CV joints, Why they come off Options

[Bartlett 914]

I have experienced CV foints coming off my earlier car. This is an issue that many cars go through. I see threads on this topic from time to time. Most replies involve using new Schorr washers or wiring the bolt head. Some have used longer screws with jam nuts on the back side of the output flange. I do not see the actual cause being discussed and I would like to offer an explanation here.

What I have found is when the CV joint works loose. the stress on the bolts pulling on the output flange of the transmission distorts the flange. This results on a smaller contact surface between the CV and the flange. I corrected this using my lathe years ago an I have never had the work loose again. They worked loose in the past within a year.

Yesterday a local 914 member came over and we addressed his CV joint. One had come off. It did damage the the shift rod and caused considerable him grief. We replaced the shift console bushing and the shift rod. We then installed a replacement CV joint. I removed the output flange and put it in the lathe. Here are 3 shots. The first shows the condition before doing anything. You will see the flange surface is not clean. If it had been in full contact it would have looked better. The second shot shows a very slight clean up cut. Here you can see there is a 10 mm wide section where the flange cleaned up. Around all the bolt holes you can see a slight cleaning. What this means is no matter how tight you make the bolts that this is the only surface that will make contact! This results in allowing the CV to wiggle slightly under load. This leads to the screws backing out. So only a very small percentage of surface is making contact. For the correct results you need 100 percent contact between the two surfaces. The third shot is the final cleanup. I removed about .005" to get there. This metal is not real hard and cleans up easily..

If you do not have a lathe, then a flat surface and abrasive paper may work. Cutting a small countersink at each hole may also be helpful. You want to make certain that the threaded holes are hot high spots on the flange. New washers are advised.

[Bartlett 914]

Here is the finished flange

[ctc911ctc]

Every new job is an opportunity to buy new tools!

Now, how big is a lathe?

[Superhawk996]

Nice job with clean up. (IMG:style_emoticons/default/piratenanner.gif)

As you state, the flange most likely was distorted by the wobble induced by a CV joint that has loosend and was driven with it loose.

However, proper torque and new Schnorr washers are mandatory.

When you get to root cause of the CV issues, it comes down to loss of clamp load. Way too many guys out there not using torque wrenches. A torque wrench is the price of entry to doing things properly and reliably.

Since you have a lathe, I'll assume that having a torque wrench is a no brainer. (IMG:style_emoticons/default/smile.gif)

[Bartlett 914]

Nice job with clean up. (IMG:style_emoticons/default/piratenanner.gif)

As you state, the flange most likely was distorted by the wobble induced by a CV joint that has loosend and was driven with it loose.

However, proper torque and new Schnorr washers are mandatory.

When you get to root cause of the CV issues, it comes down to loss of clamp load. Way too many guys out there not using torque wrenches. A torque wrench is the price of entry to doing things properly.

Since you have a lathe, I'll assume that having a torque wrench is a no brainer. (IMG:style_emoticons/default/smile.gif)

I think you missed the point I was making. No matter how tight you make the bolts, the Joint will work loose. The gap created by the flange not being flat will allow movement between the surfaces, If there is no movement, then the bolts will not have any forces causing them to work loose!

[Superhawk996]

Nice job with clean up. (IMG:style_emoticons/default/piratenanner.gif)

As you state, the flange most likely was distorted by the wobble induced by a CV joint that has loosend and was driven with it loose.

However, proper torque and new Schnorr washers are mandatory.

When you get to root cause of the CV issues, it comes down to loss of clamp load. Way too many guys out there not using torque wrenches. A torque wrench is the price of entry to doing things properly.

Since you have a lathe, I'll assume that having a torque wrench is a no brainer. (IMG:style_emoticons/default/smile.gif)

I think you missed the point I was making. No matter how tight you make the bolts, the Joint will work loose. The gap created by the flange not being flat will allow movement between the surfaces, If there is no movement, then the bolts will not have any forces causing them to work loose!

I get the point.

I've done joint design and fastener testing for years. You are correct that a flat surface to make a "hard" joint is the starting point for a reliable joint. However, a flat surface with improper clamp load applied still won't be reliable.

My point was that surface started it's life flat when it left the factory. If it has been distorted, it is mostly likely a symptom of the joint having wobbled loose previously. There are way too many CV joints loosening up for that many to have left the factory with warped flanges. Way more likely that they aren't being torqued properly when guys make repairs.

I think you did a great job cleaning the damage back up and getting back to a solid joint as a starting point.

If the fasteners have been damaged by being torqued beyond yield previsouly (not likely with a grade 12.9 fastener), are not torqued properly, and new Schnorr washers aren't used, you're likely to see the joint loosen again despite the great job you did on returning the face to flat.

[Bartlett 914]

Nice job with clean up. (IMG:style_emoticons/default/piratenanner.gif)

As you state, the flange most likely was distorted by the wobble induced by a CV joint that has loosend and was driven with it loose.

However, proper torque and new Schnorr washers are mandatory.

When you get to root cause of the CV issues, it comes down to loss of clamp load. Way too many guys out there not using torque wrenches. A torque wrench is the price of entry to doing things properly.

Since you have a lathe, I'll assume that having a torque wrench is a no brainer. (IMG:style_emoticons/default/smile.gif)

I think you missed the point I was making. No matter how tight you make the bolts, the Joint will work loose. The gap created by the flange not being flat will allow movement between the surfaces, If there is no movement, then the bolts will not have any forces causing them to work loose!

I get the point.

I've done design and fastener testing for years. You are correct that a flat surface to make a "hard" joint is the starting point for a reliable joint. However, a flat surface with improper clamp load applied still won't be reliable.

My point was that surface started it's life flat when it left the factory. If it has been distorted it is mostly likely a symptom of the joint having wobbled loose previously. There are way too many CV joints loosening up for that many to have left the factory with warped flanges. Way more likely that they aren't being torqued properly when guys make repairs.

I think you did a great job cleaning the damage back up and getting back to a solid joint as a starting point.

If the fasteners have been damaged by being torqued beyond yield previsouly (not likely with a grade 12.9 fastener), are not torqued properly, and new Schnorr washers aren't used, you're likely to see the joint loosen again despite the great job you did on returning the face to flat.

I agree that the CV had come loose for a variety of reasons improper torque included. The issue of loose CV joints keeps coming. Once one has loosened, the flange will be distorted. It needs to be trued up in the repair. It is this message that is missing on every thread concerning CV joints coming off.

[Superhawk996]

Nice job with clean up. (IMG:style_emoticons/default/piratenanner.gif)

As you state, the flange most likely was distorted by the wobble induced by a CV joint that has loosend and was driven with it loose.

However, proper torque and new Schnorr washers are mandatory.

When you get to root cause of the CV issues, it comes down to loss of clamp load. Way too many guys out there not using torque wrenches. A torque wrench is the price of entry to doing things properly.

Since you have a lathe, I'll assume that having a torque wrench is a no brainer. (IMG:style_emoticons/default/smile.gif)

I think you missed the point I was making. No matter how tight you make the bolts, the Joint will work loose. The gap created by the flange not being flat will allow movement between the surfaces, If there is no movement, then the bolts will not have any forces causing them to work loose!

I get the point.

I've done design and fastener testing for years. You are correct that a flat surface to make a "hard" joint is the starting point for a reliable joint. However, a flat surface with improper clamp load applied still won't be reliable.

My point was that surface started it's life flat when it left the factory. If it has been distorted it is mostly likely a symptom of the joint having wobbled loose previously. There are way too many CV joints loosening up for that many to have left the factory with warped flanges. Way more likely that they aren't being torqued properly when guys make repairs.

I think you did a great job cleaning the damage back up and getting back to a solid joint as a starting point.

If the fasteners have been damaged by being torqued beyond yield previsouly (not likely with a grade 12.9 fastener), are not torqued properly, and new Schnorr washers aren't used, you're likely to see the joint loosen again despite the great job you did on returning the face to flat.

I agree that the CV had come loose for a variety of reasons improper torque included. The issue of loose CV joints keeps coming. Once one has loosened, the flange will be distorted. It needs to be trued up in the repair. It is this message that is missing on every thread concerning CV joints coming off.

Fair point.

I would suspect deformed flanges would be the exception rather than the rule but truing them up certainly isn't a bad thing. Didn't mean to imply that your point about flatness wasn't valid.

My mantra continues to be start by using a Torque Wrench as the fundamental premise. Guessing doesn't work. I can't tell you how many professional mechanics I've run into over the years that think they have a calibrated "feel" for fastener torque. I'm not buying it for a second.

I also wonder how many guys are running around without the proper shear pins in place on the CV's. Without those shear pins, you won't be able to put enough clamp load on that joint with the 4 standard fasteners alone to keep it from loosening.

[914werke]

not to throw fuel on the fire but to ignore the locating roll pins in this assembly is folly. Those two opposing points need to be TIGHT. If you can unscrew the 4 12 point fasteners & your CV just comes off, its time to renew your roll pins, CV, address the hub, or some combination

[R Dub]

That's my output flange in the pictures and it definitely was not flat before it got put on the lathe. Just wanted to give a huge thanks to Mark (Bartlett 914) and Glenn for all of their help and expertise. I learned a lot from them and needless to say the car wouldn't be running without them. Just can't thank then enough. Looking forward to hanging out with them again and learning more and more about these cars.

[R Dub]

*duplicate*

[Bartlett 914]

That's my output flange in the pictures. Just wanted to give a huge thanks to Mark (Bartlett 914) and Glenn for all of their help and expertise. I learned a lot from them and needless to say the car wouldn't be running without them. Just can't thank then enough. Looking forward to hanging out with them again and learning more and more about these cars.

(IMG:style_emoticons/default/beerchug.gif)

[wysri9]

All read and understood - very helpful. Further question though. How tight is tight for those roll pins?? I cannot get anywhere close to engaging the roll pins by hand. Should I need to torque up the bolts to pull those in? I am about to make up my drive shafts and it would be very helpful to know how that should go....

[Bartlett 914]

All read and understood - very helpful. Further question though. How tight is tight for those roll pins?? I cannot get anywhere close to engaging the roll pins by hand. Should I need to torque up the bolts to pull those in? I am about to make up my drive shafts and it would be very helpful to know how that should go....

It is a spring pin. It does fit tight as this is a big portion of the driving force. It should start but you will have difficulty engaging fully by hand. I sure can't. Make sure it is started and use the bolts to pull it together. Do this evenly. Don't do this all from one screw. Go from side to side somewhat

[wndsrfr]

Ummmm.....no mention of the use of the gasket in this thread. Uhh...wouldn't the gasket compressing take care of the lack of perfect finish? What am I missing here??

[wysri9]

All read and understood - very helpful. Further question though. How tight is tight for those roll pins?? I cannot get anywhere close to engaging the roll pins by hand. Should I need to torque up the bolts to pull those in? I am about to make up my drive shafts and it would be very helpful to know how that should go....

It is a spring pin. It does fit tight as this is a big portion of the driving force. It should start but you will have difficulty engaging fully by hand. I sure can't. Make sure it is started and use the bolts to pull it together. Do this evenly. Don't do this all from one screw. Go from side to side somewhat

Thats really helpful (IMG:style_emoticons/default/beerchug.gif)

[Porschef]

This is very interesting; years ago when I swapped out engines I put CV gaskets on where there hadn’t been ones. I too, had one loosen and detach from the flange (drivers side); fortunately I was going only about 25mph and started feeling the wobble. Minutes prior I had been going over 75...

My question is about removing the flanges from the tranny, I can’t remember how it’s done. If it’s simple enough I’d like to do this when I have the trailing arms out.

Also, are the gaskets mandatory? Been fine for years without them.

I do have a lathe.

Thanks in advance (IMG:style_emoticons/default/beerchug.gif)

[brant]

So my theory is that the gasket actually causes the joints to come loose

It’s a very thick gasket and can deform a small significant amount
Thus when a new gasket compresses it can allow the bolts to loose proper tension and loosen

I have not ran gaskets for about 20 years now
I do run some silicone, new washers every time
My cv”s no longer come loose

[Porschef]

Interesting. Thanks for the reply; I wonder how many others are running without the gaskets... (IMG:style_emoticons/default/beerchug.gif)

[Spoke]

Ummmm.....no mention of the use of the gasket in this thread. Uhh...wouldn't the gasket compressing take care of the lack of perfect finish? What am I missing here??

Doesn't the gasket fit in a recess in one of the surfaces? So that the gasket wouldn't interfere with the mating of the flange and the CV?