The Flywheel & Clutch Project

CPMaverick said:

For multi-disc clutches?

Click to expand...

I missed the ACT comment and replied to the wrong question.

Besides that all those clutch companies seem to sell the same clutches painted in different colors, blue, yellow, white… the offering is near identical I don’t think they make them (with the exception of the purple Xclutch from Australia which seem different), and no one seem to know the most basic characteristics of their clutch, like the torque rating which somehow is a “secret” or the minimum release force.

Then we don’t know what steel is used for the friction surfaces (flywheel, pressure plates) and the metallic discs (cerametallic pads or sintered) cut through them like butter. They don’t last.

Their steel flywheel are so soft then cannot cut the ring gear in them and affix a separate ring gear that comes with lots of its own problems, for example if the clutch is really worked the flywheel expands and can crack the ring gear, and I’ve seen various attempts to secure the ring with screws or with manual welding.

Also some of those companies just make too many applications. One of them has 89 pages of applications on their website with 20 applications per page, that’s a lot of design and inventory for a small company. They make flywheels and clutches for more than 1000 cars models, including for 40-years old cars that never ever made it to the US market. They says “made in USA” but all they show on their TikTok account is riveting the discs. No blanks forging, no CNC machining, no sheet metal stamping, no CAD, just rivets.

I start to think they all get the same clutch parts from some company somewhere, with small design variations and colors, and do the last assembly operation in the country they are based to claim they are made there.

The 7.25” are also mostly the same across all the companies that offer them, and they are all clones of the original AP Racing design from circa 1990, but again we don’t know what materials are used in those copies.

They all want their clutch to feel “like stock” and that’s probably why the torque rating is such a secret: you cannot significantly increase the torque capacity without significantly increasing the clamping force, and using hard pads against suitably hard friction surfaces requires a lot of clamping force, so they use soft (and probably cheap/common grade) steel to make the abrasive pads bite with a relatively low clamping force, the grab is terribly harsh, and the flywheels wears out quickly.

One company says their flywheels are made of “incredibly strong” 1045 steel, but 1045 is a medium-carbon common grade with a Brinell hardness around 170, which is not hard enough for metallic clutch discs.

The flywheel is designed for the TTV Racing 4488-03-00x clutch series, which comes in 4 variants with different diaphragms rated from 772Nm (560lb-ft) to 1295Nm (955lb-ft) minimums - a range that covers everything from a fully maxed out stock turbo up to a seriously big one.

The flywheel should also work with the AP Racing CP7373 with the CP2822-23FM3 1x23T gear hub, but this clutch might require a modification of the release bearing as the AP clutch is not as tall (the setup height is different.)

It really does not make any sense to talk about horsepower when discussing transmission components (clutches, gearboxes, diffs and shafts really only ever see the torque and nothing else) but to frame the ideas, I did the following: I took the rated minimum torque of the clutches and assumed it was the peak torque of the engine, multiplied that value by 0.75 to get an estimate of the engine torque at peak RPM (6500) then computed a horsepower figure at that torque and speed.

Again, clutches are not rated by horsepower, and TTV (and AP Racing, and Sachs Performance) does not publish any horsepower figures for their clutches, only torque ratings, but since some aftermarket company do it, here you go:

772Nm (560lb-ft) ~ 579Nm (427lb-ft) @ 6500 ~ 528bhp
886Nm (654lb-ft) ~ 664Nm (490lb-ft) @ 6500 ~ 606bhp
1113Nm (821lb-ft) ~ 835Nm (616lb-ft) @ 6500 ~ 762bhp
1295Nm (955lb-ft) ~ 971Nm (716lb-ft) @ 6500 ~ 886bhp

These are really almost senseless (and conservative) estimates: if the turbo is so big it can hold the boost to the 6500 redline and the engine really makes 772Nm all the way there, the horsepower figure for the first clutch becomes 704bhp, 808bhp for the second, then 1016bhp the the third, and finally 1182bhp for the last.

This is why it makes little sense to rate a clutch by horsepower without saying at what engine speed/torque it corresponds. Again for the first clutch, if the peak power RPM was 7000 instead of 6500 and the turbo could keep up with the boost up there, the horsepower "rating" would jump to 758hp, or even 813hp at 7500 RPM, if the engine kept making that much torque up there, so that's why its best to stick to torque and torque only when rating transmission components of all kinds.

I don't know why aftermarket companies (tuning as we all know it, as opposed to motorsport companies working for manufacturers in motorsport programs) are so secretive about the most important characteristic of the clutches they sell.

Note that the torque rating of serious clutches is the minimum torque they hold over their entire service life. They all have wear tolerances that can be measured when servicing the car and while under the maximum rated wear, the clutches are guaranteed to hold at least those numbers, so they should be seen as conservative as they really are the numbers they can withstand at the end of their life.

If anyone wonders, I am just solving my personal clutch problems and wishes. I have tried every clutches under the sun (well, almost) and ruled out some outright on total weight concerns, or increased inertia on the input shafts frying the gearbox synchros and slowing down gear changes, or because they seemed identical, save for the color.

I’ve tried blue clutches, black clutches, white clutches, stage 2+, 3+, and 5 with dual sintered 7.25” plates and they all failed except one, which I removed early.

The main reason I see is rapid wear of the flywheel and pressure plates. I attribute this to the use of common grade steel which is not suitable for friction surfaces, let alone sintered metal loaded with silicon dioxide (quartz) as an anti-wear agent, aka cerametallic pads.


As a hint, most aftermarket steel flywheels have an affixed ring gear. Either it’s to save some of the manufacturing costs (the rings can be purchased separately) or more likely because the steel used for the flywheel isn’t hard enough to make the ring gear. In any case the steel used wears very quickly.

The other day I saw a tiktok where a guy showed a pair of used sintered disk with the tool marks still visible on the surface (so barely broken-in) but his flywheel was ground and the pressure plates too.

He didn’t insist much on that since it was some sort of promotion for the clutch company. The guy said he did 50 runs so 12.5 miles if I count correctly, with 50 launches, and the clutch was visibly shot.

This is not motorsport hardware, just cheap lookalikes sold at a high price, and I fell for that too.

The steel used for the friction surfaces should be considerably harder than the friction pads, otherwise you aren’t wearing out the clutch but grinding the flywheel and pressure plates instead, with abrasive discs that barely sweat doing so.

What they appear to do is use a common grade material for lower production cost and soft diaphragm springs to maintain the OE pedal feel. Then they warn you the clutches may last less than OE “depending on your driving.”

The friction coefficient of the pad material is higher than an OE organic disc and they rely solely on that for the advertised (but often only vaguely specified) increase in torque capacity, not much on additional clamping force. In the metallic pad case, the pads are so hard they bite easily into the friction surface and the clutch is like an on/off switch.

Then it judders badly if your feather it because the pads hook easily into the steel they use.

The steel wears out quickly no matter how gentle you are, and the spring pressure decreases rapidly as the tension reduces with the wear of the pressure surfaces. If you lose 1mm total on a sintered clutch, so 0.25mm (or 1/100") per face on a dual disk setup, the clutch won’t hold its advertised torque anymore. 2mm total wear and you can barely move the car.

My last clutch is still harsh af but slips in first gear now, it still judders and grabs like hell and then slips right after. I get gapped by Priuses from a dig and I cannot accelerate harder than a bicycle as the thing slips immediately🤦🏻‍♂️

The only combo that I’ve seen survive a little bit is a steel insert on an alloy flywheel mated to a Kevlar face, but the other side of the hybrid (2+) disk had pucks that ground the pressure plate in no time.

A stage 3+ pucked clutch started slipping in 5th and 6 after a few months, and the dual-disc sintered metal clutch lasted seven months with 0 (zero) launches and one autox event. It started slipping briefly in 3rd gear on the second autox event. I was in winter tires and the ground was a little wet so I first thought of wheel slip, but it was the clutch.

I have been through 6 clutches in total thus far, I think. One is still somewhat okay and I keep it as spare.

The stock clutch that came with the car is also okay but the DMF flywheel is broken: the damping part is destroyed and it just has a massive play.

To break things down I’ve done 100’000km with the stock clutch and, save for the DMF damper, it’s still OK. The steel friction surfaces are a bit worn (less than 0.5mm) and the disc is maybe halfway. It never showed any sign of slipping.

Then I went through five aftermarket clutches from four different suppliers in the next 90’000km, four of them are trash/slipping, and one, the lower spec’d one, has about half its life left.

I typically do more than 100’000km with any quality clutch on a high-torque engine.

My conclusion is: stick to the Ford clutch for as long as you can. If you have a high-torque build and the Ford clutch slips or you tend to fry it (the smell is terrible) then… I don’t know.

If you see a steel flywheel with the ring gear separate from the flywheel body, and they sell it for use with pucked / ceramic / sintered metal discs, run.

If the clutch company refuses to give a minimum torque rating and wear spec (minimum disk thickness or total wear at which the torque spec is still guaranteed) - run!

If the clutch company rate their clutch by horsepower, run!

If the clutch company gives a fantasist range for the torque spec, like 400-800lbft with a stock pedal feel, instead of a single number they can be held accountable for, run!

People had some success with those purple clutches but they are large-diameter and use a double-sprung disk which adds a ton of inertia, both to the gearbox input shaft and to the crank. They look solid but just the number of bolts they put at the circumference, where they impact inertia the most, kinds-of turns me off. I hear they drive « like stock » which I can easily believe, and in organic trim they should last a long time.

Many have reported problems with their synchros shortly after installing these, though, which is attributable to the large diameter dual-sprung-discs inertia that the synchros have to accelerate or slowdown during each gear changes. The stock disc is single and unsprung so adding the spring mechanism and doubling the number of discs while maintaining the OE diameter is certain to add a lot of inertia and stress/wear on the synchros. This is inescapable.

Also they are just plain heavy. The weight I found on the web is 18.8kg (41.5lbs) which is more than the stock’s 17.3kg (38.1lbs) so it’s a step back in acceleration performance and engine response no matter what the users or the vendor says. It is bound to be slower than stock because it’s heavier with an emphasis on the circumference, where it hurts the most, and the discs are heavier too.

This might still be the best choice for many users.

I want to reduce inertia without going crazy and add the ability to hold a solid 700Nm (516lb-ft) minimum which I know my engine can peak at in cold weather. @TomekRST also has shown 720Nm (531lb-ft) with his new turbo setup, so that’s a ballpark within the realm of present reality. Needless to say, the only torque a clutch is ever going to see is the torque at the flywheel as it literally bolts on it.

The clutch package I’m putting together at great cost and risks with that home-designed flywheel should meet those goals. The total weight is around 9kg (19.8lbs) which is much lighter than stock and all aftermarket combos without being crazy light, and I have a choice of 776 or 882Nm (560 or 654lb-ft) minimum torque capacity within the release load I’m willing to tolerate on a road car. The reduced diameter from 240 to 184mm also has a big impact on inertia beyond the weight reduction itself.

I control the fulcrum diameter (the lever effect on the clutch’s spring) through a custom adapter ring that I make for the stock release bearing, so I can reduce the pedal load to something I know will be OK for road use.

I’ve driven cars where the hard launch speed (stage start) is 8500RPM and where you are guaranteed to stall if you get to the friction point with less than 4000RPM on the engine.

A good launch is when you feather the clutch for about two seconds at 7000-7500 until it grabs and you must never dip below 6000 or the engine bogs down. These engines idle at 2000 and the quickest blip of the throttle in neutral reaches 6000RPM like we reach 1500 when blipping the throttle in our cars.

I know how that feels and this is definitely not what I’m aiming for, yet for some reasons those cars are less atrocious to drive in slow bumper-to-bumper traffic (on road sections, which are the most parts of a rally, and in and out of cities depending on where the service park and the “parc fermé” are) than some of the aftermarket clutches I own.

I accumulated approx. 8’000km behind the wheel of such cars, WRC2 (207 S2000 Evo), R3 and Gr. A over the years.

I also spent countless hours in a Maxi Gr N (that won PWRC at the Monte-Carlo 2001 with my engine map, see a pic below, ahead of four Ralliart Evos) and an ex-works World Rally Car running another of my engine maps. In particular, that other car led a European championship event for a whole day, ahead of an M-Sport Focus WRC.

I worked on a ex-works Escort Gr. A (GSE shell, MTR engine) where I did a 2WD traction control and gear change strategy for a newly installed 6-speed sequential gearbox, and worked on the maps of other Gr N who won national and regional events in Switzerland, France and Italy. This is to say I know a thing or two about rally engine management, ALS, launch control, traction control, gear change strategies, and winning cars at national and international level, in addition to be an occasional driver myself. I don’t make that clutch completely out of the blue sticking random parts together without knowing what result I want to achieve.

I think/hope from past experience I should be able to feather the clutch at 1800-2000RPM with my light combo for mild launches that already pull ahead of most normal vehicles on flat roads, and do slow traffic starts around 1500 RPM with the idle speed at 1000.

The only everyday caveat is the need to use the handbrake on steep uphill starts. The hill-assist feature is calibrated for the heavy stock flywheel/clutch combo and it holds too much and for too long with a light rotating assembly, making it somewhat easier to stall because of the brakes being applied, so it must be disabled: you feather and the car stays stuck and you stall. I already experience this sometimes with my current setup which weight about 11.2kg (24.7lbs)

I should get the new combo in 2-3 weeks now, I’ll know soon enough how good or bad it is for my usage.

If it’s « The One » (haha) and I like it very much as a sporty low inertia 184mm (7.25") option for those like me who want to maximize acceleration and gearshift performance together with serious torque capacity, I’ll think of making it available as a kit, with the direct hydraulic line, a new modified Ford release bearing, the 3-plate TTV Racing clutch, the monobloc flywheel cut in the right material for the pads, all small hardware, and the required ARP bolts, as they will be on my car.

If that happens there will be four torque options with increasing pedal loads. The first one should only be moderately stiffer than stock combined with my custom adapter ring that increases the lever effect as much as the clutch specs allow. Nothing crazy to be expected on the pedal side yet it covers most builds already.


Some of my past clutches (five) and flywheels (four, OE not shown) in the « lab » to illustrate the suffering. There is one more set still on the car right now, and clutch number seven is on order. I hope it will work OK and last.