Double Helical Gear Performance Results in High Speed Gear Trains Dr. Robert F. Handschuh Aerospace Engineer Army Research Laboratory, Vehicle Technology Directorate NASA Glenn Research Center Cleveland, Ohio 44135 [email protected] Ryan Ehinger, Eric Sinusas, & Charles Kilmain Bell Helicopter Textron Inc. Fort Worth, Texas 76101 ABSTRACT The operation of high speed gearing systems in the transmissions of tiltrotor aircraft has an effect on overall propulsion system efficiency. Recent work has focused on many aspects of high-speed helical gear trains as would be used in tiltrotor aircraft such as operational characteristics, comparison of analytical predictions to experimental data and the affect of superfinishing on transmission performance. Baseline tests of an aerospace quality system have been conducted in the NASA Glenn High-Speed Helical Gear Train Test Facility and have been described in earlier studies. These earlier tests had utilized single helical gears. The results that will be described in this study are those attained using double helical gears. This type of gear mesh can be configured in this facility to either pump the air-oil environment from the center gap between the meshing gears to the outside of tooth ends or in the reverse direction. Tests were conducted with both inward and outward air-oil pumping directions. Results are compared to the earlier baseline results of single helical gears. INTRODUCTION In the work to be described in this paper, a double helical gear train was tested. The test facility has gears that are bolted High speed, heavily loaded and lightweight gearing to the shafting (in practice this interface is avoided to remove components are found in propulsion systems for rotorcraft. fasteners and shafting complexity). This permitted the gearing The high pitch line velocity that is part of these systems to be mounted from either side of the gear inner flange and makes the thermal aspects of the gear system design very thereby the system could be tested two ways. One way was to important. Transmission systems using helical gear trains in have the gears pump the air-lubricant environment (from tiltrotor aircraft provide the proper spacing between the lubricating and cooling jets) outward towards the axial ends of parallel engine and rotor shafts as shown in Figure 1 the teeth. The second way was to have the gear mesh pump (Ref. 1). These gear trains can induce additional thermal the environment inward from the ends of the teeth towards the problems as the idler gears in this system receive two center gap. The double helical gears were also tested with and meshing cycles (on opposites sides of the gear teeth) per without shrouding as used in the prior studies. Temperature revolution. Therefore, weight optimized aerospace drive data from axial thermocouple rakes and tooth end arrays will system components can have difficulty when it is required to also be provided. Finally, the double helical results will be operate with a primary lubrication system failure. directly compared to prior studies from the same facility (single helical gear results; (Refs. 2 to 7)). In past studies using this gear train system, testing has focused on basic operational characteristics (Refs. 2 and 3), TEST FACILITY AND TEST HARDWARE comparison of analytical predictions to experimental results (Ref. 4), and the affect of superfinishing (Ref. 5). The results Test Facility from these studies have shown and quantified the effect of The test facility used for this study is shown in operational conditions on the power to rotate the gear system Figure 2. The facility is a closed-loop, torque-regenerative at high speed and load. testing system. There is a test gearbox and slave gearbox that are basically mirror images of each other. Each gearbox has In prior testing the individual thermal/energy loss an input gear, three idlers, and one bull gear. The gearboxes mechanisms were separated into gear meshing, bearing, and are joined together through the input gears and bull gears via windage losses. The gear meshing and bearing losses are shafting. fairly well understood (Ref. 6) and some design modifications to the gears and bearings can provide some A 500 hp DC drive motor powers the facility and its benefit to increase performance. The gear windage losses, output speed is increased using a 5.7:1 speed-increasing however, are the least understood, but can be a source of gearbox. The output of the speed-increasing gearbox then dramatic tiltrotor performance improvement once this passes through a torque and speed sensor before connecting mechanism is better understood. to the slave gearbox. Figure 1. Layout view of typical tiltrotor nacelle (airplane mode). Figure 2. NASA high-speed helical gear train test facility. Each gearbox has separate oil supply and scavenge bull gear shaft) needed in the single helical gear arrangement pumps and reservoirs. All flow rates have been calibrated at were removed. The double helical gear configuration does various temperatures and pressures prior to installation for not produce an axial load from the gear forces; therefore the accurate flow rate measurements. Lubrication system flow thrust bearings are not needed. A photograph of the double rate is controlled using the supply pressure. Temperature is helical gears in the inward pumping gear arrangement controlled via immersion heaters in the reservoir and heat without the shrouds is shown in Figure 3. A photograph of exchangers that cool the lubricant returned from the the shrouds from earlier tests is shown in Figure 4 gearboxes. Each lubrication system has very fine 3-μm (shrouding was configurable for either single or double filtration. Nominal flow rate into the test or slave gearboxes helical gear trains). at 80 psi is approximately 15 gpm. Table 1. Basic Gear Design Data for The lubricant used in the tests to be described was a Double Helical Gear Train. synthetic turbine engine lubricant (DoD-PRF-85734). This Number of teeth input and 2nd idler 50 lubricant is used in gas turbine engines as well as the drive Number of teeth 1st and 3rd idler 51 systems for rotorcraft. Number of teeth bull gear 139 Axial face width each side, mm (in.) 39.6 (1.560) Test Hardware Gap between tooth flank sides, mm (in.) 41.1 (0.555) The test hardware used in the tests to be described is Normal plane module, mm (normal 0.254 (10) aerospace quality hardware. The basic gear design plane diametral pitch (1/in.) information is contained in Table 1. The input and bull gear Normal plane pressure angle, degree 20 shafts typically have a combination of roller bearings with Transverse plane module, mm 3.10 (8.192) ball bearings to contain the resultant thrust loads, whereas Helix angle, degree 35 the idler gears only have roller bearings. For the double Gear material Pyrowear 53 helical gear tests, the two thrust bearings (input shaft and Figure 5. Comparison of rake probes for single and double helical gear train studies. Figure 3. Photograph of test gearbox with shrouds removed. Figure 6. Array probe used at the tooth end—meshing position. Figure 7. Locations of array and rake thermocouples within the test gearbox. Data Acquisition Figure 4. Shrouding used during testing. The test facility data system monitors three important TEST INSTRUMENTATION AND facility parameters during operation. Speed, torque (supplied DATA ACQUISITION torque and loop torque), and temperature measurements were made during all the testing conducted. The test system Test Instrumentation loop torque is measured on the shaft connecting the bull The instrumentation used in these tests was the same as gears from the test and slave gearboxes. A telemetry system in prior studies except for the axial fling-off temperature was utilized in this location to measure loop torque. rakes. With the double helical gear mesh tests, another set of probes was fabricated so that the fling-off temperatures from The data recording system used in this study has the the various axial locations of the gear could be measured. capability of taking data from all parameters at a rate of one The probe is shown in Figure 5 along with the one from sample per second (tests in this study recorded data every prior single helical testing. Also in these tests, axial arrays 2 sec). The data is displayed to the test operator in real time. (9 thermocouples) were used at the mesh exit regions Data is stored in a spreadsheet format and each sensor can be (Figure 6). The temperature probes were at locations in the viewed at any time during a test and when post processing gearbox as shown in Figure 7. the results. The test procedure for collecting the data to be inward pumping configuration, the inward pumping had a presented was the following. For a given set of conditions much higher temperature at the gap region when the (speed, torque, lubricant pressure and lubricant oil inlet shrouding was in place. temperature) the facility was operated for at least 5 min or until the temperatures of interest had stabilized (~±2 °F). The next comparison to be made is shown in Figure 9. In this figure the maximum thermocouple rake and array EXPERIMENTAL RESULTS temperatures are plotted for the two speed conditions (12500 and 15000 rpm input shaft speed), three bull gear shaft In this part of the paper, just the results attained with the torque levels (33, 67, and 100 percent of maximum), double helical gears will be discussed. Comparison to the outward or inward pumping configuration, and with or single helical gear tests will be discussed in the next section. without shrouds. All data shown in Figure 9 was taken with The gears were operated at various speed, torque, lubricant the standard inlet conditions mentioned above. In the figure flow rate and lubricant inlet temperature conditions. For all two bands of data are shown due to the two speed the results shown in this section, the lubricant inlet conditions. The rake or array maximum temperatures temperature for the test gearbox was held at 200 °F. Also, increased linearly with applied torque. At either input shaft the slave gearbox conditions (lubricant temperature and flow speed (band), the inward pumping gear arrangement rate) remained unchanged for all of the test conditions of the produced the highest (shrouded, rake probe) and lowest data to be presented. For the experimental results to be (shrouded or unshrouded, array probe) temperatures. For the described in this paper the nomenclature shown in Table 2 inward pumping arrangement, a limited amount of lubricant will be used. from the meshing gears would be assumed to strike the array Table 2. Figure Nomenclature probe in comparison to that flung from the gear mesh radially and pumped axially towards the center of the face OP Outward Pumping width of the gear. IP Inward Pumping S Shrouds The last comparison with respect to just the double NS No Shrouds helical gear tests is shown in Figure 10. In Figure 10(a) and SH Single Helical (b) the standard conditions (200 °F lubricant inlet DH Double Helical temperature, and 80 psi lubricant jet pressure) were provided ISF Superfinished Gear Flanks for two speeds and three load cases as described earlier. In GRND Ground Gear Flanks Figure 10(a) drive motor power (amount of power to rotate the entire test rig) and in Figure 10(b) lubricant temperature The first comparison is made between temperatures rise (outlet minus inlet temperatures) is plotted versus bull across the gear face width and the direction of gears gear shaft torque. From Figure 10(a) or (b) it can be seen pumping of the air—lubricant mixture, as shown in Figure 8. that either the amount of power or the temperature rise was Four different sets of data are shown in the figure for the 2nd minimized using outward pumping configuration with and 3rd idler gear location. Thermocouple probes 1 and 6 shrouds. Outward pumping also required the most power or were near the ends of the teeth and thermocouple probes 3 had the highest temperature rise if the shrouds were and 4 are near the center gap of the double helical gears. The removed. When the gears were run with them pumping outward pumping arrangement had the lowest temperature at towards the gap, there was less of an effect with or without the gap locations. The inward pumping configuration had an the shrouds. increase in temperature towards the tooth gap. Also, for the Figure 8. Rake temperature as a function of location across the face width, 200 °F lubricant inlet temperature, 15000 rpm, 5000 hp (2nd to 3rd idler location). Figure 9. Maximum rake and array temperatures as a function of torque, speed, and shrouding (inward and outward pumping). Figure 10. Effect of shrouding on drive motor power (a) and lubricant temperature rise (b). that by lowering the jet pressure, the jets cannot penetrate COMPARISON TO PRIOR TESTING deep enough to the roots of the teeth so the level of loading In this part of the paper the results from double helical was kept to 33 percent of maximum. The trends were that gear testing will be compared to prior experimental results higher pressure (more cooling flow) reduced the temperature from single helical gears. As was shown earlier in this paper, change from inlet to exit of the gearbox. Also the double two shaft speeds of interest will be shown with varying helical outward pumping arrangement produced the lowest amounts of torque applied. The two speed levels pertain to temperature increase for all conditions tested. the two tiltrotor flight operating conditions of hover (high shaft speed) and forward flight (lower shaft speed). In Figure 14 the next comparison is made between the Comparisons will be made for the amount of power to rotate temperature change as a function of two speeds and three the gear train in the test facility, amount of power absorbed torque levels for six different gearing configurations. Once by the lubricant (thermal efficiency), inlet lubricant jet again the double helical, outward pumping configuration pressure (lubricant flow), lubricant temperature increase with shrouds had the lowest temperature rise for the across the gearbox, and thermocouple information from the lubricant between inlet and exit of the gearbox. Nearly all radial and axial probes. the data showed a linear dependence on the torque applied. The first comparison, Figure 11, will be between the The last two Figure 15 and Figure 16 will show the various configurations with respect to the amount of power temperature probe data for the single and double helical gear to rotate the system at two rotational speeds with varying configurations. The thermocouple array maximum amounts of bull gear shaft torque. The single helical (SH) temperature differential is shown for two speeds and three gear results are shown for superfinished (ISF) and torque levels, all at 200 °F lubricant inlet temperature. The conventionally ground (GRND) gear surfaces. All data is for temperature difference was found from the maximum shrouded (S) gear conditions. (In subsequent figures NS— thermocouple array measured minus the lubricant inlet means that the shrouds were removed for the test temperature (~200 °F). For this data, one can see three bands conditions.) For the conditions tested at the two different of data. The single helical gears have a much higher rotational speeds and three levels of bull gear torque, the temperature differential from the array probe thermocouples double helical gears configured in the outward pumping, than the double helical gears. The double helical, inward or shrouded arrangement provided the lowest power loss at all outward pumping gears, had a similar result for temperature conditions. difference at 15000 rpm as the single helical at 12500 rpm. Also the double helical gear configurations at 12500 rpm The next comparison is for power loss to the lubricant, were the lowest. An explanation for this is fairly Figure 12, between the various configurations shown. The straightforward. Since a single helical gear has the entire level of losses (hp) were determined via the flow rate, face width to pump the entrapped lubricant—air lubricant properties, and the temperature differential from environment, it would be expected to have a much higher lubricant inlet to exit. In this figure, once again the double exit (array) temperature as opposed to one only moving helical gear train had the lowest power loss for both speed across approximately half the face width. and all loading conditions. This data is also for all configurations using shrouds. The last comparison to make will be with the rake probes that measured the air- lubricant being flung radially Next the effect of lubricant system jet pressure from the gear teeth. In Figure 16 the temperature differences (volumetric flow rate) is presented in Figure 13. The tests between the maximum rake temperature and lubricant inlet were all run at 33 percent of maximum torque and a 200 °F temperature is plotted versus applied torque for the two lubricant inlet temperature. As before, the data is contained speed conditions (12500 and 15000 rpm). All test data in two “speed bands”. Lower jet pressure reduces the shown was taken with shrouds in place. At the 12500 rpm amount of lubricant available to keep the gear surfaces fully input shaft speed, all data was within ~7 °F of each other. At flooded and provide adequate cooling. In this facility the jets the higher speed conditions, both double helical gear feed lubricant into and out of mesh. The lubrication system configurations were better with the outward pumping double operates in a dry sump mode—all lubricant is jet fed and helical configuration being the best at all torque conditions. scavenged away by lubrication pumps. There is also concern Figure 11. Drive motor power versus torque, two speeds at 200 °F oil in temperature. Figure 12. Power losses versus torque, two speeds at 200 °F oil—in temperature. Figure 13. Temperature change of the lubricant versus pressure at 200 °F, two speeds, 19000 in*lb torque at bull gear shaft. Figure 14. Temperature change of lubricant versus torque, two speeds at 200 °F. Figure 15. Temperature of array thermocouple maximum difference from inlet temperature versus torque at two speeds, 200 °F lubricant inlet temperature. Figure 16. Temperature change of oil inlet to maximum rake probe data versus torque, two speeds, 200 °F lubricant inlet temperature. CONCLUSIONS showed that the most significant difference was at the 15000 rpm condition with the double helical gear, Based on the results found in this study and those outward pumping arrangement being the one with the conducted previously the following observations can be lowest temperature increase between the oil inlet concluded temperature and the maximum rake probe temperature. 1. Double helical gear trains that outwardly pump the REFERENCES air—lubricant mixture axially produced the highest performance when compared to other tested 1. Kilmain, C., Murray, R., and Huffman, C.: V-22 conditions that included inward pumping double Drive System Description and Design Technologies, helical gears and single helical gear trains (ground American Helicopter Society 51st Annual Forum, and superfinished). May 1995. 2. Handschuh, R. and Kilmain, C: Preliminary 2. Double helical gear trains that have an inward Investigation of the Thermal Behavior of High-Speed pumping arrangement produced similar performance Helical Gear Trains, NASA/TM—2002-211336, results with or without shrouds. The results from ARL–TR–2661, March 2002. these tests were in between the outward pumping arrangement being better than outward pumping 3. Handschuh, R. and Kilmain, C.: Efficiency of High- without shrouds and not as good as outward pumping Speed Helical Gear Trains, NASA/TM—2003- with shrouds. This was found in both the drive motor 212222, ARL–TR–2968, April, 2003. power needed to rotate the entire test facility as well 4. Handschuh, R. and Kilmain, C.: Preliminary as the temperature change of the lubricant across the Comparison of Experimental and Analytical gearbox (difference between inlet and outlet lubricant Efficiency Results of High-Speed Helical Gear temperature). Trains, ASME 2003 Design Engineering Technical Conference, September 2003, Chicago, IL. 3. The double helical, outward pumping shrouded arrangement was also the best when lubricant 5. Handschuh, R. and Kilmain, C.: Experimental Study pressure was reduced, providing the lowest of the Influence of Speed and Load on Thermal temperature increase at nearly all conditions. Behavior of High-Speed Helical Gear Trains, NASA/TM—2005-213632, ARL–TR–3488, July 4. Temperature probes were used to measure the radial 2005. and axial air—lubricant environment as the gears 6. Handschuh, R. and Kilmain, C: Operational Influence were tested. The array probe temperatures were much on Thermal Behavior of High-Speed Helical Gear higher for the single helical (ground or superfinshed) Trains, NASA/TM—2006-214344, ARL–TR–3969, in comparison to either inward or outwardly pumped November 2006. configurations. In the case of a single helical gear the array probe will measure the air—lubricant that is 7. Handschuh, R.; Kilmain, C.; Ehinger, R.: Operational pumped some distance across the face width in Condition and Superfinishing Effect on High-Speed comparison to the double helical gear system that Helical Gearing System Performance, NASA/TM— might only be half the distance for the case of 2007-214696; ARL–TR–4099, June, 2007. outward pumping. The data from the rake probes