Belt Failures

Will an R22 or R44 fly with one drive belt missing?

Through the years Robinson has had several issues relating to drive belts, especially in the R22. A belt slowly coming apart, breaking completely, stretching then slipping in its groove or rolling completely out of its groove have all been occurrences R22/R44 pilots have had to deal with in flight. So, the obvious question is: if a belt comes off in flight, will the remaining belt(s) continue to drive the rotor? The short answer is “yes, definitely.” But before I explain the short answer, let’s look at the history, design and emergencies of the belt drive system.

Of course, the purpose of the drive belts is to transmit power from the engine to the helicopter drive train, through the two gearboxes then on to the main and tail rotor systems. Originally, the R22 had four individual V-belts to accomplish this task, each with their own groove in the upper and lower sheave. The “V” shape causes the belt to wedge tightly into the groove, increasing friction and allowing increased torque to be transmitted. In 1986, two individual belts were banded together by a common rubber backing or “tie band” creating two sets of banded V-belts for additional lateral stiffness. But through the 1980s and early 90s many small problems kept cropping up. For example, the upper and lower sheaves were originally made of aluminum and when the belts were loose in the grooves during startup and shutdown sharp edges were being worn into the sides of each groove potentially cutting the belt. That’s why Robinson initially recommended putting the clutch switch to the “engage” position as soon as possible after engine starting and wait no more than 30 seconds to pull the mixture on shutdown. This limited the time the lower sheave was being driven by the engine with loose belts. Eventually, this problem was solved by going to stainless steel sheaves. There was no belt guard initially and the importance of maintaining sheave alignment and/or belt slack was not well understood in the field so one belt rolling onto another belt or jumping out of its groove was not uncommon. Frank Robinson actually had this happen in an R22 in 1980 on a flight into Long Beach Airport. He made a precautionary landing in the parking lot of the Long Beach Convention Center about two miles from the airport, called the factory from a phone booth and the belt was quickly replaced allowing Frank to depart before the press showed up. However, probably the biggest problem was poorly manufactured belts. Belts breaking, delaminating or stretching were common occurrences and escalated to the point where during the flight portion of the Pilot Safety Course, after each training flight, the belt condition, tension and alignment was checked. In 2010 Robinson changed to drive belts manufactured by Mitsoboshi Belting Ltd (Revision-Z belts), a Japanese company that manufactures the belts in Ottawa, Illinois. Since then, there have been nowhere near the problems with the revision-Z and later belts as there had been with earlier belts.

If there is one emergency procedure R22/R44 pilots need to be very familiar with it is what to do when the “clutch” light comes on. The light coming on for just a few seconds is normal as the belts stretch during flight or as they break in. But, the “clutch” light staying on for a certain amount of time is an indicator of several possible problems with the drive line, including the drive belts. Through the years Robinson has varied the amount of time the pilot should wait before taking action by pulling the clutch circuit breaker when the light illuminates (thus stopping all power going to the clutch actuator). Originally, the Pilot’s Operating Handbook said 6-9 seconds, then 7-8 seconds then in 2012 we revised both the emergency procedure in section 3 of the Pilot’s Operating Handbook and Safety Notice #28 to a flat 10 seconds and also give the pilot a better explanation of the recognition factors and proper execution of the procedure if the light comes on during flight. The emergency procedure tells the pilot to “Reduce power and land immediately if there are other indications of a drive system failure” whereas the Safety Notice now delineates what those indications are, i.e. “smell of hot rubber, noise or vibration”. In two of the belt failures I have had the smell of hot or burning rubber was the first indication, 15-20 seconds before the light illuminated. This is caused from the frictional heat created if the belt slips substantially on the sheaves prior to breaking off. In the case of noise, understand the entire belt area is surrounded by sheet metal: fuel tanks, cowlings, decking etc. Image the pinging noise bits of belt spinning off at high speeds then hitting sheet metal would make or a more banging noise if larger pieces of belt hit, for example, the main fuel tank. Increased vibration could result from one belt rolling onto another belt or a section of belt peels away and hits the adjacent steel frame tubes with each revolution. The emergency procedure then states “be prepared to enter autorotation”. I frequently joke about this statement in our Safety Course because when, in any helicopter, should the pilot not be prepared to enter an autorotation? However, pilots have misinterpreted this statement. There have been a number of accidents where a belt came off, the light came on, the pilot pulled the clutch circuit breaker, overreacted and immediately entered an autorotation then spread the skids at the bottom. The Safety Notice now clearly explains when to enter an autorotation: “If tachometer needle split occurs, enter autorotation”. So, an autorotation should be the last resort–only if you can no longer maintain rotor RPM. However, the needle split will differ from what a pilot is used to seeing in practice autorotations. In this case since nothing is driving the rotor, the rotor RPM will decrease, but the engine has not failed and since it’s no longer driving the rotor, the engine RPM will increase rapidly. One footnote on pulling the clutch circuit breaker. Remember where that breaker is located, under the passenger legs. Depending on the size of the passenger and how they are positioned in the seat, the route to that circuit breaker might vary, that is, between their legs, under their legs or perhaps even around their legs. Once the pilot has already waited 10 seconds, now is not the time to explain to a passenger the need to go between, under or around their legs. So, include this situation as part of a normal passenger briefing.

Many factors affect how the belts wear and therefore the ongoing airworthiness of the belts. Environmental conditions, such as extreme temperatures and sandy/dusty conditions, improper sheave alignment or sheave condition, excessive belt slack and exceeding gross weight/power limitations can dramatically reduce the life and performance of the belts. Therefore, the preflight and 100 hour/annual inspections are vital in determining continued belt airworthiness. During the preflight the pilot should carefully check the belts for signs of cracking or damage on the V side of the belt, tie band problems such as fraying, peeling or splitting and excessive belt slack. The Australian Transport Safety Bureau published a report (AI 2009-038) that provided some excellent examples of what to look for when pre-flighting the belts.

To determine proper belt slack, press in on the belts just above the fan scroll. The belts should deflect approximately 1.5 inches (4 cm). Additionally when the clutch is engaged on start-up, the rotor should turn within 5 seconds. For further information review Safety Notice #33.

Now to explain the “short answer”. We have conducted formal ground and flight tests at Robinson in both the R22 and R44 with one belt removed and the aircraft loaded to maximum gross weight. In the R44 flight test, the belt was removed then, during rotor engagement, the belt tension actuator was allowed to run the exact time as a previous normal start (78 seconds). The clutch circuit breaker was then pulled freezing the actuator in the previous position when there were four belts installed. I hovered the helicopter for a period of time then pushed the circuit breaker in allowing the actuator to apply additional tension as would happen with the real belt failure emergency. After 10 seconds, I pulled the clutch circuit breaker per the emergency procedure instruction. I then moved the collective up and down repeatedly, hovering at different altitudes, accelerated to 80 knots down the runway, turned around and terminated back at my departure point with a normal approach to a hover. At no time was there any sign of a problem: no smell of hot rubber, no indication of belt slippage and no unusual vibration.

In the R22 flight test the belt had been removed when the clutch was fully engaged (The clutch had not been disengaged after the previous flight.) to simulate an inflight failure. Following a special protocol, the aircraft was started with the clutch fully engaged, the belts tight (not recommended) and the clutch circuit breaker pulled in the “out” position. After a few minutes of hovering the tachometer needles showed a difference of 2-3% (engine at 104%, the rotor at 101-102%) indicating a slight slippage of the remaining belts. I then pushed in the circuit breaker, allowing the actuator to apply tension to the remaining belt then pulled the circuit breaker out at 10 seconds, again, per the R22 emergency procedure and continued to hover. The tachometer and manifold pressure remained the same and there were no other abnormal indications such as noise or smell.

As a result of these tests and my own experience with belt failures, I feel completely confident that if a pilot were to have a belt come off in flight, then execute the proper emergency procedure, it will be a fairly straightforward matter of calmly finding a place to land and conducting a normal landing.

Tim Tucker

November 2019