The Society shall not be responsible for state-

ments or opinions advanced in papers or in dis-

cussion at meetings of the Society or of its

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/WA1965-GTPapers/80081/V001T01A003/2569129/v001t01a003-65-wa-gtp-9.pdf by guest on 14 January 2023

Divisions or Sections, or printed in its publications.
65-WA/GTP-9
Discussion is printed only if the paper is published
in an ASME journal.
$1.,50 PER COPY Released for general publication upon presentation
750 TO ASME MEMBERS Copyright © 1965 by ASME

Elimination of Turbine Erosion in the T56

Turboprop Engine
J. G. TOMLINSON
Supervisor, Combustion Design,
Allison Division, General Motors
Corporation, Indianapolis, Ind.

L. N. MONTGOMERY
Senior Project Engineer, Allison
Division, General Motors Corporation,
Indianapolis, Ind. Mem. ASME.

Turbine erosion was encountered in service operation with the T56 turboprop engine.
Combustor carboning was suspected to be the cause of the problem. No direct evi-
dence of carboning had been observed during development or in more than one
million hours of service operation. An accelerated investigation was initiated to estab-
lish a quantitative measure of carbon particles in the combustor exhaust gases. A
unique test method was developed to collect carbon particles on a component com-
bustor test rig. Evaluations were conducted with the carbon collector to determine the
effects of fuel nozzles, fuel type, operating conditions, and combustor configuration
on carbon particle output. The combustor configuration was found to be the most
important factor in the control of the problem. Combustor carboning was established
as the cause of turbine erosion and a combustion liner modification was developed
for service release. Service tests were conducted to correlate development test results
and to verify that turbine erosion was eliminated with the modified combustion lines.

Contributed by the Gas Turbine Power Division for presentation at the Winter Annual Meeting,
Chicago, Ill., November 7-11, 1965, of The American Society of Mechanical Engineers. Manu-
script received at ASME Headquarters, August 5, 1965.
Written discussion on this paper will be accepted up to December 13, 1965.
Copies will be available until September 1, 1966.

THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS, UNITED ENGINEERING CENTER, 345 EAST 47th STREET, NEW YORK, N.Y. 10017
 TOROUEMETER ASSEMBLY

AND TIE STRUT POWER SECTION

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/WA1965-GTPapers/80081/V001T01A003/2569129/v001t01a003-65-wa-gtp-9.pdf by guest on 14 January 2023

REDUCTION GEAR ASSEMBLY
ACCESSORIES DRIVE
HOUSING ASSEMBLY

Fig. 1 T56 engine

Elimination of Turbine Erosion in the T56 Turboprop Engine

J. G. TOMLINSON L. N. MONTGOMERY

The purpose of this paper is to present the eral more cases were reported and by inspection of
results of a development program conducted to elim- the parts from these engines the problem was de-
inate a turbine-erosion problem encountered in ser- fined sufficiently to begin a development program.
vice with the T56 turboprop engine. The eroded vanes were characteristically worn in
the band just downstream of the vanes, and the
BACKGROUND blades were eroded on the leading edge and knife
seals. Typical examples of erosion are shown in
The T56 engine consists of a gas-turbine pow- Figs.2 and 3. The erosion pattern in the turbine
er section connected to a reduction gear box and section appeared to have been caused by carbon
propeller. The general engine arrangement is produced in the combustor, although no carbon de-
shown in Fig.l. The power section is a single- posits had been observed in T56 combustors during
shaft unit consisting of a 14-stage axial-flow development or service operation totaling more
compressor, a can-annular combustion section hav- than one million hours on this model. It was sub-
ing six cans or combustion liners and a four-stage sequently determined that a minor change on pro-
axial-flow turbine. The T56-A-7 engine is rated duction combustion liners brought out a latent
at 4050-equivalent-shaft hp and operates on JP-4 combustor carboning problem which accelerated tur-
or JP-5 fuel. It is the powerplant for the C130-B bine erosion. The first positive evidence of com-
and -E aircraft, which are in worldwide use by bustor carboning was noted when an engine was shut
the Military Air Transport Service (MATS) of the down at altitude and deposits were found when the
United States Air Force and the United States engine was disassembled. The deposits were hard,
Coast Guard. A tanker version, the GV-1, is also amorphous and almost 1/2 in. in the largest dimen-
used for in-flight refueling of Navy and Marine sion, which was large enough to cause the observed
Corps aircraft. Other models of the T56 engine turbine damage. The flow path which carbon parti-
are in use on other military and civil aircraft. cles followed in the turbine and the location of
eroded areas is shown in Fig.4.
PROBLEM
DEVELOPMENT PROGRAM
In August 1963, two incidents of severe tur-
bine vane-band erosion on T56-A-7 engines were re- A program was established to develop a com-
ported from service activities. By November sev- bustion section modification to eliminate turbine

2
 Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/WA1965-GTPapers/80081/V001T01A003/2569129/v001t01a003-65-wa-gtp-9.pdf by guest on 14 January 2023
Pig. 4 Flow path of solid particles through the turbine

velopment for some time. This design had a dome-

shaped like half of a torus, and it was called the
toroidal-dome combustor. Development experience
included both test-stand and flight-service opera-
Fig. 2 Erosion pattern on turbine blade tion. The toroidal-dome combustor was consistent-
ly cleaner than standard dome designs from both a
smoke and carbon output standpoint, and performance
∎ REA' OF EROSION
,
results were as good as with standard combustors.
Arrangements were made with the using services to
conduct a service evaluation of this configuration
in the type of service where severe turbine ero-
sion was being encountered. Comparative results
were assured by utilizing a four-engine aircraft
with standard combustors installed in two engines
and the toroidal-dome combustors installed in the
other two engines. A test period of 400 hr of nor-
mal operation was established, long enough to pro-
duce measurable erosion of the vane bands (erosion
problems were significant at from 800 to 2000 hr
Fig.3 Erosion on turbine vane band
normal service) and short enough to obtain results
in two months if flights were expedited with a
minimum turn around time. Sufficient parts were
erosion in the T56 engine. To meet this objective, made available to conduct evaluations with MATS
more precise data were required on erosion rates (Air Force), Navy, and Marine Corps aircraft, and
as encountered in service, for both production programs were established with all three services.
style and experimental combustor designs. Also,
a laboA.tory program was required for the develOp, COMPONENT DEVELOPMENT
ment of new burner designs and a method of meas-
uring carbon output from them. A low smoke com- The problem of measuring combustor carbon out-
bustor design was available which had been in de-put was not straightforward, and test equipment

TABLE 1
Engine Average Vane Blade Leading- Shroud
Combustor Type Position Erosion Depth - In. Edge Condition Wear

Standard 1 0.028 1/32 reduction Moderate

of chord

Standard 2 0.017 Leading edge Slight

roughened

Toroidal Dome 3 0.007 No change Slight

Toroidal Dome 4 0.000 No change Slight

3
 400

a 300
S.
b11

c.

O. 200

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/WA1965-GTPapers/80081/V001T01A003/2569129/v001t01a003-65-wa-gtp-9.pdf by guest on 14 January 2023

O
2
100
C

lmprov
ORW
0
0
100 200 300 400
Fuel flow rate per liner, lb/hr ,

Fig.5 Carbon deposits on the dome of a stand- Fig. 8 Reduction in carbon emission with the improved
ard liner liner

Nozzle

Solid particles haust appearance were very clean as expected, and

Flow the test results made this combustor appear prom-
Combustor Exhaust
gas diverted
ising as a solution. However, retrofit to this
design would have required considerable modifica-
tion of the standard design. It was desired to
establish a simple field modification for the cur-
Test rig schematic rent service parts, if possible, to have a fix
Sample collector which could be immediately applied throughout the
Fig. 6 Flow paths of particles and gas through the carbon fleet without waiting for parts delivery as would
collector be required in the case of a new combustor design.
Further development effort was therefore expedited
with the standard combustor as a basic design.
development was required to obtain the desired re- It was learned early in the component devel-
sults. While this program was in process, compo- opment program that carbon deposits formed con-
nent combustor evaluations were conducted to de- sistently at the same locations on the liner dome.
termine the location of carbon accumulations in The locations for a given liner configuration were
the combustor and to observe the combustor exhaust found by borescope inspection after a quick shut-
through a view port for evidence of carbon output. down of the burner rig to simulate an engine feath-
The toroidal-dome combustion-liner design which er shutdown. A view of the deposits in the dome
was in service testing in the field was checked of a standard liner after a shutdown is shown in
on the view-port rig. The smoke density and ex- Fig.5. These deposits and the pattern of forma-

DOME AIR
INCREASED
EXTRA AIR STAGING
LOCATION OF ADDED TO PRIMARY ZONE
CARBON
DEPOSITS

Fig. 7 Design features of the improved combustion liner

 Table 2 Comparison of Carbon-Emission Rates --
Standard and Improved Liner

Turbine Carbon
1‘4 Inlet Emission
Combustor Altitude (ft) Temp (F) (gr/hr)
Standard 10,000 1560 109
20,000 1560 90

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/WA1965-GTPapers/80081/V001T01A003/2569129/v001t01a003-65-wa-gtp-9.pdf by guest on 14 January 2023

Modified 10,000 1560 7
IMPROVED DESIGN

20,000 1560 5

sults with combustor development results on the

component test rig, later developments made these
data extremely valuable in establishing a service
fix on the combustor. The service evaluations
which were in process with Navy and Marine Corps
aircraft with the toroidal-dome combustors had
Fig. 9 Carbon deposits collected been started at a later date and were not completed
at this time. However, the results obtained from
the completed programs were the same as the MATS
results and they proved the combustor-erosion re-
tion were studied to determine the mechanism of lation in varied service operations.
formation. The deposits were formed from liquid
fuel drops which impinged on the liner dome and CARBON-COLLECTOR TESTS
walls and cracked before they evaporated. The lo-
cations of deposition were limited to areas where The successful development of a carbon col-
an inadequate air supply was available to sweep lector for component combustor testing was com-
away the fuel droplets. It was necessary to in- pleted shortly after the first flight-service eval-
crease film air to these areas and increase the uation was finished. The collector worked by in-
total air supply to the primary zone. Combustor ertial separation and consisted of a sharp turn
modifications were evaluated which included in the combustor exhaust gas stream, which turned
changes in air delivery at these locations. The at approximately the same mean angle as the gas
N effect of the modifications on temperature-pattern flow in the turbine inlet vanes. The collector
stability, lean-extinction limits and ignition was therefore a reasonably close model of the en-
limits was checked during testing. Several com- gine at a particle concentrator. A diagram of
bustor changes were evaluated in this phase of the the gas and particle flow through the collector is
program. shown in Fig.6. At least 85 percent of the par-
ticles between 0.06-and 0.15-in. dia were collected
SERVICE EVALUATION RESULTS as determined by a functional check of the collec-
tor. The collection efficiency dropped rapidly
The first service evaluation to be completed for smaller particle sizes, but these were less
on the toroidal-dome combustors was on aircraft important to the immediate problem and could be
operated by MATS out of Charleston, South Carolina , ignored.
Air Force Base. All four turbine sections from Test results with this carbon collector show-
the evaluation aircraft were disassembled and the ed a carbon output from the standard service com-
amount of vane erosion was determined by measuring bustor of as much as 450 grains/hr at sea-level,
the penetration into the vane band where all cases take-off conditions. The carbon particles collec-
of vane erosion had been observed. The results of ted were hard and up to 1/2 in. in the largest
these measurements are shown in Table 1. These dimension, similar to the deposits which had been
service test results demonstrated that a complete observed in the service engine. Collections were
solution to the erosion problem was possible and made over the complete operating range of the en-
confirmed the relationship between the combustor gine to establish the carbon formation rates.
and the turbine-erosion problem. Although it was Carbon emission was found to be primarily a func-
not possible at this time to correlate these re- tion of fuel flow rate and only a small effect of

5
 Table 3 Summary of Service Testing
Number of
Evaluations Activity Location Aircraft Evaluation Hours
3 1608 ATW, MATS 0130 768
Charleston, S. C. Transport

1 VMGR 352, USMC GV-1 401

El Toro, California Tanker

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/WA1965-GTPapers/80081/V001T01A003/2569129/v001t01a003-65-wa-gtp-9.pdf by guest on 14 January 2023

1 NATC, USN P3A 536
Patuxent River, Patrol
Md. 17 05

air flow rate and pressure variation was observed. stalled on two engines of a C130 for service eval-
There was a significant difference in output be- uation. The normal use rate of the aircraft was
tween early and more recent production liners, accelerated where possible, and the evaluation was
because of a slight relocation of primary air completed at 389 hours flight time. No measurable
holes. There was also some variation between in- or visual erosion was found in the engines which
dividual liners and fuel nozzles of the same de- had the modified low carbon combustors installed.
sign. The toroidal-dome combustion liner which The baseline engine which had standard combustors
was already in service evaluation had a low rate, installed showed characteristic vane-band and
so there was a good correlation between erosion blade erosion in the turbine with vane-band pene-
in service operation and carbon-collection rates tration in excess of 0.075 in. on all the vane
determined on the component combustor rig. segments. It was concluded that the configuration
Modification and rig testing of the standard evaluated on this service test was the solution
liners was continued, and a field-type modification to the turbine erosion problem and the necessary
was developed which could be accomplished on parts steps were taken to modify the service liners to
already in service. This configuration had a car- this design.
bon-collection rate that was less than 10 percent The service evaluation phase of this program
of the rate from the standard combustor and in required more than 1700 hr of flight time on four
the same range as the output of the toroidal-dome aircraft operated by the using activities of the
combustor. In the modified design, an extra stage Air Force, Navy and Marine Corps. A summary of
of primary air was added downstream of the existing the flight evaluations is shown in Table 3.
hole pattern, and the flow of film-cooling air on
the liner dome was increased locally. The combus- CONCLUSIONS
tion liner changes are shown in Fig.7. To demon-
strate the effect of the modification, a set of 1 Combustor carboning was determined to be
service liners was obtained from an engine which the cause of severe turbine erosion in the T56
had experienced severe turbine erosion. These com- turboprop engine. The carbon deposits resulted
bustors and their fuel nozzles were run as receiv- from the impingement and cracking of liquid-fuel
ed on the collector rig, then reworked to the pro- droplets on the combustor wall.
posed configuration and run again on the collector
2 An inertial-type separator was developed
rig. The modification reduced the carbon-emission
to separate carbon particles from the combustor
rate by as much as 18 to 1 on these parts. The re-
exhaust gases at the high flow rates, temperatures,
sults of this test are shown in Fig.8. A visual
and pressures encountered in gas-turbine engines.
comparison of the carbon collected from one of the
Comparative measurements of combustor carbon out-
combustors before and after modification is shown
put with this device provided a good measure for
in Fig.9. A tabulation of comparative carbon-
the turbine-erosion problem.
collected data obtained from one of the test com-
bustion liners is shown in Table 2. 3 The carbon deposition was determined to be
primarily a function of fuel delivery rate. Com-
FINAL SERVICE EVALUATION bustor carbon output was reduced by a factor of
ten or more with a combustion liner modification
Two sets of the improved combustors were sup- which could be accomplished at the field mainten-
plied to MATS at Charleston where they were in- ante level. The change consisted of increasing

6
the flow of film-cooling air in the dome area ACKNOWLEDGMENTS
where the carbon deposits formed.
4 The success of this program was dependent Acknowledgment is made to Colonel E. B. Berry,
upon close cooperation between the using military Lt. Colonel J. K. Massie and Captain J. M. Snel-
services, contractor service representatives, and grove of the 1608 Air Transport Wing, MATS Charles-
the manufacturer's engineering staff. An impor- ton AFB; and Captain W. B. Blose of MATS Headquar-
tant requirement was service evaluation of pro- ters for their cooperation in conducting the ser-
posed design modifications and demonstration of vice evaluations which made this program a success.

Downloaded from http://asmedigitalcollection.asme.org/GT/proceedings-pdf/WA1965-GTPapers/80081/V001T01A003/2569129/v001t01a003-65-wa-gtp-9.pdf by guest on 14 January 2023

the final release. The field evaluations were com- Acknowledgment is also extended to other MATS,
pleted without hindrance to any aircraft operating Navy and Marine Corps personnel who aided in col-
sdhedules. lecting data and operation on this program.