11.9 MPa) for the Rosemount Twisted Square. These RMS values can be converted to peak values
of 68.0 ksi (469.1 MPa) for the conventional and 2.4 ksi (16.8 MPa) for the Rosemount Twisted Square.
The maximum peak dynamic stress on the Rosemount Twisted Square is much less than the fatigue limit
of 13.6 ksi (93.8 MPa) given by ASME PTC 19.3-TW for a 300-series stainless steel Van Stone thermowell.
It is also significantly less than the lower fatigue limit of 5.4 ksi (37.2 MPa) for a threaded thermowell. The

Rosemount Twisted Square Thermowell 9

Rosemount Twisted Square Thermowell Technical Note
March 2018 00840-0100-4592, Rev AA

Rosemount Twisted Square thermowell in this test situation does not pass steady-state drag stress
calculations for water velocities over 27 ft/s (8.2 m/s), and dynamic stress and wake frequencies are not a
concern for the process ranges allowed by Rosemount Twisted Square calculations.
Both thermowells show decreasing transverse vibration as the velocities increase past the peak vibration
point. It is worth noting that a similar test set-up was used for 19-in. thermowell samples in the past, with
water velocities up to 50 ft/s (15.2 m/s). The trends were also consistent with recent tests with velocity
ranges up to 28 ft/s (8.5 m/s). Vibration amplitudes were shown to increase steadily for velocities past 30
ft/s (9.1 m/s) without any peaks. The stress values at 50 ft/s (15.2 m/s) were still lower than the
maximum values near transverse resonance.
A video of the thermowell vibrations in this test can be found here.
The video is a powerful tool for understanding the severity of VIV that can be experienced by a
conventional thermowell while highlighting the impact of the Rosemount Twisted Square design to
damp vibration.

6.0 Third party testing—Rosemount Twisted Square drag

coefficient
(Derek G. Dahl, 2016)

6.1 Test methods

Emerson has worked with FluiDyne AeroSystems to conduct a third-party investigation to determine the
steady-state drag coefficient (CD) of the Rosemount Twisted Square. This dimensionless parameter
quantifies drag or resistance of an object subject to fluid flow and is necessary for conducting
steady-state stress calculations on thermowells. It is known that CDis dependent on geometry as well as
orientation and increases with the addition of strakes, so the coefficient could not be assumed for the
Rosemount Twisted Square without testing.
The test set-up (Figure 11 on page 11 and Figure 12 on page 11) consisted of a 3/4-in. circular cylinder
and a Rosemount Twisted Square machined from 3/4-in. round bar stock. Each cylinder was mounted in a
wind tunnel using externally mounted load cells on either side to measure drag force. The wind tunnel
was 12 inches (304 mm) wide and 9.125 inches (232 mm) tall, leaving a blockage ratio of less than 8.4
percent. Data was taken at velocities of about 30 to 270 ft/s (9.1 to 82.3 m/s) resulting in Reynolds
numbers (Re) of 1.1 x 104 to 9.7 x 104.
CD is formed by normalizing the measured drag (D) by the product of fluid kinetic energy and the
cylinder projected frontal area (AT) as shown in the following equation where ρ is the fluid density and U
is the fluid velocity. The diameter of the circle that circumscribes the Rosemount Twisted Square was
used in calculations for Re and AT during this investigation and must be used consistently with CD for
thermowell steady-state drag calculations.

10 Rosemount Twisted Square Thermowell

Technical Note Rosemount Twisted Square Thermowell
00840-0100-4952, Rev AA March 2018

6.2 Experimental results

Results show that the drag coefficients for both cylinders were about the same and steady within the
tested range of Re (Figure 13 on page 12). The circular cylinder had an average CD=1.0017, and the
Rosemount Twisted Square had an average CD=1.0023.
The report concludes that the drag coefficient does not need to be adjusted when evaluating
steady-state stress on the Rosemount Twisted Square thermowell, and the value of CD=1.4, used by
ASME PTC 19.3 TW-2016, is suitable for calculations.

Figure 11. Blower-driven Wind Tunnel Schematic

Test Article Test Section

TOP VIEW Blower

Atmosphere FLOW

Static Pressure, P Load Cell Control Valve

Blower
SIDE VIEW

Atmosphere FLOW

Load Cell Test Article NOT DRAWN TO SCALE

Control Valve

Figure 12. Isometric View of Wind Tunnel

(Left) isometric view of wind tunnel used for drag coefficient test at FluiDyne; (top) 3/4- in. circular cylinder;
(bottom right) 3/4-in. Rosemount Twisted Square.

Rosemount Twisted Square Thermowell 11

Rosemount Twisted Square Thermowell Technical Note
March 2018 00840-0100-4592, Rev AA

Figure 13. Drag Stress Results Comparison

Steady-state Drag Coefficient of Drag Measurements of
3/4" Diameter Thermowell Cylinders 3/4" Diameter Thermowell Cylinders
Velocity [m/s]

1.20 0.0 15.2 30.5 45.7 61.0 76.2 91.4

6.00
26.69
1.15
5.00 22.24
1.10 Circular
Drag Coefficient, CD

Circular

Pound Force [lbs]

1.05 Twisted 4.00 Twisted Square 17.79

Force [N]
Square
1.00
3.00 13.35
0.95
2.00 8.90
0.90

0.85 1.00 4.45

0.80 0.00
0.00
0.00E+00 2.00E+04 4.00E+04 6.00E+04 8.00E+04 1.00E+05
0.0 50.0 100.0 150.0 200.0 250.0 300.0
Reynolds Number, Re
Velocity [ft/s]

Results comparing drag stress of a circular cylinder to the Rosemount Twisted Square, conducted at FluiDyne Aerotest Laboratory.

7.0 Rosemount Twisted Square thermowell calculations

The Rosemount Twisted Square thermowell minimizes the complexity of thermowell calculations
compared with conventional thermowell calculations per ASME PTC 19.3 TW (Figure 14), saving time
and money during the design process. This standard defines the dimensional requirements for
thermowells to fit within the standard and specifically states that thermowells with designed surface
structures are beyond the scope of the standard due to the difficulty of providing design rules. The
challenge with providing design rules for unique geometry comes with the complicated prediction of
vortex behavior for wake frequency ratios and dynamic stress. The Rosemount Twisted Square design
relies on testing to confirm sufficiently damped dynamic stress, eliminating the need for wake frequency
and dynamic stress calculations because the thermowell will fail from steady-state drag stress before a
fatigue failure will occur. Emerson calculates the steady-state stress and hydrostatic pressure limits by
following the strategy used in ASME PTC 19.3 TW with conservative adjustments to account for the
Rosemount Twisted Square shape and a steady-state drag coefficient verified by third-party testing.

Figure 14. Thermowell Calculations

Conventional Thermowell Calculations

Twisted Square Calculations
to Satisfy ASME PTC 19.3 TW
Frequency Limit - Ensure the thermowell will not lock-in to
Frequency Limit
resonant frequencies.
Dynamic Stress Limit - Ensure the vortex induced dynamic
Dynamic Stress Limit
stress will not exceed the fatigue limit.
Static Stress Limit - Ensure the steady-state drag stress from
Static Stress Limit
the fluid will not exceed the allowable working stress.
Hydrostatic Pressure Limit - Ensure the thermowell will not
Hydrostatic Pressure Limit
rupture from the process pressure.

Left shows the different calculations done for a conventional thermowell to satisfy ASME PTC 19.3 TW; the Rosemount Twisted
Square design eliminates the need for wake frequency and dynamic stress calculations.

The frequency limit is the most difficult calculation to pass, and the iterations to the thermowell
dimensions are undesired, resulting in reduced length and increased diameters. Without the challenge
12 Rosemount Twisted Square Thermowell
Technical Note Rosemount Twisted Square Thermowell
00840-0100-4952, Rev AA March 2018

of wake frequency ratios, calculations can be done quickly with low chance of dimensional changes to
the thermowell. Rather than modifying a conventional thermowell design to meet requirements of
conventional calculations (Figure 15), a single Rosemount Twisted Square thermowell can work in most
situations, satisfying multiple measurement points or process changes.

Figure 15. Calculation Adjustments to Thermowells

Process 1 Process 2 Process 3

Process Details

Process Description Heavy Oil Heavy Gas Medium Gas

850 kg/m3[53 60 kg/m3[3.75 7 kg/m3[0.44

Fluid Density lb/ft3] lb/ft3] lb/ft3]
80 m/s [262.5
Fluid Viscosity 9 m/s [29.5 ft/s] 20 m/s [65.5 ft/s]
ft/s]

Process Pressure 1.5 Mpa[220 psi] 10 Mpa[1450 psi] 8 Mpa[1160 psi]

Pass Frequency Ratio

Fail Fail Fail
Criteria?
Calculations

Pass Dynamic Stress

Fail Fail Fail
Criteria?

Pass Static Stress Twisted Square

Pass Pass Pass
Criteria? One size fits all
without
Pass Hydrostatic adjusting
Pass Pass Pass
Pressure Criteria? dimensions

Rosemount Twisted Square Thermowell 13

Rosemount Twisted Square Thermowell Technical Note
March 2018 00840-0100-4592, Rev AA

Process 1 Process 2 Process 3 Process 1, 2, 3

Unsupported Length 235 mm [9.25 140 mm [5.51 305 mm [12

305 mm [12 in.]
“L” in.] in.] in.]
Solution: Thermowell Adjustments
155 mm [6.1
Shielded Length “S” 155 mm [6.1 in.] 155 mm [6.1 in.] 155 mm [6.1 in.]
in.]
150 mm [5.9
Immersion Length “I” 150 mm [5.9 in.] 80 mm [3.15 in.] -15mm [-.59 in.]
in.]

Root Diameter “A” 30 mm [1.18 in.] 35 mm [1.38 in.] 40 mm [1.57 in.] 19 mm [.75 in.]

Tip Diameter “B” 20 mm [.79 in.] 25mm [.98 in.] 30 mm [1.18 in.] Twisted Square

Example of adjustments to a conventional thermowell design to pass calculations at different process conditions. The Rosemount
Twisted Square design would work for all processes allowing for a single thermowell design without adjustments to the dimensions
as well as a single sensor length. The temperature used for analysis was 392 °F (200 °C).

8.0 Conclusion
Emerson’s patented Rosemount Twisted Square thermowell breaks the mold of traditional thermowell
design limitations. The robust stem design has been proven through testing to eliminate VIV issues and
streamline thermowell design. This allows it to be used in higher flow velocity applications resulting in
more responsive and accurate temperature measurements, faster calculations and reduced inventory.
The design is available in the Rosemount 114C Thermowell platform with threaded, flanged, Van Stone
and socket weld mounting styles in any material available through the Rosemount 114C Thermowell.
For more information, please see the Rosemount Twisted Square Product Data Sheet.

14 Rosemount Twisted Square Thermowell

Technical Note Rosemount Twisted Square Thermowell
00840-0100-4952, Rev AA March 2018

References
Blevins, R. D. (2001). Flow-Induced Vibration Second Edition. Malabar, Florida: Krieger Publishing
Company.
C. Scruton, D. E. (1957). A Means for Avoiding Wind-Excited Oscillations of Structures with Circular or Nearly
Circular Cross-Section. National Physical Laboratory.
Derek G. Dahl, K. L. (2016). Wind Tunnel Drag Measurements of Emerson Thermowells with Circular and
Twisted Square Cross Sections. FluiDyne AeroSystems.
Gerrard, J. (1964). The mechanics of the formation region of vortices behind bluff bodies. Journal of Fluid
Mechanics, 401-416.
Htsuo Ishizaki, H. H. (1984, October). The efficiency of helical strakes for the suppression of
vortex-excited oscillation of steel stacks. Engineering Structures, pp. 334-339.
Igarashi, T. (1984). Characteristics of the Flow around a Square Prism. Bulletin of JSME, Vol 27, No. 231,
1858-1865.
J. F. Wilson, J. C. (1989). Vortex Load Reduction: Experiments in Optimal Helical Strake Geometry for
Rigid Cylinders. Journal of Energy Resources Technology, 72-76.
Lienhard, J. H. (1966). Synopsis of Lift, Drag, and Vortex Frequency Data for Rigid Circular Cylinders. Pullman,
Washington: Washington State University.
McClean, J. F. (2014). An Experimental Investigation of Aspect Ratio and Incidence Angle Effects for the
Flow Around Surface-Mounted Finite-Height Square Prisms. Journal of Fluids Engineering.
Peter Bearman, M. B. (2004). Experimental studies of passive control of vortex-induced vibration.
European Journal of Mechanics B/Fluids 23, 9-15.
R.F. Huang, B. L. (2010). Time-averaged topological flow patterns and their influence on vortex shedding
of a square cylinder in crossflow at incidence. Journal of Fluids and Structures 26, 406-429.
Raed K. Lubbad, S. L. (2011). Experimental Investigations of the Efficiency of Round-Sectioned Helical
Strakes in Suppressing Vortex Induced Vibrations. Journal of Offshore Mechanics and Arctic Engineering.
Raghavan A. Kumar, C.-H. S. (2008). Passive Control of Vortex-Induced Vibrations: An Overview. Recent
Patents on Mechanical Engineering, 1-11.

Rosemount Twisted Square Thermowell 15

 Technical Note
00840-0100-4952, Rev SS
March 2018

Global Headquarters
Emerson Automation Solutions
6021 Innovation Blvd.
Shakopee, MN 55379, USA
+1 800 999 9307 or +1 952 906 8888
+1 952 949 7001
RFQ.RMD-RCC@Emerson.com

North America Regional Office

Emerson Automation Solutions
8200 Market Blvd.
Chanhassen, MN 55317, USA
+1 800 999 9307 or +1 952 906 8888
+1 952 949 7001
RMT-NA.RCCRFQ@Emerson.com

Latin America Regional Office

Emerson Automation Solutions
1300 Concord Terrace, Suite 400
Sunrise, FL 33323, USA
+1 954 846 5030
+1 954 846 5121
RFQ.RMD-RCC@Emerson.com

Europe Regional Office

Emerson Automation Solutions Europe GmbH
Neuhofstrasse 19a P.O. Box 1046
CH 6340 Baar
Switzerland
+41 (0) 41 768 6111
+41 (0) 41 768 6300
RFQ.RMD-RCC@Emerson.com
Linkedin.com/company/Emerson-Automation-Solutions
Asia Pacific Regional Office
Emerson Automation Solutions
1 Pandan Crescent Twitter.com/Rosemount_News
Singapore 128461
+65 6777 8211
Facebook.com/Rosemount
+65 6777 0947
Enquiries@AP.Emerson.com
Youtube.com/user/RosemountMeasurement
Middle East and Africa Regional Office
Emerson Automation Solutions
Google.com/+RosemountMeasurement
Emerson FZE P.O. Box 17033
Jebel Ali Free Zone - South 2
Emerson Terms and Conditions of Sale are available upon request.
Dubai, United Arab Emirates The Emerson logo is a trademark and service mark of Emerson Electric Co. Rosemount
+971 4 8118100 is a mark of one of the Emerson family of companies. All other marks are the property
of their respective owners.
+971 4 8865465 © 2018 Emerson. All rights reserved.
RFQ.RMTMEA@Emerson.com