Cog Belt Break

Up until recently I was convinced that the cog belt breakage issues that were experienced by one manufacturer of an aftermarket cog belt drive system for the Rotorway main Rotor system was solved. The AP cog belt drive system uses a hydraulic cog belt tensioner to apply the appropriate amount of tension to the cog belt on the slack side to keep the cogs firmly seated in the driving cog wheel.

One problem with a cog belt system is that if the cog belt is not run under enough tension, the cogs can ride up on the teeth of the driving cog wheel. The loads on the cog teeth of an undertensioned belt are then transferred closer to the tips of the cog teeth instead of at the base of the teeth where they are attached to the belt webbing.

The cog belts get their strength from either Kevlar or Carbon cords that run through the belt webbing and provide very high tension strength. If, however, the cog belt is mishandled by bending the belt into a smaller radius than it's minimum allowed radius (such as playing with it once you take it out of the box, turning it inside out, flexing it between your hands, etc.) the cords in the web can be damaged to the point that a belt failure will be imminent.

If the cog belt is run under-tensioned the individual cog teeth will ride up on the driving wheel where they are bent back in turn (beyond the minimum allowed radius) which will also damage the Kevlar or carbon fiber cords in the web to the point of eventual failure.

The photo below shows the AP cog belt tensioner system and the carbon fiber cog belt that is supplied by the factory. The blue material is mold release from the belt. This photo shows a new system with about 1 hour on the installation.

In October of 2008 Larry S brought his exec 162F to my training center for his phase II flight training. He had purchased his cog belt drive from the factory and had a very experienced builders helper/instructor install it prior to beginning his phase one flight training with that instructor. Following his phase one instructing and solo endorsement to hover, Larry hovered his ship until it had aquired a total time of 21 hours on the cog belt. He then called me and scheduled his phase 2 flight training.

When Larry arrived he had his ship ready. We did some minor work on several issues that I discovered on my safety inspection and then boarded the helicopter to begin his phase 2 training. I picked the ship up and hovered for around five minutes to get the feel for the controls and rigging. I then gave Larry the controls and had him hover for around 15 minutes to get him accustomed to the weight of both of us in the ship.

It appeared that all systems were working properly so we took off into the pattern to see how the ship handled above ETL. The wind was very gusty and with each gust the nose would tuck down giving me a very uneasy feeling as the cyclic bumped the aft stop. I aborted the flight and we returned directly to the helipad to check the horizontal stabilizer orientation. The HS was set for a pitch up of the tail so we brought the angle down to parallel with the tail boom center line-- This is all part of wringing out a new ship. The instructor who assisted Larry in his phase 1 training had not taken the ship to altitude so he did not experience the nose down pitching to be aware that it needed adjusting.

Following the re-setting of the horizontal stabilizer angle, we lifted off and began our take-off run. With the helicopter nose slightly lowered, the skid height between five to seven feet above the grassy surface, and a forward ground speed of around twelve miles per hour, Larry on the controls with me, instantly the nose of the helicopter did a 90 degree pivot to the right and the ship dropped. I instinctivly applied opposite anti-torque pedal, applied full up collective, and full cyclic in the opposite direction that the helicopter was pitching.

The ship hit the surface at about 90 degrees to the direction that we were traveling accross the ground. Larry had just installed my extended reinforced skid modification and it was directly onto that left extended reinforced skid that the weight of the ship dropped. With the control inputs that I applied the ship leaned left in the direction of travel and yet somehow pivoted around to 180 degrees from the direction of our original travel and slid to a stop upright.

The photo below is of the ship pointing back in the direction from which we were departing. You can see the two yellow landing lights to the left of the photo and the helipad that we departed from is between them. Notice the left extended skid that all of the helicopter's weight was transferred to as it pivoted around. The skid and landing gear was bent but did not collapse allowing the helicopter to remain upright. If you have not installed the extended reinforced landing skids, go to the VPS, web site by clicking this link. As the ship skidded to a stop sliding backwards the extended skids, tail rotor, stinger, and tail boom took all of the forces stopping the ship and preventing it from rolling over.

The photo below shows how the tail rotor dug up the grass as the ship pivoted and slid to a stop tail first.

The below photo shows the type of damage that can occur to the tail boom if it is not properly reinforced. (check out the tail boom area on this web site -hints and tips- regarding how to strenthen this area of your tail boom) and the distorted skids and gear legs on the following photos.

In the below photo you can see the grass that the tail rotor tore up as well as the bent tail stinger.

This ship did remain upright so the damage was minimal compared to what would have resulted if I had not been able to keep the ship upright.

The below photo shows the ship after we towed it back to where we started, on the helipad in front of the hangar.

When we removed the panels, we immediately saw what had happened, this is how the cog belt was laying on the frame following the break.

This photo shows the cog belt as we found it on the large cog wheel.

The below photo shows an up close view of the actual break in the cog belt. You can clearly see the broken carbon fibers that were damaged by some improper forces applied to them, the question is what caused those forces.

I need to point out that following the belt break, I carefully rechecked all of the belt system alignments and everything was perfect. I then called my friend and fellow CFI who installed the system for Larry to inform him about what had happened and to let him know this failure had nothing to do with his workmanship, which is always impeccable.

About a week following this break, I recieved a phone call from Bruce in Main. He informed me that he had also experienced a similar cog belt break on his Kevlar belt running on the same AP cog belt system. We now know of two such failures with this drive system. There has been an active discussion on ROG about the reason that the belt may have broken.

Surprisingly, after my initial call to the factory to report this failure, they have not attempted to contact me for any follow up or findings, suggestions, or assistence. Since I have received no feed-back from the factory, I turned to Andrew Burr of Vertical Performance helicopter products to help with a solution--I just did not feel comfortable flying this system again in this configuration following this break without knowing what had actually caused the cog belt to break.

I tend to lean toward the fact that the cog belt is not brought up to proper tension until the engine is actually running and oil pressure is acheived to activate the tension mechanism. VPHelo, LLc is testing a much stronger belt and a totally different elastomeric tension system that does not depend on the engine oil pressure to keep the belt tensioned at all times.

The VPS, LLc cog belt tensioner and stronger belt are shown in the below photo.

As one of only two pilots that has experienced an in-flight cog belt failure in flight, I have put a lot of thought during sleepless nights trying to figure this one out. Besides doing some damage to Larry's skids, tail rotor, and denting of the tail boom when the helicopter landed backwards, the outcome could have been a double fatality had it happened a few seconds later while higher, faster, and near the lake.

I am not sure that the check valve in the oil lines that some have suggested would be the answer only because I usually see some oil seepage around the plunger. Since the oil volume inside the plunger is very small, it would only take a very slight seepage to release the tension held on the belt even if the check valve held firm.

On the first AP system that I installed in PA quite a few years back the plunger did not extend at all. We called Al and he agreed to send a new tensioner ASAP but due to customs issues enroute from Canada, it was still going to take several days. Since I was there to conduct flight training, I suggested to my student that we modify the tensioner to work under spring tension. We went to the local Home Depot and purchased a stiff spring that was designed as a screen door closer.

We cut the spring to length, installed it over the plunger so that it provided a constant 10 pounds pressure on the tension unit, and began hover testing. We closely monitored the tensioner at start-up, idle, and at full rpm. It worked well with the exception that the piston pulsed like the engine oil tensioner due to the lack of dampening that the elastomeric tensioner exibits. The spring did give the added advantage of providing the proper tension to the cog belt during engine start pulses. It did have more movement during operation than the VP elastomeric tensioner.

At that time I did not know how important it was to keep the belt properly tensioned during the start sequence because I had no Idea that these belts could break due to hypertensioning (high tension transfered to individual cog belt segments due to a momentarlily loose belt riding up on the drive cog sprocket). I am now convinced that the quick fix that we did in order to get the ship flying was most likely a vast improvement over the original design because even when the engine was not running the spring held the cog belt under a tension load. I am sure that AP had no idea that these belts would break or that the hydraulic tensioner may actually be the cause due to no tension during the start sequence.

We now know of two cog belts that broke. One, a Kevlar belt from AP, broke a week before the carbon fiber belt broke with Larry and me. The factory was aware of the first break but did not know the reason. If we had known that the cog belts could break it may have caused this discussion earlier and stimulated the same thought processes. Since two belts have broken now, each of a different material, there is a common thread. The one that broke on Larry's ship was installed by Mark P. and it was installed perfectly and was at the proper tensions that are called for on the RW instructions.


Since we have 4 Rotorway helicopters at our Sho-Me Helicopters, LLc traning center, Larry and I decided to run a couple of tests. We pulled two Rotorway 162F ships out of the hangar at the same time. Both cog drives were at the same temperature in the same location. Both cog belt systems were installed correctly and the cog belts were set to the exact same tension as directed by the factory. Both drive cog wheels were set with the required .1 degree aft and left tilt.

The ship that the VP tensioner is running in is Larry Stuckman's who had the cog belt break in flight here at our training facility. Larry installed new landing gear and skids, and new tail rotor belts, TR shaft and bearings, and new TR blades. He installed the high-load cog belt that VP is now supplying along with the tensioner system that Andrew built for Larry to test.

Below Larry stands next to his ship during his phase II flight training at our training center. We also were flight testing his new VP cog belt and tensioner following the repairs to his ship after the cog belt break.



We shot a short video of each cog belt tensioning system during the start sequence. On the VP system the start pulses are absorbed by the elastomeric tensioner as it is designed to do. Note on the AP/Factory engine oil tensioner the tensioner does not even apply any tension to the flopping cog belt until the engine is actually running and builds oil pressure.

Also note that the oil pressure tensioner never seems to stabilize during operation. The "axel" bolt on the cog wheel also continues to spin in it's aluminum bracket, that will cause wear and slop in time.

Check out the video links below to see each system in operation:

To view the AP engine oil cog tensioner in action: CLICK HERE

To view the VP Elastomeric Cog tensioner in action: CLICK HERE

For more on Cog Belt Breaks CLICK HERE