915 Transmission Rebuild - Thoughts on Micro or Superfinishing Gears & Shafts

[Gonbad]

I am in the process of getting the parts I need ordered and double checking the condition of parts I will be re-using.

The used reverse gear set I got from Germany and the 5th/Reverse slider I found are in good condition but the finish on them isn't smooth. This is also true for the parts I am keeping, particularly the gear sets and bushings. I don't see that it would hurt the parts as micro-polishing doesn't change the dimensions or tolerances of the parts.

So has anyone had the gear sets, bushings, main and pinion shafts, ring gear micro-polished or super finished? Thoughts on the results? Shops used and experience with them?

[Matt Monson]

It’s a common misconception that it doesn’t change the spec or remove any material when it’s micropolished. I’ve seen many good used parts ruined by trying to improve them.

One of the worst was the time a guy took a cup car crownwheel and pinion and REM finished it, but failed to tape it off below the gear head. Once the unit was finished the bearing fell right on. It’s supposed to be an interference fit.

[30westrob]

“The enemy of good is perfect”. Rob

[H-viken]

The guy rebuilding my 915 shotpeened the new gears, saying it would strengthen them

Remains to be seen

[PeteKz]

"Ball-peened"? Did he mean shot-peened? That would make more sense.

[Matt Monson]

I’m Picturing gears in a vice being hammered on…

[stownsen914]

I'm sure someone, somewhere has done that, sadly

[H-viken]

Sorry shot-peened

Now corrected

[lateapex911]

So, we do such things in racing. In a car like a Miata, it's art of a suite of things that can help you stay in the draft and eeek out at the end of the straight.

It's expensive, of course. And you need the right firm to do it, as mentioned above.

Some have said that the parts wear into that level of finish through use over time, so the benefit is not as great (and it's already not big), which makes it a questionable call.
And the right lubricant, and amount are also part of the package.

If you think about it reversed, going back to my comment about the lesser benefit on used parts, some people think you're taking life out of the parts.

For street use, I'd say if you have a friend who does it cheap as a gift, or you feel the new parts are deficient as they currently sit, then it's a consideration.

Will it be noticeable on a street car? That's a bit like wines...some people can't tell a Shiraz from a Malbec, where others will tell you that the barrel was white oak, not red oak.

Most people won't be able to feel it in the seat of their pants, but there are plenty who want to check it off the box of possible things to do.

Me? Haven't won the lottery, and it's waaaayy low on my to do list for my street driven and occasionally tracked car.

My SCCA EP MX-5 racer?....gulp, I'll be looking at the parts where I can, but even then it's down on the list, there are bigger gains to be made for less $...it's not low hanging fruit.

[Gonbad]

I like your response and agree the performance gain is probably not felt. My question is more in researching increased part durability through reduced friction and wear on expensive if not already no longer available parts. The expense doesn’t seem unreasonable as I have seen shops that will do a motorcycle transmission for around $500. Not saying I’ll do that but it’s in the ballpark when I am already looking at $7000 in parts to rebuild my transmission like I want it.

I have a couple of people in mind if I do this. One is Rodney at RPM transmissions, I haven’t spoken to him about Porsche parts in particular. He is more into the domestic high performance world but he may know of other options if he isn’t able to.

I always appreciate others experiences and thoughts when it comes these cars. They are unique beasts and take a different level of care and thoughts when changing things.

Originally Posted by lateapex911

So, we do such things in racing. In a car like a Miata, it's art of a suite of things that can help you stay in the draft and eeek out at the end of the straight.

It's expensive, of course. And you need the right firm to do it, as mentioned above.

Some have said that the parts wear into that level of finish through use over time, so the benefit is not as great (and it's already not big), which makes it a questionable call.
And the right lubricant, and amount are also part of the package.

If you think about it reversed, going back to my comment about the lesser benefit on used parts, some people think you're taking life out of the parts.

For street use, I'd say if you have a friend who does it cheap as a gift, or you feel the new parts are deficient as they currently sit, then it's a consideration.

Will it be noticeable on a street car? That's a bit like wines...some people can't tell a Shiraz from a Malbec, where others will tell you that the barrel was white oak, not red oak.

Most people won't be able to feel it in the seat of their pants, but there are plenty who want to check it off the box of possible things to do.

Me? Haven't won the lottery, and it's waaaayy low on my to do list for my street driven and occasionally tracked car.

My SCCA EP MX-5 racer?....gulp, I'll be looking at the parts where I can, but even then it's down on the list, there are bigger gains to be made for less $...it's not low hanging fruit.

[MBEngineering]

Hi
If you are having the gear set's treated, make a note of the matching No's and fit Cable Ties (Red/Blue/Green/Black) as sets,

also if the dog teeth are to be changed, leave the old ones fitted for the treatment and change after,

if you can cover all wear face's for bearings and thrust faces, as Matt pointed out about the pinion shaft bearing.

regards mike

[RogueEngineer]

Gear designer here, I feel compelled to chime in!

I design gears for industrial and aerospace applications. We occasionally superfinish gears and nearly always shot peen them

Superfinishing can improve the surface finish on the gear flanks - that is the part of the tooth in contact during meshing. The surface roughness of the flank is important as it (in part) dictates the required oil viscosity needed to keep an oil film separating the tooth flanks. Higher roughness requires thicker oil. Thicker oil sucks power through windage and oil churning. Super finishing can allow the use of thinner oils safely, or running at higher operating temperatures (i.e. as oil viscosity drops). Both of these things can be important in high performance applications. We would take ground gears from a 12-16 u-in surface roughness to 2-4 u-in surface roughness. We don't do it too much anymore as it definitely removes -some- material from the gear flanks - and sometimes it's not perfectly homogenous. This can screw with my carefully designed microgeometry, so I'm a little wary of it. It still has its uses, though.

Shot peening can absolutely improve the bending and contact fatigue strength of gears. This is well studied and is applied often in critical applications. However, it is a highly controlled process with a lot of variables - shot intensity, shot size, shot hardness, coverage...and so on. These variables are generally driven by the root radii of the gears and by the desired induced compressive stress zone. For an automotive application I'm guessing the shop performing the peening would have their own special formula for this - but honestly not knowing the details would make me nervous.

If the intensity is too high or the shot media too hard it can absolutely destroy any gear micro geometry present from the factory. I've seen it happen.

All that being said, when applied correctly you can get a "better" set of gears with these processes, presuming they are applied appropriately.

[PeteKz]

RogueEngineer, thanks for jumping in with your expertise. Have you designed gears for your applications that were lapped together for precise fit? I understand that many years ago in industrial equipment, cutting the gears slightly large, then running them together with lapping compound to final size was supposed to produce gears that fit very precisely, and therefore had higher load capacity. I was aware of this process in steel mills to finish very large gear sets.

[Matt Monson]

Originally Posted by RogueEngineer

Gear designer here, I feel compelled to chime in!

I design gears for industrial and aerospace applications. We occasionally superfinish gears and nearly always shot peen them

Superfinishing can improve the surface finish on the gear flanks - that is the part of the tooth in contact during meshing. The surface roughness of the flank is important as it (in part) dictates the required oil viscosity needed to keep an oil film separating the tooth flanks. Higher roughness requires thicker oil. Thicker oil sucks power through windage and oil churning. Super finishing can allow the use of thinner oils safely, or running at higher operating temperatures (i.e. as oil viscosity drops). Both of these things can be important in high performance applications. We would take ground gears from a 12-16 u-in surface roughness to 2-4 u-in surface roughness. We don't do it too much anymore as it definitely removes -some- material from the gear flanks - and sometimes it's not perfectly homogenous. This can screw with my carefully designed microgeometry, so I'm a little wary of it. It still has its uses, though.

Shot peening can absolutely improve the bending and contact fatigue strength of gears. This is well studied and is applied often in critical applications. However, it is a highly controlled process with a lot of variables - shot intensity, shot size, shot hardness, coverage...and so on. These variables are generally driven by the root radii of the gears and by the desired induced compressive stress zone. For an automotive application I'm guessing the shop performing the peening would have their own special formula for this - but honestly not knowing the details would make me nervous.

If the intensity is too high or the shot media too hard it can absolutely destroy any gear micro geometry present from the factory. I've seen it happen.

All that being said, when applied correctly you can get a "better" set of gears with these processes, presuming they are applied appropriately.

Gear manufacturer here, Porsche gears specifically ,and with these comments I almost completely agree. We use shotpeening as part of our manufacturing process, prior to heat treat. We do not micro finish our transmission gear sets. We do rem finish the bevel and spider gears in our LSDs as they are a much more coarse tooth profile and cut at a much higher feed rate than our racing gearsets.

That said, my comments above were specific to this topic, which is about used gears. Techniques used in new manufacturing aren’t generally appropriate for used gears, IMO

Proprietary sequential 996/997 Cup Car gearbox shown.





Regards,

Matt Monson
Guard Transmission LLC

[Matt Monson]

More pictures of things from a previous thread on a similar topic.

What you see here is an NOS QP 4th gear next to one of our own QP 4th gears. I used an NOS gear so that there would be zero wear and you can see the dramatic difference in tooth profile and design. Things have come a long way in the last half century.


A good view of our gear tooth profiles with tip relief. As rogue engineer has mentioned, this is where the real power is made(saved).

Spider gear for LSD

[RogueEngineer]

Originally Posted by PeteKz

RogueEngineer, thanks for jumping in with your expertise. Have you designed gears for your applications that were lapped together for precise fit? I understand that many years ago in industrial equipment, cutting the gears slightly large, then running them together with lapping compound to final size was supposed to produce gears that fit very precisely, and therefore had higher load capacity. I was aware of this process in steel mills to finish very large gear sets.

To answer your question: I have not. I'd say 95% of the gears I've designed have been finish ground as the last machining step.

I am at least tangentially aware of lapping gears together, though. I've heard of it being used for very large gear sets where precision final machining may be impossible. Modern gears typically have tip/root relief (small adjustments in the involute shape of the tooth flank) or lead correction/crowning (adjustments along the width of the gear face) in order to facilitate good contact under load as teeth/gear bodies/housings/shafts shift and flex. These tiny adjustments would be destroyed when lapping.

[Gonbad]

I wasn't specifically going to have the parts REM finished. Now with Matt from Guard putting in his experience, if I did I would probably wouldn't have the shafts REM treated.

Another thought:

What about having the parts vibro/tumbled in a polishing media?

Use a polishing wheel and different compounds to polish the shafts and other suitable parts but obviously not the gear sets or sliders.

I am not trying to over complicate things and I am not looking for every bit of power. This just seems a good way to reduce friction thus reducing wear and heat. It doesn't appear to be that expensive either. I also already have the transmission pulled apart. I live in Texas and spirited drives in the hill country or at COTA can create very high transmission temps. When I pulled the transmission on my C5 Corvette after a long weekend in the hill country and at COTA you could tell from the metal coloring that it had gotten hot.

[Matt Monson]

Money is better spent on transmission cooling. A Wevo or CMS spraybar is the RSR replica approach, but $$$

Generally a Tilton pump, heat exchanger and an extra liter or so of fluid is fine for the kind of use you intend.

[shoooo32]

I agree with Matt 100%. Oil temps are the the Achilles heel of the 915. Has anyone tried this cooler setup sold by Cogs Cogs?

https://ebay.us/m/d8X86Z

It looks like a knockoff of the FVD kit at half the price:

https://www.fvd.net/us-en/915300911550/transmission-oil-cooler-915-930-transmissions.html

[Matt Monson]

I wouldn’t buy anything from German Joe, just on principle. He’s the motornesiter of the gearbox world.

Patrick has one for $815.