Vector and force analysis

[Bill Musarra]

I have a problem with my 930 clutch settings and maybe you guys can help me out. I am trying to set the clutch slave cylinder and pivot bar. For simplicity I have labeled the pivot bar as the long free floating arm, short central shaft rotating arm and central shaft.

The long arm lines up with the slave cylinder housing at 90º without the slave in place. The slave has a throw of 30mm. The shaft of the slave is held in the full open position by an internal spring with a force of 14.4# of pressure.

This force is countered by a balance spring on the opposite side of the long arm.

With no hydraulic fluid in the line and no central shaft rotating arm it holds the long arm at 15mm.In order for the clutch to release the central shaft rotating arm must move 7-8mm. The ratio of the long freee floating arm to the central shaft rotating arm is 2:1. Many of you may remember the Excel spreadsheet that Tom did for the calculations on setting up the clutch mechanism on his G50. They apply for the 930 as well. His reference can be found here: http://www.gt40s.com/forum/gt40-build-logs/13742-drb-5-a-7.html at post 123. .My setup is exactly the same as Tom's. My master cylinder is .75. My master cylinder movement is the same, pedal movement the same and the slave is a standard Porsche with the exception of the internal spring.The pedals are Tilton floor units with a ratio of 5.5:1 in place of the 5:1 that Tom has.
When hydraulic fluid is added, the numbers don't quite add up. With the long free floating arm extended to 30mm and then released to allow it to go to a neutral position(15mm), it stops at 18mm. If the Central shaft rotating arm is placed on the shaft so it can be rotated through its play(wiggled), the arm will go on back to approximately the 15mm mark. If the long arm is compressed into the slave and released it will not go back to the 15mm mark with out wiggling the central shaft.
The central shaft has been wiped clean and the long arm shaft has been cleaned and a drop of 90 wgt. transmission fluid applied to give it a slippery surface to slide. Notice the position of the long arm in the following pics.

When all is hooked up the adjustment screw is set. On the 930 it is to be set with 1-2mm of gap between the arms. Sometimes when installing the short central rotating arm a large gap will be present and that screw is to take up that room. The 930s should have no force on the TOB. When put into action, the long free floating arm will not rotate enough for the clutch to release. It seems to go about 10-12mm. Even when there is 0 gap. What this tells me is that the slave is not able to get the long arm to the 30mm mark even though it is at the end of its extension. The only thing that has happened with all this adjusting is that the clutch pedal has developed a zone of free play of one to two inches in its first movement before any resistance is noticed in the pedal. This is with or without any gap of the arms. Sort of like the free play in a mechanical setup. I noticed that the long free floating arm moves about 1 mm and that the fork on the TOB has moved more snuggly against the TOB so that the "ears" of the TOB will not move, whereas they would at the neutral position. This may be nothing?
Here is where the vector stuff comes into play. I need you guys to tell me if it is nothing or something that needs to be worked on or my problem is something else. It is my guess to the problem.
At the first of the post I mentioned the long arm was at 90º to the slave or better to say the slave mounting. When at the 15mm mark the angle is at 78º and at flull extension(30mm) the angle is 68º. The balance spring is not at a true 90º to the short end of the long free floating arm. It is attached to a bolt on the trans adapter that is as close as I could get and it is close to 90º.
I know that there is a lot of physics going on here(Hooks Law, vector/force analysis etc.) so this is where I need some guidance. From what I can figure from the internet, with the angles at less than 90º some of the force exerted in the original direction is lost or rather reduced or redirected and or the distance traveled in the original direction is reduced and the amount of force needed to get to the original point(30mm) is increased. I hope that makes sense. I think what should happen is that the slave needs to be at 90º at the 15mm mark. That way at the 30mm mark the angle will be at 78º and less will be changed. The other idea I have is that the master cylinder needs to be moved in a slight bit to compensate for the reduced travel to get the arm to get to the 30mm mark. I realize this is not a rigid system in that the rod on the slave will rotate some to adapt to the slot in the long free rotating arm, so the angles will be a little off. So if you have an idea chime in, or ask a question

Bill

 

[Jac Mac]

Bill you sure you havent got the third & fourth pics swapped around with regard to the position of free floating arm? From your description I think you need to reposition the slave cyl in order to get a better pushrod to free floating arm operating angle at cylinder/pushrod location as per the attached drawing [ pushrod/cyl centerline in red ]. With your current setup it appears that you are losing leverage at the full extension point.

 

[Seymour Snerd]

...with the angles at less than 90º some of the force exerted in the original direction is lost or rather reduced or redirected and or the distance traveled in the original direction is reduced and the amount of force needed to get to the original point(30mm) is increased. I hope that makes sense. I think what should happen is that the slave needs to be at 90º at the 15mm mark. That way at the 30mm mark the angle will be at 78º and less will be changed.

Bill --

As we saw from a previous thread on your 930 clutch mechanism, I don't really get how it works so I won't mess up the thread again. However, I can help with the physics/trig question....

The really short answer is "torque magnitude = r F sin(theta)". r is the length of the lever, F is the magnitude of the force you apply, and theta is the angle between the arm and and the pushrod.

The longer but more specific answer is that when you push on a lever that is at anything other than a right angle to your push, as you said you increase motion but reduce torque proportional to the sine of the angle between force and lever. So for your 78-degree case, sin(78) is about 0.98. So you will move the lever through 2% more rotation, but you will be applying 2% less torque to the shaft. IOW, tilting levers like that is a way of "gearing up" the mechanism, as you correctly observe above; you're trading torque for displacement.

Just to finish your examples, with a 68-degree lever-tilt sin(68) is 0.93, so here you are gaining 7% in rotation but you're applying 93% of the torque you would with a 90-degree angle.

Now, keep in mind that the above numbers apply only to "instantaneous" angular positions. If you consider moving the lever through a 30-degree arc starting at 10 degrees and ending at 40, it's probably obvious that at the beginning of the arc you are applying most of your torque and moving the lever slowly, but by the end the lever is moving relatively faster, but with much lower torque. IOW, the pedal pressure is increasing. So for best overall mechanical advantage over the slave cylinder you want the halfway point of clutch pedal travel to correspond to the 90-degree position on any levers. But that's probably obvious. And if you play around with sine on your calculator you'll see that the effects don't really get serious until around 60 degrees and they then develop rapidly as the angle falls.

Hope this helps.

 

[Jim Cowden]

Bill have you had this working before or is it all new.

Im thinking the same as Jack, you may not have the slave in exactly the right spot as I noticed the mount is hand made.
Rotating the central rotating short arm on the spline anticlockwise looking from the top then pushing the slave all the way home.

The dead spot you talk about is that the free play or have you got air in the system.
I have to use a mighty vac on mine or i will never get a clutch, same on Lotus elise they are hard to bleed the air because on the length of the line front to rear.
You think you have it all out but you dont.
If you dont have a mighty vac you open the bleed and someone in the car has to hit the cl/pedal as hard as they can to push the air out, you then lock it off and repeat the process over.
Are you using a hard line front to rear or a braid, braid can swell just enough to be a pain.

Hope this gives you something to look at.

Jim

 

[EGLITOM]

Dear Bill
Jim is basically saying the same as i in my email:
"Im thinking the same as Jack, you may not have the slave in exactly the right spot as I noticed the mount is hand made.
Rotating the central rotating short arm on the spline anticlockwise looking from the top then pushing the slave all the way home."

Forget about the 15mm mark. adjust your rotating arm in a way that you have the slave pushed back at least 22- 24mm. Leave the rest of 5 - 6mm for the clutch disc thining due to usage. ( this will pull your long part of the free flaoting arm further back).

In any case the preload of your TOB is just the load of the internal slave spring. Which will take out all the play of the release leverages. This preload is standard on each porsche clutch setup and does not hurt anything. The balance spring you installed is counterforcing this preload and through pulling on the short arm pushing back the long arm. thus creating "spread tolerances " on the leverage system. In other words pulling the frea floating arm away from the rotating arm and increasing the dead play.

As mentioned in my email. I would get rid of the balancings spring, and adjust the rotating arm in way to push back the slave by 22- 24mm. If you start at the 15mm mark you will not get enough travel to release the clutch with the dead play incorporated due to the lever adjustment you have and the balance spring you have.

This should work. To avoid overacting the pressure plate, adjust your pedal stop in a way that the max positiv movement of the slave is 16-17mm from the point of time no dead play can be felt. This limits the TOB travel to 8,5mm which is fine to release ( usually at a travel of 6,2mm) but not to overstress the diaphragm springs.

TOM

 

[flatchat(Chris)]

This is my system Bill --works great -- we do use an in line residual valve (2 lb)
'cause the slave is higher than the master -- theoretically to reduce drain back so's you don't have to pump up the gap again

 

[zollis]

Sculptured art, pleasing to look at . pic # 3

Z.C.

 

[Bill Musarra]

Got held up on trying to implement some of the ideas that you guys have put forward. My wife is finishing off the basement and my garage looks like a hoarders paradise!!
Tom: I took your advice and did away with the balance spring. I think that it has a reason for being there as I saw it in the blowup schematic at the Porsche dealership. So if it eats the TOB I will find a way to make it work. Be that as it may, I will run it without the spring for now.

Jack, Allan, and Jim: You guys reinforced what I thought all along in that the start has to be at 90º. Later on I may go a little further and let it start at more than 90º and let the 90º mark be at the point of release. Something to think about.

So here is the approach I took. I took the slave mounting block off and checked it over. It is a thick piece of aluminum so altering its back side to get to 90º would not work. Plus the mtg. holes would have to be altered and it would put another variable into play with trying to line everything up.

In my first post I said,"The long arm lines up with the slave cylinder housing at 90º without the slave in place." After I thought about it, it made sense to use this surface to set the free floating arm to. The problem was that the slave rod when contracted, sticks out beyond the surface of the block. So I introduced a spacer to move the slave back enough to mount the arm against the surface of the block. I just happened to find a piece of scrap aluminum that was 16mm thick, and bored it out to accommodate the slave.

In mounting the Central shaft rotating arm I found that if I moved the free floating arm out ~13mm it lined up perfectly with the back side of the long arm requiring no adjustment from the adjustment screw.
I bled the system again with a Mightyvac I found at Pep Boys. This didn't work so well as the bleed screw has too much wear in it and it continually drew air no matter how I set it up. Even with grease around the fitting. So I went back to the one man system of using the pedal and opening and closing the bleed screw. I also took a look at how I was adjusting the pedal to get the "numbers" that I needed. I was still coming up short. My aproach was all wrong. I was adjusting the stop screw in the back of the pedal to get the pedal to move further forward. That was my mistake. I needed to unscrew the master cylinder plunger some from the pedal. This moves the pedal further back. Thus more travel of the pedal. With the system bled and the short arm off I set the master cylinder throw at 1". Tom's spreadsheet says .96" which is ~1/16" short of 1". My reasoning was that if the angle isn't 90º, then the master had to travel a little further to get the movement desired of the TOB. So I installed the short arm and set the clutch pedal to a little short of the stop screw(~5mm or less). I measured the distance the slave moved with a set of calipers the best I could. It measured ~13-14mm. So I wasn't expecting it to release. When I checked to see if the wheels turned while in gear, I was amazed to find that they turned. So the clutch is now releasing!!!!! This was last night and it was late, I was tired, and I was on call all day today. My back was killing me from all the bending I had done over the last week or so working with this problem, so I decided to leave it alone and have the "First Drive" on Sunday.

I want to thank each of you guys for helping with this problem. This is my first ever build and without help from guys like you I don't think I would have made it. This forum is really great. I am hoping it is the last problem in my quest to get this car on the road.(We know better don't we)

Bill

 

[Jim Cowden]

Glad to here it works.

When the bleeder sucks air ,use bluetac or around the base of the nipple.

Jim