Since 1989 I have built a few drives from 6061-t6, the drive on the older Nessa II is modular design, the hub operates on 62 mm outer race tapered roller bearings, the spindle is an aircraft wheel axle, the bearings are serviceable and may be examined (more then 700 hours on same bearings) usual annual consists repacking and replacing axle seals and new cotter pin. Due to the modular hub/sprocket assembly over the years I have been able to test the following ratios:

On my latest project Nessa IV, I settle on 72/34 due to weight, it is the lighter sprocket combination, less then 14 lb for all of the reduction system and hardware, also holds the starter on the engine. This results in some compromising of performance, five more lb would allow the 2.5 ratio the best performance ratio for the Geo 1.3.

Less then 14 lbs for this 2.12 ratio drive, it took over fourty hours to machine the upper sprocket/hub as one, started with 26 lb of round 6061-t6 and the result was 5.375 lb part. The prop extention is an option.

same drive bolts on the DOC 100 hp Older system is modular, hub and sprockets interchange

Using a precision dividing head mounted on my milling machine, I was able to create the sprockets.

Let me share with you my experiences with vibration but more important torsional vibration as I am able to understand, remember I am just an automotive tech.

The simple way of understanding what it is, in most aircraft the process of starting the engine and shutting down causes the airplane to shake, this is a form of torsional vibration. You will read about some one failing/braking mounts, brackets, castings and other parts to re make them stronger and so many hours later the same part failed again, in most cases this is probably caused by torsional vibration.

My first and only experience (relating to vibration failure) with repeated failed parts was the flex plate/ring gear for the automatic transmission installed on the 3 cyl. 1.0L.

Every 25 to 26 hours one of the 8 welds from the ring gear to the flex plate would develop a crack, then in less then a couple hours of flying the rest would follow. One time the ring gear separated with the engine over 5200 rpm, I did not know what had taken place (lots of noise) landed and I was sick from what I had experienced. I had seen the crack with in a couple of hours before but I ignored.

One day while testing the maximum timing advance available some sparks came from behind the flex plate as the engine was being gently cycled from 4000 rpm or so, to an idle, upon investigating, found the flex plate was moving almost .500" and contacting the starter drive that was stationary. I had replaced a few flex plates by now. After this a standard transmission flywheel was installed, problem went away.

Today on new installations I always run the engine and cycle from full rpm to idle and look for blurring moving parts (with prop off). The HTD belt will start flapping and cause a double line of travel as the cycle rpm causes waves in the belt. I have examined timing belts in engines running at work, some do the same at high rpm settings. But this is torsional vibration caused. Same phenomena may be observed in serpentine belts operating the accessory pulleys in some modern engines.

With the 3 cylinders also experienced surface cracks in two different propellers. By now I was getting the idea that this was not a happy installation, so I tried a 4 cylinder 1.3L on the first start up it felt smooth my airplane was not rattling or buzzing, it just felt good and sounded happy.

The only part that I ever broke on the 1.3L was a crankshaft, my fault and ignorance. On a routine annual at about 700 hours of logged flying I replaced the HTD belt with same hours because fraying on one edge of the belt. The sprocket guide/flange (I call it the belt keeper) was made of 6061 t6 and it had worn over time causing some wear to the edge of the belt, to day I make the same part from 4130. I had ordered a new belt; the new belt replacement was a HTD PowerGrip GT 2 the updated replacement for my original. On installation I had a hard time to assemble the new belt, it was tight, the belt was smaller in circular pitch diameter just a few thousands, back then the center to center distance was fixed no adjustment available. The new belt was squealing/wining this high pitch sound should have alerted any one that this was not a happy working part. People would comment that the engine sounded like a mini-turbine. Well I kept ignoring the noise.

One needs to understand that the sprockets become larger in diameter with heat, they end up to be same temperature as the operating engine, the center to center distance also increases, as engine block expands. The Kevlar belt seems to shrink with heat (I was told) all this caused a lot of pulling tension on the crankshaft, with time it caused wear to the rear upper main bearing, once the wear took place the bending moment induced to the crankshaft caused it to snap. This happen at about 80 hours after replacing the belt on a day that I logged over 7 hours of hard flying using my GPS to adjust the power setting to maintain the eta of destination (sun set), last part of the trip the engine operated 3 hours at 90% power setting with coolant temp just above the 230° f on a very hot humid day. Less then 10 minutes from home the turbine sound quit, the engine stayed at same power/rpm but the high pitch squeal/wine was gone, I turned back to an airstrip that I had just flown over, landed, taxied up to the ramp, shut off the engine, I was sure that the drive had failed, no, it was the crankshaft, no oil leaks just my new 80 hour belt had lots of play in it.

I am confident that I caused this! If the belt had been properly set in tension this should not happen!

I believe that when the system is cold the belt should slide back and forth along the axis of the driven sprocket with some resistance using finger pressure that is my way of setting tension on the belt.

I will examine the main rear upper bearing when I get to 500 hours on the replacement engine.

Back to torsional vibration, every time a cylinder fires, the connecting rod hammers the crankshaft, turning and deflecting/twisting. This twist/rebound cycle is known as torsional vibration. A four cycle four cylinder will generate two pulses per revolution of the crankshaft (two power strokes), with the engine operating at 5200 rpm this pulsating device will create 2600 pulses or 2600 times the crankshaft will change speed during the power strokes while the steady reading of 5200 rpm indicate on the tack, but your propeller or mine wants to remain constant because of its inertia. Some twisting/flexing takes place in the reduction drive, in my case I believe to be the belt, this causes belt flapping. The flapping belt becomes a form of vibration damper.

Second note to failure of crankshaft, I also believe that the excessive tension imposed on the replacement belt did not help absorb or dampening the torsional vibration from the crankshaft adding to the destruction.

Automotive engines use harmonic balancer to control vibration not necessarily to balance the engine rotating assembly.

Four-cylinder crankshafts results in a major critical speed of well over 9000 rpm that may be why we do not see torsional damper in most four cylinders. The Suzuki 1.3L DOH four valve engines do have a form of elasticity balancer but the engine is rated to 7500 rpm and operates two camshafts.

Amplitude of the pulse, increase the size of displacement, bigger piston, longer stroke, higher compression, turbo charging, all this will amplify the pulse because it is a bigger hammer on the crank.

Idle the engine too low it will start to shake the airframe, causing resonance. The Geo/Suzuki 1.3L seems to have an area between 1100 and 1200 rpm that starts to get your attention, some people add the heavy flywheel to solve this (do you see heavy flywheel on modern designed motorcycles?) but in our application there is no need to bring the engine below 1400 rpm, with a ratio of 2.21 the prop turns at 633 rpm it is ok. You should know that with this type of installation on a light airplane, final approaches to landing, one uses about 3000 rpm on the engine, lower then this, the propeller becomes an air brake device with so little kinetic energy (light plane) you will bounce the landings, took me a long time to figure out why most times I had bad landings but landed very short.

I have changed between a standard flywheel machined to 8.5 lb balanced and the auto flex plate/ring gear on the 1.3L within hours of each other, not able to tell or feel any changes, due to weight I am flying with the auto flex plate, but to be wise it should have a stiffener plate of aluminum.

My installation is as smooth as the best, on Final inspections we ask builders to start and run the engines (amateur-builts) I always put my hand on the airframes to identify any issues. So I have compared hundreds of installations to mine. At cruise speeds when the headsets electronic noise canceling system is activated after wile I am able to detect a pulsating flapping rhythm sound in both of my installations that took hundred of hours to pin point to be the FLAPPING BELT possibly absorbing torsional vibration from the crankshaft, but it is minute, it is the type of thing that you pick up on super smooth calm day before entering an area with out a place to land like a lake. I could add more weight and more parts to try to isolate it but why, the last time I got in an airplane with a Lycoming after a few minutes I wanted to land the pounding and the shaking/roughness that I had forgotten made me un-comfortable, never stopped me from flying before but now that I fly a smooth power plant I am able to compare.

Remember life is full of compromises, if looking for 80 hp the Geo/Suzuki 1.3L is an excellent compromise to the alternative 912 but some people pay the big bucks and get the 912 it is all about choices, be happy! Want to compare weights got bottom of engine page

jim@nessaaircraft.net

select a page

Nessa II Nessa IV Engines Cooling Reduction Drive Exhaust

Fuel Ignition Fuel Injection Farm Builder

Home Page