Abstract

AU: Furrer RE

Date: 06/28/2007

Job Director: Shi X

SUBJ: Performance & Endurance Evaluation of the CPA (Sauer-Danfoss) Gear

Pump Series

Abs: A set of Sauer-Danfoss gear pumps was sent from H&HS to TC-G for product
evaluation. A test matrix was developed and two pumps were tested per this
regime (113cc/rev & 81cc/rev). A noise/sound evaluation was then done on one
pump (113cc/rev; CPA1007 #1). These results are contained REDI Document
No. DN1052829, 2007. A cyclic endurance test was run on this same pump; the
intention was to endurance test for 1 million cycles. However, a cracked cover
developed and terminated testing at 310,000 cycles. Following this event,
another Sauer-Danfoss pump of the same displacement and configuration
(CPA1007 #2) was performance tested and then endurance tested, using the
same parameters. The outlet port threads, on the flange port adaptor, fatigued
and failed at 750,000 cycles. Also, a similar crack to the first failure was found in
the cover of this unit. It was not determined if this crack had not broken through
to the exterior of the part, due to the condition of the facility after the outlet flange
failure. Both pumps endurance tested had material failures that led to test
termination.

Keys: Gear Pump

Sauer-Danfoss
REDI Report DN1052829

Retention: Indefinite

NOTE:

Caterpillar Confidential: Yellow 1 For Internal Use Only

 Technical Document
Date: 06/28/2007

Author: Furrer RE

Charge No: 11-06910

Test Period: 08/30/2006 to 01/09/2007

Source Div: 881-01, TC-G

SUBJ: Performance & Endurance Evaluation of the CPA (Sauer-Danfoss) Gear

Pump Series

Introduction:

A set of gear pumps was provided for evaluation at TCG. The primary objectives were to
determine the key operating efficiencies (volumetric, mechanical, and overall) as well as the
sound/noise characteristics and the endurance life of the pump. This information was to be used
as part of product review for purchase, or manufacture.
Objectives:

1. Discover & report any design related issues encountered that require discussions with
Sauer-Danfoss.
2. Determine the overall operating efficiency for this pump. This would also require
determining the other primary efficiencies (volumetric and mechanical) as part of the
process.
3. Determine the noise characteristics for this pump. This work was done by Yuanhong
Guan and is covered by REDI Report DN1052829, 2007.
4. Determine if the pump could withstand 1 million endurance cycles at 22,700 kPa.

Conclusions:

1. The load seal design arrangement would seem to present several opportunities for
improvement. Also, the pump assembly bolting arrangement would also present
opportunities for improvement.
2. The average efficiencies, at 93°C fluid temperature, were actually very good for a gear
pump product. Overall efficiencies averaged from 60% (1000rpm/18000kPa) to 80%
(2000rpm/24,820kPa). Drawing specific conclusions relative to efficiency are difficult
due to the issues surrounding the load seal extrusion seen. The results provided above
appear to be higher than what experience would indicate to be normally expected for a
gear pump. All instrument calibration was checked before, and after testing, and found
to be within specification limits. Future test data used for comparison to this data, on this
same series of pumps, should keep this information in mind.
3. Noise/Sound work was conducted by Yuanhong Guan and the finding are reported in
another report (REF: REDI Report DN1052829, 2007)
4. Neither of the pumps that were endurance tested survived the prescribed 1 million
cycles. The endurance results were as follows:

Caterpillar Confidential: Yellow 2 For Internal Use Only

 Furrer RE
06/28/2007

a. CPA1007 (#1)- Completed 310,000 cycles at 21,500 kPa, at which point the cover
developed a crack that permitted external leakage. Testing was then terminated on
this pump
b. CPA1007 (#2)- Completed 750,000 cycles at 21,500 kPa, at which point the cover
flange threads fatigued and cover developed a crack. Testing was then terminated
on this pump.

Recommendations:

1. Discussions should be held on how the seal arrangement, load seal assembly, and bolt
assembly methodology, can be improved. Also, suggest that consideration be given to a
fixed low speed (600rpm) /high load pressure (24820kPa) bearing endurance be
conducted following a successful cyclic endurance test. (These issues have been
discussed, and a molded load seal design is now being considered. Also, validation
testing is being planned that will encompass the low speed bearing testing.)
2. Recommendations on improving the pump efficiency are hard to determine at this point.
Due to the issues surrounding the load seal, it makes evaluating the mechanical
efficiency results especially difficult. (Since this issue is included in the load seal design,
a new design seal and validation testing will address this issue also.)
3. Sound/Noise recommendations are deferred to Yuanhong Guan and his report.
4. Suggest consideration be given to changing the flange and cover material to iron (of
appropriate grade and section) in place to the aluminum that is currently in use. The
cover cracks and thread failures seen on the endurance tests would be eliminated. (This
has been discussed, and changing to an iron material for the components mentioned, is
planned when pump is resourced to China. Validation of this material is also being
planned as part of the overall plan.)

Significant Results:

1. There was an issue with stability when pump was run under load speed (below
1500rpm) and high load pressure (above 15000kPa) conditions.
2. The average overall efficiencies, at 93°C fluid temperature, were as follows:

CPA1007 (#1)
1000rpm/18000kPa 2000rpm/24,820kPa
a. Volumetric 63.9% 83.0%
b. Mechanical 94.5% 96.6%
c. Overall 60.4% 80.2%

CPA1007 (#2)
1000rpm/18000kPa 2000rpm/24,820kPa
d. Volumetric 85.3% 88.4%
e. Mechanical 96.9% 95.6%
f. Overall 82.6% 84.6%

Caterpillar Confidential: Yellow 3 For Internal Use Only

 Furrer RE
06/28/2007

CPA1025
1000rpm/18000kPa 2000rpm/24,820kPa
g. Volumetric 69.3% 81.0%
h. Mechanical 97.3% 89.0%
r. Overall 67.4% 72.0%
3. The load seal and load seal back-up design appears to have been a contributor the
instability at low speed/high load situation. During teardown inspection, it was found that
loading was not constrained to the normal loading area, but across the entire face of the
load plate. This was thought to be a result of the load seal extruding underneath the
plate. The purpose of the load seal is to confine pressure to a design limited area, and
when it fails, this area is greatly increased. The result is overloading of the load plate on
the internal pump gears.
4. In all of the pumps examined, the load seal had extruded. This situation appeared to be
related to the backing material and the assembly method used.
5. Also noticed in all pumps were assembly paint marks on all main bolts. This indicates
that there may have been concern with making sure that all bolts had proper torque
applied. This method/design of providing clamp load to the pump could present
assembly issues in making sure that adequate torque is applied to the individual
fasteners. This is a concern when this methodology is still being used on a mature
product.
6. Significant low speed testing could not be accomplished, to determine bearing life, due
to the load stability issues mentioned.
7. The assembly of the pump was difficult due to the clamp bolt arrangement. A wrench
and socket are required to provide assembly torque.
8. Inconsistent wear seen on gear teeth of CPA1007 (#1) after endurance test teardown
inspection. This appears to be from a geometry error in gear manufacturing.
9. High loading of shaft seal onto output drive shaft seen after teardown inspection of
CPA1007 (#1). No leakage noticed at the shaft/seal interface, but the endurance test
only ran 31% of required cycles.
10. Condition of shaft journals, and bearings, on CPA1007 (#2) were in surprising good
condition, even though exposed to a water/oil fluid mixture, high temperature, and
apparently ran for several minutes without fluid in the reservoir. Photos are attached in
report body.

Randall E. Furrer
881-01, Hydraulic Component Development, TCG
(309) 578-8559

Caterpillar Confidential: Yellow 4 For Internal Use Only

 Furrer RE
06/28/2007

Discussion:

A set of pumps was delivered to TCG to evaluate as part of product review for a potential
manufacturing venture with Sauer Danfoss. This evaluation was to include performance testing,
endurance testing and an evaluation of the overall worthiness of the pump package. Originally,
this testing was designated to run on TCG Facility 202-2E (Figure 13/IM&TE), which is a constant
200HP dynamometer. During the course of running the first pump, CPA1007, it was determined
that torque limitations existed with this facility that didn’t allow completion of this work. So, a larger
facility, 208-1S, was selected and the pump was mounted on it per the schematic below (Figure
1). Instruments were installed to measure the critical performance parameters per the attached
IM&TE listing below for TCG 208-1S (Figure 2). Photographs were taken of the pump installed on
both facilities (TCG 202-2E & 208-1S; Figures 3-5) and are also attached below.

Performance testing was conducted after this without any interruptions. Data has been tabled and
is attached below (Figures 6-7). Endurance testing was done, on this pump, only on TCG Facility
208-1S. A trace of the pressure, for the endurance test, is also attached below (Figure 8). This
pump survived at total of 310,000 cycles before a crack in the aluminum cover created an external
leak, and the test was terminated. The cover was Xyglo checked for crack examination. This
method shows the location and the extent of the crack involved, but the best results are usually
seen with ferrous materials. This cover is aluminum, so the results were not optimal. Photos were
taken after this work to document the crack/cover condition, and are included below. (Figures 14-
15) Also, photographs taken of the other major internal components, during the teardown
inspection, are included (Figures 16-35).

Another pump, CPA1025, was also installed on TCG Facility 202-2E. This pump’s performance
testing was done almost exclusively on this facility. However, the low speed (below 1500rpm) and
high load pressure (above 18000kPa) conditions would cause the dynamometer to stall. This also
indicated an overdraw torque condition, which this pump shouldn’t have been able to do. After
seeing the load seal failures that occurred in the previous test pump, it was decided that the pump
was overloaded and requiring greater than expected drive torque. It is for this reason that this
pump was also moved to TCG Facility 208-1S (600HP) and was retested at the low speed/high
load pressure conditions. The data that was taken, while installed on 202-2E, is tabled below.
(Figure 12)

After failing the endurance unit, CPA1007 (#1), it was decided that another pump would be
endurance tested. The pump selected was the same as the first endurance unit, CPA1007, and
will be designated in this report as CPA1007 (#2). This unit was performance tested prior to being
endurance tested also, and the results are below (Figure 9-10). Upon completion of the
performance work, this pump was endurance tested using the same test regime as the first. The
representative endurance test cycle was captured and is below (Figure 11). This second pump
was to be run for 1 million cycles, but it only completed 750,000 cycles. At that point, the outlet
port threads fatigued and allowed the outlet port flange to separate. This separation provided an
opening for the test fluid (SAE10W; CT5347-USED IN ALL TESTING DONE) to escape, and
eventually drain the reservoir before the facility could be shutdown. It should be mentioned that
the draining of the reservoir created high internal temperatures, and due to the facility sprinkler
system being set-off, water ingestion into the hydraulic system was the eventual result. This
combination of high temperature, and an oil/water mix, is usually fatal for pump components.

Caterpillar Confidential: Yellow 5 For Internal Use Only

 Furrer RE
06/28/2007

However, the bearings and shafts of this pump looked to be in reasonable condition. Though
blued, due to the high temperature, the shafts were not badly scored. The bearings still had the
original “orange-peel” texture to the inside diameter surface. They had also seen high
temperature, but had not degraded to any great amount visually. Due to the catastrophic nature of
the failure, a major external leakage created by fatigued threads, the testing was terminated. Most
of the internal seals were melted, due to the high internal temperatures mentioned, seriously
damaged, or compromised. Photos were not taken of the internal seals for this pump (CPA1007-
#2); most of the major components were photographed and are below (Figures 36-57).

Another item that should be discussed is the cover of the second endurance unit (CPA1007 #2).
Even though the outlet port fatigue was the root cause of the failure, it was decided to die
penetrant check this cover to determine if any cracks were present. The results were surprising in
that a crack was occurring, and it was located in the same location as the previous failure. Also,
the length of the crack appeared to be approximately the same length as the crack in the previous
failure. Photos were taken and are installed below, both by with this cover by itself, and with the
previous cover for comparison (Figures 40-44).

Caterpillar Confidential: Yellow 6 For Internal Use Only

 Furrer RE
06/28/2007

Figure 1. Circuit Schematic Used for Testing Sauer-Danfoss CPA Pumps

Figure 2. IM&TE Instrument Log (Set-up for Facility 208-1S)

Caterpillar Confidential: Yellow 7 For Internal Use Only

 Furrer RE
06/28/2007

Figures 3-4 CPA 1007 Pump Installation on Facility

Solenoid Operated Load

Valve – Same Valve
Used Regardless of
Facility

Figure 3- CPA1007 (#1)

Installed on TCG 202-2E Figure 4

Figure 5. CPA 1007 Installation on TCG Facility 208-1S (Shown w/Sound Measuring Instruments)

Figure 5

Caterpillar Confidential: Yellow 8 For Internal Use Only

 Furrer RE
06/28/2007

Figure 6. Output Flow versus Speed Data (Derived Displacement)- CPA1007 (#1)

Figure 6

Figure 7. Performance Data for CPA1007 (#1)

Figure 7

Caterpillar Confidential: Yellow 9 For Internal Use Only

 Furrer RE
06/28/2007

Figure 8. CPA1007 (#1)- Endurance Test Trace

Note: Drive speed was set at 2700rpm at “No-Load” and the

“automatic” feature was engaged to actuate the solenoid operated load
valve. This, in turn, created the load pressure wave shown.
Figure 8

Caterpillar Confidential: Yellow 10 For Internal Use Only

 Furrer RE
06/28/2007

Figure 9. Output Flow versus Speed Data (Derived Displacement)- CPA1007 (#2)

Figure 9

Figure 10. Performance Data for CPA1007 (#2)

Figure 10

Caterpillar Confidential: Yellow 11 For Internal Use Only

 Furrer RE
06/28/2007

Figure 11. CPA1007 (#2)- Endurance Test Trace

Figure 11

Caterpillar Confidential: Yellow 12 For Internal Use Only

 Furrer RE
06/28/2007

Figure 12. CPA1025 Performance Data (Taken on TCG 202-2E)

Figure 12

Figure 13. IM&TE for CPA1025 (TCG 202-2E)

Figure 13

Caterpillar Confidential: Yellow 13 For Internal Use Only

 Furrer RE
06/28/2007

Figures 14-15. CPA1007 (#1) -Photograph of Cover Crack (Xyglo Results)- After 310,000 cycles

Figure 14.

After Xyglo inspection a crack could

be seen with the naked eye along
this line. “Green Haze” area created
by penetrate bleeding out of the
deep crack into casting pores.

Figure 15.

Caterpillar Confidential: Yellow 14 For Internal Use Only

 Furrer RE
06/28/2007

Figure 16. CPA 1007 (#1) –“As Removed” from Facility 208-1S after Endurance Testing

Note the indexing paint

marks that were applied
to the clamp bolts. Also,
note that it is impossible
to use anything but a
wrench to remove/access
these clamp bolts, due to
the position of the
mounting flange/casting.

Figure 16.

Figure 17. Photograph of Pump Disassembly –As the Cover & Gear plate is Lifted

Notice that a “gap” exists

between load plate and the
front flange. Suspicious that
load seal may be pinched.

Figure 17.

Caterpillar Confidential: Yellow 15 For Internal Use Only

 Furrer RE
06/28/2007

Figure 18. Photo of Cover Being Lifted During Disassembly

Notice the same

approximate “gap” on this
loadplate also. Also, would
be an indicator of a pinched
load seal.

Figure 18.

Figure 19. Loadplate Condition – Cover Side- At Disassembly/After 310,000 cycles

Cover Side Loadplate- Notice the witness

marks from the gears is quite widespread
and heavy. It is not completely confined to
the “loaded” area as would be expected
by a healthy & active load seal. This
appears to be overloaded.
Figure 19.

Caterpillar Confidential: Yellow 16 For Internal Use Only

 Furrer RE
06/28/2007

Figure 20-21. CPA1007 (#1)-Additional Photos of the Cover Side Loadplate – After Endurance

Figure 20. Figure 21.

Figure 22. CPA1007 (#1)-Cover Side Plate Close up- After Endurance

Inlet Core

Figure 22. Pressure Core

Caterpillar Confidential: Yellow 17 For Internal Use Only

 Furrer RE
06/28/2007

Figure 23. CPA1007 (#1)- Extruded Load Seal/Cover Area- After Endurance

Load Seal Back-Up starting to

extrude, which lifts Loadplate;
load seal appears to be short and
not long enough.

Figure 23.

Figure 24. CPA1007 (#1)- Loadplate/Cover Side- After Endurance

Figure 24. Loadplate appears to be overloaded and loaded

area appears to be too large. Also, material
appears to have material being removed at top
of part from overloading.

Caterpillar Confidential: Yellow 18 For Internal Use Only

 Furrer RE
06/28/2007

Figure 25. CPA1007 (#1) – Gear Faces After Endurance

Gear faces show

signs of overloading
from bearing against
Figure 25. Loadplate.

Figure 26. CPA1007 (#1)-Shaft Journals-After Endurance

Shaft Journals from cover end of

gearplate. Notice that there is
some apparent minute scratches
on idler journal.

Figure 26.

Caterpillar Confidential: Yellow 19 For Internal Use Only

 Furrer RE
06/28/2007

Figure 27. CPA1007 (#1)-Flange Side Loadplate- After Endurance Test

Notice that the Loadplate appears to be

overloaded. There is also material being
removed from Loadplate at the outlet
port side. Finally, note that the loaded
area appears to be oversized. This is
Figure 27. probably due to the extruded load seal
condition found.

Figure 28. CPA1007 (#1) - Flange- After Endurance

Flange- Notice 2 items:

1. Load Seal Back-up is extruded on
one side.
2. Load Seal appears to be too short
and will not stay in groove.

-Not Highlighting/Obvious in Photo

Figure 28.

Caterpillar Confidential: Yellow 20 For Internal Use Only

 Furrer RE
06/28/2007

Figure 29. CPA1007 (#1) - Load Seal Back-Up- After Endurance (Close-Up)

Figure 29.

Figure 30. CPA1007 (#1)- Load Seal Back-Up- After Endurance (Close-up & Removed from
Groove)

Figure 30.

Caterpillar Confidential: Yellow 21 For Internal Use Only

 Furrer RE
06/28/2007

Figure 31. CPA1007 (#1) - Gears- After Endurance (Close-up)

Unusual/uneven wear pattern seen

on flanks of gears; suspicious that
there is a geometry error in
manufacturing creating this wear.

Figure 31.

Caterpillar Confidential: Yellow 22 For Internal Use Only

 Furrer RE
06/28/2007

Figure 32. CPA1007 (#1)- Gear – After Endurance

Notice in this photo:

1. Alternating nature of
unusual/uneven wear on
gear flank
2. Striations on the shaft
journal; indicates some
Figure 32. wear.

Figure 33-34. CPA1007 (#1) -Gearplate Photos- After Endurance

Figure 33. Figure 34.

Both photos are showing the inlet side of the gearplate, and the
contact area with the gears. Contact appeared heaviest in this local
but wasn’t usual for pumps of this type from this observer’s POV.

Caterpillar Confidential: Yellow 23 For Internal Use Only

 Furrer RE
06/28/2007

Figure 35. CPA1007-Post Endurance- Close-up Photo Showing Shaft Seal Bearing Area

Shaft Seal Area on Driveshaft- Lips

appears to have been bearing hard
in contact area. It is unknown what
the darkened area between seals
exist.

Seal Boundary Area – Seal Boundary Area –

Inboard Seal Outboard Seal

Figure 35.

Caterpillar Confidential: Yellow 24 For Internal Use Only

 Furrer RE
06/28/2007

CPA1007 (#2) PHOTOS

Figure 36. TCG Facility 208-1S After Failure

Sprinklers were set off

due to the oil jet impact
ceiling at attachment point
of sprinkler. Water
discharge from sprinklers,
along with oil from the test
pump, covered entire
facility. Notice rust on
facility bed.

Figure 36.

Caterpillar Confidential: Yellow 25 For Internal Use Only

 Furrer RE
06/28/2007

Figure 37. Installed Pump on TCG Facility 208-1S –Photo Showing Fatigued Outlet Port
Threads/Extrusion Gap

Gap Created by Fatigued Threads Allowed

Movement of Outlet Flange away from
Sealing Surface on Pump

Figure 37.

Figures 38-39. Photos Showing Close-up of Extrusion Gap Created by Fatigued Outlet Flange
Threads

Figure 38 Figure 39

Caterpillar Confidential: Yellow 26 For Internal Use Only

 Furrer RE
06/28/2007

Figure 40. Cover Crack Found on CPA1007 (Unit #2)

Figure 40

This crack was found upon disassembly of pump. Though the primary failure of the pump was the
outlet port thread fatigue, this cracked also developed. It would have also lead to an eventual
failure. The thinking is that this crack was subsurface during the operation of the pump, and then
opened when the threads failed. The thread failure, and the corresponding extrusion gap, through
which the facility reservoir was drained, resulted in very high internal temperatures. This additional
increase in temperature is thought to be what caused the crack to open at this point.

Caterpillar Confidential: Yellow 27 For Internal Use Only

 Furrer RE
06/28/2007

Figures 41-42. Failed Covers/Both Units –Cracks Shown for Comparison

CPA1007 –
#2 Endurance
Unit
Figure 41

CPA1007 –
#1 Endurance
Unit

Figure 42

Caterpillar Confidential: Yellow 28 For Internal Use Only

 Furrer RE
06/28/2007

Figures 43-44. Failed Covers/Both Units –Covers Rotated for Crack Comparison

CPA1007 –
#1 Endurance
Unit

Figure 43 Photos were rotated to

place the cracks seen in
both failed covers in the
same orientation. As can
been by these photos, the
cracks are approximately
the same length and in
the same location

CPA1007 –
#2 Endurance
Unit
Figure 44

Caterpillar Confidential: Yellow 29 For Internal Use Only

 Furrer RE
06/28/2007

Figures 45-46. Close-up Photos of Fatigued Outlet Port Threads

Both of these photos are

presented to show the top view
Figure 45
of the threads. The failed thread
material formed “thread
columns” and can clearly be
seen in these views.

Figure 46

Caterpillar Confidential: Yellow 30 For Internal Use Only

 Furrer RE
06/28/2007

Figures 47-48. Close-up Photos of Fatigued Outlet Port Threads

Thread material
columns were
moving out of the
cover, and the clamp
load originally
provided by the
flange bolts was
non-existent

Figure 47

Thread columns
separating from
main cover casting

Figure 48

Caterpillar Confidential: Yellow 31 For Internal Use Only

 Furrer RE
06/28/2007

Figures 49-50. CPA1007 (#2)- Load Plate & Shaft Journal Condition – After Endurance

Note that the surface

condition, though
completely blued from
high heat, did not have
material gouged from it.
A complete seizure did
not occur

Figure 49

It should be noticed
that the condition of
the shaft journals
was reasonably
good considering
the exposure to high
temperatures.

Figure 50

Caterpillar Confidential: Yellow 32 For Internal Use Only

 Furrer RE
06/28/2007

Figures 51-52. Internal Part Condition at Test Termination – CPA 1007 (#2)

Both of these photos show

the presence of water in the
oil. This resulted in the rust
Figure 51 seen. The water most likely
came from the facility
sprinklers partially filling the
reservoir after the outlet port
flange thread failure purged
the oil from the tank.

Figure 52

Caterpillar Confidential: Yellow 33 For Internal Use Only

 Furrer RE
06/28/2007

Figures 53-54. Internal Part Condition at Test Termination – CPA 1007 (#2)

Notice the surface condition

Figure 53 of the wear-plate and
bearings. They are blued
from high internal/fluid
temperatures, but there is
no gouging, or material
removed from either item.
This would be expected
from the lack of lubrication
seen.

Figure 54

Caterpillar Confidential: Yellow 34 For Internal Use Only

 Furrer RE
06/28/2007

Figures 55-56. Drive Gear Journal Condition at Test Termination - CPA 1007 (#2)

These photos show the

Figure 55 condition of the shaft
journals and the shaft seal
area. The shaft journals
appear to be in good
condition considering the
amount of exposure to
heat. The shaft seal area
is rough and has material
pulled from surface. This
is also consistent with
high heat exposure.

Figure 56

Caterpillar Confidential: Yellow 35 For Internal Use Only

 Furrer RE
06/28/2007

Figure 57. Gear Condition at Test Termination- CPA1007 (#2)

Figure 57

Caterpillar Confidential: Yellow 36 For Internal Use Only