Lizard928

Tom,

Short answer, don't wait as when the spring fails, it can take the box out with it.....

atb

Here's more pics, pal nut removed with lock nut backed off and pertinent page in WSM:







I followed the WSM spec and backed the stud out to 20mm, and it was a lot of tension when engaged. The shift was very firm. It felt nice and solid, but I couldn't understand why the spring would need so much tension when engaged, so I backed it out a couple mm's. Haven't decided on a final setting yet but will post when I do.

Lizard928

Adam,

keep in mind that my spring is of slightly different dimensions, so the factory 20mm measurement is not really required.
Its just to push it in till it touches, add maybe 1 turn, and lock it down.

GregBBRD

Bad example. Not spring steel, just steel.

Want proof?

Pull on it in the middle with a chain and watch it bend.....tow truck drivers do it on a routine basis.

I've been doing a lot of research on this shifting spring....I'm down to my last few new ones and have perhaps only 20 used ones left. I've been asking a lot of materials/mechanical engineers about welding on spring steel....and I can't find a single one that will say anything but "Only if you want to go to hell or have hellish results." I can't even get the Chinese to do this....and they will do virtually anything you ask them to do!

I respect Colin, but I'm still shaking my head on this one. I absolutely don't see how it can work, long term.

If not broken, I wouldn't sell or throw my old one away....

atb

That's where I am out now, just want to measure the stud length before I lock it down.

Seeing the spring in action really makes it easy to see why it breaks. When you select 1st or reverse, it does take a pretty heavy load.

Lizard928

When subjected to enough force, spring steel will bend and stay bent.
However when this bar is reefed on, the inner/outer welds are bending and the welded portion is holding it in a fixed position.
If this was a regular mild steel, it would not return to its fixed position. The moment it is forced back, it would stay back, or rely on the front rubber bushing to pull it back. The fact it returns to the same position for years on end indicates that it is a form of spring steel.

A simple test is to take a rear lower control arm off the 928, and push on the inner bushing which connects to the cross member. It returns to its original position (granted that you dont bend it too far, if you do it will flex the inner weld and remain bent). If you take the same length of mild steel and clamp one end and bend it, it will retain some of the bend (nearest to the clamped end), almost regardless of how little you bend it. In fact if the piece is long enough it will bend itself.

My #1 goal with this design was to prevent the catastrophic (potential) failure that the stock spring allowed. Where the stock ramp is pushed through a square hole (major stress risers), and then punched from the back pushing sideways on the square hole is the main failure point. I looked at every possible way of making these without welding. At the time I made them, I couldnt find even 1 way of making them without either welding them, or having stress risers...
As stated previously, the welding of them does produce heat affected zones, which does harden the (affected area only) spring, however all this does is prevent that area from flexing, instead it remains fixed, until too much force is exerted and the HAZ area will simply bend. You could take the spring I made, clamp the round area in a vice, and with a fair bit of force bend it over to a 90 deg angle and it still will not break off and go through the gears.

I am very curios as to how you predict that my spring will fail? I have a couple spares and would be more than happy to do any tests to see if I can make one fail!

atb

It would be interesting to test a stocker to destruction as a baseline. Basically jig it up and bend it back and forth like what happens inside the tranny until it breaks. I wonder how the heat generated in the gear box is a factor, i.e. are the LSD boxes more prone to this failure?

Lizard928

I have not noticed a difference in the failure from LSD to non LSD boxes......

You just took your stocker out, I have another one of those springs here, feel free to come test away to destruction. I will not be liable for any physical injuries you sustain in the process though.......

GregBBRD

You have to realize that the stock springs survived for many, many years, heat cycles, and flex cycles before any failures occurred. It's not like these things failed under warranty....the huge majority of them are still in service!

Tough for you to duplicate these conditions or even simulate them.

However, the results of a broken spring can be terrible and I think it is something you need to do, as your personal legal "exposure" for a spring failure has to be really ugly....especially on a part which has known potential for ruining gearboxes and possibly "locking" them up at high speeds. And if someone gets injured over that gearbox locking up.....wow!

You could pretty easily make a fixture with an electric motor and a cam that would flex the spring through the entire flex cycle....even simple to make a counter for the number of cycles.

Tougher to induce hydrogen embrittlement and heat cycles at home.....but the "flex test", alone, would go a long way.

DKWalser

I've been following the debate about whether or not the part can be produced by welding spring steel. Colin believes that it can be done. (Let's hope he's right.) Others have consulted experts who say it cannot. I'm interested in the debate, not because I have anything to add, but because my son is a mechanical engineer who is finishing his masters degree in materials science. I asked him if the part could be produced using "friction stir welding" (the focus of his research). Here is his reply:

Yes, that is definitely something that could be friction stir welded. If it were done correctly, the weld zone could be stronger than the base metal. There will still be a heat affected zone outside the weld area, but it will be smaller than in a fusion weld, and less severe as well. Traditionally, to do the weld with Friction Stir Welding, if the part looks like an "L", with the long part of the L being butted up against the top of the small part of the L, they would clamp the piece and weld up through the small part of the L into the other plate. However, when I was at that conference in San Diego, one of the presentations was on a new procedure that allowed welding at a diagonal, into the the inside shoulder of the L, which had some benefits too.

Friction stir welding isn't something a small shop could do on its own. But, if the part run were large enough, it might be economical (not the same as cheap) to have the part welded by someone with the proper equipment.

Alan

It does seem that for this application - having all the return/bias springs outside the gearbox would be the easiest solution... surprised after the first failures & collateral damage that Porsche didn't re-engineer it that way...

Alan

GregBBRD

Porsche considers certain parts to have a life expectancy. This was certain one of those pieces. These cars had to be 10 years old, before I saw my first one break.....and I've never seen one break in a "low mileage" gearbox.

I've only ever seen two gearboxes where the pieces got run through the gears....and both of these transmissions had the spring broken for a long time, before this happened.....just not that common. The pieces are so heavy that they generally fall to the bottom and stay there.

People having panic attacks over this spring breaking are a lot like the people that think the stock belt tensioner is poor.....it's not. It's actually brilliant and does a great job.

One of the interesting concepts that you pick up, working on vehicles your entire life, is that some people expect metal bits to last forever. Never going to happen.

"If it is made out of metal, it is eventually going to fail." Absolutely true.

When these springs were available, I replaced this every time I rebuilt one of these transmissions. Now I need to be "more selective", until something to replace it is made and proven.

DKWalser

I don't want to steal the thread, but this is an important point to me right now. Are you saying that you would not replace the stock belt tensioner with a PorKensioner if you had the chance? Or, are you saying you wouldn't tear apart a perfectly fine engine just to install a PorKensioner? It's important to me, right now, because my car is torn apart awaiting a PorKensioner to be shipped when it become available. I could use the stock unit (and drive my car much sooner) if the PorKensioner if of no material benefit.

ROG100

David,
You may be better off starting a new thread and asking what oil you should use 8>).
Roger

DKWalser

Sorry. I didn't realize it was a delicate or controversial question.

Me? I use extra virgin olive oil on my salad. The extra-extra has too strong a taste and the regular olive oil is too bland for my palate. Like all things (including belt tensioners), YMMV!