1、05FTM16CH47D Engine Transmission InputPinion Seeded Fault Testby: J.P. Petrella, J.S. Kachelries and S.M. Holder, The BoeingCompany and T.E. Neupert, U.S. ArmyTECHNICAL PAPERAmerican Gear Manufacturers AssociationCH47D Engine Transmission Input Pinion SeededFault TestJoseph P. Petrella, James S. Kac
2、helries and Scott M. Holder, The BoeingCompany and Thomas E. Neupert, U.S. ArmyThe statements and opinions contained herein are those of the author and should not be construed as anofficial action or opinion of the American Gear Manufacturers Association.AbstractThis paper summarizes an Engine Trans
3、mission Input Pinion Seeded Fault Test that was accomplished as aportion of the validation process for the Boeing Rotorcraft Transmission Vibration Diagnostic System (TVDS)Analyzer. The test specimen was a high speed CH47D (Chinook) engine transmission input pinion with aknown defect (i.e., seeded f
4、ault) machined into a high stress area of a gear tooth root. The specimen wasoperated at “overload” conditions in order to initiate and propagate a fatigue crack from the seeded fault.During the testing, the TVDS analyzer monitored the test pinion real time and successfully provided asufficient warn
5、ing time of the impending failure. (Note: No metallic chips were generated during the test).Following testing, the TVDS data (acquired real time) was evaluated along with a post-test fractographicevaluation of the fatigue crack. During the post-test fractographic evaluation, arrest lines and fatigue
6、striations were analyzed to develop crack propagation data as a function of the number of applied load cycles.This data was then correlated with the collected TVDS data to better understand the potential warning signalsthe TVDS system could provide about the impending failure of a helicopters primar
7、y drive systemcomponents. In this test, the TVDS analyzer provided a minimum of eleven minutes warning time (at higherthan normal operation loads) of the impending failure, which would allow the pilot enough time to unload thesuspect engine transmission and continue to fly on the remaining engine.Co
8、pyright 2005American Gear Manufacturers Association500 Montgomery Street, Suite 350Alexandria, Virginia, 22314October, 2005ISBN: 1-55589-864-5Page 1 CH-47D Engine Transmission Input Pinion Seeded Fault Test Joseph P. Petrella Senior Design Engineer Drive Systems The Boeing Company, Philadelphia, Pa.
9、 (610) 591-2230 e-mail: James S. Kachelries Technical Fellow Materials Engineering The Boeing Company, Philadelphia, Pa (610) 591-7577 e-mail: Scott M. Holder Senior Design Engineer Materials Engineering The Boeing Company, Philadelphia, Pa (610) 591-4227 e-mail: Thomas E.Neupert Senior Aerospace
10、 Engineer Aviation Program Executive Office U.S. Army, Huntsville, Al. (256) 955-6490 e-mail: tom.neupertus.army.mil INTRODUCTION Boeing Rotorcraft of Philadelphia (BRP) has been involved in the development of an airborne failure detection system designed to monitor major load carrying helicopter tr
11、ansmission gears and shafts. In 1991, BRP created a specification to develop and build a Transmission Vibration Diagnostic System (TVDS) Analyzer based on Stewart Hughes Limited Mechanical System Diagnostic Analyzer (MSDA). That same year, Stewart Hughes Limited was contracted by BRP to build and qu
12、alify a TVDS analyzer under BRP Independent Research therefore, if four or more techniques were exceeded at one time, all four of the warning light indicators would be illuminated. The analyzer utilized an IBM compatible PC, operating under the XENIX and MS DOS operating systems, as a ground station
13、. The ground station was used to configure the analyzer for testing, long term storage Page 2 of the data transferred from the analyzer, a platform to review analyzed data utilizing a set of graphic displays and a medium to transfer test data. Figure 1 is a picture of the BRPs TVDS Analyzer and its
14、support PC ground station. Figure 1 - TVDS Analyzer and its support PC ground station CH-47D ENGINE TRANSMISSION DESCRIPTION The CH-47D aircraft contains two engine transmissions mounted on the front of the left hand and right hand engines, see Figure 2. The engine transmission consists of a single
15、spiral bevel mesh with an overall gear reduction of 1.23:1. The transmission transfers the power received from the engine through a quill shaft and redirects the power 90 to the combiner transmission via shafting. Each engine transmission has a continuous twin engine design power rating of 3,750 hp
16、and a single engine design power rating of 4,600 hp with both ratings at an input speed of 15,066 rpm and an output speed of 12,263 rpm. The transmission also incorporates a one-way clutch, which allows the drive system to overrun the engines during autorotation or a sudden reduction in engine rpm.
17、Each of the two engine transmissions are the same except for the location of some external lubrication system components. Either transmission can be converted to one for the opposite side by exchanging the location of these components. Figure 2 also shows a sectioned view of the engine transmission.
18、 DESCRIPTION OF TEST SPECIMEN The test specimen was a CH-47D engine transmission spiral bevel input pinion. The specimen was visually and magnetic particle inspected prior to inducing the stress riser (seeded fault). Following these inspections, Electric Discharge Machining (EDM) was used to produce
19、 a notch in the drive side root fillet of one tooth, as illustrated in Figure 3. The actual notch size, which was developed from BRPs stress data on the pinion, was 0.100 inch long by 0.010 inch deep by 0.006 inch wide. Figure 2 Schematic of Test Pinion (Blue) installed in an Engine Transmission TES
20、T STAND DESCRIPTION The CH-47D Engine / Combiner Transmission Test Stand, located at BRPs transmission facility, was used to conduct this testing. The bench test stand is a closed loop torque system, which has the exact gear ratio as the CH-47D engine / combiner transmission set. The test stand has
21、the capability to change torque and speed while the transmission is operating. Shown in Figure 4 is the test Right Hand (R/H) Engine Transmission, Left Hand (L/H) Engine Transmission and Combiner Transmission installed into the Engine / Combiner Transmission Test Stand. Page 3 Figure 3 Test Specimen
22、 with EDM Notch INSTRUMENTATION The parameters required to be monitored during the test were as follows: 1.) Oil sump temperature * 2.) Oil cooler out temperature * 3.) Primary oil pressure * 4.) Auxiliary oil pressure * 5.) Sync shaft speed * 6.) Five (5) transmission accelerometers (one for each e
23、ngine transmission and three for the combiner transmission). 7.) Transmission cross shaft torques* 8.) Phase accurate 1/rev sensor on the combiner transmission input pinion shaft 9.) Time code generator The parameters identified with an asterisk (*) were documented every 15 minutes on transmission r
24、un sheets by the mechanic. The remaining items were continuously recorded on a Metrum RSR 512 data recorder and Honeywell 101 analog recorder. Oscilloscopes and a spectrum analyzer were utilized to monitor (real-time) the accelerometers and 1/rev signal of the sync shaft. During all bench testing, w
25、ide frequency range (5-20,000 Hz) magnetic tape recorders were utilized to record accelerometer, 1/rev sensor, and time code data. The Honeywell 101 recorder was configured in a “continuous record loop“ mode and was recording continuous real time data throughout the test. The BRP TVDS analyzer was u
26、sed real time to detect any crack propagation in the seeded fault specimen. The TVDS analyzer utilized the vibration signal from the accelerometers and the phase accurate speed sensor to monitor the seeded fault pinion and its mating spiral bevel ring gear during the test. The locations of the accel
27、erometers and speed sensor are shown in Figure 4. All electronic time code generators/loggers (TVDS and Metrum RSR-512 internal and Honeywell 101 external) were set to the real time and run time at the test console. This was done with the objective of having the capability to accurately develop a cr
28、ack growth analysis versus the operating conditions. TEST PROCEDURE The TVDS testing of the engine transmission pinion specimen was to be accomplished in three phases. Phase 1 - Stabilization: This run was accomplished in order to stabilize the signature of the transmission. It has been BRPs experie
29、nce that when a transmission is completely disassembled and then reassembled, its vibration signature could change and that typically following an initial run on the transmission, the vibration signature would become “stabilized.” This run was also utilized to create an initial baseline and to verif
30、y the set-up for the TVDS analyzer. The duration of the run was 80 minutes at 65% Twin Engine (TE) design torque at 100% input speed. The following start-up procedure was used forCross-Section Notch EDM Notch Drive Side EDM Notch Within Red Oval Page 4 Figure 4 Locations of Accelerometers (Arrows 1-
31、5) and Speed Sensor (Arrow 6) all phases of the testing: 6-10% Twin Engine torque was applied to the L/H (non-test) and R/H (test) Engine Transmissions. The transmissions were brought up to test speed. The load on the L/H Transmission was increased to 65% TE Torque. Once the 65% TE condition was obt
32、ained on the L/H Transmission, the load on the R/H Transmission was increased to the specified test condition. Accumulated test time began when the R/H Transmission achieved test condition torque. Phase 2 - Crack Initiation: The TVDS analyzer was initially configured with the threshold values establ
33、ished from the stabilization run while new threshold values were developed for this new load condition. During this phase, the transmission was operated at an elevated load (135% TE design torque at 100% rpm) to insure crack initiation. Once a clear indication was obtained from the analyzer, the tra
34、nsmission was shut down and partially disassembled to verify that a crack had initiated from the EDM notch. If a crack was confirmed, the transmission was reassembled for Phase 3 testing. Phase 3 Four Hour “Get Home Flight”: Once the TVDS analyzer had identified that a crack had initiated, the trans
35、mission was operated at 65% TE design torque at 100% rpm to simulate a typical heavy gross weight condition for the CH-47D. The goal of this phase of the test was to demonstrate that the transmission could operate for four hours, ensuring that the Aircraft could “Get Home” safely. Note: This four ho
36、ur goal was the objective for non-redundant primary drive components. This was considered the maximum time required for a “get home flight” for an Aircraft that had a fault indication while it was on a mission. However, the 4 hour requirement is less important for the engine transmission, since the
37、CH-47D Aircraft has the capability to fly using only one of its two engines. Therefore, the minimum test goal in this case is to detect an impending failure with sufficient warning time for the pilot to unload the suspect engine transmission, since he could continue to fly safely using the remaining
38、 engine. During all phases of the testing, if any of the following conditions occurred, the test stand would be immediately shut-down and a full investigation of the anomaly performed: 1.) Abnormal test stand occurrence. 2.) Test stand transmission abnormal noise, smoke, vibration, etc. Page 5 3.) T
39、est specimen abnormal vibration, etc. 4.) Significant TVDS indications. 5.) Transmission chip light indication. Throughout the entire test, any significant events were monitored and manually recorded. All electronic real and run time code generators / loggers were set to the same time as the test st
40、and console so that it would be possible to accurately develop any post-test analysis. TEST RESULTS The initial 80 minute stabilization run at 65%TE torque (Phase 1) was accomplished to establish the TVDS signature. No anomalies were noted. The engine transmission was then operated at 135% TE torque
41、, which is equivalent to 110% SE torque, to begin the “crack initiation run.” The transmission was operated at this test condition for 13 minutes, 7 seconds when the TVDS analyzer indicated that a crack had initiated and was propagating. The test stand was shutdown for inspection of the pinion. Duri
42、ng the inspection, very tight hairline cracks were visually observed (30X) emanating from both sides of the EDM notch. The cracks were oriented parallel to the root fillet of the pinion gear tooth. With both a positive indication from the analyzer, and visual conformation of the crack, the final 4 h
43、our “get home” phase was initiated. The transmission was operated at 65% TE torque for Phase 3. The transmission operated at this test condition for 9 minutes. During this time, numerous algorithms from the TVDS analyzer were exceeding their threshold limits, visual changes of the raw signals being
44、displayed on the oscilloscopes were noted, and within the last seconds at this test condition an audible pitch change in the transmission were detected by test personnel. Once the audible pitch change was noted, the test stand was immediately shutdown. A summary of the test time, along with cumulati
45、ve cycles at load, are contained in Table 1. After termination of the test, the pinion was removed from the transmission and submitted to BRPs Material Engineering Laboratory for fractographic evaluation. It should be noted that at no time during this test did the chip detectors, electrified screens
46、, temperature probes, or pressure transducers indicate that a problem was present in the test transmission. POST TEST EVALUATION OF TVDS DATA Post-test evaluation of the TVDS data revealed that 10 of the BBPS Techniques (RMS, WHT, SOA-TE0-G1, SOA-TE0-P1, SOA-TE0-G2, SOA-TE0-P2, COMP-TE0-G, COMP-TE0-
47、P, COMP-TE0-PK and IDENT-T1) and 6 of the SGAV Techniques (RMS, SO1, SO2, FM4B, ACH and SCH) monitoring the pinion had continually exceeded their warning levels once the crack had initiated to a detectable size. In order to develop an extremely conservative approach to evaluate the TVDS data, the po
48、st-test exceedance levels were established to be the point at which the TVDS analysis technique data exceeded the mean plus five standard deviations continuously. The data utilized to create the mean plus five standard deviations was the initial 80 minutes of the test. In using this methodology, the
49、 warning times from these techniques ranged from 20 minutes 56 seconds (SGAV FM4B) to 7 minutes 23 seconds (BBPS IDENT-T1). A typical display of the data from one of the BBPS Techniques (SOA-TE0-P1) and one of the SGAV Techniques (RMS) are shown in Figures 5 and 6 respectively. To summarize the warning times for each of the analysis technique packages, a summary chart displaying warning times prior to test termination for each of the analysis techniques that meet the requir
