After going through all this, I'm a little unclear about the non-M030 hubs.

Are there any failures from non-M030 hubs? It looks like everything is from the M030's.

 

For all the MO 30 hub racers, these are the facts. Many months ago originally when I was contacted by the super cup people the price was estimated to be close to $600 retail per unit that was with the first machinist. With a new machinist in a better working relationship, the price set retail now is $475 per hub and that includes bearings and studs, the stud size is the customers option The other hub being offered is without studs and is made of lighter and has the same cut outs front and back that the original hubs and in between those cut outs is were the old hubs failed it was very thin material in that area between cut outs. What some people are not figuring into the equation is you either to change your old studs over to the new hub yourself if you have the means in your garage at home, or you will have to pay to have the studs removed from your old hubs and installed in your new hubs by your technician or buying new studs and have them installed either way this could possibly cost $100 or more. I also e-mailed Jeremy who is with the 944 super cup people and offered the hubs at $475, and with a group buy even less. As Gary says be safe check them often and have fun racing.

Thanks for your time
Dan
Stuttgartmotorsports

 

Seems easy if all you got to do is remove a couple mm. Then again, I'm not a machinest. lol

 

A930Rocket, yes we will make the non-Mo30 hubs but they will have to be with a larger order more than just two or four hubs, or possibly added to a larger order with the Mo30 hubs.

Stuttgart motorsports

 

As far as the hub situation and the postings, I have been following along and have responded to some people privately who asked questions on the forum. I have very mixed emotions on posting on the forums as they sometimes become inflammatory. I believe some of the information that is being posted to the forum is incorrect and I want to set the record straight. The assumption that the failure happens due to the pocketing on the front face is not correct. This is where the crack usually becomes outwardly visible, but the crack starts much further inside at the seat of the inner bearing race. I have looked at these units carefully and as a degreed mechanical engineer have been taught how to identify and analyze these types of failures. It is not a guess, it is fact.

A crack needs a "reason" to start and the seat area of the inner bearing offers a stress concentration that over time propagates a crack. When I did my design I did it in SolidWorks, which is basically a 3 dimension CAD program. I modeled the factory original design and then an alternate one that I felt could be machined (all this was done before I did any computer stress analysis). The neat part is that with this 3-D model you can put the model/design into another software program called Cosmos that basically does a finite element analysis of the piece. I have done this on the stock piece and my piece. I also sent factory hub samples for analysis so I knew exactly what the material is, all its properties and what hardness it is etc... With this info I was able to do a very detailed analysis of the model in the Cosmos program. Interestingly enough, the Cosmos program predicts the start of the failure exactly where we see it in the real world; in the corner radius of the rear bearing race. The crack then propagates outward at a 45 degree angle until it breaks through the surface.

When I then tested my preliminary design in Cosmos and compared it to the factory design, I did not achieve much longer life because all I did was basically use a better material. But through computer analysis and a little ingenuity I have now designed a piece which is predicted by the software to last approx 6.5 times longer than the factory piece. The pocketing in the front does nothing to degradate the life of the part (this area is still blue, or lightly loaded, in the color coded Cosmos result model) and basically is there just to offer weight reduction and more surface area for heat dissipation. There is little doubt that a properly designed billet replacement will last longer than the factory piece by a wide margin, but unless we know where the problem areas are, we cannot properly design a replacement piece to take full advantage of the billet material strengths.

Designing a piece, especially an aluminum piece that is subjected to cyclical loading is not an easy task. It is all about managing the "flow" of stress throughout the piece so that "lines of stress" flow evenly and smoothly. I am taking a few liberties with this analogy to keep it simple, but basically this is correct. It should also be noted that Aluminum, unlike steel, does not have an endurance limit. In simple terms, an endurance limit lets you design a part for infinite life in a fatigue inducing environment. This means that a steel part, properly designed, will not fail due to fatigue, it will have infinite life. Aluminum will fail, there is no question about it, it's only a question of when. The more load, the less cycles of load it will last,... the better design the longer the life,... the less stress concentrations the longer the life, etc... Most people who are not engineers do not understand this concept.

Now, in terms of testing a part, lets think about this,... a typical tire is appox 6.4 feet in circumference which means the hub will go around about 823 times per mile. Just 20,000 mile is over 16 million revolutions and we all know that factory hubs last a lot longer than that when used on the street. Track service however is much more severe than street driving and subjects the hub to greater forces for a variety of reasons. So how much track testing is necessary?? If we do about 235 hours of track testing (20000 miles at an 85 mph average) we might be able to say the unit is tested. I use 20000 miles because I know of an M030 car that has just about 20000 track only miles on it and he just experienced his first hub failure due to fatigue so 20000 miles might indicate a life equal to the factory design.

The factory design is a good one; it is not as if you are going to put on a new factory hub and have a failure. We are talking about years of use and many thousands of miles of track punishment before you would have a problem with a factory hub (there is always the possibility of a casting flaw causing an earlier failure but barring that…). Factory hubs however are over $900 each and not many people want to change them out unless they need to. My idea had always been to design a hub that will give significantly more life then the factory unit at a very good price and at $375 each I believe I have done that.

With regard to the stud issue, we have chosen to leave the studs off of our hubs because of the many choices racers have in terms of stud length. Rather than install one size that may not be correct for the individual buyer, we chose to give the buyer his choice. I had mentioned in a prior post that I will sell and install a 52mm stud for $6 per stud. This is not an issue.

I don't want to start a flame war here, merely tell you why we did what we did. Many of the folks who have posted here are our customers. I make a good product and I honestly cannot think of a true failure any of my products have ever had. My hub has been thoroughly analyzed and I know from the Cosmos analysis that we were right about the source of the failure and I have a much better design then the factories by over a factor of 6, again as evidence by the Cosmos analysis.

As an aside, I will soon have non-M030 hubs available. They are in jeopardy of failing as well in the same failure mode. I have done the design (did it at the same time I did the M030 hubs) but I wasn't sure what the market would be since there are so many more standard hubs available in the junk yards, so I didn't make them. I have now decided to make them and should have them in a few weeks approx.

If you have read this far, I appreciate your patience. Thanks for reading. I would be more then happy to discuss any of this material, just give me a call if you are interested.

Thanks,
Karl Poeltl
Racer's Edge Inc.
865-675-8285

 

This one appears not to fit in your analyst program, the bearing area is quite intact?

 

Actually the picture above is a perfect example of the failure I am referring to in my earlier post. If you were to remove the rear seal, and bearing race of the failed hub pictured above, you would see that the fracture starts at the bearing race seat. The crack then progresses at 45 degrees from the source of the crack and shows itself on the outside of the hub, just as the picture above shows.

Below is a picture of a failed hub that I dissected to get a better look at the insides. The pieces that are missing are the ones that I sent to the lab for analysis so I cannot reconstruct it, but if I did it would look almost identical to the picture in the above post as you can tell from what is remaining.

If you look very closely at the race seat area that remains, there is a very visible crack that goes around the entire seat area in the corner radius. I posted a second picture below that shows the intact backside of the hub in this area as the crack had not yet propagated out before the hub was taken out of service. This is why this is such a dangerous failure mode, it happens from the inside out, and it only becomes apparent after you have a major problem, unless you are performing regular inspections of your hubs on the inside in the bearing seat area.

Karl Poeltl
Racer's Edge Inc.
865-675-8285

 

Karl,

From what I see above I would agree on the failure initiation site. The radius seems quite small for a cast part in a racing application. Add in wider, stickier and taller tires and the picture looks bleak with age/cycles.

I'll assume that the lab is going to do some micro work along with the chem/physical work. When I looked at the cylinder walls on a 944 turbo I was amazed at how much porosity I found in the castings. Also, for everyone that thinks the piston rides on a layer of exposed Si particles it isn't happening. The area fraction of exposed Si is on the small side. Apparently it doesn't take much.

Alan

 

Could this also be caused/exacerbated by internal stress set up by the tight fit of the race in the seat? eg., if there were .0005 more diameter in the hub for the race would there be less ongoing stress?

 

The interference/friction fit of the bearing races within the hub bore is what keeps them seated. If the fit was much looser, there would be a risk of spinning the race within the hub, which could be potentially damaging to the hub (galling).

But I would guess that yes, the pressing or pounding the bearing races in and out on the cast hubs could increase the probability or potential for the failures. It would be interesting to know the specifics on the failed hubs, overall mileage, track/race miles, and how many times the bearings have been replaced.

Maybe there is no connection to failures, but I would think that pounding several sets bearings in and out of a hub has to have some effect on the life of the aluminum casting.

 

You need a certain amount of pretension fit in order to have the two very dissimilar materials stay snug through a wide thermal range. A larger radius at the root (and in the outer bearing race as well) or better machining to lesson stress risers might help. I don't have any numbers but I bet the internal preloaded stress is only a part of the overall loading values. The hub I have at home and cannot find at the moment shows the same failure as the dissected one.

 

Karl,
This is precisely the way mine failed. The bearing seat was cracked through >180 degrees, and the crack propagated both in both directions - across the bearing seat to the outer [installed] side and through the flange area. The pics are in an earlier post. Kurt has my failed part, which you are welcome to have if it helps to have more failed examples to examine.

 

Karl,

One more variant to the mix... Do you (can you) plan on including a non-M030 hub for the smaller bearing, '86 spindle in your product line???

I know these aren't the "recommended" spindles for racing, but, as you note, there are a ton of them out there, and for me (and I expect others), swapping 'em (and everything else related to them and a different offset) would be prohibitively expensive at this point.

 

Is anyone up for a group purchase? I'd like to get a set of these for my '88 951S.

 

That is simply not true the hub I posted had no cracks inside or outside on the bearing area, if there had been a cracks started in that bearing area that portion would had exploded out as well as the other areas that exploded outward. Many do crack in the rear bearing area, when they crack in the bearing area they explode out from that area, to say they start cracking in bearing area but explode out in the center of the hub in the pocket areas is simply not true the bearing would break out as well. To say that "all" of the M030 hubs start cracking in the bearing area is not true, and I leave it that, it's obvious they do break!

Have a nice day

We agree to disagree.