TR6 flywheels & clutches

>Since the question of clutch & fly wheels has arisen, I'm
> wondering if anyone has tried putting together a small button clutch
> (Tilton or Quarter Master) with a flex plate,on a TR-6 ? I can only
> imagine the rotating mass coming down to near nothing.

>David C. Wingett

Having had just a big bunch of history with the TR6 engine and flywheels and clutches flying off, making an installation with a light flywheel and the lightest smallest clutch makes a significant difference in the durability of the crankshaft and the available rpm. When making the flywheel (aluminum) have the center hub as deep as possible with about a .002" interference fit. This hub should then go onto the crank almost touching the seal. The problem is that the flywheel actually wobbles as the vibration of the crankshaft is transferred to it and this movement is then trying to PRY the flywheel off the crank. When you lighten or reduce the size of the clutch you reduce this movement by about one half. I measured all this on the dnyo with a oscilloscope and potentiometers.

For the fitting you then heat the flywheel and it will slide onto the crank end. I also drilled and taped the end of the crank for FOUR more bolts. Use bolts that are HARD not just strong. By hard I mean bolts that have a very limited stretch before breaking. Bolts that are strong and stretch allow the flywheel to slide around and wobble but the bolts don't break. Too bad, cause you are still a DNF.. Machine bolts such as used for hold downs in shop machinery such as a mill are hard and do not stretch.

Kas Kastner

> Kas
> Thanks for your response. Yes, I have also had a fly wheel
> come lose but, not disconnect ( last lap of the last race, fastest lap
> of the season). Now, with the ATI balancer, and all the other
> balancing work done, shouldn't that take care of any odd
> vibration...? The Tilton and Quarter Master button clutches are very
> state of the art & balanced to a Bone. The only problem I would see
> is finding a source for the flex plate, using a stock ring gear (of
> course it would have to be CUSTOM made, every thing else is) and
> calibrating a throw out bearing to match the clutch.

>David C. Wingett

The problem is twofold. It is weight and it is size which of course is surface speed. The larger the diameter the higher the surface speed. To make the best of the bad rules and possibility of keeping everything together I used an aluminum clutch from a Porsche and an aluminum flywheel. I made a steel hub to fit over the crank with an interference fit then bolted the flywheel to the hub. This combination gave a TOTAL weight of 13 pounds. Were I to do this again I would go to a multiple disc clutch. Then you would have the light weight AND the small diameter.

To reduce the largest diameter weight, I even removed the ring gear teeth in specific spots as I knew the engine always stopped in certain areas and I could probably get away with it. Lots of work to not much benefit. We then turned the engine to 7500 to with pretty good reliability. The rpm range of the torsional vibration in the crank can be moved around but is always there. The crank in the TR6 has no overlap and thus is a whippy devil. I made the weights as light as possible and this gave a problem at about 4400 RPM again at 6100 RPM and then the last vibration was over 8000 and was not of importance. This third vibration by the way was so severe it was off the page on the pen recorder. The biggest problem we had was holding the engine revs at the proper amount on a yellow flag or the pace lap to not be in the range of the vibration. I figured out that light was good by the simple theory that if I tooled off the flywheel to nothing, then what could fall off.

NOTHING. Therefore light is good. I would like to add that when we whipped the clutch flywheel problem then the torsionals went to the other end of the crank and broke the timing chain, we fixed that with a double row, then it actually sheared the camshaft bolts to where we had to add another bigger bolt to keep it all intact. Worth doing though, we WON. Made 252 bhp at 7900 rpm in 1972. Boy this stuff is old isn't it.

 Kas Kastner

> I pulled the transmission on my TR6 IT racecar Tuesday. The flywheel was held on by two bolts(barely),
> and the other two bolts were broken. The crankshaft looks like there is metal from the flywheel molecularly
> bonded to it! Has anyone ever experienced this before or have an explanation for this phenomenon?
> The PO must have run it a while with the flywheel loose.

This is the same trouble that I have sent severeal messaages on. The TR-6 engine is prone to this because the crankshaft is a whip ( no overlap). Measured on a scope with potentometers on each side of the flywheel at the ring edge, the flywheel moves back and forth over .500" at 6200 revs. What this does is actually stretch the bolts, then the flyweel is loose and the bolts break in shear and all is lost. The fix I found was the lightest combination flywheel and clutch,( mine was total of 13 pounds) then put 7/16" FLYWHEEL bolts in the crank and add two more. The bolts must be hard not just strong....not grade 8 or stainless or most of the aircraft stuff. They are stong but will stretch. The NASCAR shops have the ticket. Hard bolts is the ticket, such as they use for hold down clamps on a mill.

 Kas Kastner

> Yes, I thought I'd switch to ARP bolts, anyway. I still don't know how to
> evaluate whether a bolt is "hard" or "strong", but at least I'm more
> comfortable that my flywheel bolts won't break... :-)

Hard versus strong is easy to explain, but I'm not sure what physical parameters denote it, certainly not grade 5, grade 8. A strong bolt can stretch but won't break, a hard bolt won't stretch but when it reaches yield it cracks rather than stretching. The ARP bolts seem to be both. From my limited metallurgy reading (mostly associated with knife making) I think the "psi" number is strength, that is, they stick a sample in a measuring tool that tries to pull it in two. Hardness is a different number, which as I recall is measured by pressing a graduated diamond point into the metal and seeing how deep it penetrates with a specific force applied. Sometimes you see knives with little dimples where they did that. There are a bunch of other tests, like Charpy V-notch that tests the resistance to shear forces. And Brinell that measures another aspect of
hardness. But it all distills down to "buy some really good bolts", and that basically means "pay ARP what they ask".

Bill Babcock

I, too, got into some metallurgy when making knives. I think the essential quality required of the bolts mentioned by Kas is "toughness," which is a combination of the qualities of tensile strength, to a great degree, and to a smaller degree, resistance to deformation and cracking.

The relationship between tensile strength and elasticity is determined by Young's modulus, which is, effectively, the multiplication of stress and strain (stress being the applied force, and strain the amount of stretch in the material for that applied force). Virtually all alloys of steel are very close to one figure -- 2.9 x 10^7 psi. So, the lower the tensile strength at the yield point, the more the bolt will stretch before yield without permanent deformation (known as elongation). The higher the tensile strength, because of Young's modulus, the less elongation before the yield point. A grade 2 bolt will, under load, stretch by 15-20% before yielding, and still come back to its original dimensions when the load is removed. A grade 8 bolt will take a much higher load, but will elongate maybe 4% before the yield point.

Bolt hardness comes into play particularly in situations like flywheel bolts. Surface hardening a bolt does make that small amount of the surface material more brittle, but also helps prevent very small nicks, such as can happen when the edge of a bolt hole bangs into the bolt under big loads. Under large pulsating loads, nicks grow into cracks, causing the bolt to fail.

The relationship of pulsating or cycling loads (good examples are flywheel mountings and cylinder heads) to the fasteners used is that the tensile strength before yield of the fastener must not only be greater than the peak load, the fastener must also be able to be torqued to a high enough value to pre-stretch the bolt so that the load applied on the bolt when torqued exceeds the peak load by a comfortable margin. That way, in the case of a flywheel, the bolt absolutely restrains the mounting area of the flywheel from moving at all. And, because the shear loads at the flywheel are very high, the bolt also has to have sufficient strength to resist shear.

The problem is complicated in the areas mentioned above by the fact that the block and the crankshaft are cast items, and the tensile strength of those castings is not equal to that of the bolts used. That means there has to be sufficient receiving thread area to spread the torqued load so that the threads don't yield, but are still put in tension enough to resist movement under peak load.

Getting back to metallurgy, bolt cracking is always of concern with any steel alloy, because of the cubic geometry of iron's crystal lattice structure. That makes for lots of flat planes to slide against each other if there's sufficient force to create a displacement (the starting point of a crack). By contrast, titanium alloyed with very small amounts of iron creates a hexagonal crystal structure, and those hexagons lock into each other (something like cells in a honeycomb), making displacement in all planes much more difficult.

Unfortunately for us, commonly available titanium alloys do not always have the ultimate tensile strength of, say, a grade 8 bolt, so a larger bolt must often be used, and they are quite expensive and require some careful handling and protection. Certain alloys of titanium are very susceptible to environmental damage. When the first versions of the SR-71 spy plane (the airframe of which is virtually all titanium) were being assembled with titanium bolts, workers would assemble a portion of the frame, and come back the next day to find all the bolt heads laying on the floor. After some investigation, it was discovered that they were using cadmium-plated sockets to install the bolts, and the cadmium transferred to the bolt heads during installation was enough to cause a chemical change in the alloy which reduced its strength to next to nothing. (!) Thereafter, every tool kit had only sockets with a black oxide finish.

Michael D. Porter

Return to the page top