The issue of a diaphragm spring clutch has received some attention in various online forums (fora?) over the years, and just lately on this forum, some discussion about pushrod length and bearings in relation to clutch action was discussed. In that discussion, regular correspondent Bill described how he machined his pressure plate so that an increased deflection of the spring would result in an increase in clamping force in the clutch-engaged condition (let me know if I’ve misunderstood this Bill). I was somewhat uneasy about this and it niggled me for some days and so I decided to research the issue. Here is what I’ve come up with.

With a regular coil spring clutch, such as fitted to early Morris 850, the spring characteristic is linear. That is, to deflect a spring a certain amount requires a specific force (dependent on the elastic properties of the metal) and to deflect it twice that amount, requires double the force.

For a diaphragm spring clutch, things are not so straight forward. The geometry of this spring means that even through the elastic properties of the steel are linear, the actual response of the spring is not. This is used to advantage in an automotive clutch because the non-linear characteristic can be used to compensate for wear of the clutch linings. There are other advantages as well, such as more consistent distribution of mass, thinner mechanism, and more.

It is worthwhile having a look to see what is going on in both cases.

Figure 1 shows the situation with a normal coil spring clutch. The vertical axis is the clamping force. Consider line No. 1. With foot off the pedal, we begin with a certain clamping force shown by the round marker on the vertical axis, and as pedal is applied, the amount of force needed to move the pressure plate increases. The movement of the pressure plate is on the horizontal axis. You can see that it takes more pedal effort to completely disengage the clutch (arrowed line moves upwards direction).

As the clutch lining wears, we might be on Line No. 2. The clamping force with no pedal applied is less, and we need to give more displacement of the pressure plate to move it off the friction plate.

As the clutch wears more, at full pedal, we may be at the point where the clamping force is zero (as desired at full disengagement), but we are just off the lining.

With a completely worn out clutch, full pedal does not disengage the clutch and if you’ve ever done this on a car with a worn out clutch, you’ll find that there is still some friction there which in the worst cases, can inch the car forward a little if it is in gear.

Figure 2 shows the situation with a diaphragm spring clutch. The same four examples are shown, but because of the non-linear, and even negative characteristic of the diaphragm spring, the clamping force initially increases as the friction lining wears. However, we do get to a point where the clamping force decreases and we approach the coil spring situation where eventually not enough pedal can be applied to disengage the clutch.

Fig. 3 shows the situation in which the clutch has been modified. For the sake of convenience, I’ll call this “Bill’s clutch” in the hope that he doesn’t mind too much.

The effect of the modification is essentially to thicken the friction plate so that everything else being equal, the diaphragm spring is more compressed with the clutch fully engaged (foot off the pedal).

But note carefully where curve No. 1 is now at. The clamping force is now less than it was before, and indeed, equivalent to that of Curve No. 2 on the standard clutch. As Bill’s clutch wears, the clamping force increases, but eventually we get to the point where it is completely worn out as before, and it becomes impossible to release the clutch. The significance of this is that while Bill’s clutch gives less clamping force when “new”, it retains or gives an increase in clamping force for longer as the clutch wears.

Quite an interesting modification, but it taken too far, then the “as new” clamping force may be so low as to result in slippage – which paradoxically will improve as the clutch wears.

Regards, Tony

I found this useful in explaining the diaphragm/pressure plate setup and configuration.
https://youtu.be/C3c5EcpYiP8?si=634LDwFUiotsaIr_

There is also a part 2.

I think your figure 2 is assuming a perfectly matched set of components where the spring deflection is already optimised. Having added a lightened flywheel with material removed from the strap/plate mating area, and with a pressure plate that had its face machined, i had to go through the process of measuring and adjusting the pressure plate ‘horns’ to bring everything back to an optimal setup. So by machining the plate its figure 2 you would aim for. (Or add material, but less likely)

_________________
1966 Mini Deluxe - “Gabby” 34 years under restoration and counting...
1970 MGBGT
1959 MGA

Yes that's right. It's curve #1 on figure 2 that should be the starting point. Any modification to the parts involved should result in things coming back to curve #1 so that there is room for wear without sacrificing clamping force. If you are too far to the right, then you lose clamping force (but regain it as things wear). If you are too far to the left, then you lose clamping force as things wear and there's no going back. The only real way to increase clamping force is to get a different diaphragm.

For a coil spring clutch, you are already on the slide down to the left and you lose clamping force as things wear. In this case, you can shift the static position upwards by compressing the spring more (limited by coil binding).