of a Mustang II Experimental Airplane
MAS 603 -Aircraft and Spacecraft Development, Section 1
November 19, 1999
This paper briefly describes one individual's experience in the building, modification and flight testing of a Bushby Mustang II experimental airplane. Some of the thought and decision processes as well as technical challenges and psychological obstacles are discussed. Major sections include: Background, The Building Process, Modifications and Features, Certification, Flight Testing and Summary. It is intended that a current or prospective builder of such a consuming project as building an experimental airplane may learn by another's experiences, encounters and mistakes so as to anticipate or avoid the same in their own endeavors.
The Midget Mustang was designed in 1948 by David Long and was one of the first all metal high performance homebuilts. In 1959 Robert 'Bob' Bushby bought the design rights to the Midget Mustang and in 1965, he designed and built a two-place aircraft based on the Midget called the Mustang II. In 1992, Bushby sold the design rights of both aircraft to Mustang Aeronautics, which now sells the plans and kits with numerous fabricated components (Mustang Aeronautics, 1999).
"Experimental aircraft" actually encompass several different types of airworthiness certificates issued by the Federal Aviation Administration (FAA) under Federal Aviation Regulation (FAR) 21.191 for such purposes as Research and Development, exhibition and air racing aircraft, etc. However, paragraph (g) "Amateur-built aircraft"- provides for "Operating an aircraft ... assembled by persons who undertook the construction project solely for their own education or recreation" (Code of Federal Regulations, 1995). It is under this paragraph that this particular aircraft was licensed.
The term "amateur-built", "homebuilt" or "experimental" aircraft, as these aircraft are also frequently referred to, may conjure up a vision of an aircraft that does not inspire confidence in quality, thus the term "custom built" aircraft is preferred by some (Alexander, 1999).
The builder's skill level, learning curve and rationale for certain modifications may be put into perspective by explaining that he was a first time builder, holds an Airframe & Powerplant (A&P) mechanic license with several years experience working on light piston aircraft, and is a professional pilot with a general aviation background.
RATIONALE TO BUILD
It is so important for prospective builders to assess thoroughly and objectively their reasons for building, the gains they expect to make, and the purpose the aircraft will serve. It requires a definite commitment of money, hard work and time - on the individual's part and on the part of the family (Alexander, 1999). The significance of the commitment required can not be overstated.
"In building your own airplane, there is an unmatched feeling of accomplishment" (Bingelis, 1979). It is a satisfying and rewarding experience to know that you have constructed the airplane you are flying. Additional reasons include saving money in comparison to purchasing a factory built aircraft with the same performance, the opportunity to learn the mechanical aspects of the airplane with the related safety benefits, and the ability to maintain the airplane without an A&P license. You also become a member of an elite group of individuals (Alexander, 1999).
INITIAL DECISIONS and ESTIMATES
First was the selection of the type of construction, the choices being: all metal, steel tube and fabric, wood, or composite (fiberglass). The builder did not wish to assume the risks involved with having a primary structure of fiberglass and the health hazards involved in its construction. The performance advantages and durability of all metal construction was preferred over the other materials in spite of the additional tools and slightly greater skills that would be involved.
The desire to be able to carry a passenger left the two-place all-metal model choices of: Van's RV-4, the Thorp T-18, the Smythe Sidewinder, and the Bushby Mustang II. It should be noted that now the RV series has additional models available, but the builder's desire for a thin, sleek, laminar flow wing would eliminate them as options.
The original estimated build time by Bushby Aircraft for the Mustang II was 1,200 hours in two years at an estimated total cost of approximately $13,000 (in 1983 dollars). It is recalled that at that time, the prospective builder's father - a Program Manager for 30 years - (R. Henry, Jr., personal communication, September, 1983) said to "triple any time or cost estimates", which turned out to be an amazingly accurate statement.
Today, it is quite interesting to hear prospective builder's estimates for their project completions. This project's actual numbers were - 3,500 hours in 7 building years (15 calendar years) and approximately $30,000 (in 1998). Many of the reasons for the differences are discussed in this paper, including the numerous features and modifications that were added. The builder is unaware however, of any living Mustang II builder that claims to have spent less than 2,500 hours on its completion. In fairness to Mustang Aeronautics, the current kit manufacturer's estimate of 1,000 hours for the "fast build" basic airframe kit (with completed wings) seems quite realistic. The kits now available include many welded and finished parts including the spars and landing gear, significantly reducing the time required from the "plans built" version discussed here. It should be pointed out though, that once the basic airframe is complete and appears "90% complete", as the saying goes, "you're almost half-way there". The installation of the engine, cowling, canopy, systems, instruments, avionics, interior, fairings and paint will take a minimum of another 1,000-1,500 hours.
A trip to Oshkosh to talk with other Mustang builders and one to Washington's Air and Space Museum library in August, 1983 were made to learn as much as possible about the Mustang II. The decision to build the Mustang II was made August 28, 1983. The plan was to fly it to Oshkosh in 1985.
THE BUILDING PROCESS
The plans arrived September 15, 1983, and the "kit" (materials package) on October 4, 1983.
Construction began October 21, 1983. From the beginning, complete documentation in the form of a builder's log, time log, financial records, and photographs were kept. Today, a summary is available on-line at: http://www.experimentalairplane.com.
ROLE OF THE BUILDER
The builder acts as a one-person aircraft manufacturing design-build team and actually serves many roles: as planner, designer, engineer, manufacturer, mechanic, welder, painter, and finally, pilot. He must become proficient in tooling, sheet metal, fiberglass, Plexiglas, upholstery, avionics, electrical, engines, hydraulics, plumbing, and most importantly - quality control. Technical writing, legal and certification issues, procurement, logistics, finance and insurance are additional areas encountered by the homebuilder. Assistance was required at a few points - when riveting the forward fuselage, tailcone and wings. 8.2% of the building time was accomplished by assistants. Many builders elect to subcontract welding, painting and at least some of the other builder functions.
The requirement to develop a level of competency in all of these areas may prove to be an insurmountable challenge to some that embark on the process without considering all the skills and learning that is involved. Probably the most notable element of the process - as was pointed out by several observers - is the need to be a designer/engineer many times during the aircraft development. It really is a process of development, not just assembly, since every amateur-built airplane is unique. The plans do not spell out all the details. In many cases, for example, how and where to run the plumbing or electrical lines is entirely up to the builder. The FAA's Advisory Circular 43.13-1A (Federal Aviation Administration, 1972), along with the series of Sportplane Construction books by Tony Bingelis (Bingelis, 1979, 1983, 1986, 1995) are invaluable in solving design and installation problems to airworthiness standards.
"The project was not one of building an airplane so much as one of building parts. By breaking the work down into individual parts the project is not so overwhelming" (Cox, 1975). These words continued to provide encouragement, since the greatest challenge in a project of this magnitude is to not quit. There were many times when it became more of an ordeal than the recreation or fun as originally envisioned.
Several learning curves were encountered; the two most time consuming being metal working and riveting techniques. Other skills, including steel working and welding, took time to master as well. Today, the kits available include most welded parts. Error correction was another significant challenge. Errors in plans and building errors, such as the need to replace poorly driven rivets and scrapped parts, made for considerable difficulties near the beginning of the project, but tapered off as proficiency increased.
"Builder block" is a phenomenon of boredom, confusion or discouragement commonly experienced by builders, which can slow or halt the building process. When experienced, it was overcome by changing to working on a different section or system. Additionally, the recognition of the difference between "perfection" and "satisfactory, safe and airworthy" was crucial to making progress. EAA Technical Counselor Don Eaves (personal communication, February 3, 1995) provided encouragement when he said about his airplane: "It's made for go, not for show", putting into perspective the unimportance of minor aesthetic imperfections.
Other psychological challenges included the balancing and proper prioritizing of one's personal, family and professional life with the project. The project frequently was the lower priority and consequently was abandoned in mid 1985. (Project total time - 450 hours.) The project was sold and shipped to St. Croix but was never worked on. In late 1990, after it survived Hurricane Hugo, the project was reacquired, and shipped to Tennessee. In 1991 and 1992 other airplane restoration projects were accomplished, and in 1993, steady work on the Mustang II was restarted.
A "project management" approach was taken and a timetable plan listing expected section completion dates was made and posted in the workshop. First flight was projected to be January 1, 1997. The actual date was 18 months later, mostly attributed to unanticipated employment priorities. Also posted were "to do" lists that broke the major sections or assemblies down into outline form listing the individual tasks as manageable milestones. This resulted in the psychological accomplishment of an item getting a line drawn through it after every ten to twenty hours of work. Attending airshows during the building process resulted in several productive bursts of motivation.
Damage assessment, corrosion removal, and the preservation of aluminum and steel was the emphasis for the first few months after reacquiring the project from the Caribbean. The center section was constructed in a vertical jig, and the flap and control stick assembly and flight control tubes were completed. The center section was then put into a horizontal jig and the forward fuselage components were fabricated and installed.
In April 1994, the forward fuselage was complete and the tailcone was started. (Project total time - 1070 hours.) On January 21, 1995, the empennage was complete, the airframe removed from the horizontal jig, and the aircraft placed on its landing gear for the first time. (Project total time - 1400 hours.) The first EAA Technical Counselor visit and inspection took place February 3, 1995. The instrument panel and systems were designed, and wing construction began after building their jigs. The builder's Twin Comanche was damaged by a windstorm and was salvaged in the summer of 1995.
It was during this time when alternating between sheet metal work, systems installations and Flight Manual development was found to keep the project interesting and considerable progress was made. On October 16, 1995, the engine was hung on the airframe, on the airplane's fourth engine mount. (Project total time - 1698 hours.) Wing tanks were sealed, flight and engine instruments and electrical and avionics wiring was installed. Fuel system components were installed and tested. On May 25, 1996 the airframe metal construction was complete. (Project total time - 2003 hours.)
Propeller installation, oil cooler assembly and the remainder of the engine accessories, plumbing and instrumentation was completed. On October 20, 1996, the engine was run for the first time - disappointingly, only four of the seven engine gauges worked. By the third engine run a month later, all gauges were working. All avionics and antennas were installed and tested satisfactorily. The Twin Comanche cowling was modified to fit the Mustang II and an engine cooling plenum chamber designed and built.
The second EAA Technical Counselor inspection was on January 7, 1997. Flight controls were painted, cowling completed and engine with cowling was run in June 1997. (Project total time - 2666 hours.) The canopy and windshield installation, tailcone baggage compartment modification and interior installation took place the last half of 1997. (Project total time - 2986 hours.) Numerous "small" tasks such as fiberglass fairings, fresh air vents, heater system and glareshield installation filled the first quarter of 1998. On March 31, 1998 the last rivet was set in a wheel fairing and the Mustang II fabrication was complete. (Project total time - 3202 hours.)
The aircraft was primed, sealed, painted and final assembled except for attaching the outer wing panels. After many hours were spent masking, the ChromaLusionTM stripes were applied and painting was complete. (Project total time - 3444 hours.) On June 3, 1998, the airplane was transported to the airport and final assembly begun. The pitot-static system, altimeter, transponder and encoder were certified, and with the help of the EAA Technical Counselor, the aircraft was weighed, CG determined and final inspection conducted. (Project total time - 3502 hours.)
Incorporating the use of several different tools resulted in marked production efficiency. A pneumatic squeezer for both riveting and dimpling vastly improved quality and speed while allowing work to be accomplished by one person almost silently in the middle of the night. Additional tools found to be very helpful were a frame riveter, eight-foot rulers, and a digital level. A radial saw with a 10-inch carbide tip blade was used to cut aluminum angle and bar stock, while a ten inch metal cutting blade was used to cut all steel - with very smooth edges. A hacksaw was rarely needed. Switching to 3/32 inch rivets in place of 1/8 inch rivets for the wing skins made riveting much easier - even after considering that 75% more were required. A disaster turned into a positive result when the builder's Twin Comanche was totaled by a windstorm. Much of the salvage from it was used in the Mustang, including the flight instruments and avionics, engine and its accessories, propeller, and most of the cowling. Using these parts simplified those installations.
MODIFICATIONS and FEATURES
Retractable Mustang II builder John Thompson said "Either stay with the plans or design your own airplane" (Kessler, 1975, p. 56). Any changes to the plans will have an impact on project completion time, and if modifying the original airframe design, may likely involve unanticipated problems or additional redesigns later. Any structural change needs to be carefully considered. In spite of the potential problems and extra time required, modifications or added features do have their advantages. The distinction should be made - for the purposes of this paper - that an "airframe modification" is a structural change to the original plans, a "performance modification" is intended to reduce drag or increase engine efficiency, while a "feature" is an additional "nice to have" item.
Kent Paser, the author of "Speed with Economy" (Paser, 1975) says, "A cool-running engine and accessories promotes safety, reliability, and long life", thus, certain mods were highly desired. The performance mods below were incorporated from his book:
Engine plenum chamber (instead of baffling), crossover exhaust shield for forward accessories, cowl flap, oil cooler cowl flap, wing root fairings, VOR and GPS antennas inside the canopy, and removable wing tiedown rings. Six months after the aircraft was flying, 100 hours was spent constructing fiberglass gear leg and brake fairings. Unfortunately, only a two-knot cruise speed increase was realized. Following is the list of additional features, modifications and deviations from plans:
T-18 canopy; roll bar positioned 2 inches aft of plans location; redesigned canopy latch; redesigned IFR instrument panel layout, 2 inches aft of plans location; C-150 seats recovered in leather on Cessna seat tracks with fore and aft limit locks; tailcone baggage compartment (6 cubic feet); flap control of July 1980 newsletter design; control stick assembly made from rectangular tubing (as in Feb. 1980 newsletter); wet wings, 18 gallons each (plus 25 gallon main tank) Total fuel = 61 gallons; Sky Sports capacitance fuel quantity system; fuel selector system to transfer wing fuel to main tank or pump directly to engine; modified Hoerner style wing tips with strobe and nav light units; landing and taxi lights- (rectangular automotive Halogen bulbs -GE H4701); Cleveland "super heavy duty" 5.00 X 5 magnesium brakes and wheels; Piper Twin Comanche cowling (highly modified) with cowl flap; stainless steel and asbestos exhaust tunnel shield; fresh air inlet NACA ducts and eyeball vents; Piper Arrow power quadrant.
Upholstered armrests, carpeted interior; soundproofing insulation; modified Dennis Ashby (Comanche) glareshield with integral lighting; Piper external power plug; heated Piper pitot / static probe.
INSTRUMENTS AND AVIONICS:
Full gyro panel; Piper true airspeed indicator - remarked with Mustang II speeds in knots; Narco MK-12D TSO with GS, ID 825 TSO Nav indicator; King KN-62A TSO DME; KING KT-76 TSO transponder; Aerosonic encoding altimeter; Piper Marker beacons; PM-2000 stereo intercom; stereo CD player (installed on aft bulkhead); Hobbs meter, speaker, and stereo pilot/copilot headphone jacks; AstroTech LC-2 digital chronometer; avionics bus circuit breaker / master switch; Glasair control stick grips with push-to-talk switches.
ENGINE AND PROPELLER:
Lycoming IO-320-B1A (160 HP) engine. Engine gauges: tach, MP, fuel flow, EGT, CHT, oil pressure, oil temp, volt meter, amp meter; Hartzell HC-E2YL-2 (72") constant speed propeller (from the Twin Comanche).
DuPont® VariPrime self-etching primer; DuPont® Prime-N-Seal primer sealer; DuPont® Imron® Polyurethane Enamel paint - White; DuPont® ChromaLusion "Absolute Purpleen" for stripes and Chameleon - (Chromatic paint appears to change color depending on light angle and viewing angle).
It is readily apparent that incorporating these forty modifications to the original Bushby plans added considerable time and expense to the building process. The extra weight was, of course, a consideration, but determined to be worthwhile to make this a comfortable, practical, long-range IFR cross-country airplane.
As mentioned above, EAA Technical Counselors were called upon for three inspections during the building process. The FAA was consulted at the beginning of the process and to determine the local office's requirements for certification. The FAA was issuing a special book "Amateur Built Aircraft Reference Material" (FAA, 1997) at Oshkosh, 1997 which included numerous Advisory Circulars and FAA forms relating to the certification and registration of amateur built aircraft. A meeting with the FAA before the final inspection to discuss the certification went very smoothly, probably due to the extensive documentation provided. FAA final inspection occurred on June 18, 1998 and the Special Airworthiness Certificate and Operating Limitations were issued.
The EAA Flight Advisor Program was utilized and the Flight Advisor agreed to assist during the first flight.
The 25 hour Flight test program was developed based on ideas and information obtained from the FAA's Advisory Circular AC-90-89 "Amateur-Built Aircraft Flight Testing Handbook" (Federal Aviation Administration, 1989), "Flight Testing Homebuilt Aircraft" (Askue, 1992), and the Mustang II Construction Manual. Preparations included the meeting with and briefing airport fire / rescue personnel and briefing the two ground flight test assistants on emergency procedures, including specific directions to and telephone number of the nearest hospital. In the days before the first flight, a lot of time was spent thinking about it. The mental attitude taken was as if someone else of unknown skills had built this airplane - while pretending to be a disinterested, skeptical, hired test pilot. An objective approach was taken to inspect and try to find anything unsatisfactory - with the requirement that any discrepancies found be corrected before flight. This frame of mind reduced the emotions involved to a tolerable level.
The test program plan included a series of high-speed taxis down the runway leading to "land backs" or very short flights of just a few seconds as suggested by Vaughan Askue in "Flight Testing Homebuilt Aircraft" (Askue, 1992). The EAA Flight Advisor strongly disagreed with these procedures and suggested simply a few high-speed taxis, including raising the tail, followed by hard braking to glaze the semi-metallic brake linings. This was accomplished immediately before the first flight.
The EAA Flight Advisor agreed to fly "chase" in his experimental RV-6. All final ground checks were made, checklist completed (twice) and the chase plane took off first to assist by clearing the area of traffic. The initial take-off was extremely quick and actually quite surprising. Gross weight was 1,517 pounds. The take-off roll as recorded on video was approximately six seconds and 500 feet long. Initial climb was approximately 1,800 feet per minute at 80 MPH. Briefly, while passing through 800' AGL halfway down the runway, the sensation, realization and significance of what was happening sunk in. Flight was finally taking place in an airplane built by one's own hands - "I built those wings out there...". The Flight Advisor broke the silence and contemplation with "Is everything alright?" "Yes, it's handling fine, and [engine] gauges are in the green." Some gentle maneuvering was performed, followed by an in-flight inspection by the chase plane for any abnormalities or leaks, then a comparison of indicated airspeeds between planes showed that all was satisfactory. Two distracting problems were immediately apparent though - a loud squealing air leak at the canopy/windshield seal and insufficient elevator trim control friction. The noise was temporarily silenced by a piece of tape over the area, and aft stick needed to be held since the trim was being forced to neutral by the air loads.
A few power-off stalls were performed to verify the airspeed calibration. Indicated stall occurred at 51 MPH [later calibrated to be 55 MPH] thus the approach speed was calculated to be a minimum of 66 MPH - 75 would be used for the first landing. Forty-five degree steep turns felt excellent and the bubble canopy visibility was outstanding. After 30 minutes, the Flight Advisor cleared the area again and the first landing was successfully accomplished. The 3,500 hours of work to this point suddenly seemed worth it.
Other minor discrepancies that were found were that the engine idle was too low and the fuel flow gauge read too low. The idle was later adjusted and the new fuel flow gauge ordered that was later installed.
SUBSEQUENT FLIGHT TESTS
The elevator trim lever was tightened and the canopy seal modified to remove the air leak and squeal. The four-inch tailwheel did not want to "break away" to allow the tailwheel to swivel, so it was replaced with a six-inch tailwheel. Clean stalls were accomplished at 60 MPH IAS [later calibrated to be 62 MPH]. The airplane was still found to have inadequate nose up trim when at the forward CG limit when below 120 MPH. The original MPH airspeed indicator was later replaced by one displaying knots.
The third flight test was marred by a rough running engine. A fuel injector nozzle was found to be 50% blocked. After cleaning all injectors, the engine ran smoothly. Fortunately, despite the hot conditions, engine cooling was never a problem, probably due to the engine cooling plenum chamber. On later flights, cruise data was collected at optimum altitude. 182 MPH/158K was the 75% power cruise speed. On the sixth flight (10 hours TT), the wing tanks were used and a leak was found in the left wing when filled above one half. Fuel transfer and the "direct to engine" selection were tested during takeoffs and landings. The left wing was removed and "sloshed" with tank sealer and reinstalled.
On the seventh flight (12 hours TT), full fuel was loaded plus ballast for a max gross weight takeoff of 1,850 pounds. Speed was taken to Vne plus 10% (253 MPH) in a power dive, and aileron, elevator, and rudder control input doublets were induced. Fortunately, no flutter or control problems were encountered. Rubber edge strips on the wing attach fairings came loose and continued to be difficult to keep attached. Months later they were replaced with new wing attach fairings without rubber strips. With the CG at 70 inches (mid-range), trim control was fine, even at approach speeds. Ten pounds of ballast was installed in the tail making trim available at approach speeds when near the forward CG limit.
A difficult to troubleshoot radio squelch problem turned out to be interference from a bad GPS battery pack. On flight nine (17 hours TT), accelerated stalls, power on stalls, and spins to the left and right were performed. An open Coke can made for quite a mess during the initial spin entry. They were banned on future aerobatic test flights.
Grass field landings, ILS approaches, and no flap landings were performed. On flight 12 (20 hours TT), aerobatic flight testing was performed at 1,500 pounds. Three turn spins, wingovers, barrel rolls, aileron rolls, loops, and brief inverted flight was accomplished.
Four hours remained, and due to the very hot weather (95-100°F surface temperatures) during the test month, the remaining hours were flown at 9,500 feet and l0,500 feet around the test area. In full sun conditions, the large bubble canopy acted like a greenhouse and cockpit temperatures were running 30-35°F above outside air temperature. Stick-on green plastic sun shields cut the suns glare fairly well, but only lowered the temperature slightly. Even at l0,000 feet with an OAT of 52°F, the cockpit was at 85°F most of the time. Lower altitude tests saw in-flight cockpit temperatures of 110-120°F.
On July 18, 1998, 25.l hours was logged, the appropriate entries made in the logs, and flight testing was complete. The builder's extraordinarily patient wife was the first passenger later that day.
One week later, the first flight out of the test area - 560 NM to Oshkosh "AirVenture 1998" with the builder's oldest son was made, joined by the rest of the family via the airlines. A week of perfect weather, a great parking spot and camaraderie with family, friends, and thousands of other aviation enthusiasts made for a wonderful conclusion to the goal of the lifetime - to build and safely fly your own airplane to Oshkosh!
Yes, it was worth it.
As of this writing, the Mustang II N727RH has flown 138 hours, has 176 landings, and has been to 14 states and the Bahamas.
Alexander, R. (1999, October). Getting Started in Aircraft Building. Sport Aviation, 48 (11), 109-113.
Askue, V. (1992). Flight Testing Homebuilt Aircraft. Ames, Iowa: Iowa State University Press.
Bingelis, T. (1979). The Sportplane Builder. Austin, Texas: Tony Bingelis.
Bingelis, T. (1983). Firewall Forward. Austin, Texas: Tony Bingelis.
Bingelis, T. (1986). Sportplane Construction Techniques. Austin, Texas: Sportplane Builder Publications.
Bingelis, T. (1995). Tony Bingelis on Engines Oshkosh, Wisconsin: EAA Aviation Foundation.
Code of Federal Regulations, CFR 14, Parts 1 to 59. (1995). Washington, DC: U.S. Government Printing Office.
Cox, P. (1975, July). A Pair of Mustang IIs. Sport Aviation, 24 (7), 30-34.
Eaves, D. (1995, February 3). Personal communication.
Federal Aviation Administration, (1972, April 17). Acceptable Methods, Techniques and Practices -Aircraft Inspection and Repair. (Advisory Circular 43.13-1A). Washington, DC: U.S. Government Printing Office.
Federal Aviation Administration. (1989, September 18). Amateur-Built Aircraft Flight Testing Handbook (Advisory Circular AC-90-89). Washington, DC: U.S. Government Printing Office.
Federal Aviation Administration. (1997). Amateur-Built Aircraft Reference Material - Oshkosh 1997 AFS-600. Washington, DC: U.S. Government Printing Office.
Kessler, M. (1975, August). Retractable Mustang II. Sport Aviation, 24 (8), 56.
Mustang Aeronautics [Online]. Mustang II History. Available: http://www.MustangAero.com/Mustang%20II/Mustang%20II%20History.html [1999, November 10]
Paser, K. (1994). Speed with Economy. Littleton, Colorado: Paser Publications.