Experimental test pilot ranks are no place for cowboys
“Beginning autorotation with left turn.”
“Roger. Autorotate to left.”
After a few moments, during which the MD 600N helicopter plummets toward the ground engaged in a sweeping left turn, the first pilot’s voice comes over the radio. “Recovering.” He recovers at approximately 100 ft over the desert floor.
“Roger. Recover.” This voice belongs to the pilot of the chase aircraft, also an MD helicopter.
“Making 180-degree turn, climbing to 1,500.”
“Roger. We’ll stay back.” It takes the single-engine helicopter less than two minutes to complete this task, and when the aircraft is at the right altitude and flying at precisely the correct speed, the next task begins.
“Beginning autorotation with right turn.”
“Roger. Autorotate to the right.”
Such is a typical test flight for Greg Ashe, senior experimental test pilot for Boeing’s rotorcraft division based in Mesa, Ariz. Ashe, who has been test flying the egg-shaped helicopters and their brethren since they bore the Hughes brand more than 20 years ago, was a natural choice to assist MD Helicopters in its certification testing after the company purchased the old McDonnell Douglas commercial line of helicopters from Boeing in 1997 (Boeing is providing engineering test support to MD Helicopters). Ashe flew the certification tests for the MD 600’s new yaw stability augmentation system in June.
Thanks to movie portrayals, most people envision two things when they think of test pilots: daredevils and the military. On the contrary, most experimental test pilots work quietly behind the scenes, using their engineering degrees to fully participate in various facets of the aircraft design years before the aircraft flies. When the aircraft does fly, the experimental test pilots carry out several sets of individual maneuvers based on exacting weights, altitudes, speeds and other factors. Precision, then, is key in the experimental test pilot’s routine flying.
“I consider flying fairly boring,” said Ashe, who has more than 12,000 hr in more than 40 types of rotary- and fixed-wing aircraft. “But typically our test flights are short, and there’s something to do every minute. You have to set up each maneuver precisely, speeds within one or two knots, power settings within a percent or two, altitudes plus or minus 50 feet, and hold all of those variables before starting the maneuver. The goal is to see how many test points you can hit without having to redo any.”
Know Thy Airplane
By the time an experimental test pilot straps on a new corporate jet, he or she has been part of the engineering and design effort for multiple years.
“I have been intimately involved in every part of the aircraft design,” said Carroll Beeler, chief test pilot for San Antonio-based Sino Swearingen, a relatively new company still working on certifying its first business jet, the SJ30-2. Beeler flew the first flight of the SJ30-2 in November last year, more than three years after he was hired in 1997. “I was involved in selecting and placing virtually every knob and button on the flight deck, and even researched and bought some off-the-shelf equipment like the primary flight display.”
Beeler said that part of the company’s mission for the SJ30-2 is to provide “intuitive operation for single-pilot operations.” To that end, he helped design a color-coded circuit breaker panel to allow easier identification and prevent errors due to flipping the wrong circuit breakers. Making design suggestions based on experience and human-factors recommendations soon becomes second nature to the experimental test pilot.
“We look for human factors issues in the cockpit design and layout,” said Beeler. “Can a switch be labeled incorrectly or misunderstood? Are there switches that look alike right next to each other, or any items that could confuse foreigners who don’t use the same kind of English that we use? We bounce these types of question off each other to make sure that the layout is the most intuitive for all pilots.”
Often the experimental test pilots delve so far into the design work that they temporarily leave their wings on a shelf and pick up the engineer’s calculators. Following the development of the MD 900 Explorer, Ashe and fellow experimental test pilot Chan Morse were awarded a patent on their integrated engine control system.
“We participate in the design across the board,” said Morse, who spent four years on the MD 900 Explorer program before making the helicopter’s first flight. “Our job is to improve performance and handling, help design the crew stations and represent the operator in what they will need in flying the aircraft.”
Being intimately involved in the development of the aircraft is beneficial for a number of reasons beyond the actual improvement of the design. Not only do the pilots know what to expect in terms of engine and flight control performance, but they also know the cockpit layout like the proverbial “back of their hands” in case the aircraft does not perform as expected and the pilots have to get out of an emergency situation.
“The pilots are pretty intimate with what’s there and what we should expect from the aircraft,” said Morse. “From being involved in the overall design, we’ve had a chance to second-guess the design decisions and know what the trade-offs are from making those decisions.”
Ashe agreed, “We spend a lot of time with the aircraft as it’s being built, understanding the machine from the drawing stage, participating in the structures lab, flying the ground-test vehicle and overall getting an understanding of how everything’s going to work.”
Practice, Practice, Practice
Before the first flight of any new aircraft, the program test pilots have usually spent some time flying either a simulator or performing ground tests. Morse said that fixed-wing pilots are luckier in this regard in that simulators are easier to program and more closely resemble the actual flight characteristics of the aircraft. “Helicopters are hard to predict in terms of handling and control characteristics,” he noted. “On fixed-wing programs they can build and fly a simulator and it gives good feedback data. We usually don’t get full-scale simulators.”
“We are never just given a test card and told to go fly the test,” said Tom Carr, chief test pilot for the Raytheon Hawker Horizon. Carr pointed out that even though all of the experimental test pilots from the Horizon program are participating on multiple programs in the Raytheon spectrum of aircraft, time served on those other programs ranges from a few days for testing new RVSM instrumentation packages on production aircraft to a few months for other test flying of more complicated modifications. In any case, the pilots are involved with designing and planning the tests well before they step into the aircraft to fly the test card.
But it wasn’t always this way. According to Pete Reynolds, vice president of flight test for Bombardier Aerospace, when he started as a test pilot for Learjet more than two decades ago pilots spent an average of only 18 months on a flight-test project.
“There has been an increasing trend over the past few years to get test pilots more involved earlier in the program, and this has had some positive effects on the industry as a whole,” said Reynolds. “Earlier involvement of the pilots, along with the new computerized development tools, provides a positive way of getting products developed more quickly, but also more safely. Pilots get a better feel for the aircraft before they go flying, enhancing the safety of the test program.”
The earlier involvement of pilots, Reynolds believes, has improved the forecast accuracy of today’s computer tools and encouraged the expanded use of simulators and test rigs, few of which were available 20 years ago. These tools allow greater predictability of the aircraft design and permit the pilots to provide input to both the design and the test plans earlier in the program. “We are conducting testing today more thoroughly than has ever been,” he said. “This has resulted in a tremendous increase in aircraft safety and an improvement in safety rates over the past few years.”
First Flights and Close Calls
“The first flight of a new model is quite a thrill,” said Reynolds, who has had the honor and responsibility of making first flights in a number of Bombardier types, including several Learjets and the Global Express. “Even besides the technical aspects of first flight, there’s an emotional side to this major event, a real thrill to be associated with the team of people who bring a project together. At Bombardier, we’ve had 12 new aircraft in 12 years; it’s been amazing.”
Making a first flight in a new aircraft denotes the crème de la crème of the experimental test pilots’ work. Hundreds of test pilots never get the opportunity to make that first flight in a newly designed aircraft, although more opportunities are opening up as new aircraft designs roll off the drawing boards.
“I’ve been fortunate to have the opportunity to make as many flights in new helicopters as I have,” said Morse, who has made first flights in various models of the Notar-equipped MD helicopter series during his 17-year career at McDonnell Douglas/Boeing. “It’s been a good time to be in the industry. To go through three or four first flights is pretty fortunate.”
Part of the reason that Morse may be called on to test new designs is his self-imposed emphasis on training for emergency situations. He has survived several close calls during his career using flight control skills gained during training flights.
“We’ve had a couple of interesting moments in autorotation testing where we’d run out of cyclic or collective,” recalled Ashe, who has worked with Morse on most of the Notar programs. “Chan had seen that before, running out of cyclic and having to use collective for longitudinal control, and it became a training issue. We’d start making training flights and concentrate on using only one method of control to fly the aircraft.”
Often it doesn’t take anything as dramatic as the separation of the tailboom off an MD 600N to prompt a design change, but one of Morse’s close calls came from an autorotation at max gross weight data point that resulted in the main rotor striking the tailboom. “The blade hooked the control cable and ripped it into a full-left condition, which put too much strain on the weakened boom.” explained Morse, who had to use the collective to bring up the nose after the tailboom departed the aircraft and cushion the landing. “Among the design changes we made after that incident was to move the control cable to the bottom of the boom.”
“Every test pilot has had things not go as well as anticipated,” said Jim Dwyer, manager of flight-test operations for Bombardier’s Learjet division and project test pilot on the company’s new Continental. Among Dwyer’s first flights are the Continental on Aug. 14, 2001, and accompanying Reynolds on the first flight of the Learjet 45 in October 1995. “Part of being a test pilot is being able to handle those types of things. You have to be prepared for the unexpected and use your skill, knowledge and judgment to handle the situation.”
First flights generally do not result in close calls due to the extensive preparation and the light workload of the flight. For large OEMs, the program test pilots may test in a simulator or similar company aircraft. But for smaller companies like Sino Swearingen, where the first flight of the program is the first flight of the company, training for this flight may involve borrowing resources from other sources.
“When we were getting ready for the first flight in the SJ30-2, we went to FlightSafety and borrowed the CitationJet simulator for an hour,” said Beeler. “We did high-speed taxi and emergency procedures in the simulator to approximate what it should feel like in the SJ30-2.”
Beeler’s first flight in the SJ30-2 on November 30 last year lasted approximately 45 min, flying north out of San Antonio. “It was real middle-of-the-envelope flying,” said Beeler. “We kept a chase plane close by to make sure that there were no brown spots on the aircraft from leaks.” The maneuvers on the SJ30-2’s first flight included 360-deg turns in both directions, stall approaches at 8,000 to 10,000 ft, simulated landing and go-arounds at altitude and pulling each engine back to idle.
“We did not take any data points as such,” said Chuck Wall, who was the copilot on the SJ30-2 first flight. “We were just there to check out functions on the aircraft, such as how the flight controls responded.”
Raytheon’s Hawker Horizon spent a little longer in the air on its first flight. With a copilot and flight test engineer on board, Carr flew the new jet for 2 hr 29 min on August 11 from the company’s plant in Wichita. “We had planned for a two-hour 15-minute flight,” said Carr. “But it’s an indication of how well things went that we flew even longer. We would have stayed up longer if we had had more fuel.”
Besides some air-to-air photography work, Carr’s first flight consisted mostly of hitting aerodynamic data points on the test cards. “We did some speed expansion work at 225 knots to the max angle of attack for various flap settings,” said Carr. “We looked at handling qualities for steady-heading sideslips, maneuvering stick forces per g and longitudinal and lateral stability for a variety of flap and gear configurations.”
Part of the reason that the Horizon crew could plan on such a long and productive first flight was that the crew had been training in an integrated systems development facility (ISDF), which allows all of the Horizon systems to be integrated, tested and refined, and in a simulator with aerodynamic and engine models programmed specifically for the Horizon.
“We practiced for the first flight in the simulator,” said Carr. “We flew emergency procedures and rehearsed the first-flight test cards thoroughly in the sim before going out to the aircraft.”
Working with the FAA
It nearly goes without saying that certifying an aircraft or a modification to an aircraft requires a great deal of interaction with the FAA. On some test flights, particularly those involving stall and handling characteristics, the FAA is directly involved using its own test pilots. On other tests, the FAA delegates its
authority to Designated Engineering Representatives (DERs) or Delegated Option Authorities (DOAs), depending on the weight and type of aircraft being certified. The DERs or DOAs have the same sign-off authority as an FAA pilot, and are tightly regulated by the FAA, which can revoke a DER or DOA certificate at any time.
OEMs take different stands on how many DERs or DOAs they try to maintain on
staff, depending on corporate philosophy and the company’s relationship with the FAA.
“Not all of our test pilots are DERs, and we don’t have any goals to make them all DERs,” said Dwyer, who manages 15 pilots in the Continental program. “Because our aircraft are certified in Canada, we do have a number of pilots on staff who have the Canadian equivalent of a DER. Not all of our flying is done by company test pilots, either. The FAA also actively participates in our test programs and flies our aircraft on audits.”
Raytheon, on the other hand, uses a large number of test pilot DERs and DOAs. “We’re blessed by having a wide variety of aircraft, including Bonanzas, Barons and King Airs, in addition to the jets,” said Carr, who explained that most of the Raytheon test pilots working on the smaller aircraft (those with mtows of less than 12,500 lb) are DOAs under FAR Part 23, while the test pilots on the larger aircraft and jets are DERs under Part 25.
“On the Horizon program we are still in development, so the FAA has not become involved yet,” said Carr. “Currently we are still making sure that the aircraft is certifiable. We won’t bring in the FAA until we are sure that the aircraft will pass all of the certification tests.”
A Lengthy Path
For the larger companies such as Raytheon Aircraft, having varied aircraft and certification programs allows the company to home-grow its experimental test pilots. “We have a mixture of test pilots, some who are graduates of military test pilot schools and some who have only civilian flight experience,” Carr noted. “We’ve had people start out with only their private pilot certificate and spend five to seven years as a data analyst, gaining experience while moving up through the ranks. But all get in-house training before becoming experimental test pilots.”
At large OEMs, potential experimental test pilots start their careers at the entry-level position of data analyst. A non-flying position, the data analyst extracts the important aerodynamic, structural and control results from the raw data obtained during the flight tests. This requires engineering skill and knowledge, some of which is learned on the job. According to Carr, “Engineering skills are really more important than flying.”
After several years working as a data analyst, the candidate would be promoted to a flight test engineer (FTE). Although still not a pilot position, the FTE at least gets to fly in the aircraft, running the data acquisition equipment after having designed and instrumented the aircraft according to the test cards. FTEs learn what “good” data looks like, such as if the data is within tolerances for that test and could be used for certification.
Carr said that some experienced Raytheon FTEs start flying part time as production test pilots. This is the next step in the pursuit of becoming an experimental test pilot. Production test pilots fly each aircraft as it comes off the assembly line to ensure that it flies properly and all controls are working before delivery to the customer.
According to Bombardier’s Dwyer, the production test pilots need to be as precise as experimental test pilots. “Production test pilots determine how the aircraft is performing against how the aircraft is supposed to perform,” says Dwyer. “They have to have good piloting skills and knowledge, just like the experimental test pilots.”
The production test pilots may report to a different department than the experimental test pilots, who usually are part of the engineering division, but may be loaned out to engineering to perform flight testing on small modifications, such as a new avionics package or in-flight entertainment system. This would be the final stepping stone to becoming an experimental test pilot in a large OEM, a process that can take as little as five years and as many as 10, depending on the pilot’s background and the company’s needs.
Smaller OEMs that either don’t have the number of programs or the time to train their own experimental test pilots often have to hire pilots who have gained their base engineering knowledge and flight-test experience elsewhere. Sino Swearingen currently has an opening for a “DER-qualified test pilot” posted on its Web site to join its two-pilot team of Beeler and Wall. Beeler’s experience was gained from serving as an FTE and production test pilot for McDonnell Douglas and an experimental test pilot for engine certifications at General Electric. He holds a master of science in systems management and has 11,000 flight hr in more than 50 aircraft types. Although Beeler is a Vietnam War veteran, he has not attended any of the military or civilian test pilot schools.
In contrast, Wall is a graduate of the U.S. Air Force Test Pilot School and was an experimental test pilot in the service before becoming a chief test pilot at McDonnell Douglas, working on the MD-80 and C-17 programs. Wall has accumulated more than 8,000 hr in some 30 different aircraft types, both military and civilian.
“We’re a small company certifying our first airplane,” said Beeler. “Our entire flight-test crew consists of two pilots, two FTEs and perhaps up to a dozen data analysts.”
To TPS or Not to TPS?
You might expect that the business aircraft experimental test pilot comes out of the military and has attended one of the six military or civilian test schools in the Western hemisphere. But although graduating from a test-pilot school seems to be an advantage, the truth is that the majority of civilian experimental test pilots have not attended such schools.
At Bombardier, test-pilot school is not a requirement, according to Dwyer. “Having military test-pilot school and experience is definitely a feather in [an applicant’s] cap, but not required,” he said. “We do send people to the National Test Pilot School on occasion, but it is fairly expensive. Instead we do some home-grown training at the company.”
The National Test Pilot School (NTPS), based in Mojave, Calif., offers several courses ranging in length from two to 50 weeks to train test pilots and FTEs on both engineering and flight facets of experimental test piloting (see box on page 43). But the training comes at a hefty price: the 18-week performance and flying qualities professional course costs $100,000 for each FTE and $225,000 for each pilot. The 50-week test pilot/flight test engineer professional course must be individually quoted for each company and/or pilot. The high cost of the school may be one reason that attending test pilot school is not a requirement for civilian test pilots.
“We just don’t have the resources to go to NTPS,” pointed out Sino Swearingen’s Beeler, citing the limited funding and personnel resources available at the company. Beeler and Wall are instead associate fellows of the Society of Experimental Test Pilots, and use their SETP contacts and professional relationships to learn more about testing methods and human-factors issues encountered by other test pilots. “We can keep up with the latest methods and trends by attending SETP symposiums, reading the symposium proceedings and talking to others,” he said. “The things that have changed over the years are the data acquisition systems–most of the maneuvers that we fly haven’t changed since Chuck Yeager.”
“Think about it–could anyone teach you how to do your job in a year?” asked Horizon test pilot Carr. He has never attended a test-pilot school and does not require any of his crew to attend either. In his opinion, even the one-year course is “not enough time to really learn the job. We would still need to teach them how to do our tests; even experienced test pilots are required to complete in-house training for Raytheon’s experimental test-pilot programs. In the civilian world, the number of experimental test pilots graduated from either a military or civilian test-pilot school is in the minority.”
The Profile of an Experimental Test Pilot
Experimental test pilots must be highly experienced and knowledgeable. They typically hold degrees in aeronautical or electrical engineering, have thousands of hours of flight time and have flown many aircraft types. Test pilots also must be in control, not only precisely controlling the movements of the aircraft to achieve various test points, but also of themselves, using concentration and discipline to get the job done.
“It’s an adrenaline rush,” admitted Dwyer. “Taking the aircraft to where the average pilot doesn’t is a challenge. You have a knowledge base of what the test is and you can anticipate what the aircraft will do. If something doesn’t feel right, you have to decide immediately to stop or take it a bit further.”
Experimental test pilots are really engineers with wings. They must have the methodical patience necessary to hit each and every test point with accuracy and in anticipation not only of what the aircraft should do, but what the aircraft could do and how to recover from the unexpected.
“People think that test pilots just do hot-dog maneuvers,” said Dwyer. “But a good test pilot knows the envelope and has a good knowledge base. Preparing for the unexpected is the source of the test pilot’s knowledge. We must think of the contingencies and have a way out in case something happens.”
Sino Swearingen’s Beeler explained, “Test pilots are always looking for trouble. We’re always searching for something that doesn’t look right or feel right. Are there any characteristics on landing that might be a difficulty for other pilots? We’re always looking to make sure that the average pilot doesn’t find himself backed into a corner.”
Reynolds said the experimental test pilot must be such a good pilot that “flying is almost second nature. A test pilot shouldn’t need to worry about the mechanical part of flying, and must be willing to make tough judgment calls when it comes down to it.”
Sean Roberts, NTPS director, sees some particular traits from the students who graduate from his programs. “An experimental test pilot cannot be the wild guy shown in the movies,” he noted. “The pilot must be an excellent, thorough engineer with good flying capabilities. He must meticulously plan the test and fly the plan on a safe basis, and be able to talk to engineers to find a solution to any problems. There is no room for cowboys among test-pilot ranks.”
Of course, there is some room for fun as an experimental test pilot, especially when you reach the top of the test-pilot ranks. Reynolds recounted the time in 1978 that he set the time-to-climb record in the Learjet 28 with Neil Armstrong as his copilot. “Neil was on Learjet’s board of directors, which is only one of the reasons that he was on the flight,” he explained. “We reached 15,000 meters, just under 50,000 feet, in about 12 minutes.” Reynolds also set an official speed record on the way to the 1997 Paris Air Show in the then-uncertified Global Express. He said it is common for companies to use official speed and time-to-climb records to highlight the performance of an aircraft, and that if the aircraft is new, often the test pilots get to set the records. “It is such a thrill to show off the aircraft,” he added.