Clutches 101: Let me start off with THE BASICS :
1)What does a clutch do?
The clutch engages and disengages the torque from the engine to the transmission. This engagement/disengagement period is a crucial element of the clutch. Without this engagement phase where the clutch slips and carries only a portion of the torque, the need for a clutch wouldn't exist.
2)There are two
main components to the clutch, the pressure plate and the clutch disc.
The pressure plate bolts to the flywheel and clamps the disc against the flywheel with a prescribed load (clamp load) generated by a diaphragm spring working through fulcrum points (pivot diameters) within the pressure plate.
The clutch disc connects to the input shaft of the transmission and once the clutch is engaged carries the engine's torque to the transmission. The linings on each side of the disc are made of various friction materials. Springs are incorporated in most street clutch discs to dampen torsional vibrations to reduce transmission noises. A marcel (wavy metal) is placed between the linings and the drive plate to help make engagement progressive and reduce chatter.
3)The
formula for the torque capacity of a clutch is :
T = NxRxPxF
T is Torque in ft. lbs., N is the number of surfaces , R is the radius of gyration (mean radius of the friction material in feet), P is the pounds of clamping force, and F is the friction coefficient.
Let's look a little deeper:
From the formula for torque capacity you can see that if the number of surfaces double (such as going from one disc to two) the torque capacity is double. If the disc size is double then the torque capacity is double, and so on.
With each variable to increase clutch torque capacity there are benefits and drawbacks, whether it is changing N,R,P or F and there are different methods to achieve these changes. This is where the plot thickens and the expertise (or sometimes voodoo) takes place.
Surface area: Notice from the formula that there is no variable for surface area, only size. This is because surface area has little to do with torque capacity. As surface area goes up, the clamp load per square inch goes down. So, the two in essence cancel each other out. More friction surface, less pressure. Surface area has everything to do with heat capacity and inertia, and not much to do with torque capacity. However, if the clutch gets too hot from having too little surface area, then it does affect torque capacity.
Number of surfaces (N) : Not usually a practical option. A double disc clutch will carry twice as much torque as a single disc clutch that is the same size and clamp load and friction. Expense, transmission clearance, floater rattle and inadequate disengagement are drawbacks that influence the twin disc setup. Two production cars that I am aware of (Acura NSX and Porsche 928) went back to a single disc clutch when the double disc proved too troublesome. It is really best suited for a racing transmission that doesn't have syncros because there is nothing to separate the clutch disc from the mating surfaces. This clutch drag can hinder fast shifting.
Radius of gyration (R): Increasing the size of a clutch will help to carry more torque and also dissipates heat over a greater area. The drawback is that also usually increases mass or inertia so the assembly is most likely heavier and finding a bigger clutch that will fit is a challenge. Centrifugal force is higher on the disc as well which means slower shifting. Unless there is an optional size already available then considering adapting to a bigger clutch is not usually practical.
So what's what two variables are left? : Clamping Force (P) and Friction (F).
I. Pounds of clamping force (P) : This is a variable that is played with quite a bit. There are
four different methods listed below:
A. The first is to simply alter the geometry of the clutch by
changing fulcrum points. This method is probably the most common because only the casting has to be removed and altered to achieve the change in pivot diameter. Since this method increases clamp load and not spring load the pedal effort remains stock and obviously the clutch will not put additional force on the thrust bearings. The tradeoffs are that the pressure plate requires more travel to operate and although the pressure plate exerts more force it is over a shorter range. In other words, the clutch will feel "soggy" and light and the clutch will wear out faster than stock, proportional to the clamp load increase. The larger the change, the more severe the negative symptoms.
B. The second method is to
weight the fingers. This will gain benefit from higher clamp loads only at higher RPM's. Since the torque curve is usually relatively flat you may not have the force when you need it. Because of the flexible nature of diaphragm spring fingers the affect of the weights is kept very limited. Because centrifugal force increases exponentially, If allowed to add only 20% more clamp load from centrifugal force at 4000 rpm then at 8000 rpm the increase would be over 100% and the clutch would have trouble operating.
C. The third method is to
stack two diaphragms on top of each other. This is effective at increasing the clamp load 100% but is somewhat crude. The design benefits are that the stress on the diaphragm spring is not increased and the additional thickness of the second diaphragm will take the additional loads placed on the fingers as well. The drawbacks are extremely high pedal effort, vague feel, and an inability to make smaller increases in clamp load. Since the second diaphragm rubs against the whole surface of the first diaphragm in an environment of clutch dust, heat, etc. friction between the diaphragm surfaces becomes very high. This additional friction is felt in the pedal.
D. The last method is to
increase diaphragm spring load thus increasing clamp load. This method is the most widely used by automakers when using the same clutch size to handle more power. The diaphragm spring is replaced with a thicker, wider or taller diaphragm to produce more spring load. The benefits are that the geometry does not need to be altered from the original clutch, the engagement is predictable, small to large increases can be engineered, and typically there is an increase in working range (clutch life). Naturally, there are drawbacks. These are: higher stresses placed on the diaphragm, higher loads on the thrust bearings, and higher loads placed on the clutch linkage. The higher loads on the diaphragm must be carefully calculated and considered in order to prevent breakage and finger flex. There is no change to engagement quality like one competitor claims. Otherwise, automakers wouldn't be using this technique.
At ACT we engineer our diaphragms to gain a generous increase in spring load and then typically make a small geometry change for additional benefit. Since we have increased the working range by use of a stronger spring, the small change in geometry doesn't significantly take away from the clutch life.If we have a 30% stronger diaphragm and have a 5% increase in geometry you get actually get a 36.5% increase, not a 35% increase because the increase is amplified.
Pressure plates are really simple in design. There are few tricks, but a lot of engineering considerations. Since we have the ability to design and manufacture diaphragm springs, we feel we have broader design options than most of our competitors.
II. Friction (F): Usually, here is where the voodoo happens. To put it in basic terms, the higher the friction, the higher the wear and harsher engagement (chatter). The lower the friction the smoother the engagement and lower the wear. This is a general statement, not intended to be true in all cases. For instance, when using solid mounts clutch chatter usually will not occur even with very high friction coefficients. Also, some inferior materials will wear quickly even though the friction is not really high and some superior materials will not wear quickly even though the friction is high. Keep in mind that friction is not abrasion Friction is the two mating materials rubbing each other, not biting into each other. We calculate most stock organic friction materials to have a coefficient of around .25, while our sintered copper ceramic material is around .32 or slightly higher. There are some sintered iron materials that are higher. We consider the sintered materials for racing only.
Notice that automakers don't have sintered materials in street cars even though it can take more abuse and handle more power with less clamp load. Even the makers of high dollar sports cars don't use these materials. The reason is because of engagement quality. Automakers know to keep the friction within a certain range in order to maintain smooth engagement. If you are willing to give up this smooth engagement, a racing clutch is an option, but usually not necessary.
Full surface or pucks?: Because most materials except sintered materials are bonded with a resin, they have a much lower melting point than sintered materials. This means that they are better suited having more surface area to distribute the heat to keep them from overheating. Since the material is fairly light weight the additional material is not a problem. With sintered materials, the binder is metal so the melting point is higher and so is the weight. Therefore it makes sense to have less material to do the job. Excessive disc weight can cause problems for syncros. once again, the surface area has little to do with torque capacity.
That's some of the basics. There is much more to it than that but it's a start. For really in depth clutch tech. I would suggest a book called Manual Transmission Clutch Systems (AE-17) published by the Society of Automotive Engineers, Inc., ISBN 1-56091-923-X. For more information on ACT clutches, see our website at
www.advancedclutch.com .
