The Flywheel & Clutch Project

The two custom flywheels arrived today together with the clutch. So far, so good.

Here is the final weight of everything:

• Flywheel: 4.4kg
• Clutch mechanism with 3 pressure plates and 6x M8 bolts: 3.8kg
• Driven plates (3x 6-spoke metallic pads): 1.8kg

Total clutch + flywheel package weight: 10.0kg

Besides the weight of the elements, one important parameter is the mass moment of inertia of the driven plates, as it directly affects the synchros and the gear change speed.

I assume the 3 plates are one uniform solid disk of diameter 184mm and weighting 1.8kg. Its rotational moment of inertia is 0.03047kg·m²

The OE single disk weights 1.2kg new, and its diameter is 240mm. Under the same assumptions, its rotational moment of inertia is 0.03456kg·m²

Despite having three discs, despite the three discs being heavier in total than the single OE disk, this clutch reduces the rotational inertia on the gearbox input shaft by approximately 12%

I do not know the exact figures as the weight of the discs is not uniformly spread over their radius, but I am fairly certain the inertia is reduced compared to stock.

It is also certain that the disc inertia is greatly reduced compared to single sprung aftermarket discs which weight the same (1.8kg) as those three discs combined. The reduction in diameter from 240mm to 184mm makes a huge difference.

I have checked all dimensions and everything looks good except one thing: I'm not sure how close the input shaft comes to the crankshaft flange when the gearbox is bolted on the engine.

It was difficult to predict exactly where the discs would be situated in the clutch without having the parts in hand. I will measure everything when my gearbox is removed but right now I'm not certain that the 3rd plate (closest to the flywheel) sufficiently engages on the gearbox shaft, as the whole clutch has moved away from the gearbox to accommodate the taller mechanism.

I know by measuring other flywheels and discs I have (looking at the engagement marks on the splines of the used discs) that I should have at least 6mm of engagement on the 3rd disc's hub. At least because on the other discs I see the wear marks go all the way through, so its more than 6mm, I just don't know exactly how much.

The second disk on the CM FX725 dual-disc clutch has a thin hub that engages just 8mm on the shaft, and I want to be sure that my 3rd disc engages at least as much. Worst case I have to design slightly different riveted hubs for the discs, to reach out a bit farther from the flywheel, so the shaft grabs all three sufficiently.

big jim said:

4.4 kgs is very light

Click to expand...

That's one of the key points

big jim said:

have you upgraded the gear[box] syncros?

Click to expand...

Yes but the (light) weight of the flywheel does not relate to the synchros. All the synchros see is the inertia of the input shaft when the clutch is open and the flywheel disconnected from the gearbox. I mentioned above this clutch is guaranteed to reduce inertia on the shaft compared to OE, therefore it reduces stress on the synchros by some percent and enables faster gear changes.

What fries synchros is adding inertia to the shaft. Using a heavier disc e.g. an aftermarket sprung disk very commonly found in most clutch kits for this car, in the same 240mm diameter as OE, increase inertia and therefore increases stress on the synchros and makes gear changes somewhat slower.

Using dual sprung discs in near-OE diameter (230mm) substantially increases inertia and therefore wear and tear on the synchros, as many have found out

Also be wary of the marketing wording like "peak torque" and "lightweight" thrown in lightly (haha.) At least one very well-known vendor sells a "lightweight" performance clutch for "very high horsepower vehicles" that is heavier than the OE's DMF & clutch in every respect, and the peak torque figure the day of the installation does not means much, what matters is what minimum is left at the end of the clutch's normal service life.

The proof of the pudding is in the eating 😅

I have one major unknown to figure out (if the spline engagement is enough to sufficiently grab all three discs, in particular the one that is against the flywheel) plus a couple of more minor things, such as the bearing throw which might be too much (I may have to limit the pedal travel) and in that case I cannot tweak the pedal sensor calibration (not exposed in the COBB software, I asked them if they can add this 🤞) and I might need to tweak the « safety switch » that prevents cranking if the pedal is not fully depressed.

Other than that I don’t expect major surprises.

I did a test-fit of the flywheel on an actual crankshaft and it mates perfectly, and the clutch also fits with no hand-perceptible play as far as concentricity: the step that centers the clutch mechanism on the flywheel is precision-machined and there won’t be any perceptible runout here.

In French we say it’s a « greasy fit » which is when you have about 0.01mm difference in diameter, or about 5 microns (0.2mil) of play all around with a nice surface finish. Trust me when I say this is not the case with some other clutches in my collection: some have as much as a half-millimeter of play, meaning the centering of the clutch on the flywheel is random.

I also made the tolerance of the flywheel’s center hole very tight (about 0.04mm or 1.5mil, in diameter) which is much tighter than any aftermarket flywheel I’ve seen and should take care of centering the whole assembly: the max runout is 0.02mm (0.75mil) as-is and with just a tiny dab of super thick high-pressure grease (like CMD#3) the flywheel will center itself on the crank almost perfectly and the six ARP bolts will keep it there.

Here too, some flywheels in my collection have ten times more play in the center hole, making their centering on the crank also random.

If both the clutch and the flywheel runouts happens to be aligned, thanks to gravity when installing them on the car, the combined half-millimeter-plus runout of the whole assembly is guaranteed to shake the entire car at high RPM, and to give a hard time to the gearbox input bearing. The flywheel bolts aren’t helping with the centering as the holes in some of the other flywheels I have are 12.7mm in diameter, conveniently 1/2", while the ARP bolts are just 11.9mm at their largest diameter and 12.0mm holes are sufficient.

This is to say that many aftermarket performance products cannot be trusted regardless of the nice marketing and sponsored testimonies, and some are guaranteed to cause issues down the road.

The premise is likely that people running « stage 5 » (or whatever) clutches usually have much bigger problems to deal with and the cars/engines don’t survive long enough for those issues to become prevalent. This shows the gap between « aftermarket performance » and OEM, as well as « motorsport, » where the bar is much higher.

Back to the project, in the best case the car is back on the road in a few days, worst case I need to design new disc hubs and it will take a few weeks to make them.

Rainy afternoon, car down (the pulleys I also want to fit aren't there, maybe next week) so I prepared for the worst case where the gearbox shaft isn't long enough to sufficiently grab the 3rd disk, closest to the flywheel. I designed new hubs and for that I had to model the discs and the intermediate plates to see how the stack sits and what clearance I have etc. I also made a couple of high-res renderings, 8K, 5K and 4K to give my new GPU a workout and they came out stunning. Idk is this forum lets me upload a 8K PNG?

Here is a cutaway of the disk and intermediate plates stack. The new hubs "reach out" 7mm further towards the gearbox, and allow for 1.75mm of "crush" before they bind, meaning the whole stack can crush 5.25mm which is much more than what the pressure mechanism allows. This is not necessarily the final hub design (and I don't yet know if I need them or not) but it's fun to do.

Progress was made. I’m ONE bolt from dropping off the gearbox but last time, when I though it was the last time before a long time, I cut off a nylon strap that I use to hang the gearbox after putting it back and promised myself to replace it, and guess what?

Headed to Home Depot. I’ll resume tomorrow.

Besides that I replaced the fuel vapor evap valve and pipe. With the battery and tray removed that was fairly easy. One side is hidden behind my upper catch can and piping, or but I could get (and destroy) the clip with a small screwdriver. The new pipe comes with new clips and fits super-easily.

I also replaced the spark plugs. Nothing to report they were mostly equal and nothing to worry about except the gap was easily >1mm. They (ITV24) come at 0.8mm new so it was a good time to swap. What I had “in stock” was a set of new ITV20TT “double precious metal” I.e. there is a 0.4mm iridium central electrode and a 0.7mm platinum tip on the ground electrode. Given their very small quenching area, they come gapped at 1mm and I regapped them to the usual 0.8mm (0.0315”)

The heat range is as stock. We’ll see.

Confirming the hypothesis: the 3rd party clutch I had, that started slipping after five months of relatively moderate use with zero launches, just crunched the flywheel and pressure plates. All surfaces have a marked step and the discs look almost pristine. It’s a grinder, not a clutch.

Picture this: the clutch is slipping in first gear as you see it, I can barely accelerate on the lightest throttle and the wear pads on the friction discs show almost no wear at all. All that is gone are the flywheel and the pressure plates. I don't know who really makes this hardware, but it's untested, and not fit for the purpose. This seems to indicate that the metal grade used wasn’t up to the job.

Also one can see the flywheel does not have enough clearance for the socket required to tighten the bolts. This is just poor design coupled with poor material choice.

Last weekend’s plan:


I was able to make all final measurements and I decided to go ahead and install the clutch as-is. I calculated the engagement of “good spline” on the 3rd disk to be 7.3mm which is a hair less than I hoped but will be enough (each disk only sees 1/3 of the torque on this clutch, so the per-disk engagement on the spline can be less than a 2-disc or 1-disc clutch.)

I also calculated that I have a grand total of ~4mm of “wiggle room” before the 3rd disk hub would start touching the non-rotating part of the release bearing.

I could make new hubs and reach out about 3mm further towards the gearbox to get an even 3 x 10mm spline engagement on each disc instead of 7, 10, 10 as on this prototype setup. I’ll talk with the manufacturer and see if that makes sense.

I have max. 35mm of good useable spline on the shaft accounting for the release bearing and chamfers. I could design new hubs that use the full useable splined length of the shaft, accounting for clearances and wear: the whole stack shrinks and shifts toward the flywheel as the clutch wears, and so does the clearance between the disc hubs.

If this turns into a commercial package the fit will be perfect. It will be an off-the-shelf and relatively low cost racing clutch adaptation (compared to a top-end AP Racing clutch) with a custom flywheel and (maybe) custom disc hubs.

The assembly is the lightest/lowest inertia on the market by far, yet not crazy light, and with four minimum torque options that covers all builds from big to crazy. Crazy would be a 140mm clutch and an even lighter flywheel. Is is possible to cut the total weight probably in half still, but street driving would become very impractical.

Step 1, clean everything 😅




The clutch is much smaller and lighter than the OE clutch.


The flat diaphragm requires an adapter ring on the release bearing. This one comes from my previous clutch but I’ll design a slightly different one to match the minimum fulcrum diameter of the TTV clutch, to reduce the pedal force as much as possible. This one is not optimal for this clutch but it’s only for my proto setup. The Ford release bearing is new. They fail sometime so, as a rule, replace it when you get a chance.


The recess around the bolt’s heads is necessary to accommodate the length of the bolts vs the available thread on the crankshaft. I get full thread engagement with the bolts protruding ~1/2 turn (~0.5mm) at the back of the crankshaft’s flange.


The lightest EcoBoost flywheel to my knowledge (4.4kg) in its new home This is lighter than the alloy flywheels.


For those who wonder about the safety and reliability of all this, I do full contact stress simulations so I know what minimal spline engagement I need and how stressed the contact area is, as well as how the potential new hub will cope.

I use a large safety factor and the models are exact, for example the splines on the models are actual ANSI B92.1 30° involute splines as used on the car and modeled in KISSsoft, then exported to SolidWorks where I merge them to the parts. These are not cosmetic approximations! The assembly is then exported to Ansys Mechanical for meshing and structural analysis.


One can see the stress isn't uniform along the entire spline contact area. This is because the hub twists the shaft.

I also spun the flywheel to 12'500 RPM and I see the peak stress is a good order of magnitude below the material abilities. At some point I'll simulate the whole assembly with the clutch on to get a sense of the max safe RPM of the whole. For sure it will be well above any speed the engine will ever see. I also test on my own car, so I'll be the first to find out if anything's wrong 😌

I'm not a ME but those parts aren't completely random

The gearbox is back in, which is the heavy lifting of the reassembly. The diff output seal (towards the PTU) started leaking so I replaced it, and the PTU itself also started leaking at the usual place.

I ordered the parts on Sunday night and Ford delivered like clockwork on Tuesday morning. This is really a fantastic service. I’ll finish the reassembly probably next weekend.

I have a few the points still open: the pedal travel may need a physical stop, and if it does I’m not certain the engine start safety switch will be triggered, so I may need to tweak that.

I also replaced the crankshaft pulley with something of my design, and that comes with its own set of open questions.