Intermediate shaft out of round
 

Hi Guys,
I’m rebuilding my 86 3.2 engine and just finished checking the intermediate shaft for roundness. Assuming I am doing this correctly it looks like it is 0.07mm out of round. I have been buying and hoarding engine parts for years so I already have all new replacement gears for the shaft. Does anybody have a good intermediate shaft they could sell me? I know I can’t get a new shaft only and I don’t trust EBay for something like this Any suggestions are appreciated.http://forums.pelicanparts.com/uploa...1687283629.jpg
http://forums.pelicanparts.com/uploa...1687283629.jpg
http://forums.pelicanparts.com/uploa...1687283629.jpg

That’s the way I check the shafts. They cannot be straightened. My outside run out was .03-.04mm. In the dozens and dozens I’ve checked I’ve found .00 maybe twice. Get as close as you can with a used shaft. When the gear is installed confirm by backlash numbers. As long as u have any backlash you’re good as engine heats clearance grows. I’m not around my stash, will look later.

Your indicator is riding around on the outside edge of a flange, I don't think that's a primary datum, doesn't the gear center up on the minor diameter of the shaft. The flange is there to provide the space for the bolts, the bolts provide clamping but not locating.
That may be a "good" shaft but your measuring in the wrong spot.
David
EngiNerding

I’ll check the shaft again where the gear rides. The place I checked is where Kurt in the Klassic videos checks his,

Also, the condition of the aluminum gear concerns me, seems like a lot of wear for an engine with 78k mileshttp://forums.pelicanparts.com/uploa...1687312093.jpg

WYDBODY , you are measuring correctly , and I've essentially had the same results as gled49 .
So if you want .02 or .03 runout you may be measuring numerous shafts , the reality of it .

https://youtu.be/azn1oymRZjA

Go to 4:00 minute mark; procedure noted by wydbody1986

I just checked the workshop manual, but nothing found about. When I look at it, it's not clear the measuring point you choose is correct. It could also be the smaller one (inner of the gear) as I doubt both seats are same tolerance. Only one could be the one who centers the gear, question is which one. So I wonder what result you get when checking at that inner radius.

Am I overthinking this? How much runout is too much. Besides the video from Klassic I can’t really find any other info. The way I read the indicator I have .07mm runout, .which is .02 positive plus .05 negative as I rotate it through 360 degrees.

The correct way to read any run out on the shaft is between centers. The way shown in the video is inaccurate and incorrect. The position of the dial indicator was on a diameter that was not a ground surface. That diameter could have a high machining tolerance as it "holds" nothing.

Re: wear on the aluminum gear, how are you assessing? Look for pitting on the teeth, it's pretty common. If it shows any signs of wear, just replace it - not worth tearing down again for that. I'm pretty sure the metal I was finding in my filter (the reason my race engine is currently apart) was from the aluminum gear - I never found anything else showing significant wear.

The diameter being measured is the register for the intermediate shaft gear so I believe it is important.
I replaced a gear on an engine recently and it is making an awful noise. I am wondering now if the gear might not be accurately machined. Anyone have experience with this?

Absolutely!

Where do you found that? Not in Porsche manuals I bet. This is what you find:

http://forums.pelicanparts.com/uploa...1687357681.jpg

http://forums.pelicanparts.com/uploa...1687357681.jpg

Neil is correct here. I don't ever recall seeing the I-shaft measured like that in any of the workshop manuals (although, to be fair, they also consider the I-shaft assembly "one piece" and replaced if anything is out of tolerance). If you're off, maybe then I'd move the gear on the I-shaft to find a happy place.

What really counts here is that, with the crank and I-shaft installed, the backlash between the gears remains constant throughout rotation. If there's no binding and the backlash is around 0.02mm in several spots, you're good. If it varies from tight to loose or binding in spots, then I'd address the issue. Mark and move until backlash is acceptable.

FWIW, shafts are "straightened" with a hammer and a C-shaped drift/chisel that fits into the radius. A few taps and recheck until within tolerance. Talk to any crank grinder or cam grinder and you'll hear the same thing. With a hammer, you hit the low side (opposite of what you'd think). If you use a press, you press on the high side. Trial and error and keep repeating until the shaft is back into tolerance, and often better than the factory produced. But first things first; check the runout properly, with V-blocks and the dial indicator on the shaft. That flange hanging off in midair will skew your readings and make the shaft seem worse than it actually is.

I assumed that since the aluminum gear is press fit to that surface then bolted to it that it was a good measuring point. I’ll try another surface. If that is inconclusive I’ll guess I’ll assemble the new gears to the shaft and the crank and measure backlash.

RE: the gear I was looking at the wear on the face around the teeth. The teeth themselves looked fine, no pitting or gouging, same with the crank gear.

To add to what I said, the original drawing of the shaft would show a diameter tolerance for size and an out of round tolerance. The OD of the flange could show +/- 0.005" and an out of round tolerance of +/-0.002". I have no idea. But to measure where the video shows without knowing the drawing tolerances is very inaccurate.

You would put the shaft between centers as the bearing surfaces would have been ground with and measure the shaft run out each end and end to center. It all starts there. The shaft could be bent but the OD tolerance of that flange could mask the issue. Go with what you know, not what you assume.

As shown in the video, the bearing shells will not measure the same thickness as the tin coating will have some small differences.

I have no way to measure the shaft like that. “Go with what you know” I know what the backlash is supposed to be so I’ll measure that at several locations tomorrow after I assemble it and the crank gears and I how the shaft feels when I spin it in the bearing shells which is smooth.

At the end of the day, we care that the OD of the aluminum gear is not out of round when mounted on the shaft. I have no idea how accurately machined those gears are. If we measure only the shaft, we assume the gear is perfectly concentric. Maybe it isn't. In line with what dannobee suggested above, would it make sense to mount the gear and measure runout ... and clock the gear on the shaft to minimize the out of roundness of the overall assembly?

With regards to all the engineers, solutions need to be found. The shaft runs in bearings not on centers in the engine. After buying a new intermediate shaft as per Porsche, and have super bad backlash numbers I tried rotating the gear as mentioned, had the same backlash numbers. When I checked the register where the gear indexes, I whitenesses runout on that diameter. I started digging around in my old shafts, I found one with .01mm runout, checking it like the OP, installed the gear and backlash numbers were near perfect. I never bought another complete shaft. I defy anyone to straighten one of those. Porsche also said u could only buy 915 gears if shift hubs were shot. I’ve certainly replace hundreds of shift hubs that became available in the aftermarket, with great success. Everyone needs to come up with solutions to problems like this to your own satisfaction, good luck.