So I need some advice on this one:

2 years ago a CV went on me during an autocross, drivers inside (tranny side). The bearing race cracked and well, u know what happens then, no power transmitted means flatbed tow truck home. Replaced all 4 CVs with the kit from PMB performance.

All was well.... fast forward to today, the exact same joint blew. I haven't dissasembled for a post mortem but it isn't pretty. Luckily i flat tow the car to events now so getting home wasn't a problem. The car definitely took a beating today, put about 45 total runs (autox school x 2 drivers) on the car before it let go on Avon Tech RA tires, a few areas of the course induced some slip/grip stress.

Motor is relatively stock, probably putting out about 100-110hp crank. AutoX definitely put exess wear and tear on the part, running semi-sticky tires.

So what should I do, replace with another kit? OEM? Beef it up with 944 or heavy duty CVs? Any of you fellow autoxers have a solution that won't cost me a fortune?

Thanks!

I've put 400+HP through the Renegade Hybrids setup. Sticky street tires and AX, track time and lots of street miles.

Down side to big CV's is, the Trans is the next week link.

BJH (Former renegade employee just being up-front)

We definitely need some pictures and a postmortem write up. I lost a CV and I'll be curious to see if this is the same failure.

Me too.

Did you get your CV from that Shea guy Eric?

Yes, Eric took care of all 4, great kit that helped get me back on the road. I'm wondering if generally speaking those CVs aren't built to stand up to autoX use... will break down and take pics today. The contents on the CV might be shredded since i flat towed it home after. But i'll post pics to compare..

Does that CV have the axle go to extreme angles a lot?
If so, then stiffening the rear to keep it flatter in turns may help.

Yarin

Sorry about the CV. What pressures are you running in those tires? I have a set too.

Dave

Here are pics. The bearing race cracked, exact same failure mode as last time (2 years ago).

Dave - running 24-26 psi hot.

Can anyone think of why the same CV failed? Is there something specific to torque distribution or the way the diff works that would cause the left side to go first? Could it be generally related to the clockwise nature of the autox course?

Strange.. Anyhow, i feel like if i replace it with another kit, it will fail again, just a question of when.

I'm sure this wouldn't be a problem for a street car, R compound tires + autox put a fair amount of strain on the joints. The lots we run in tend to have slight humps leading to the rear end getting light and tires slipping under accel. Combine that with a few spins here and there... boom.

Oh, Eric Shea??? Are you out there???

The rear is pretty flat during cornering, suspension is rather stiff. I don't believe excess sway is the issue.

Also tires are 205/50/15, not running some massive sticky tires here either..

Correct me if i'm wrong, but isn't this a common failure on these cars?

Mark and I had been investigating an incedent where, on his way out to RRC last year, his CV inner cage snapped. These are the only two failures we've seen out of now roughly 100 CV's. Mark's situation was a little "interesting" but not beyond what others could encounter. He was pulling out of an entrance apron which allowed one wheel to extend while accelerating.

When the gang arrived here I gave him a new CV and we looked at the extension of the inner race and cage. Andy, being fresh off the Rubics CV puzzle said he thought they extended further than he had seen before.

Those CV's were disassambled/machined and then reassembled. We wondered if the inner race may have been installed the same upon reassembly. I don't believe we ever got the answer to that but, after digging into it we felt it was just a fluke.

We now machine the CV's assembled and tape the entire surface to avoid metal shavings etc. I've checked the CV's out of the box and they don't seem to hyperextend.

Bottom line: With a 2% failure rate and given the application of serious autocrossing with slicks etc. I'm not sure which way to go here... I want to know what happened so we can asses if this will be a future problem but, I don't see this as somthing that will affect others.

There have been some really good answers herein as well. You could get larger CV's (Bus axles, 911 axles w/spacers etc.) but, CV's are relatively cheap compared to transmissions.

PM me your info and I'll send another CV kit out today N/C. Make sure you use all the grease and torq everything properly etc. Sounds like this may have been another one-off (hoping).

E.

P.S. Silence Jeff! Back to that welding thing you do... now!

Yes and No. Some cars still have their stock CV's.

Are 914 CV's "robust"? Not really. Again, given your application, I'm not sure which way I'd go. I would think you could easily go 5-6 years w/o a CV failure while the normal driver could go 10-20 (or longer).

(Mr. Read... how'm I doing? )

Makes sense...

Just sent u a PM. Thanks for chiming in! Hope this is just a fluke, looks like i'll be able to get back on the road within a week.

Thanks Eric!

These are areas of concern that I see:

Looks like the inner race was rubbing on the cage:
Click to view attachment

Looks to be confirmed here:
Click to view attachment

Eric Shea;

Great reply. I know how important your reputation is to you, and before anyone else started guessing what happened, I thought you should chime in. We all look forward to the improvements you and the other 914 innovators make to modernize our 35 year old components. So, have you considered manufacturing two-levels of CV joints? One for the daily driver and a beefed-up set for the racer/AXer? Or does the racer/AXer just exaggerate the DD forces, so that a single CV type will meet the needs of both?

Thanks for your speedy reply, and have a great holiday weekend.

Keep in mind that after the CV failed, it was flat towed (4 wheels on the ground) in neutral about 50 miles. So the drive shaft was spinning within the broken bearing race. That may have causeds the signs of wear that you see. Made a not so friendly metallic failure noise all the way home...

Couple of thoughts..

The beetle cv's don't have as great an angle of operation as the 914 cv's...

That is, the up and down travel of the wheel is less with the beetle cv's.

If you measure the axle length and the amount you can move the wheel end of the axle up, you can do a little trig and get an angle of operation. If the angle measures more than the allowable operating angle,there is a clue.

Another thought- looking at the crack, inward loading of the shaft on the cv race might load the slot in the race to produce the crack as shown.

It does look like a fatigue crack that propagated. Look for nicks at either end of the crack- that's where they start. Starting end of the crack can have wear from rubbing together

The other though is around modern crap. Some of the products made today are NOT the products of yesterday. Those parts were made made with fanaticism to produce a reliable part. Today it's nick on the part? use it... Heat treat? who cares if it's a little off, Poison in the drugs? Ship it... Poison in the pet food? Ship the product...

Ken

We don't mfg. the joints. We machine 2 of the 6 bolt holes to fit the roll pins.

I currently have 4 new bus axles here that we are now looking for a new machine shop to manufacture the adapters. This kit was going to be around $499.00 complete (axles, adapters etc.). One set is actually pre-sold to a very patient customer who has been waiting for the adapter machining process.

Once these are done this will be a bolt on solution for the more agressive 914 or higher HP applications.

That makes a ton of sense... forgot about that. OK. CV kit is already packed with a label on the box. Enjoy. Sorry for the problem.

are those Lobro CV's you are using?

if they are the Empi ones then you will break them.
a known fact in dune buggy circles. some get less than an hours use before failure.

Is there any wheel hopping on the course? Something that would put high instantaneous loads on the CV? I wonder if the cryo process would have any benefit? The metallurgy of the parts may have some part in the failure.

There is really no aggrevated "hopping" as in aggressive wheel spin on launching. However, the when a corner unload on the throttle that wheel will spin a little and catch upon settling the weight distribution.

I was discussing heat treating the parts before installation with a few engineers, beneficial??

Those are German "Meyle" not EMPI. We compared the Meyle and the Lobro side by side in the first few batches and didn't see any differences.

I would think that cryo might help and there's a guy in your neck of the woods (Mass) that does it. Can't think of where they are but the guys name is Robin if I recall and very nice guy to work with.

Meyle is a German company that does not produce their parts there any longer.
Its better than Empi, but it is still an aftermarket part. Made in Turkey and beyond.
I've used their parts on my Benz and the difference in quality is immediately apparent from OEM. They worked well, but still aftermarket.

I'm not knocking your work Eric and definitely keep offering them for sure!
not like there are other options for our cars.


I would Install a German Lobro and see how she goes.

I can switch to Lobro no problem if that's deemed the problem but with 2 failures in almost 2 years, I'm not altogether sure there "is" a problem.

Lobro would be about $15-20.00 more.

Eric, I have been.... and priming,powdercoating and painting suspension parts.
All sorts of good stuff! Better living thru chemicals!

I will chime in anyways as I have exposure to CV failure in a broad area.
Lots a racing background earlier in life. Have seen my share of axle blow ups over and over again.

The stock 914 CV is fine for a stock car with normal day to day driving loads as it was designed. An acceptable failure percentage is all part of the equation in mass produced parts. If 2% or even 5% failed that is usually acceptable to the manufacturer. That’s why new cars have warranties!

Many years ago I was working with some Formula Fords and Vee cars. Nothing special. No high horsepower just really light cars that were capable or very high loads for what they were. I have also worked with high HP 911 based race cars and those loads increase exponentially.

CV's, Guibos, Tripod joints are all amazing mechanically. They do two things. They allow radial and axial mis-alignment to transmitted rotational energy. Easy enough to comprehend. We can all visualize how it works.

Now the complicated part. The design of a CV joint can only contain so much radial stress subject to torque. Radial stress is stress towards or away from the central axis of a curved member (this would be the double ball design of a CV joint) For each axis (3 on a 914 CV) we have 2 balls positioned in a V pattern. One right and one left. Without the double ball V pattern creating equalization the joint would unscrew itself to death. The forces on the outer housing are mechanically manageable. The weakness is the 4 bolt pattern. Roll pins will squish a bit and allow the outer housing to expand and twist even just 1/1000th of an inch! . Sometimes this will cause a failure..Sometimes not.

Now we need to look at "Hoop stress". HS is mechanical stress defined for rotationally-symmetric objects being the result of forces acting circumferentially (perpendicular both to the axis and to the radius of the object). Along with axial stress and radial stress, it is a component of the stress tensor in cylindrical coordinates. I know everyone has lost me here. So get a beer and read on.

The 914 CV housing is flame heat treated at the bearing contact area but not the entire housing. This is so the bearings don't wear away at the contact area. The rest (outer part) of the housing is not heat treated. This area remains "soft". With only 4 bolts holding everything together the area at the roll pins are outside the "rectangle of strength" are a weak point. No clamping pressure is present to contain the forces dead center between 3 balls radially around the circle .

Now let’s look at the bearing cage. It's thin. If the machining of the cage is not an ideal spacing to the outer housing unequal forces get generated. I know, I know it’s just a CV joint but really look at it closer. If the cage does not match the outer housing precisely within its 360 degree circumference problems develop. If the ball bearings are not equal all the way around the joint unequal forces are exhibited. Measure your balls with a caliper! Centrifugal load forces become unequal at 3 points. With torque and rpm's something is going to give. This is usually the result of the bearing cage expanding beyond its elastic limits (ala cracks) and ultimate failure. Matching the cage to the outer housing and uniform bearings are paramount to durability as with any rotating assembly.

Taken a step further. Smaller engines tend to have higher gear ratios. The lower gears are very high which get the vehicle "out of the hole" without to much effort but essentially shock the drive train in racing conditions. Even without big horsepower this still takes it toll on CV joints via wind up. The first 2 gears will have highest load on CV's. The wider the gear spacing the higher the "shift shock". High acceleration and continuous hard braking exaggerate this even more. A tight CV will behave differently than a loose CV. The closer the tolerance between housing, cage and balls will be more durable in the long run.

Another issue is angle of misalignment. The higher the angle the more stress is placed on the CV joint axially. Add the range of suspension travel in a 914 to this. The higher the angle the higher incidence of failure will occur. Choices? Decrease suspension range of travel or increase CV joint strength for given range of motion.

What can be done? I know by experience...You can hone and polish the outer housings to make everything slick & slippery! Any sharp edges can be radiused and polished. Equally done to the bearing cages. All your balls must be matched to the 1/1000th for racing! Failure rate will decrease or you can go with a more robust joint. Tripod joints have a much less rate of failure than standard CV's due to inherent design. A six bolt pattern versus a 4 bolt pattern equalizes housing forces better. A wider outer housing does wonders with an increase in torque.

INTERMISSION: I have a big bag of Reese's Pieces I am tearing into.

I am back. Being that I am going with late Carrera axles,hubs with spacers stock 914 is not so much of a concern for me. For others I maybe able to help. What is the bolt pattern for a stock 914 CV joint in inches or MM's, also I need the stock 914 axle spline #? The reason is Loebro and other suppliers have been making a 6 bolt racing CV for years that may be workable on the 914. An option is lightweight billet aluminum housings with replaceable steel cages for less than you would think.

Ok Eric. I am going back to the welding thing I do now!

I think the materails being used these days are not what they used to be.
Not saying I put the most abuse on my but more then 700 autocross runs and 30,000 miles in 5 years. the joints that were in the car were old but lasted until the bigger motor was installed. I had loose joints so I replaced them with with old spares. these spares have been in the car for several years. Big slicks and driving the wheels off it.

Like to give a shout out to Eric for being an upstanding guy.

I think the good old joints are ging to be worth even more in the near future.

Being a physicist, I can understand what Jeff is talking about. Hence my questions about suspension travel and hopping. It may be that Eric will have to offer two types of CV's. One for the street cars and another for the racers. I think an AXer will be harder on the CV vs the track guys.

Even a little slip and grab will be hard on the CV especially if the axle is at an angle. The Loebro may be well worth the extra $$ if it reduces the failure rate farther; say 4 years instead of 2 years use. More heat treatment or perhaps cryo treatment may also help extend life. Basically, the CV will always fail, it is a matter of time. It may be worth inspecting the cages for cracks on a regular basis, and replace them before they fail. I suspect the CV was failing (crack present) for a while before it broke.

Alright, alright... so you know everything! (anybody else hate this guy? Welds like he does, knows all this $hit, has all the cool tools... can devour entire bags of Reeses Pieces in a single bound ) You should write a book...

Seriously, thanks Jeff. That's about as good-a-info we've gotten so far.

What I got, plugging in Yarins autocrossing as the general use, is:



Not too many autocross courses that get beyond those two gears.

Thanks Joe... I'm wondering (goofy thought); could it be the older CV's that Joe put in have "polished themselves" a bit through the wear cycle of normal use?

Thanks for all the info guys! Joe has put way more autox time and street miles on this stock CVs than I have without any issues. He is also running MUCH more rubber and stickier rubber. The sample size of this experiment is too small, not controlled and has too many variables to come to any real conclusion. I'll put the new joint through its paces and see what happens, odds are you probably won't hear from me for a few years...

Hands down, Eric is one of the best vendors I've ever dealt with. Thanks!!

Side question - What's the best method of installing the inner race on the splined shaft? I'd prefer to install it without removing the other CV from the hub, but i doubt that is possible without inducing excess movement into the other CV. So, once removed, what technique do you guys recommend? It's a bitch to get on and off. Some sort of large press with pads?

Eric,

CV joints are a lot like ball and roller bearings, they don't manufacture them rough and expect them to polish themselves.

Like bearings, it's a finely balanced mix of tight tolerances, near perfect machining, really round balls, surface finish, and hardened surfaces....

Ken

Thanks Ken...

I was just going off Jeff comments:



My line of thinking was years of use may have caused some micro polishing...

Back to being a dummy now.

One more thing. The 914's sheetmetal suspension ears are very prone to lateral movement under cornering loads. If the ears are allowed to teater in and out on side load the entire axle assembly will follow. What does this mean? Axle moves in and out. High range of misalignment is designed into the CV assembly although the cage itself does not allow for that much lateral play. The result may be the ball trying to escape its confinement and popping the cage apart. So why did it pop?

Erics your analysis on the prior page shows a good eye for the wear found on the inner race. The cage showed very little contact wear at the cracked area. This can be from a lowered car on the bump stops while corning. Throw in some lateral movement on the trailing arm and kaboom. I do also see two burrs, one on the outer housing to the point where the edge is rolled. The other is on the inner cage at the crack. Notice is right between the bearing reliefs on both parts? Need to figure out primary cause and secondary damage from the failure. I tend to sway towards trailing arm moving in and out under load. More so as it was the inner CV that went away.

To help understand the dynamics of a CV in action the first picture shows all parts at rest in a straight line. No stress's. Doesnt really matter whether is an inner or outer CV as the dynamics are the same.

The second pic shows an exaggerated angle of attack for the axle shaft and inner race in relation to the cage and outer housing. This would apply to a lowered car. If excessive lateral movement of the outer housing occurs (trailing arm goes in/out on cornering/ acceleration load) the cage will pop.

Hey Jeff, thanks for your in depth analysis.. but also don't forget the CV rolled for 50 miles after it failed and likely experienced significant rotational wear due to the cage failure. I would say it would be difficult to extract the root cause of the failure in this situation because additional damage/wear caused by towing after destruction.

I don't hate Eric....

Anybody who does, is doing him wrong...

Eric has stepped up to the plate and is offering us a bolt in replacement for NLA parts and I respect him for it...

Personally, I like dialogue that identifies a problem, and a collective group of minds that figures out a problem and makes it go away.

Here's a thought....

Hows the bushings in the trailing arm?

Tight or loose?

If the bushings have slop and yer really doging cones, you might get some inward movement of the ends of the trailing arms which causes inward movement of the axles and REALLY high CV loads. Think that could lead to a CV joint failure????

It makes sense for a 30 year old car to have more than one problem that surfaces at another part........ Gawd, I hope we don't have to do a DOE!

So, what kind of bushings and how much slop is in the specific trailing arm????

Load the crap outta it sideways,and measure the deflection If yer really dodging, the inward deflection could be quite high.....

Ken

WOW! Welcome to CSI-Jeff Hail!!!

Until I read this I figured the Butler did it with a Candlestick in the Library...

Thanks Ken

But I think Jeff misplaced an endquote in the html, which led to the same in Eric Read's post. I fixed them.

I was just taking a at Jeff cuase he's so damn smart and he welds like a Motha &^%$er. Make no mistake, he's madly in love with me (he may not know it yet but... thems is the facts).

Thanks all. This has been some pretty good analysis for sure. Big thanks to Jeff. I don't know how you pick this stuff up. I'm still trying to figure out where the little bally looking thingy's go.

All great points once again! How can I manually check the side deflection of the trailing arm? I would imagine I need to induce a rather signifcant force to get any form of deflection. I'll take a look though. Bushings are on my list of stuff to do this winter.

Easy to do visually and by grabbing a feel.

Bushings, wheel bearings are easy to inspect- just jack up a corner and grab the tire in mid air, rocking back and forth with hands placed 180 degrees horizontally. See what moves. Now change hand position to 180 degrees verticle and try to rock back and forth. If the road wheel assembly still has the same slop or an audible metallic sound wheel bearings are suspect. If only moves in the horizonal grasp look at the bushings.


If something structural is suspected measure the inner sidewall or rim to wheelhouse distance directly above the axle line with the car flat on all fours. This can be done rough with a tape measure.

Jack the car up about 20 inches on one side enough to get both tires off the ground on the same side. If your car does not have much rear suspension travel (lowered) all the better because it will be bottomed forcing deflection making it easier to identify an issue. Check measurements again on the loaded side. What do you see?

Lower jack & repeat procedure for the other side and re-measure.

This will only provide a light loading but should identify if something is moving that shouldnt be.

If sheemetal is moving from fatigue or fracture (inner console ears) the measurement will close by a lot. You need to figure out what the lateral (sideways) movement is? From here you can get a trusted buddy to slowly raise the jack while your head is under the car focusing on the inner ear as you go from level (static) to load.

Tomorrow is Eric's Birthday

Make sure the tranny is centered between the trailing arms.
Also, with the rear off the ground and the trailing arms at full droop grab the axles and make sure you have plunge in both directions. Next remove the springs and run the trailing arms to full bump, (stopping on the way up at ride ht) then recheck the axles for plunge.
The fact that the two failures occurred to the same joint position has me wondering if the tranny is off center.
If the drive train is moving in the car in hard corners it can shift toward the outside wheel and jam the axle. Look in the inside of the tranny drive flange for evidence that the axle has been jammed into it. Use only Lobro joints. If everything is installed correctly it will take more torque to break a CV than you'll get out of a normally aspirated 2.0l. A lot more.

Eric sent me a new CV no charge, got it a few days later. I pulled everything apart and came across something a bit odd during installation. The new CV slides easily over the splines of the shaft with no resistance at all. Furthermore, there is a tiny bit of slop between the splines of the drive shaft and the inner race of the CV. I took a few shots and snapped a quick video. Let me know what you guys think. I was hoping to autocross tomorrow, but I decided to cancel due to this area of concern. I've installed five of these joints before, all require significant force to press the inner race onto the drive shaft. In order to remove the old CV, I needed a bearing puller. Something is strange... is this a concern or am i crazy?? Is the driveshaft too small, or CV inner race too big?? My first guess is that the CV is the issue since the previous CV was a tight fight.

more..

Here is a quick video I shot illustrating the concern I have http://video.google.com/videoplay?docid=51...39578&hl=en

All of mine slip on without pressing. That does look a little large, but then again, the newer parts ain't what they usta be.

And don't look a gift CV in the mouth...

Hows the tranny and engine mounts, and the training arm bushings???????

Ken

The tranny mounts are the solid hard mounts, engine mounts are good as well. I'll look at the trailing arm bushings next time.

Any comments on the slop of the inner race on the drive shaft spline? There is a small amount of slop as seen in the video, however the CV itself is tight. Run with it or find another CV? Five of the last CVs I installed were extremely tight, this one slides on by hand and can be rotated ever so slightly.

Bump.. anyone care to share their CV installation stories? How much rotational slop is OK?

I was about to reply that mine typically slip onto the splines without much force, but after watching the video, I changed my mind.

I can't tell exactly from the video how much slop is there, but it looks like the inner race twists on the axle shaft splines. I don't think it should do that. Have you tried a different axle shaft for comparison?

Could pulling your old CV inner race off the shaft have damaged the shaft splines?

I wonder if the inner race was smashed by the old broken joint to the point where the splines were crushed and now the new joint is now a loose fit on the splines.

I wish I had a definitive answer for you, but I do not. I would try a spare shaft with the new joint.

Good luck with it.

John

I have a lifted Vanagon Syncro with 230 HP Subaru SVX motor running the original CV's. I drive it hard and the CV's are still holding. They have tons of rotational slop. Other Vangon guys seem to snap em constantly. Same with my 914. 20 years of beating on the street and the autocross track the still holding up!

I think its a mystery why some fail and some last forver. "knock on wood"

Daron

Just thinking out loud here...

Do you run on a particularly grippy surface? Concrete? Race track quality or hard, coarse asphalt? One thing to possibly suspect is flex in the trailing arm or the mounting points. If you're placing a significant side load on the car, and the chassis or suspension is flexing excessively, it's possible you could be placing too much of an axial load on the CV.

That it seems to be occurring on only one side implies a possible external cause.

If this makes sense, I'd check the trailing arms and trailing arm mounts for any signs of flex/fatigue/wear/cracking. Either that, or you're running your courses on part of an oval track, and use the same banked turn at most events...

Just a thought...

-Josh2