Institutional Archive of the Naval Postgraduate School

Calhoun: The NPS Institutional Archive DSpace Repository

Theses and Dissertations 1. Thesis and Dissertation Collection, all items


Power plant and drive train improvements of the NPS Hummingbird remotely piloted helicopter

Conway, Robert E.

Monterey, California. Naval Postgraduate School

This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.

Downloaded from NPS Archive: Calhoun

Calhoun is the Naval Postgraduate School's public access digital repository for th D U DLEY research mate rials and institutional publications c reated by the NPS community. «iit : 3 Calhoun is named for Professor of Mathematics Guy K. Calhoun, NPS's first IHN KN Ox appointed and published scholarly author. a ] LIBRARY Dudley Knox Library / Naval Postgraduate School 411 Dyer Road / 1 University Circle Monterey, California USA 93943




Robert E. Conway

September, 1996

Thesis Advisor: E. Roberts Wood

Approved for public release; distribution is unlimited.



CO a | es: | REPORT DOCUMENTATION PAGE | ewsemonsomns cons |

Public reporting burden for this collection of information is estimated to average | hour per response, including the time for reviewing instruction, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, ee at ee nn ba Ag tee

11. AGENCY USE ONLY (Leave blank) DATE REPORT TYPE AND DATES COVERED : September 1996 Master’s Thesis


16. AUTHOR(S) Conway, Robert E.

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESSES) Naval Postgraduate School Monterey CA 93943-5000


11. SUPPLEMENTARY NOTES The views expressed in this thesis are those of the author and do not reflect the official policy or position of the Department of Defense or the U.S. Government.

12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE Approved for public release; distribution is unlimited. 13. ABSTRACT (maximum 200 words) Originally designed as a target drone for the U.S. Army, the NPS Hummingbird has undergone several modifications to convert it into a reliable research platform. The 165 pound remotely piloted helicopter (RPH) | is powered by a Weslake Aeromarine Engines Limited (WAEL) 342 two stroke, twin cylinder, 25 hp, gasoline engine. An engine failure due to cylinder overheating halted research efforts until investigation as to the cause and subsequent corrections could be made. Costing approximately $3000 per engine, another failure is unacceptable. The tasks undertaken in this thesis were to investigate the cause of the overheat failure and improve the engine cooling system. Cooling system corrections required total redesigns of the engine cooling | and engine start systems. Additionally, research of the RPH’s history revealed a need for a torsional shock absorber to be incorporated in the drive train to increase component life. The changes made to Hummingbird | provide a decrease in empty weight, minimal center of gravity change and, most importantly, an increase in user | safety furnishing the Department of Aeronautics and Astronautics with a dependable vehicle for rotary wing


® | 16. PRICE CODE |


SUBJECT TERMS: Helicopter, Radio-Controlled, Power Plant, Drive Train, Unmanned Aerial Vehicle


Unclassified Unclassified Unclassified

NSN 7540-01 -280-5500 Standard Form 298 (Rev. 2-89) Prescribed by ANSI Std. 239-18 298-102



Approved for public release; distribution is unlimited.


Robert E. Conway Lieutenant Commander, United States Navy B.S., United States Naval Academy, 1985

Submitted in partial fulfillment of the requirements for the degree of






Onginally designed as a target drone for the U.S. Army, the NPS Hummingbird has undergone several modifications to convert it into a reliable research platform. The 165 pound remotely piloted helicopter (RPH) is powered by a Westlake Aeromarine Engines Limited (WAEL) 342 two stroke, twin cylinder, 25 hp, gasoline engine. An engine failure due to cylinder overheating halted research efforts until investigation as to the cause and subsequent corrections could be made. Costing approximately $3000 per engine, another failure is unacceptable. The tasks undertaken in this thesis were to investigate the cause of the overheat failure and improve the engine cooling system. Cooling system corrections required total redesigns of the engine cooling and engine start systems. Additionally, research of the RPH’s history revealed a need for a torsional shock absorber to be incorporated in the drive train to increase component life. The changes made to Hummingbird provide a decrease in empty weight, minimal center of gravity change and, most importantly, an increase in user safety furnishing the Department of Aeronautics and Astronautics with a dependable vehicle for rotary wing



ho LIN SU DLLCGS ye nn oe 1 Il. BACKGROUND INVESTIGATION .......... ccc cece ccc c cscs cccccces 3 Il. BASIS FOR MODIFICATIONS ..........-cccccccccccccccccccccccces 5 A. STARTER ..... Baa ca ssa! alaleeioletetstaia!'c's states save steers co etenevonereroerene 5

1. Original Configuration ..........ccc ccc cccccccccccccees 5

72 TATLEIMEFCMCIOMEIES 500s cc cc cscs cccs sc ecsceeewens oe oe

B. ENGINE COOLING SYSTEM .......... ccc ccc cece cece cence 7

1. Original Configuration .............ccccccccccccccceces 7

z Cooling System Deficiencies ..........cccccsssccccscces 8

C MP eanRy Im PLURRO AMIN o244 "s-axsisie-s:s ovaxerexenessnavexeie « 6 sininpe » 0 sleleiarens 6 a eeiereie s 13

i. Original Configuration ............ccccccccceccccccces 13

2: Drive Tram Deficieii€tes. ..........cccccsscccccccesene 14

IV. ENGINE BREAK-IN AND TESTING ....... 0... cece ccc cece csc ceneee 17 A. MINGHINE, PES TING SET-UP 2.2... ccs ccwccscccscvccsecsvces 17

1. imneennie WSs Gate 5c. sre aie o's 5 0 aisle «ch ete eerie si oe uae 17

3 Hand-held Engine Starter ............ ccc cccccccccccces 19

B. FINGINGE TEST PROCEDURES . . . . ... «5 sicisicssisuataleia eleieya eee oe cues 21


C. ENGINE TEST RESUEGSIH: 9. 826.500. cece c cece sccecess 22

V. BESIGN VIG IIICATIONS |. . oo 050000005 50 6c aiuieleieieisieiieisiers|eistelelee rere 27 A. OI eer nC 27

1. Design Considerations ...........ccccccccsccccsccccces 27

2; Starter Design Implementation ........ iets a ennpainoueuseenseneee 28

3. Starter Operation ...........ceeee. ee ee 33

B. ENGINE COOLING SYSTEM .......... ccc ccc ccc ccc ccccces «33

1. Design Considerations. .........cccccccccccccccccccces 33

he. Cooling System Design Implementation .............. 2. 34

a. CCV UO TRING io saisincpeisisiaiaieasers.wcigiaielss 4 <igs.« 0 014 34

(1) Upper Engine Mount. ............0.scccceee 36

(2) Adapter Installation. ................ ee 36

b. Component Removal ...........ccccececccccece 33

C; Coolim® Systt im iscciiein caps epee es seins « 0 o's.00 0 0 0 39

C. PRGA IML ERENCRIN ofc < (cle to ele aja /aanleselele'='« cists se ais leisieieisic's's oe «s/s sie's ¢ 41

1. Design Comsideratioms. « ... 6 /n:sjojn:e:«:0)0isisiiaisia)aieieiaiaisieieieie <0,010Ks 41

2. Shock Absorber Design .........cccccccccccccccccccces 41

IE ec ER MO NIC MIDEUSSMULINGS ~ 5.0 5 6 cai 5 6 0:0 's sicpauessintanein. « esse Misialeiegicicsaieces © + wicisia's « 45 VH. CONCLUSIONS AND RECOMMENDATIONS .........c-ccccccccccs 47


RECOMMENDATIONS ... 2... ccc ccc ccc cccccccsccccccccecs 49 1. Lelio Fis CN rod ae ee. re ant eae 49 2. AVL aC CO I oso sioaicicicccccccccccccweties 49

3. Complete Implementation of Torsional Shock Absorber ...50

4. Bower Plantes. ...2cccccccccccccccs CT me 50 5. Configure Hummingbird for Forward Flight ............ 50 6. Main Rotor Head Design .............. oe eee $1 7 NOTAR?® Research .......... WO Ss «See, OS eee Oat 51 Ai EINIDIXCAS TISTIOR SUPPLIERS 2.1... ccccccccccccnccccscccscssecen 53 APPENDIX R: POINTS OF CONTACT 3.2... cccccccccccccscccevccccsccs 55


ELECTRICAL SCHEMATICS ......... 2. cece ccc ccc cece 57


SEES 21 OOD ie ce aia ee wos a ee 65



@®e@eeeeeseeeeeeepeeeeeeeeeeeeeeoeneeoeeoeeeesaeoaweneaeaeeeee 8

@ee@eeeee¢ceceseseeeeeeeoweeneaweaeeeaeaseoeeeoeeesneoaneese*e7eeoeeee 6 @


Figure 1. Original Starting System Configuration .............cccccccccccces 6 Figure 2. Layout of Engine Cooling Components ..............cccccccssccccs 8 Figure 3. Existing Engine Cooling Cowling ............cccccccscccccccccces 10 Figure 4. Exhaust Manifold Position ..........cccccccccsecccccccccccccees 11 Figure 5. Right Side View of Hummingbird Without Ventilation Port .......... 12 Figure 6. Main Transmission ...........ccceee. ee a Batetare sanetetetets 14 Figure 7. Engine Test Apparatus ...........ccceeeees peer e ee ee cece sc eme Figure 8. WAEL 342 in the Test Configuration ........ Se 19 Figure 9. Hand-held Starter ..............000. ooo. Ree 20 Figure 10. CHT Comparison for Large and Small Cylinder Heads. ........... 25 Figure 11. Top View of Starter. ..... eee ee ee 28 Figure 12. Bottom View of Starter. ........ccccccccccccccecccccccccccccces 29 Figure 13. Starter Mounted to Landing Gear. ...........ccccccccccccccccecs 31 Figure 14. Starter Coupled to Engine. ............0cscccccccecccccccuccees 32 Figure 15. Size Comparison of Cylinder Heads. .............0-0-cccccccces 35 Figure 16. Upper Engine Mount. ........ 2. ccccccccccccccccccccccccccccess 37 Brea ES ACER TOON) siace « « oicunse.+ © ocgeie's' 6 6 Sxagagepigeemumueugys’s << <6 cus © meemagels oleae 38 Figure 18. Cooling Fan and Drive Configuration. ...............ccccccccceee 40 Figure 19. Intermediate Shaft of the Main Transmission. ................... 42 Figure 20. Layout of Shock Absorber Components. ...............ecceeceees 44


I wish to acknowledge several key individuals without whose help this project would not have been possible. First and foremost I would like to thank God for the strength and perseverance to get through these last nine months. I would also like to especially thank the Aerospace Engineering Technician Mr. Don Meeks. His knowledge and expertise in UAV’s was invaluable to me not having any background at all in model aviation. His dedication shown by working after hours until my work was completed was uplifting. Also deserving my gratitude is Dr. Wood. After all the setbacks and aggravations he was always there to help me gather myself back together, sort out the problem and move on. I would also like to thank my family for providing me with encouragement and understanding during the trying times of my thesis research. Finally, a general thank you to all those who I pestered

as I used them as sounding boards for my ideas and theories.



Recent achievements of the Naval Postgraduate School’s Aeronautics and Astronautics Department in helicopter research have emphasized the importance of the work done here. The outstanding performance of NPS design teams in the American Helicopter Society’s annual helicopter design competition , valuable student and staff research and exposure in local media and professional publications have positioned the Aeronautics and Astronautics Department as a leader in helicopter research. Among the many resources of the department is the NPS Hummingbird, part of the remotely piloted helicopter (RPH) research program. The Hummingbird is a unique rotary wing aircraft that possesses characteristics well suited for scale model research in Higher Harmonic Control (HHC), NOTAR?® , and other rotary wing fields. To date there have been at least three inquiries from outside sources for future experimentation on the NPS Hummingbird. In order to comply with these and other requests, several design deficiencies have been corrected to bring the aircraft from its original configuration as a target drone to its current status as a reliable rotary wing test platform.

Thus far, necessary design modifications to the airframe, main rotor and transmission have been completed and implemented in order to have a reliable RPH suitable for quality research. Due to an overheat failure of one of the Hummingbird’s two inventory Weslake Aeromarine Engines Limited (WAEL) 342 engines, ground and flight testing was halted until an investigation as to the cause of the overheat and subsequent corrections could be

made. The WAEL 342 engine is a 342 cc, two-stroke, simultaneously firing, twin cylinder

gasoline engine produced by Target Technology Ltd. in the United Kingdom. It possesses a maximum power rating of 25 hp at a rated speed of 7000 rpm and a maximum torque of 24 ft-lb at 4000 rpm. It is an ultra-lightweight power plant designed, developed and manufactured specifically for remotely piloted vehicles (RPV) and unmanned aerial vehicle (UAV) installations. The scope of the following research is to investigate current power plant deficiencies and to provide adequate solutions for effective engine cooling and drive train reliability in order to avoid future costly delays in RPH research.

Modification of the engine cooling system included increasing engine cylinder heat dissipation and improving interior fuselage ventilation. To facilitate the cooling system improvement a redesign of the engine starting system was implemented which allowed for a significant increase in payload capability and a decrease in gross weight while, most importantly, increasing user safety. A drive train modification consisting of the design of a torsional shock absorber was also required to prolong the life of the drive train and main rotor components by protecting it from observed engine torque impulses. The ultimate goal of this study was to modify the current power plant and drive train to provide sufficient power and man airframe dependability and safety while keeping gross weight changes,

center of gravity shifts and modification costs to a minimum.


In order to effectively trouble shoot the engine cooling problems, it was first necessary to obtain information about the WAEL 342 engine. There was no supporting documentation except an engine operators manual which was included with the RPH and spare parts from Mr. John Gorham, the original designer. The engine operator’s manual had a company name, Weslake Aeromarine Engines Ltd. in the United Kingdom, but no address or other points of contact. Investigation began by tracking down the engine listing in the 1980-1981, 1981-1982 and 1982-1983 editions of Janes’ All the World’s Aircraft. Weslake Aeromarine was the company that buult the engine as of that year and a telephone call was put through to find out more, current information. It was learned that Weslake was bought out and included as subdivision of the company Normalair-Garrett Ltd. who had sold the manufacturing rights of the WAEL 342 to a company named Target Technology Ltd., also in Great Britain. Information concerning the engine’s performance, cost and support was requested from this company. and received via facsimile.

The engine documentation stated that the WAEL 342 engine was designed for external use only m air streams of approximately 150 kph or 93 mph. This definitely meant that some sort of cooling system designed specifically for the Hummingbird’s internal use was required. The fax also contained information on engine performance and a larger cylinder head that was used for improved cylinder cooling purposes. Included price

information made it absolutely clear that purchasing a replacement engine was to be a last

resort. A thirty day price quote on 25 April 1996 put the WAEL 342 at £1855.85 or $2969.36 per unit for a quantity of one to nine engines.

An American affiliate of Target Technology Ltd. was also listed in the fax. Southwest Aerospace in Tustin Ca. was contacted to find out more information about the WAEL 342. A Mr. Ian Matyear was able to provide data on purchasing the WAEL 342 engine and its components. He also provided another source of information, Mr. Ken Beckman, who had vast experience with similar size RPH’s. Mr. Beckman proved to be an invaluable source of information as he was quite familiar with the Hinmingbird’s original design. During the RPH’s initial development as a target drone for the U. S. Army, Mr. Beckman reported on the status of this project to Boeing Aircraft Company, the primary contractor, as to the progress of the subcontractor, Gorham Model Products. He mentioned that the cooling system installed was an after thought as engines before had failed due to overheating and also described many of the other deficiencies as they had existed prior to NPS student’s modifications. He also mentioned that there existed “wild” torque fluctuations in the engine and that there was a possibility of main rotor and drive train damage and, most importantly, a potential safety hazard. Reports of engine runaway and main rotor separation due to excessive shock and vibrations were among his so told cautions. He suggested that some sort of torsional shock device be installed to remedy this situation.

In all, the above investigation proved to be very necessary and of great benefit in providing adequate solutions to the current problems. Mr. Beckman’s information and advice from his first hand experience with this RPH was an inestimable value to this project.

The above sources will be of great help throughout the life of the Hummingbird program.


As previously stated, in order to successfully modify the Hummingbird’s engine cooling system it was necessary to include improvements and redesigns of the starting, engine cooling and drive-train systems. Investigation of the Hummingbird’s history, both prior to and after the Naval Postgraduate School’s acquisition, and close inspection of the drive-train layout yielded observation of several design weaknesses. The following is a

discussion of the deficiencies in the original power plant and drive train configurations.


1. Original Configuration

The original starter was a permanent attachment to the Hummingbird. The starter motor was a 12 volt electrical motor which was hard-mounted to the airframe close to the forward engine cylinder. The starter motor shaft was fitted with a 1 inch diameter sprocket which drove another 7 inch sprocket through a chain drive providing a 7:1 mechanical advantage for the starter motor. A one-way bearing on the 7 inch sprocket provided for starting force in the counterclockwise direction and freewheeling in the clockwise direction as viewed from the top of the RPH. Two DC power cables ran from positive and negative cable receptacles in the left side of the forward fuselage to the starter motor. Figure 1 shows

this configuration.


Figure 1. Original Starting System Configuration

To start the Hummingbird, external power cables were hooked up to a 12 volt automobile battery which provided power to the starter motor. Once achieving a starting rpm of approximately 1000 rpm the WAEL 342 cc gasoline engine ran independently. The external power cables then had to be physically pulled out of the receptacles from approximately ten feet away.

2. Starter Deficiencies

The starter motor was found to be under powered. Prior starting attempts required two fully charged 12 volt marine batteries connected in series (24 volts total) for the electric —_ to provide enough starting torque to the engine. This configuration ran the risk of

burning out the starter motor and causing an electrical fire on the airframe.

Removal of the power cables risked entanglement of the cables in the main rotor system as the cables were free to whip as they were pulled from the fuselage. The main rotor arc extends 5 feet from the rotor hub and sits approximately 3 feet off the ground. Entanglement of the cables is an unnecessary risk to the program and a safety hazard.

The starter motor and its associated hardware weighs approximately 8 pounds or approximately five percent of the Hummingbird’s advertised gross weight. Elimination of this weight would provide an attractive thirteen percent increase in payload capability or compensate for the weight of the modifications.

Finally, the starter which was mounted very close to the forward engine cylinder and the 7 inch sprocket mounted on the lower engine drive shaft restricted cooling air from flowing smoothly over the cylinder’s heat fins. A fiberglass cowling that directed airflow over the engine had to be modified by cutting away an approximately 3x5 inch section in order to accommodate the starter motor installation which reduced the engine-cooling system efficiency.


1. Original Configuration

The configuration of the engine-cooling system was as follows. A 6 inch diameter vane-axial impeller was mounted to the engine drive shaft which rotated at speeds between 3000 and 7000 rpm. A crudely manufactured cowling was installed over the engine and impeller with the top of the impeller exposed just above the cowling. The cowling was constructed of fiberglass and fit closely around the engine. A make-shift diffuser, similar

to those found in smaller scale model helicopter cooling systems, was molded into the top

of the cowling to provide an increase in pressure to reduce losses in the system. The air was directed out the bottom of the fuselage and into the atmosphere. Figure 2 shows the layout

of the cooling system components.

Figure 2. Layout of Engine Cooling Components

2. Cooling System Deficiencies

The original design of the Hummingbird’s engine-cooling system was proven inadequate by failure of the engine due to overheating. The engine manual states that the maximum cylinder head temperature measured at the spark plug gasket is not to exceed 482

F and maximum exhaust gas temperature 1s not to exceed 1022° F. Upon inspection of the

failed engine, deep scoring was found in both cylinders. The pistons were locked into the cylinders and unable to be removed. Deposits of metal were also found fused to the cylinder walls pointing to a massive overtemp of the engine. A partial reassembly of the power plant and drive train provided clues to the overheating problem. The engine manual specifically states: CAUTION:





MUST NOT BE EXCEEDED. The “cooling air-flow” was insufficient for the following reasons. First the cooling system was designed to draw air in to the center of the engine and then to direct it out along the cylinder. The cooling fins were perpendicular to the direction of flow causing the airflow to be disrupted as it moved further away from the center of the engme. Sufficient airflow to the cylinder head was therefore not available. A significant amount of the fiberglass cowling that directed the airflow over the engine had also been trimmed away to accommodate the starter and exhaust components. In order to mount the starter an approximately 3x5 inch square had to be removed. Other cut-outs for the engine exhaust, decompressors (small ports mounted on both cylinder barrels to aid in engine ignition) and

mounting hardware, shown in Figure 3, had widdled away at the intended design rendering

this component ineffective.

- - 7 : Gs * * i Rud

pa Ad i* Hg Bis - dpi a z J 4 oi ie. his! Lilet ot eae

Figure 3. Existing Engine Cooling Cowling

One of the main contributors to the cooling problem was the engine exhaust system. The exhaust system consisted of an exhaust manifold and a 12 inch long, 1 2 inch diameter flexible steel tube. The manifold collected exhaust from both cylinders and provided limited noise muffling. The manifold is constructed of stainless steel with an attachment for the flexible tube m the rear. The tube ran from the mnntniifoldl around the rear of the engine and exited from the bottom of the fuselage. The maximum allowable cylinder head temperature is 482 °F and the maximum allowable exhaust gas temperature is 1022 °F. As will be discussed later, average observed exhaust gas temperature with proper fuel-air mixture 1s 600 °F to 700 °F. The manifold and exhaust pipe were radiating 600 °F to 700 °F over 182.7

square inches inside a mostly enclosed fuselage with virtually little outside air entering the


fuselage interior during ground runs aside from an insignificant amount of main rotor wash.

The exhaust manifold was also positioned just 2 inches from the intake of the cooling fan

as seen in figure 4.

ie eer

Figure 4. Exhaust Manifold Position

The result was an intake of cooling air with an equal or higher temperature than the maximum cylinder head temperature. |

Prior ground tests of the Hummingbird at NPS were conducted with the fiberglass front body shell off to allow the maximum heat dissipation and ventilation possible. ritiedieieb of RPH’s similar to the Hianmingbird show ventilation holes cut in both sides of the front body shell. The Hummingbird, however, only had one hole cut in the left side.

Figure 5 shows a right side view of the RPH with no ventilation hole cut in the fiberglass



a - aoe tee ee j : iY 2: i ta4 t yk ee S | jaan | jaek = foe !

mm 8 ; §

Figure 5. Right Side View of Hummingbird Without Ventilation Port body shell and the exhaust manifold’s position just inside the fuselage. The solid fiberglass

shell on right side allowed a build up of the exhaust manifold temperature which could have further increased the cooling air temperature at the intake of the cooling fan. The object seen in the hole in the fuselage in Figure 6 is the exhaust manifold.

One final cause of overheating was an improper engine operation and improper adjustment of the eaaiiee fuel mixture in the carburetor. The engines received from Gorham Model Products had no information as to total engine time and previous carburetor adjustment settings. The damaged engine was most likely a new engine requiring a two hour

break-in period which was not accomplished. The carburetor’s high and low speed needle


jets, initially set at the factory and possibly moved during shipment, were not checked for proper adjustment. Cylinder head and exhaust gas temperatures were seen to vary a great deal (+100° for cylinder head and +150° for exhaust gas temperature) during the engine break-in period carburetor adjustments. Thus, maladjustment of the carburetor settings can

very easily cause an overheat problem.


1. Original Configuration

The drive train begins at the engine drive shaft. The vane-axial cooling fan and centrifugal coupler were mounted on a shaft extension and a drive pulley is attached to the top of the centrifugal coupler. A drive belt connects this pulley to an intermediate shaft to which a sprague clutch is mounted for autorotational capability. Another belt drive is connected to a pulley which is keyed to the main rotor shaft. The tail rotor take-off is driven by bevel gear mounted to this pulley. Total gear reduction is 10:1 from the engine to the main rotor shaft and 3.2:1 to the intermediate shaft. The tail rotor drive shaft turns at 40%

of the engine shaft speed. Figure 6 shows the main transmission layout.


Figure 6. Main Transmission

2. Drive Train Deficiencies

As mentioned in the background investigation chapter, massive torque fluctuations due to unsteady idle speeds exist. These torque fluctuations were reported to, however unverified, snap the main rotor shaft of a similar RPH. In any case, fluctuations in engine torque are magnified ten times at the main rotor shaft due to the 10:1 mechanical advantage given by the transmission. With the exception of the hard rubber belts and limited slipping

in the centrifugal coupling, there is nothing to absorb any sort of shock caused by torque


fluctuation in the unmodified drive train. The incorporation of a shock absorber was determined necessary in order to prolong drive train component life and reduce airframe




The engine tested was a brand new WAEL 342 with no operating time accumulated. The engine manual requires a two hour break-in period before it is put to any application. A break-in schedule consisting of several low power runs at short time durations (5-10 minutes in length) and varying power runs at longer time intervals was conducted. In conjunction with these muns, engine cylinder head and exhaust gas temperatures (CHT, EGT) were monitored to determine the relationship between power and the measured temperatures, allow the correct setting of the high and low speed needle jets on the carburetor and provide clues as to how to contain the temperatures while mounted inside the helicopter’s fuselage. A. ENGINE TESTING SET-UP

ne Engine Test Stand

The engine was mounted on a test stand designed to measure thrust and rolling moment of the AROD UAV. The rolling moment was the only measurement taken and allowed the determination of engine power. Thrust measurements were not needed for the performance calculations. The moment was computed by reading a force gage which was mounted at the end of a 9 inch moment arm, as seen in Figure 7, and then converted to horsepower after obtaining engine rpm by a strobe tachometer. To provide a working load during the testing and break-in procedures a 30 inch diameter birch propeller (commonly used in ultra light aircraft applications) was installed. Figure 8 shows the engine rigged for

testing. The direction of rotation of the engine was opposite to the direction of rotation of


1@50 Ib and


8.02 00.6



ES sores eretaleseree seenscersnestoteeatetmtenstecntetetgestse neta! Mona’ n". *nteP any’ a nt ater’ ove’ e naa tate a's Simin a's ss a


A te nen? # @ Max a ory

Saree ae sreP oe oekel



1@5Q Ib and 1@100 Ib Toe] THRUST FORCE GAGES

* on ata terete tat ete etetn tn as 210 's's


| | SQA


eron ere. wreterere


Figure 7. Engine Test Apparatus


Figure 8. WAEL 342 in the Test Configuration

the propeller. For this reason the propeller had to be mounted backwards. A one gallon fuel tank provided a 25:1 fuel-oil mixture to the engine. As a precaution a 24 inch diameter room fan was placed to direct airflow over the engine. This later proved to be ineffective. 2. Hand-held Engine Starter The engine was started by a hand-held starter which contained a new starter motor shown in figure 9. With the need to start the engine in a counter-clockwise viewing the

engine from the rear on the test stand, a starter with this rotation was sought. Also at this


Figure 9. Hand-held Starter stage of the research the direction of rotation of the external starter design had not yet been

decided. Therefore a starter with the capability of providing adequate torque in both directions was prudent. The Prestolite MBJ-4407 winch motor was chosen for its availability, power and dual directional capability. The performance specifications for this motor can be found in appendix C.

The coupling of the starter to the engine shaft was accomplished through the use of a hex-ball wrench on the starter and a hex-socket bolt mounted to the propeller. The hex- ball gives the advantage of providing constant torque while allowing small misalignments

of the starter shaft. The hex-ball also does not jam into the socket which becomes a large


safety feature when hand starting a 25 hp engine. A 3/8 hex-ball was pressed into a cylindrical block of aluminum. The bottom half of this 2 inch block was bored to fit the ‘ies motor’s 3/4 inch drive shaft. The wrench coupling was then pinned to the starter motor shaft. A mounting plate for the starter motor, a momentary contact switch, solenoid and two handles were fabricated and the above components assembled. A set of automobile jumper cables was modified by adding eye terminals to one set of ends of the jumper cables. The electrical connections were then wired to the assembly in accordance with the supplied wiring diagram. The hand-held starter is shown in Figure 10.

Thermocouples were placed at the exhaust pipe and sparkplug to measure the EGT and CHT, respectively. The thermocouple leads were connected to a digital readout which displayed the two temperatures on two separate channels. Tests were conducted in static air conditions with the only air flow over the cylinders being a low velocity induced flow caused

by the propeller. With the reversed mounting of the propeller, the induced air flow was drawn over the engine cylinders and thrust forward. Finally, a “kill” switch was installed to ground the electrical connection from the magneto to the spark plugs enabling controlled

engine shutdown.


For each day of engine testing, the ambient air temperature was recorded. The safety procedures in the engine manual were then reviewed. The engine starting checklist also in the engine manual was then followed for engine start. The starter rig was connected

to a 12 volt marine gel-cell battery via the jumper cables and the hex-ball wrench drive was


inserted into the hex socket on the propeller’s hub. The switch on the starter was thrown and the starter held in place until the engine fired. Once running the engine temperatures were recorded at various engine rpm.


Although ambient air temperatures were recorded for each day of engine operation, the difference in the temperatures from day to day was proportionally insignificant to the recorded engine temperatures. Therefore all engine temperature information assumes an average ambient temperature of 65 °F.

The first engine run was conducted only to start the engine for break-in and to observe the engine and engine temperature behavior. Minor carburetor adjustment was made to obtain behavioral information also. This run revealed an idle rpm of about 3100 rpm and idle cylinder head and exhaust temperatures of 235°F and 594°F, respectively. A maximum throttle setting was briefly set. This power setting showed a maximum rpm of 4400, a maximum exhaust gas temperature of 618°F and a cylinder head temperature that would have greatly exceeded the 482°F maximum if left to continue operating. This power regime displayed evidence of the overtemp experienced by the failed engine and a requirement for cylinder cooling.

Seeking a balance between temperature limits and reducing engine operating roughness, the high speed needle jet on the carburetor was set to minimize maximum power cylinder head temperature through a slightly rich air-fuel mixture and minimize roughness at idle throttle settings through a slightly lean mixture. The needle jets were adjusted by

rotating them by 1/8 th of a turn and noting the results until a favorable condition existed.


The WAEL 342 is rated at 25 hp maximum power at 7000 rpm. After calculating the brake horsepower from the torque reading it was determined that the engine was delivering maximum power due to the propeller load at 4400 rpm. Until the final carburetor settings were achieved, the CHT and EGT were seen to vary as much as 100 °F and 150 °F, respectively. Subsequent runs on the engine began to show consistent temperature behavior with proper carburetor adjustment. The chart below shows the typical results for the properly adjusted engine.

Typical CHT and EGT at Various Engine RPM

See | araiace, [Had ven 00 Ga


The maximum cylinder head temperature was unable to be contained but was reduced to a

slow creep through 482°F. The room fan use was discontinued after noticing that there was no difference in cylinder head temperature with or without the fan running. The engine exhausted the one gallon fuel supply in 35 minutes putting the fuel consumption rate at 11.7 lb/hr which is consistent with the engine performance data provided by Target Technology Ltd..

To see the effects of forced air over the cylinder a backpack-type leaf blower was borrowed from the greens keeper shack at the NPS golf course. The blower provided an

advertised 150 mph air flow through a 4 square inch opening at full throttle. At a full


throttle setting on the WAEL and a mid to full throttle setting on the blower, the CHT was contained well below 482°F at approximately 420 - 450. The results of this experiment revealed that it was possible to contain the temperature through a forced air device.

A pronounced torque fluctuation as described by Mr. Beckman was also evident. These fluctuations were load driven as they were seen to be more pronounced at low loads at lower rpm than at high loads. The fluctuations could not be determined during engine operation due to the test configuration; however, once the engine was shut down telltale scoring on the scale face showed an approximate +10 ft-Ib fluctuation at idle power and +1 to 2 ft-lb fluctuations at full power. This multiplied by the 10:1 mechanical advantage provided by the transmission can mean a +100 ft-lb fluctuation at the main rotor mast. This observation confirmed the torque fluctuation claims of Mr. Beckman.

During the engine testing phase the oversized cylinder heads were received. Several modifications to the engine had to be made in order to accommodate the increased size of the heads. The spark plug wiring had to be lengthened to reach over the new head to the spark plug. A spark plug cap retainer had to be manufactured to keep the spark plug cap from sliding off the longer plugs due to the engine vibrations. Finally, longer cylinder head bolts were installed to compensate for the increased head thickness. Once the new cylinder heads were installed, the engine was run to observe the effects of the increased area on CHT and EGT.

The engine was again tested in static air conditions. Runs initially at idle power settings showed an approximate 25°F decrease in CHT and, as expected, no change in EGT.

Runs at maximum throttle showed a maximum CHT of 435°F. The effect of the increased


area of the cylinder head reduced the maximum CHT by over 47°F in static conditions. A

comparison of the EGT and CHT observed with the two sizes of cylinder heads is shown in

Figure 10.

po MXC Standard Cylinder = ] Pe a ee

3000 3500 4000 4500 RPM






\ mi


Figure 10. CHT Comparison for Large and Small Cylinder Heads

This determination moved the cooling system design away from a high mass flow system to an interior ventilation or low mass flow system in which the induced air flow velocities

over the engine on the test stand would be matched or increased inside the fuselage.


The engine break-in was accomplished without incident and approximately 3 hours and 20 minutes of engine operating time were accumulated. The engine performance information provided by Target Technology Ltd. was also verified as being consistent with observed engine performance. Most importantly it was shown that the larger cylinder heads provided enough heat dissipation to contam the maximum CHT at full throttle settings in

static air conditions.



The following modifications were made in order to overcome the previously discussed deficiencies. The general design constraints throughout were to minimize complexity, minimize weight and CG changes, minimize cost and increase user safety.


i: Design Considerations

Looking at its airflow obstruction effects and inability to perform reliably, it was decided to completely redesign the engine starting system. A major factor considered along with overcoming the existing deficiencies was the requirement to start the engine on the UAV test stand. Most UAV engines including the WAEL 342 in a fixed wing application need a counterclockwise torque for start. However, due to mounting constraints on this test stand, a starter must be able to provide starting torque in a clockwise direction as one faces the propeller. This demonstrated early on m the design that a starter capable of providing torque in both directions would be practical.

As mentioned earlier, removal of the starting system’s eight pounds would provide an attractive 13% increase in payload weight or compensate for the weight of the modifications. This benefit coupled with an assured starter motor weight increase due to the increased power requirement and dual direction capability drove the starter system redesign to an external configuration.

Along with the external design requirement was the need to accomplish an engine

start remotely and not to have the starter interfere with flight operations. Some sort of drop


away device would be useful and would allow the starter assembly to be removed from the operating area safely without possibility of main rotor entanglement or other interference. The final design provides a starter capable of starting Hiznmingbird on the ground for flight operations or on a test stand such as the one acquired in LT Booth’s thesis [ref. 5].

2 Starter Design Implementation

The starter assembly’s final design is as shown in Figures 11 and 12. It consists of a 0.25" x 9" x 20" base plate to which the driving components are mounted. The starter motor