May 15-19, 2016

Temperature Monitoring of PEEK

Bearings

Jie Zhou, Waukesha Bearings

Content

Problem Statement
Goal
Literature Review
Test Setup
Results
Summary and Conclusion

2
Design Characteristics
PEEK

High temperature resistance with excellent chemical and

fatigue resistance plus thermal stability
Retain mechanical properties at 250+C
Heat distortion temperature (HDT) up to 160C virgin; up
to 315C with reinforcement
Good wear resistance low coefficient of friction and high
limiting PV properties
PEEK and PEEK lining can extend bearing operating limit
Leopard, A. J. "Tilting pad bearings-limits of operation."
Lubrication Engineering 32.12 (1976): 637-644.

3
Forms and Applications of PEEK Bearings

Combined polymer bearing for Polymer lined thrust bearing for Combined solid polymer bearing
water-lubricated CHP turbine oil-lubricated pumps

Polymer lined tilting Solid polymer thrust bearing Polymer lined Flexural Pivot thrust bearing for
journal pad for water pumps steam turbines

4
Problem Statement

Bearing temperature is an important industry indicator of bearing health; however,

traditional temperature monitoring methods used for babbitt bearings might not provide
sufficient warning of bearing distress when used with PEEK bearings. Non-traditional
methods of temperature monitoring is needed for PEEK bearings.

5
Goal

To identify effective temperature monitoring options for PEEK via polymer

bearing tests and determine the best method for industrial application and lab
testing.

6
Garner and Leopard, 1985: Review for Babbitt

Discuss temperature sensor types, position, installation, and

alarm/shutdown setting for babbitt (whitemetal) lined fluid film
bearings
Emphasize the importance of pad temperature, not only
lubrication supply/discharge temperature
Function of pad temperature is to provide safeguard of
gradual changes
No monitoring benefit of maximum temperature for industrial
application
Position sensors where film breakdown could occur
Suggest placement of temperature sensors below the bond
line
Garner, Denis R., and A. J. Leopard. "Temperature measurements in fluid film bearings." Proceedings of the
13th Turbomachinery Symposium, College Station, Texas. 1985.

7
American Petroleum Institute (API)

API standards address babbitt only

API standards recommend measuring bearing-metal temperature
API 616 Gas turbine 4.8.5.5
API 617 Axial and Centrifugal compressors 2.7.1.2
API 617 Integrally geared compressors 2.7.1.3
Unless otherwise specified, thrust bearings and radial bearings shall be fitted with bearing-
metal temperature sensors installed in accordance with API std 670
API 670 specifies 75% location for sensors in tilt pad bearings
6.1.8.1.2 (tilting pad journal bearings): Either one sensor at 50/75 or two sensors at 25/75
and 75/75
6.1.8.2.2 (tilting pad thrust bearings): Temperature sensor at 75/75 location (75% radially and
75% circumferentially)

8
 Ettles et al., 2003 (PTFE vs. babbitt)
Bearing 1: 8-pad TPT, 464 mm OD; Bearing 2: 8-pad TPT, 912 mm OD,
PTFE/OVA thickness: 5/40 mm spring supported; PTFE/OVA
Test conditions thickness: 2/38.1 mm
Up to 10.2 MPa and 41 m/s Test conditions
ISO VG32 oil, flooded lubrication Up to 10 MPa and 28 m/s
Temperature ISO VG32 oil, flooded lubrication
Measure lining temperature Temperature
thermocouples (TCs) in PTFE 3 Measure pad metal temperature
mm below the surface below bond line
Later on, measure fluid film Results
temperature using hole in pad
surface
Results
No significant film T difference
between PTFE and babbitt
Higher power loss with PTFE

Ettles, C.M., et al. "Test results for PTFE-faced thrust pads, with direct comparison against Babbitt-faced pads and
correlation with analysis." Journal of Tribology 125.4 (2003): 814-823.

9
 Glavatskih, 2003: PTFE vs. Babbitt 2.0 MPa. T0 -fluid film T at the bottom of PTFE
pad

Bearing
6-pad equalized TPT, babbitt and PTFE (15% glass fiber) lined
pads, 228.6 mm OD
PTFE thickness: 1.5 mm
Test conditions
Up to 2 MPa., 15003000 rpm
ISO VG68 oil, flooded lubrication
Temperature
Pad T75/75 and collar T75 T at 3000 rpm
Measure metal temperature below bond line TCs 4 mm
below the PTFE surface; 3 mm below babbitt surface
Measure collar temperature
Results
1.5 mm thick PTFE layer leads to thermal insulation up to 23C
Collar T similar for both bearings; T_PTFE slightly higher than
T_babbitt
PTFE leads to up to 8% power loss reduction
Glavatskih, S.B. Evaluating thermal performance of a PTFE-faced tilting pad thrust
bearing. ASME. J. Tribol. 125.2 (2003):319-324. doi:10.1115/1.1506329.

10
Glavatskih, 2004: PTFE vs. Babbitt
Same bearing as in Glasvatskih, 2003 Babbitt Bearing

Present a new method: Measure fluid film temperature via

hole in pad surface and with bypassing hole
. compared to conventional industrial methods of
temperature monitoring, provided higher sensitivity to oil
film temperature in both steady state and transient
operating conditions
Existing methods of temperature measurement are
inadequate if applied to PTFE-faced bearings

Glavatskih, S.B. "A method of temperature monitoring in fluid film bearings." PTFE Bearing
Tribology International 37.2 (2004): 143-148.

11
 Glavatskih, 2004: Transient Plot for PTFE and Babbitt Bearing

T_H (fluid film) and T75/75 (metal) generally follow the same trend

Fluid film
Fluid film T in hole
T in hole
Metal T

Metal T

Babbitt PTFE

12
Henssler et al., 2015: PEEK with 50% Carbon Fiber

Solid PEEK (50% carbon fiber) pads

Journal bearing
5-pad TPJ, flooded water lubrication
15006000 rpm, 0.5 MPa
Thrust bearing
8-pad TPT, flooded water/glycol lubrication
15006000 rpm with 500 rpm step, Up to 3.8 MPa
Measure fluid film temperature via hole in pad surface,
75/75
Results: pass the test

Henssler, Dieter, et al. Qualification and optimization of Solid Polymer Tilting Pad
Bearing for Subsea Pump Application. 44th Turbomachinery Symposia, 2015.

13
 Sumi et al., 2014: PEEK vs. babbitt PEEK Long-term Test
Bearing Bearing used in MHI internal plant since 2007
14-pad TPT, 727 mm OD, PEEK lined (3 mm) and 10-pad TPT, 553 mm OD; PEEK lined (3 mm)
babbitt lined Test Conditions
Test conditions 3600 rpm, 0.8 MPa,
12 MPa (steady), 20 MPa (4 seconds); 3600 rpm Temperature
Measure fluid film temperature via hole(?) in pad Metal temperature
surface (near pad surface) Photos also suggest fluid film temperature (hole
Results in pad surface)
No damage after test Results
Compared to babbitt lined bearing, no significant 631 start-up/shutdown, total 20,462 hours
temperature or power loss difference Surface looks good

Sumi, Yuki, et al. "Development of thrust bearings with high specific load."
ASME Turbo Expo 2014: Turbine Technical Conference and Exposition.
American Society of Mechanical Engineers, 2014.

14
Zhou et al., 2015: PEEK
Bearing Results
8-pad 60% offset self-equalized TPT, PEEK lined, PEEK lined thrust bearings can operate at higher
279 mm OD bearing unit loads than babbitt lined bearings
Test conditions PEEK lined thrust bearings can be designed up to
ISO VG32 oil, Directed Lubrication, 8.0 MPa for modern turbomachinerys demanding
Load up to 16.2 MPa at 6000 and 11,000 rpm load and speed requirements
Performance study at 100013,000 rpm, 0.696.9 MPa, max No significant power loss difference between PEEK
147 m/s and babbitt (1-6%)
Temperature Observed small range of temperature variation with
PEEK lined pads
Metal temperature, TC at pivot location below bond line
Recommend PEEK for high speed/high load
applications when babbitt cannot meet the need

Zhou, Jie, et al. "Experimental

Performance Study of a High Speed
Oil Lubricated Polymer Thrust
Bearing." Lubricants 3.1 (2015): 3-
13.
Zhou, Jie, et al. " Performance of a
PEEK-Lined Tilt Pad Thrust Bearing
at High Speeds with Oil
Lubrication." 14th EDF/Pprime
workshop 2015.

15 15
Literature Review Summary Temperature Measurement Method

Material Max Max Pad / Pad / Fluid film Fluid film Fluid film /
MPa m/s metal lining /hole /hole with flush with
bypass flow surface
Garner babbitt x x
API babbitt x
Ettles 1 PTFE 10.2 41 x x
Ettles 2 PTFE 10 28 x
Glavatskih PTFE & babbitt 2 28 x x
Henssler PEEK 3.8 72 x
Sumi PEEK & babbitt 16, 20 117 x (?)
Sumi (LT) PEEK 0.8 83 x? x
Zhou PEEK & babbitt 16.2 147 x

Babbitt bearing temperature monitoring:

Pad metal temperature is industrial standard practice; lab tests also use fluid film temperature
Polymer bearing temperature monitoring:
Measuring pad fluid film temperature deviates from standard practice
Measuring temperature using fluid film with instrument flush to bearing surface has not been published
No published data relating temperature to bearing distress

16
The Current Study

Present fluid film temperature in PEEK bearings using sensor flush

with pad surface
Present fluid film temperature in PEEK bearings using hole in pad
surface located below surface
Study options for indicating bearing distress

17
Test Rig

750 kW total power

ISO VG32 oil
5678 L reservoir
1136 lpm oil pump
Individual oil control to all bearings

18
 Test Trial 1:
Measuring Pad Fluid Film Temperature Via
Sensor Flush with Pad Surface

19
PEEK Lined Pocket Feed TPT

Pocket Feed TPT with 8 PEEK lined steel

pads
Bearing OD 279 mm
4 pads with TC flush with pad surface,
measuring fluid film T
4 pads with TC embedded in pad backing
material, measuring pad metal T

20
Temperature During Performance Test

100

Steady state test from 1000 90

to 13,000 rpm

Temperature rise over inlet temperature (F)

80

Both fluid film T and metal 70

Fluid Film T at 13,000 rpm
T changed with load and 60
Metal T at 13,000 rpm

speed, as expected 50

40

30
Fluid Film T at 1000 rpm
20
Metal T at 1000 rpm
10

0
0% 20% 40% 60% 80% 100% 120% 140% 160%
Bearing Load Factor (%)

PEEK lined pads

21
Performance Test: 13,000 rpm

100 200%

90 180%
Temperature rise over inlet temperature (F)

80 160%

70 140%
Fluid Film, 75/75, Pad 1

Bearing Load Factor (%)

60 120%
Bearing Load
50 100%

40 80%

Metal, 75/75, Pad 7

30 60%

20 40%

10 20%

0 0%
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Elapsed Time (seconds)

PEEK lined pads

22
Performance Test: 1000 rpm

45 180%
45 180%

Bearing Load Factor (%)

40 160%
Temperature Rise (F)

40 35 140% 160%
30 120%
25 100%
35 140%
Temperature rise over inlet temperature (F)

20 80%
15 60%
30 10 40% 120%

Bearing Load Factor (%)

5 20%
0 0%
25 2300 2400 2500 2600 100%
Elapse Time (second)

20 80%

15 60%
Fluid Film T, 75/75, Pad 1 Bearing Load

10 40%

5 20%
Metal T , 75/75, Pad 7
0 0%
0 500 1000 1500 2000 2500 3000
Elapse Time (second)

23
Temperature During Ultimate Load Test

300

Load factor (%) and Temperature rise over inlet temperature (F)
250
Load

200

150 Fluid Film T, 75/75, Pad 3

100

Fluid Film T, 75/75, Pad 6

50

Metal T, 75/75, Pad 2

0
1000 1020 1040 1060 1080 1100 1120 1140 1160 1180 1200
Elapsed Time (Seconds)

24
Summary: Test Trial 1

Temperature Monitoring
Both fluid film temperature and metal temperature tracked the gradual change of bearing
load and speed
Fluid film temperature (flush with pad surface) swiftly tracked the sudden load change (in
1000 rpm test)

Distress Indication
Caution advised if planning to use fluid film sensor flush w/ pad surface under very high load
Capability of metal temperature of PEEK lined bearing to indicate stress
Film temperature sensor resulted in test stopping

25
 Test Trial 2:
Measuring Pad Fluid Film Temperature Via
Hole in Pad Surface

26
PEEK Lined CQDL TPT with Hole

CQDL (Self-Equalizing Directed Lubrication) TPT with

8 PEEK lined steel pads
4 pads with TC measuring metal T
1 pad with 75/75 TC measuring fluid film T with hole
in pad surface
11,000 rpm

27
Performance Test: 11,000 rpm

90

80
Temperature rise over inlet temperature (F)

70

60 Fluid film (hole) T

50

40

30
Metal T
20

10

0
0% 20% 40% 60% 80% 100% 120% 140% 160% 180% 200%
Bearing Load Factor (%)

28
Performance Test: Transient Change in Oil Inlet Temperature During Warm-up

CQDL Pad Temperature at 8400 rpm & 26% Load

949
886 Fluid film (hole) T, 75/75

888 960
Metal T, 75/75

933
Inlet T

869

500 600 700 800 900 1000 1100 1200 1300 1400 1500

29
Summary: Test Trial 2

Temperature Monitoring
Both fluid film(hole) temperature and metal temperature tracked the gradual change of
bearing load
Fluid film (hole) temperature had a shorter response time than metal temperature, as
expected
Distress Indication
Caution advised if instrument exposed to fluid film pressure
Not tested yet (prior to annual meeting)

30
 Test Trial 3:
Bearing Distress Indication

31
Distress Indication

Using pad fluid film temperature to indicate bearing distress has not been demonstrated
Fluid film (flush with pad surface):
TC localized, misleading temperature reading; TC inaccurately indicated unacceptable
temperature change
Measuring fluid film (flush surface) temperature not a reliable solution for high load
application
Fluid film (hole):
Initial trial sensor unreliable
No test data indicating bearing distress
Next : Distress indication using pad metal temperature
Ultimate load test of CQDL TPT with 8 PEEK lined steel pads
4 pads with TC in metal only
Transient date from ultimate load test at 6000 and 11,000 rpm

32
PEEK Lined CQDL TPT: Ultimate Load Test at 6000 rpm

60 450%

55
400%
Temperature rise over inlet temperature(F)

Bearing Load Factor (%)

50
Load
350%

45

300%
Metal T, 75/75
40

250%
35

30 200%
6000 8000 10000 12000 14000 16000 18000
Elapsed Time (seconds)

33
PEEK Lined CQDL TPT: Ultimate Load Test at 11,000 rpm

PEEK lined pad metal temperature can track

60 400%
change of operating conditions, as in Pocket
375%
Feed TPT
55 350% Metal temperature indicated distress and the test
Load 325% rig was shut down to prevent rig damage

Bearing Load Factor (%)

50 300%
Temperature (F)

275%

45 250%
Metal T rise over inlet, 75/75
225%

40 200%

Inlet T trend 175%

35 150%

125%

30 100%
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000
Elapsed Time (seconds)

34
Summary and Conclusions
Polymer pad temperature measurement:
Two options: material temperature and fluid film temperature
Five methods: lining material, metal backing material, fluid film (flush with pad), fluid film
(hole), and fluid film (hole) with bypass flow
Both material temperature and fluid film temperature can be used to monitor PEEK pads and
track gradual change of operating condition, based on published test data
Fluid film (flush with pad surface) method offer fast response, but not suggest for very high
load/high speed application.
Fluid film (hole) method also has quick response. Distress indication to be validated via
additional testing
Pad metal temperature can indicate bearing distress, as validated by test
Recommendation
Industrial applications: pad metal temperature is a reliable method for bearing health
monitoring and an indication of bearing distress
Lab testing: combination of metal temperature method (ultimate load) and fluid film
temperature method (fast response)
35