Contaminated Fuel Testing of Engine Controls to MIL-E Using the USN Aeronautical Engine Laboratory Technique

Transcription

1 PAPER NUMBER63-AHGT-42 Copyright 1963 by ASME PUBLICATION Contaminated Fuel Testing of Engine Controls to MIL-E Using the USN Aeronautical Engine Laboratory Technique MYRLE C. ELEY Manager, Test Facilities, Woodward Governor Company, Rockford, Ill. Testing controls using fuel contaminated according to MIL-E has created problems in obtaining test equipment with satisfactory life - and also in obtaining good mixing and proper injection of the fuel contaminant mixture. The author's company has taken the recommendations of the U. S. Navy Aeronautical Engine Laboratory in Philadelphia and developed a nonrecirculating contaminate method which has met with military approval. The method uses a concentrated fuel slurry and proportioned injection controlled by actual fuel flow rate. $ I PER COPY 5 0 C TO ASME MEMBERS The Society shall not be responsible for statements or opinions advanced in papers or in discussion at meetings of the Society or of its Divisions or Sections, or printed in its publications. Discussion is printed only if the paper is published in an ASME journal. Released for general publication upon presentation THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS 345 East 47th Street, New York 17, N. Y. Contributed by the Gas Turbine Power Division for presentation at the Aviation and Space, Hydraulic, and Gas Turbine Conference and Products Show, Los Angeles, Calif., March 3-7, 1963, of The American Society of Mechanical Engineers. Manuscript received at ASME Headquarters, December 31, Written discussion on this paper will be accepted up to April 8, Copies will be available until January 1, Printed in U.S.A.

2 6 I 24 IS St.,10 a- Boocr Rum.- 3- Hear 4-3w- Dove Mcrae 5-1.7e:1 Sra, up..o.r/ t 6-4e.,e Puov, 7- C."7 Te, 74fle 8- PNad/rygr/C SPifed2 Co..vneoc 9- PowssUir/C.Vg 10- Tc/Irellfwler // A.G. 00rWeiCICONVeofirpleKW /2- Ifea,.G.GoreGGe /3- C.~-7.WM/n."4, Fic 15- CeeInectG `-)xr/aee /6- arfo.- ce flow rjarc< /7- Sp-c, Gocoe F/47-ffor-4.41,,,vor /9 - SG um,' 20- Sc u ozcy y 9corvse.c 22- Alec-xi c-..ry-cmc 23-2,, 24- Fig. 1 Schematic of contamination test setup Contaminated Fuel Testing of Engine Controls to MILE-5009B Using the USN Aeronautical Engine Laboratory Technique MYRLE C. ELEY BACKGROUND Testing fuel components using contaminated fuel usually causes more trouble with the test equipment than with the unit under test. This is true since the test equipment outnumbers the equipment under test by about ten to one. We have been working on the problems associated with contaminated-fuel testing since both under Navy contract and on our own. Originally the contaminate consisted of a select particle size dust and salt water, then iron oxide and napthenic acid was added. The final touches were sand and organic fibers. Prior to October 31, 1960, there was no standard method of contaminant distribution in jet aircraft fuel during qualification tests. In report NAMC-AEL-1670 of October 31, 1960, the Aviation Engine Laboratory of the Naval Air Material Center outlined a standard optimum method of conducting these tests. With minor modifications to suit test requirements, this outline was followed by the Woodward Governor Company in setting up the latest "B" dirt test facility. Three MIL-E-5009B 1 1 MIL-E-5009B, January 22, 1959, Qualification Tests for Engines, Aircraft, Turbojet. fuel-component tests have been accomplished successfully with a minimum of equipment failure. SYSTEM DESCRIPTION Clean test fuel (MIL-S-3136 Type I) stored in a 220-gal glass-lined sump is gravity fed to a centrifugal boost pump; from the boost pump the fuel at 35 psig passes through a water-cooled heat exchanger to an aircraft gear pump driven tandem to the fuel control and 1:78 times slower. This 1:1.78 ratio is used to match the pump output curve to the main control inlet flow requirements. The 3-hp dc variable-speed motor is cycled to meet test requirements by means of a cam timer and a servo-operated scheduling device. Pump discharge flow is passed through a 50-psi pressurizing valve, a turbine flowmeter, and across the bottom check valves of the piston-type contaminant injector pump. The turbine flowmeter produces a frequency signal proportional to fuel flow. This signal is changed to a dc voltage by a frequency converter, amplified by a simple transistor circuit, and is used to control the speed or the 2-hp, dc injector 1

3 Fig. 2 Fig. 4 tank to provide additional mixing action. Fig. 3 pump motor at a rate proportional to pump-dish charge flow. Seal pressure for the injector pump is taken just upstream of the 50-psi pressurizing valve. In this manner seal pressure is always higher than the injector-pump pressure. The injector pressure changes from 100 to 1050 psi during cyclic operation of the test system. The fuel-control metered flow passes through an orifice-type flowmeter, a nozzle simulator, and a remotely controlled stopcock. The combined flow is passed through a combination filter-water separator and is returned to the sump. The contaminated fuel slurry is stored in a stainless-steel 150-gal funnel-bottomed tank. A "squeeze tube" chemical pump circulates the slurry from the.bottom of the tank, across the top or inlet cileck valves of the piston pump, and back to the top of the tank through a deflector nozzle. An air motor-driven propeller is mounted in the PROCEDURE In accordance with Appendix 1 of NAMC-AEL the following method was used: One hour's total flow rate was calculated based on measurements taken during cyclic,operation. The required amounts of solid contaminant and Salt water were calculated, then increased by 10 percent and packaged in 2-hr batches. The contaiminant injector pump speed range, speed level, and capacity (in this case 10 gph) were adjusted to cycle in proportion to main pump output. The slurry-tank volume was determined and the contaminant was added based on the total hourly 2 NAMC-AEL 1670 of October 31, 1960, Proposed Optimum Method of Conducting Components per Specifications MIL-E-5007B3 and MIL-E-5009B. 3 MIL-E-5007B, January 22, 1959, General Specifications for Engines, Aircraft, Turbojet. 2

4 Fig. 5 Fig. 7 Fig. 6 supply in the tank. In this manner the slurry tank was "topped off" merely by adding fuel and contaminant in 2-hr amounts. The system was monitored during the test for possible adjustment due to system-calibration change. No fuel samples were taken other than sump tank samples which were tested for residual contaminant. Napthenic acid was added based on the total system volume including slurry tank, and more was added whenever the fuel system was topped off. DISCUSSION The successful testing accomplished at the Woodward Governor Company has been attributed to the imagination and inventiveness of the environmental test technicians. All of the equipment is considered shelf items; however, changes were incorporated to secure success. The injection pump, Fig.2, has been modified to obtain intake and discharge in a downward direction to obtain self-washing by reversing the check valves and spring-loading the balls as shown. The check valves were also modified to use General Electric 608 chrome-carbide balls and seats. A pressure seal was built into the piston sleeves to maintain a 50-psi clean fuel drop toward the contaminated fuel to prevent dirt entering the sleeve. The pump operated trouble-free using the foregoing configuration. Use of a 150-gal contaminate and fuel-mixture tank permitted a less concentrated contaminate solution. This coupled with a good mixing procedure and a high-velocity contaminate pump prevented contaminate pileup in the system and contributed to the trouble free injection. The contaminate injection rate is determined by the size of the contaminate slurry tank, the concentration of the contaminate and the fuel flow into the unit under test. The use of Briggs filter and water separators to clean the fuel to 1/2 micron or better dirt and 99.9 percent salt water removal eliminated sampling procedures in all areas except at the cleanfuel sump and permitted recirculation of the fuel. The fuel used in the slurry was added to the total fuel in the sump and required extra tank, space. This was drawn off for the suceeding slurry mixture. Observations of the system after 1000 hr of running time indicate that: Slurry-pump output should be at least 10 gpm with 1/2-in-id recirculation lines. The higher flowrate reduces the tendency of all the large particles to pass through the piston pump immediately after replenishing the slurry tank. The Viton "A" pump-liner life was approximately 300 hr. Inspection of the chrome-carbide seats and balls after more than 600 hr running showed no signs of corrosion or wear. By comparison tungsten-carbide balls after only 100 hr malfunctioned due to contaminant erosion and rust. 3

5 Fig. 8 Fig. 9 The same basic test equipment can be used for contaminant tests over a wide range of flow levels. This particular test setup, without major change, could be used for cyclic flow variations from constant flow up to a change ratio of 3:1. The factors limiting the flow ratio with this equipment are, the requirement of at least one piston pulse per second4 and the fact that the injection-pump gear box is red-lined at 1800 rpm or 2.5 pulses per sec. Replacement of the gear box or gear-boxdrive ratio could allow the equipment to be used over a wider flow range with some reduction in reliability as a result of higher piston speeds. Standard turbine flowmeters are available for ratios of 20:1 and a few will measure greater flow ratios. The system reliability would be reduced by using a more concentrated slurry as would be required under higher flow conditions. The slurry concentration during the latest series of tests was 0.8 g per gal. Concentrations of up to 11.9 g per gal were used during previous testing. Concentrations this great should be avoided if possible by using larger capacity slurry and injection equipment. The contaminant proportioning accuracy over a 300-hr test was extremely good. Owing to the 10 percent additional contaminant added to the hourly supply it is unlikely that during any interval of the cyclic operation the contaminant rate fell below 100 percent of the desired. The possible transient inaccuracy caused by the phase lag of the components in the flow injector pump control loop would be 2.5 percent or less and would be self-compensating during a 1- min cycle. 4 Reference footnote' p. 4. The state of the art of contaminant testing during the last series of tests had progressed far enough to allow the system to operate unattended. Suitable interlocks, monitors, and warning devices were installed. This "automatic" system resulted in considerable saving of money, and in some respects, produced more accurate test data as a result of the continuous monitoring of control parameters and output. APPENDIX EQUIPMENT DESCRIPTION: Clean fuel sump: A.O. Smith glass-lined water tank with a capacity of approximately 200 gal, Fig.3. Boost pump: Gorman-Rupp centrifugal, capacity approximately 6000 pph at 35 psig. Heat exchanger: Young Radiator Company. Drive motor, pump and fuel control: Reliance 3-hp, dc variable-speed, powered from a Reliance motor-generator and a Woodward Governor Company built control circuit. Step-up drive: Woodward Governor Company designed 1:1.78 ratio tool number T Aircraft gear pump: Pesco single-element capacity 200 pph at 1000 psi and 2200 rpm. Cam timer (Fig.4): Reynolds Electric. Pneumatic speed control (Fig.5): Woodward Governor Company tool number T

6 Pressurizing valve: Republic Manufacturing Company. Turbine flowmeter (Fig.6): Fisher Porter P/N 10C1505. Frequency converter (Fig.7): Waugh Engineering Corporation P/N FR Amplifier: Woodward Governor Company built transistor. Contaminant injector pump (Fig.8): B-1-F proportioneers model 1140-S type 0572 modifications. Piston-rod seals: The packing type seals were replaced with a fuel pressure sealed close clearance sleeve Woodward Governor Company tool number T Check valves: Seats and balls of General Electric Company 608 chrome carbide were fitted. In addition the normal bottom to top pumping action was reversed by spring loading the check valve balls. Woodward Governor Company tool number T Airframe filter: Ful-flo 74 micron single unit type. Control mounting fixture: Woodward Governor Company tool number T incorporates a by-pass port and drive shaft fuel seal. Orifice flowmeter: Barton range pph. Stopcock: Jamesbury air-actuated valve. Filter-water separators: Briggs model BFS-3-V-50L 1/2 micron or better dirt and 99.9 percent salt water removal. Slurry tank: Stainless-steel custom built. Slurry pump (Fig.9): Vanton model XB-T60A.