Just to be clear, I have no expertise in this area at all. But it seems to me that if the torque tube is vibrating in a bending mode, then those bending vibrations are getting transmitted to the driveshaft (or perhaps better put, the vibrations from the driveshaft are getting transmitted to the torque tube). I think that any bending in the torque tube can not be a Good Thing. I understand your comment that the drive shaft itself may be vibrating with larger amplitudes. I think what you're saying is that even if the driveshaft is vibrating/swinging, that the OEM dampener is designed to counter that, at least as much as possible in a certain frequency band, but that the swinging action of the driveshaft isn't really altered and thus clamp issues would not be affected by dampening. However, to be clear, are you also making assumptions about that, and the relative magnitudes and the degree of connectivity? We don't really know that, do we? It seems to me that we really don't have much (any?) measured data on vibrations, just long term observations. Is it also possible that the dampener, through its damping action, actually quiets down the vibration of the driveshaft itself? That may be the key to the whole issue here and what everyone is arguing about. I note that the Porsche patent, at 3:53, adds almost as an afterthought that the dampener "also acts as a torsion vibration damper." Anyway, I'll leave further discussion to the experts, who know far more about this than I do. I'm not sure if we have a way to resolve the puzzle without more data!

 

They are by my interpretation writing about the torsional vibrations of the housing tube, not torsional vibrations of the shaft.

I did some critical frequency calculations when trying to understand the torque tube maybe a year or two ago. But even without any computations, isn’t it self evident that the drive shaft, or really a steel spring, that transmits 750 lbf-ft of torque at 6500 rpm in my car is subject to forces and stores fluctuating amounts of energy that are many orders of magnitude larger than what the engine block or transmission vibrations are subject to? You don’t (in my opinion) need to measure anything to declare that self evident.

 

where did I make this claim?

 

The issue isn't whether the driveshaft vibrates -- the issue is what that does to the clamp, and to what extent any dampening dampens the driveshaft itself, and what precisely causes clamp creep/failure. Very interesting problem in vehicle dynamics!

 

Many of those are interesting questions.

What is an uninteresting question is whether the torque tube housing vibration damper does anything to the driveshaft vibrations and reliability. It doesn't, and magnitudes should make that self evident.

 

I'm not saying that isn't true, but that's an assertion without proof... I don't have any skin in this game but if there's evidence other than merely conclusory assertions, let's see it!

 

1. The heavy metal weight in the torque tube doesn't actually "absorb" anything....that's just a convenient way to talk about it. According to my main frequency engineer, the energy is converted to heat. (He also constantly corrects me from using the word "harmonics" and insists that this is a "frequency" problem. I use both words to describe the same phenomena.
2. The above is why the rubber pieces that support the heavy metal weight deteriorate faster (and appear to be melted) than the rubber pieces that hold the bearings. Bearings can do absolutely nothing to absorb frequencies, according to my main frequency engineer. Since they are connected to both the housing and the shaft, they will directly transfer the frequency vibration, but can do nothing to absorb it. And the mass of the bearing makes absolutely no difference, according to him, Bearings, quite simply, are not absorption devices and absolutely can not do this job!
3. The patent information is great and clears up some of the mysteries. It is very obvious that it took different weights/rubber combinations installed at slightly different positions to absorb the different frequencies generated in each torque tube, which is completely supported by the "yellow sticker" on each weight (with no two, I've ever seen, the exact same.)
4. The patent information did not tell us anything about how Porsche cured the problem they were having with broken torque tube shafts, in testing. Was this cured by hanging the battery box off of the rear of the transmission to absorb the damaging frequencies? (No heavy weights in these cars!) If so, then the change in 1980 removing the battery box and adding in a heavy counterweight must have been a similar solution, since the problem would not have magically disappeared.
5. It is not difficult to conclude that once the heavy counterweight moves....or the rubber mounts deteriorate, the frequencies generated will no longer be correctly absorbed.
6. Finding loose clamps (manual transmission cars), worn splines inside 5th gear (manual transmission cars), migrating torque tube shafts (automatic transmission cars) is impossible....for the first 20 years or the first 60,000 miles. I've been working on these cars since 1978 and I have NEVER seem one of these defects show up until many miles or many years had gone by. Therefore, I maintain that the damage associated with the above pieces coincides with the deterioration of the rubbers on the weight.
7. I have rebuilt ( or replaced with another rebuilt) many (many) engines that have had thrust bearing failure. In my "earlier" years, I did not necessarily "rebuild" the torque tubes. I installed several engines with the original torque tube (and properly torqued the hardware, BTW) and sent the cars home, with instructions to return at 2,000 miles. I had severalof these engines return in 2,000 miles with crazy flex plate preload. Potentially damaging amounts of preload (which never happened on a low mileage 928.) It did not take me seeing more than a few of these cars do this to change what I was doing. Subsequent rebuilding/replacement of the torque tubes always cured this issue. This, all by itself, has supported my years and years of insisting that the heavy weight and attaching rubbers were absorbing damaging frequencies/harmonics.
8. The above (5) makes it difficult/impossible to re-install the original weight back into the torque tube and realistically expect it to absorb the original frequencies. If the bearings are changed (common)...or the torque tube shaft is changed (becoming more common), it becomes even less likely for the weight/rubbers to do their original job as well as they did originally. Does this mean throwing the weight into the trash? Maybe the original weight!
9. Of interest, the change in the weight and rubbers that occurred in the 1991 automatic cars had one other major change in common....these weights were no longer marked with degrees and frequencies....at least I've never seen one that was.
10. I concluded that the "softer" 1991-1995 rubbers were more generic in their ability to absorb frequencies than the early "specific frequency/position" weight/rubber combination.
11. I've been installing the later "generic" weights and rubbers (or modifying the original weights and using the later rubbers), for quite some time. These pieces are readily available, from Porsche....which is extremely interesting, all by itself. Porsche supplies no bearings, no bearing housings, no torque tube shafts, no torque tube housings, yet has both the replacement rubbers and complete weights in inventory. No rebuilt torque tubes, yet they sell the weight with rubbers and the rubbers separate. What does that single thing tell you?
12. Until there is quantitative scientific testing done to prove that the weight and rubber do not prevent dangerous/damaging frequencies/harmonics to the torque tube, I will continue to install the heavy weight and rubbers (late model.)
13. Everyone else can do as they please. In terms of "job security, I actually prefer that the counterweight gets left out. I (or my son) will be here for many years, to repair the damage and do the job correctly. And since there is zero chance that Porsche will ever make enough replacement 5th gear sets to repair the damage, we will have those available in the future.


14. I do not see that I have anything else to add to this discussion...and my workload is extremely heavy, so I will "bow out" and allow the "throw the useless counterweight away" crowd to have the last words.

 

 

I don't agree that you don't have more to add to this discussion, Greg.

Questions that beg an answer: Given that the location of the weight and the amount of the weight itself are necessary to absorb the destructive frequencies,

1) if the existing weight has moved within the torque tube, how would a person determine exactly where to remount the weight to restore the correct absorption when rebuilding the TT?
2) if the weight is missing from the TT, how does one determine what the weight needs to be and what frequency needs to be absorbed?
3) Are the later "generic" weights all the same weight, which I think would mean that Porsche determined that the frequency needing to be absorbed was a singular frequency?

I have a loose weight in the TT of one of my cars. I know it needs to be corrected. How best to fix it?

 

I don't doubt the attached document is anything that any of the main players on this topic have not seen before. There is no favor for one point of this debate versus another brought on by revisiting the language shared below, in my opinion. Just sharing as pertinent information for folks that have limited familiarity.

 

Greg,

What scientific evidence have you provided over the years for all your engine modifications, replacement parts and TT rebuild methods, besides anecdotal observations and conclusions?

You have done what everyone else has done in the aftermarket community, you see a problem and, based upon your observations and prior knowledge in the area, you think of a solution, you then do more research into your assumptions, you make the parts, put them in and test them in real world conditions. Then if needed, they are tweaked to keep refining them, that is if you really care about the products you sell, some don't. I believe you do.

As you said, Porsche had millions of dollars in the R&D of the new 928 platform, but that didn't help them from avoiding 928 problems that have cropped up over the years due to poor designs that initially were thought to be good. That's where the aftermarket comes in, to work up solutions. The problem with scientific testing in the aftermarket, especially in such a small one as the 928 world, is the expense of it which would never be realized in the subsequent sales of those products.

The only debate we seem to have is the re-use of the OE vibration dampener again. You will not recognize the Super Bearings, which you incorrectly keep labeling as simple bearings, as vibration dampening devices which they really are and how I designed them to work.

The main concern I have always had is the re-use of the old vibe dampener and parts since they fail and in a short amount of time compared to the OE bearing units from my own observations as stated in my prior posts. However if new rubber holders are used and care is taken during their installation, using them in conjunction with my products will definitely be a "belt and suspenders" solution to stop harmful and comfort robbing 928 driveline NVHs.

As an aside and to no one in particular, if the vibe dampeners and rubber holders are available from Porsche, why then would someone want to reproduce them in the aftermarket?

And in conclusion, I have noticed the 928 forum is rich with individuals from many different backgrounds and occupations. Is there anyone out there who can conduct a test needed to qualify our assumptions for free or next to free? It would be interesting, no?

 

There are two almost completely separate questions being (unintentionally or intentionally) confused here.

The first question is whether the drive shaft support bearings that are isolated with elastic rubber from the torque tube housing are effective at reducing the torque tube housing vibration (a cabin NVH source). I don’t know the answer to that question. On the one hand, my car doesn’t sound much different with heavier bearings and no torque tube housing vibration dampener than what it sounded with stock bearings and the dampener (the bearing whine is gone, though). On the other hand, I see no reason why matching the damper plus bearing mass with new bearing mass would somehow cancel the same tube housing frequencies as the stock dampener, given that the rubber parts have different spring constants and the bearings support the shaft. Let’s call it as I don’t know.

The second question is whether the torque tube housing vibration can somehow damage the driveline, and in particular the drive shaft. Without any data, I’m willing to rule this out as a possibility just because of the degrees of isolation and relative forces acting on the components.

Confusing these two questions is analogous to confusing exhaust pulse tuning for power production with attaching a weight to a muffler to reduce exhaust drone. Yes, both problems are associated with formulas that have frequency and exhaust pipe length in them, but they are fundamentally different.

Other interesting questions are:

What are the spring constants of the vibration damper aftermarket replacement rubber springs? What is the resulting change in the tuned frequency of the dampener when one swaps out the old stock rubber springs and replaces them with the aftermarket springs?

Is there a well-tested aftermarket replacement rear clamp for five speed cars that would not fatigue and/or come loose over time? I’m thinking of situations in which torque put thru the driveline is 3x the stock torque and the twisting strain on the drive shaft similarly greater.

 

Some might find my attached thoughts on the vibration damper design of interest.

 

Despite countless discussions about this issue over the last 20 years a very interesting piece of info has been introduced into this thread that in the 20 years I have been interested in the subject matter I have never seen raised previously and that is the change that was introduced into the damper for the 91 model year. The interesting question being "why did they do this?". Had I known this back in 1999 when my original 90 S4 torque tube shaft ruptured I would have insisted that Porsche pay for the total repair as part of their "goodwill" programme but that is another matter.

I will not repeat what I have written previously other than I was convinced there is a resonant speed at 3050 rpms on the automatics and for the first time I see a resonant speed of 3300 quoted on credible documentation and also, that the thicker, tapered 2 bearing drive shaft is not the culprit but rather the victim of an underlying issue. I have yet to see a credible explanation of why some automatic clamps let go and some do not.

My experience aligns with that of GB about failure seemingly manifesting itself at around 60k miles however as I am concerned the problem is more complex than that. My shaft snapped at 90k km but after fitting a brand new Porsche torque tube I found that some two weeks after the car was returned to me a vibration appeared to exactly 3050 rpms [my association with a possible resonant speed] and the main dealers could not stop it from returning. Literature tells us that the damper is there to eliminate NVH associated with driver irritation but I suspect there may be much more to it than that. If I were to hazard a guess I would suspect that the Porsche engineers knew they had a problem that they could not resolve totally without the horrendous expense of a redesign and a total recall that would have have finished them off financially in the early 90's. Maybe the automatic's two bearing shaft was a half assed attempt to resolve an issue that hey discovered did not work and thus modified the damper for MY 1991?

No idea how or why Bill Ball could do so many miles on his 87[?] auto without issue given he was someone who on his own admission did not exactly "***** foot around". I have seen about 6 examples with issues and they all happened in the mileage range Greg suggests. My current GTS chassis has the original gearbox and torque tube fitted and after transplanting the S4 motor into the car the original items have now covered approaching 100k miles without issue and the front clamp has the Loctite solution applied to it- it could of course fail next week for all I know but to date it is holding strong.

So what does this lot & my experience to date with the automatic variants tell me?
I would suggest the following:
1. If the front clamp is not secure you are going to lose either the shaft or the engine and possibly both.
2. The plot may be stabilsed by the later damper design but it did not stop TBF on my car's original GTS motor prior to my acquisition.
3. The replacement new torque tube I fitted on the 90S4 would have been of the later design but it did not stop the clamp migration issue but the greater clamping force with the application of Loctite did.
4. I have no current reason to believe the later [tapered] drive shaft is suspect design wise despite the seemingly bad reputation it has.

Finally, any solution that works is fine by me. I dare say some folks think the Locitite solution does not work- my experience suggests otherwise- for sure it is not as elegant as Constantine's clamp that is how Porsche should have designed it to start with This thread has explained something that I was not aware of [the revised damper design]. It is also quite conceivable that Constantine's bearing design eliminates issues by virtue of its broader footprint providing a capability that the stock items do not- all I can say on the matter is that I have never read of or heard of a shaft kitted out with his super bearings and clamp ever having failed- this may not constitute absolute proof of functionality but by now it should be worth something significant in the performance analysis stakes.

Bottom line- it is a complex issue that as I am concerned is still not entirely understood. In the engineering world we cannot always eliminate the root cause of the problem but if we can eliminate the symptoms that will often suffice. I suspect the resonance issue is more of a problem with automatics simply because street motors probably spend more of their time at around the low 3k rpms speed range than any other speed and quite possibly therein lies the "root cause" of the issue.

 

Also, note the 22% difference in "resonate speeds" between manual and auto TTs from the shared doc above. It's hard not to ponder the correlation between this data and auto versus manual thrust bearing failure rates.

I speculate that the 300M main shafts and Super Bearings are both considerable improvements over stock. Have any rebuilt TTs with Super Bearings been associated with resultant TBFs, broken shafts, or damaged splines in the past 15 years?