Robinson Helicopter Company is built on pragmatism. This philosophy comes from Frank Robinson, the company's president and founder, who is not only a member of the Experimental Test Pilots Association but is also an aeronautical engineer who in 1973 began designing the highly successful R22 helicopter. Although considered the archetype of simplicity, the R22 was engineered for reliability and performance.
One may think that such an engineering feat was accomplished with sophisticated computer-aided design and advanced wind tunnel testing. Not so. Each part was created in Robinson's mind, assembled in a small rented hangar at the airport in Torrance, California, and flight tested by him to prove the design. Robinson's commitment to testing continues today, as every R22 that comes off the production line is extensively flight tested. Thinking that would make for some interesting flying, I visited the factory to spend the day as a production test pilot.
To understand how the helicopters are assembled I toured the company's new state-of-the-art, 260,000-square-foot factory with Doug Tompkins, Robinson's chief pilot. The area where critical components are fatigue tested was the most impressive. Here, custom-designed machines shake, stretch, and stress different components 24 hours a day until a part fails. That part is then redesigned and put back on the machine. The process continues until something else fails. Robinson considers this never-ending process essential to achieving the highest possible level of reliability.
Tompkins and I joined up with ship number 2475, a brand new R22. The first thing we did was thoroughly preflight the helicopter. We opened all the cowlings and access panels to ensure that no obvious mistakes were made on the assembly line. We paid special attention to critical drive components, flight control systems, and structural integrity. The helicopter passed our scrutiny, so we rolled it out to the flight line. Since helicopters right off the line have no airworthiness certificate and are still considered experimental, company policy requires all pilots and mechanics on board to wear helmets and fire-retardant Nomex flight suits.
After climbing into the cockpit and carefully examining each system on the checklist, we check for full range of motion on all the flight controls and flip on the master switch. With all systems operational, I turn the starter switch and the Lycoming 0-320 springs to life. We closely monitor engine oil pressure, engine temperature, and electrical loads. After performing the standard magneto check and carburetor heat checks familiar to pilots of fixed-wing aircraft, we test the helicopter's free wheeling clutch. (In the event of a power loss, this clutch allows the helicopter to enter autorotation by disconnecting the engine from the rotor system.) This is done at normal rotor operating rpm by closing the throttle and looking for a clean separation on the engine and rotor tachometers.
"Torrance Ground, Robinson Eight-One-Delta ready to taxi, flight test to the antenna site," Tompkins says over the radio. The reply comes back, "Robinson Eight-One-Delta, company traffic on short final and one hovering at the antenna site, cleared to taxi." Tompkins points to a small grassy area at the base of several red and white antenna towers and says, "Go ahead and pick it up and hover taxi over there." I slowly raise the collective control and for the first time in its life the helicopter leaves the ground. Once at the antenna site we get a feel for the vibration levels in a hover. It is a little rough — but that is typical, according to Tompkins. Next, we make rapid pedal turns and abrupt movements with the collective and cyclic controls. The objective here is to strain the helicopter's drive system and flight controls. This way if something is loose or defective and is going to break, it will happen while hovering a few feet off the ground. Once convinced the helicopter will stay together, we fly it back to the flight test hangar to be outfitted with vibration detection and correction equipment.
This equipment, manufactured by the Chadwick Corporation, has three vibration sensors. One is placed on the side of the instrument console to measure airframe vibrations, the second is bolted to the non- rotating part of the main rotor shaft, and the third goes on the tail rotor support bracket. The pilot and mechanic take the helicopter out for a quick flight to get vibration readings in a hover. Small washer-style weights are added to the rotor head to reduce the vibration levels. The amount of weight to add or subtract is decided by trial and error until the vibration drops below an acceptable level. Sometimes the vibration is so high the entire rotor head must be removed and shifted a few thousandths of an inch.
Next is the high-speed track test to ensure that both rotor blades are flying in the same plane at VNE of 102 knots. On the trailing edge of each rotor blade is a small tab that functions like the aileron on an airplane wing. With the helicopter flying at 102 knots, the mechanic uses a special strobe light to view the blade tips and later bends the tabs to make the blades fly in the same plane.
The flight control rigging needs to be checked and adjusted if necessary. A weight bag is placed on the tail boom to bring the center of gravity to the aft limit. The pilot takes the helicopter up to 1,000 feet, accelerates to VNE, and reduces main rotor rpm to the lower limit of 97 percent. In this configuration the pilot must have a minimum of three- quarters of an inch of control movement before the cyclic control touches the forward stop. We were unable to do this test together because with both seats occupied, the center of gravity can't reach the aft limit.
A similar procedure is used for the forward center of gravity test, but this test requires two people and lead weights on the forward tip of the skids to reach the forward center of gravity limit. To simulate a worst case scenario, we flew the R22 backwards at 20 knots over the taxiway at maximum gross weight, reduced the rotor rpm to the lower limit, and touched the aft cyclic control stop. Again, there must be three- quarters of an inch of control movement. This test puts the pilot in an interesting position, because if he uses too much forward stick to stop the rearward movement, the helicopter will lose altitude and hit the taxiway sliding backwards. The trick here is small, precise control movement or, as another test pilot suggested, using right pedal to spin the helicopter 180 degrees so that it would then be moving forward.
The pedals control the tail rotor and, in flight, are used for trim much like a rudder. The tail rotor also prevents the engine torque from spinning the fuselage during low-speed, high-power maneuvers. In most helicopters, adding power requires left pedal (French and Russian helicopters need right pedal because their rotors spin in the opposite direction). If the pedals are not rigged correctly, the pilot can run out of left pedal in times of high power demand. To test this, Tompkins and I took off and climbed at VY (53 knots) and, using maximum takeoff power for that day, we rolled the rotor rpm down to 97 percent. In this configuration, the R22 must be able to yaw 30 degrees to the left. Our R22 passed with plenty of pedal remaining. In a practical sense, the R22's tail rotor performs exceptionally well; I have flown a fully loaded R22 at a density altitude higher than 10,000 feet and still had more than adequate tail rotor authority.
The rotor system must be adjusted to allow the blades to continue turning should the R22's engine decide to stop breathing at the worst time. Tompkins and I did several autorotations, each time recording pressure altitude, outside air temperature, indicated rotor rpm, and gross weight. Back on the ground, the mechanic plotted the numbers on a chart and then made the necessary changes to the main rotor blade's pitch angle. Then it was back up in the air for a practical test of hovering autorotations, autorotative turns, and main rotor rpm control. Everything checked out well, so we spent some time shooting precision autos to a pier in the Long Beach harbor. Autorotations in the R22 are demanding. For the student pilot, this teaches precision flying from the start; it provides the experienced pilot with a constant challenge.
Because of the NTSB's recent letter concerning an alleged propensity for in-flight rotor system-to-airframe contact (see " Design Defect?"), I asked Tompkins about his experience with this situation. Any two-bladed helicopter is sensitive to this in a low-G condition, but I wanted to know if the R22's rotor head design made it more susceptible. "Absolutely not," Tompkins answered. "Furthermore, the accidents cited by the NTSB involved inexperienced helicopter pilots on the controls."
To prove his point he offered to demonstrate the maneuver in question. In level flight at 95 knots, Tompkins started to slow the helicopter by quickly raising the nose. I watched as our climb rate hit 2,000 feet per minute. Tompkins then abruptly lowered the nose with forward movement of the cyclic control and I immediately felt a weightless sensation. The R22 began to roll to the right. "This is where pilots get into trouble," Tompkins explained. "The instinctive reaction is to correct the right roll with left cyclic. With the rotor system unloaded, left cyclic causes the rotor system to strike the mast and can sever it, thus allowing the rotor blades to strike the fuselage." Tompkins then demonstrated the proper recovery technique, which was gentle aft cyclic to reload the rotor disc, then left cyclic to level the helicopter.
High-time fixed-wing pilots need to be especially careful, because this type of maneuver is perfectly acceptable in an airplane. The U.S. Army has identified more than 50 fatal accidents of this type in other helicopters with two-blade rotor systems. The R22 is a responsive helicopter that demands your attention. I did several low-G maneuvers and found that while things happened fast, the R22 was extremely predictable. It became obvious that the key to preventing this type of accident is proper training.
The company's tradition of pragmatism continues with its newest helicopter, a four-place version of the R22, appropriately dubbed the R44. Robinson keeps ship numbers one and two as experimental test beds; number three is a static test vehicle and R44 number four is flown almost constantly as a proving aircraft. This allows Robinson to spot any potential problems first and get the correction to his customers more quickly. The R44 uses beefed-up versions of many of the R22's proven designs, and the production test procedures are very similar — except that when the R44's flight test is completed, it is disassembled, crated, and shipped outside the United States. Robinson is still at odds with the refusal of American insurance companies to issue waivers of subrogation (see "Pilot Briefing: Robinson Limits R44 to Foreign Markets Only," February 1994 Pilot, p. 31).
With the R22 and the R44 in full-scale production, is there an R66 in the future? "No," Robinson said. "The R44 was certified days before the FAA's new and more complex certification rules took effect." However, Robinson said he will always be busy designing and testing improvements and accessories for both helicopters. And for the test pilots, that means job security.
Tim McAdams, AOPA 925518, is a helicopter CFI who has accumulated more than 4,600 hours, of which 4,300 are in helicopters. He flies for Liberty Helicopters in New York City.
Design Defect?
Robinson founder disputes NTSB allegations
BY TIM McADAMS
The National Transportation Safety Board sent a letter to the Federal Aviation Administration criticizing the Robinson R22 helicopter in July 1994. The letter detailed three fatal mast bumping accidents involving rotor blade-to-airframe contact and a subsequent in-flight break up. The NTSB implied that the rotor system contained an inherent design deficiency that allowed the rotor blades to strike the fuselage while operating within the helicopter's normal flight envelope. The NTSB recommended that the FAA immediately issue an airworthiness directive to reduce the R22's never-exceed airspeed until the problem could be identified and necessary design changes could be approved and implemented.
Frank Robinson, Robinson Helicopter's founder and president, disputed the accuracy of the NTSB's allegations, including the aerodynamic theory used to describe mast bumping. "The R22 is vulnerable to this type of accident because it is flown by the highest risk pilots: student pilots, renter pilots on joy rides, and low-time private owners," Robinson said, specifically emphasizing that mast bumping can occur in any semi- rigid (two-bladed) rotor system. According to Robinson, high-time fixed- wing pilots with low helicopter time are the most susceptible to this type of accident.
A recent FAA study appears to support Robinson's position. The study, which used accidents that Robinson identified as involving mast bumping, revealed that the pilots presumed to be flying had an average of 161 helicopter flight hours — and 5,471 hours of fixed-wing time. An FAA flight test pilot who recently returned from the Robinson factory's flight safety course said that although the FAA is still researching the situation, he personally felt that the R22 can be safely operated by a properly trained pilot.
As an interim measure, the FAA has published several operational recommendations to help pilots avoid mast bumping (see "Pilot Briefing: FAA Issues Special Flight Limitations for R44, R22 Helos," September 1994 Pilot). However, the agency noted that all pilots flying small helicopters with two-bladed rotor systems should consider these recommendations.
The timing of this probe surprises Robinson, who noted that the R22's fatal accident rate of .97 per 100,000 flight hours is the lowest it has ever been and is comparable to other helicopters and light airplanes. Also, the R22's combined mechanical and engine failure rate — 5 percent of R22 accidents during the last 10 years — is the lowest in the helicopter industry.
Regarding the unrelated July 31, 1993, crash of an R44 at California's El Monte Airport, preliminary investigations by the NTSB centered on a possible fatigue crack in the helicopter's cyclic control system. Robinson felt it was impossible to determine how or what part might have failed because many of the critical components were too badly damaged to accurately examine. Therefore, immediately following the accident, Robinson redesigned more than a dozen parts in the cyclic control system, making the system considerably stronger than required by the FAA. The entire R44 fleet was retrofitted with the new system.
In April 1994, the NTSB released a safety recommendation regarding the R44 crash. The board's final decision determined the probable cause of the accident was indeed a fatigue failure of a part in the cyclic control stick assembly. From an operational standpoint, Robinson believes this finding is moot because every R44 now has a cyclic control system that is twice as strong as required.
Related Articles
