Were the 1974-1983 rod bolts always grade 12.9?

[ahh911]

Quick question, were the rod bolts from 74 to 83 always 12.9 grade?

Thanks,
Phil

[Walt Fricke]

Phil - you buy rod bolts from Porsche by part number, not bolt specification. One assumes OEM suppliers do the same. This isn't something you get from Maryland Metric or the like. Same PN 74-83.

[Flat6pac]

Que wI’m going to say NO, this is the reason.
12.9 is high strength and rod bolts are stretch.
On Carrera, the factory went to 12.9 rod bolts several years ago and issued an up date on the new 12.9. Previously, the torque spec was like 15 ft# and 90 degree swing. With the 12.9 which are marked, the torque update is 15#ft 90 degree swing and repeat the 90 degree swing.
Unless you’re reading the top of the bolt at 12.9 I’m sure it isn’t
Bruce

[ahh911]

I was wondering more if there was a material/composition change. My 1981 Verbus has a 12.9 stamp, my new ones have a 12.9 stamp. I was wondering at the time when trying to justify why Bentley went to 20Nm+90 for the 911 SC. I figured it was worth a question.

On the same thread, Bruce/Walt, do you guys know what the 00 stood for on the 3.2 Carrera earlier rod bolt stamping (like in the tech spec that's been printed here several times)? I figured that 12.9 stood for bolt grade, but I couldn't find a 00 on the typical bolt spec pages from manufactures (maybe didn't look very hard). It sort of put the idea that 12.9 stood for grade into question for me.

Phil

[Walt Fricke]

My first rebuild I just ordered the bolts and installed per spec (which, then, I don't recall being the torque then angle). Ever after I have used Raceware or ARP bolts, which are in the range of 220,000 psi, so metric/DIN grade hasn't been of much interest and I use the stretch gauge. It appears 180,000 psi is about what M12.9 bolts are.

Did Porsche ever use a lower grade?

The 3.2s went from a 10mm to a 9mm bolt, didn't they? To achieve the same clamping force with a smaller diameter bolt, you would have to stretch it more, would you not?

The modulus of elasticity is the same, 8.8 or 10.9 or 12.9. Just that the tensile strength is ascending in this series, so you can stretch it farther/get more clamping force - before you leave the elastic, or worse the plastic, range.

From '72 through '81 the spec in the little white books was the same - 50-55 NM. Vol III of the workshop manual just lists 50 NM (lubricated). Anderson says 37-40.5 lbs/ft. Bentley just copies what they think the factory said. Wayne just says 20Nm (~14 lbs/ft)+90* for the 10mm rod bolts.

For the 3.2s, Bentley says 20Nm+90* - same as for the 3.0s. It says the "checking" torque (on your old bolts for checking big end roundness) is 50-55 Nm.

I have the tech bulletin book through '84, but no mention of different torque or methods there.

I thought I had saved some stock rod bolts to see what was on them, but I didn't, or finally realized I had better uses for that bin than a fastener I would never use.

But I ask again - why do you care. You have the spec for torque or torque plus angle, you have the bolts. You aren't building a race motor, or you wouldn't use stock bolts. If they have 12.9 stamped on them, most likely they are just that. So?

[Walt Fricke]

Phil - I just read your posting on the crappy OEM (or Porsche?) new rod bolt fat part diameters. Awful. Eli Whitney did better. Has you questioning everything, I guess.

[HaroldMHedge]

There is a TSB on changing the connecting rod bolts material and torqueing for the Carrera (MY 84), Turbo (MY 78) and C2/C4.

This indicates that rod bolts with a head marking of 00 (old version) are torqued in steps with slightly oiled threads and nut contact surface.
1. Torque nuts to 20 Nm (14 ft. lbs.)
2. Torque nuts an additional 90 +/- 2 degrees

For the new rod bolts head markings 12.9 (P/N 964 103 176 00) torque with slightly oiled threads and nut contact surface.
1. Torque nuts to 15 Nm (10.5 fl. lbs.)
2. Torque nuts an additional 90 +/- 2 degrees.
3. Torque nuts and additional 90 +/- 2 degrees.

Replace connecting rod nuts and bolts after removal with new ones.

It is interesting that the nuts (P/N 930 103 174 00) were not changed with the connecting rod bolt change.

[ahh911]

Hi Walt, Bentley actually mentions torque to 55Nm with new bolts once and angle method once. See images. Also, the Bentley has a picture of the 3.2Carrera rod bolt when you look in the front of the book where they show the different aspects of the engine. Not very encouraging. This is all for the 911SC 2000 printing.


Wayne's online update to his book says the following, though the second printing still shows a 10mmx1.25 as angle torqued when in fact these are the earlier bolts and 9x1.25mm are the later bolts and would be the correct size.

How to Rebuild and Modify your Porsche 911 Engine: Corrections & Updates

"Page 200, the rod bolt torque value lists the Porsche spec for the later-style rod bolts (911 1984-89 and Turbo 1978-89). There is a note at the bottom of the page regarding the change in the spec. Some people prefer to use the earlier spec of 50 Nm (36.2 ft-lbs), documented at the bottom of the page."


Wayne states in the following post:
Rod bolt creep

"The 1982-83 Porsche spec book lists the angel torque method as used on the late 911SC rod bolts, which are the same as the early bolts from 1972). The even earlier bolts are similar to these bolts as well. There may have been a Porsche tech bulletin superceding this method for the early cars - I can't recall off the top of my head."


The spec book he refers to mentions the 10x1.25 rod bolt as 55Nm and the 3.3 turbo incorrectly as a 10x1.25mm whereas it should have read 9x1.25mm which is angle torqued. I can produce an image of this as well, but it was printed in 1987.

All in all, not very encouraging. When I was studying this, the only two sources of the angle torque for the earlier 10mmx125 are these two sources which for this detail don't seem to come from Porsche. But oddly, I'm still planning on going the 20Nm+90 degrees. When I test tried on the crank using that method, it felt an awful like the the originals I removed, not very scientific, but none of this seems to be. I wonder how many have angle torqued the 10mmx1.25 pre 1984 bolts and are still ok with it? I'd love to hear how many tens of thousands of miles they've put on with it, hopefully over 100k.

Phil

[Walt Fricke]

Well, 10mm rod bolts torqued (lubricated) to 50-55 NM don't seem to have been a problem. I don't believe Porsche ever published a stretch dimension spec - you get that from Raceware or ARP if you use their bolts.

A guy who cared could take some rods and bolts and compare various torques and torque + angle methods to what the stretch was. And what the recovered bolt length was afterward, which would tell you if you had exceeded the elastic limit. If you knew the required minimum clamping force needed, you could calculate the stretch needed to achieve that. You could also probably calculate where the elastic limit of a 10mm bolt was.

I suspect Porsche was just being cautious in adjuring us not to reuse these bolts, but at least in past times it wasn't worth reusing them (what you purchased fit). Going to the thinner bolt on the 3.3Ts and 3.2s must have caused some rethinking. But what would be important is the clamping force, and its relationship to yield. And you have 10mm bolts.

[ahh911]

Walt I did do testing as best I could because I had plenty of new oem bolts to do it with.
Here's what I found, these final degree numbers are for the nut final positions. The bolts always started with the tang in the same locked position.

20Nm to 55Nm (oil dumped on or lightly lubed) the nut final position was between ~20 and 30 degrees shy of the 20Nm+90 degree lightly lubed approach. (I would call this a large degree of measurement error or variation using the torque only method and it depends on the the finish of the nut and contact area.)

When going 20Nm+90 degrees and marking the position then seeing what kind of torque was needed with very lightly lubed bolt (dampened with oil cloth including nut face) to reach the same nut position: 75Nm.

Taking a torque wrench with an oiled bolt and pulling the torque wrench without sensitivity will yield the bolt at 75Nm and it will permanently elongate.

Going 20Nm-90 degrees with lightly lubed oil and nut face over and over again saw the bolt's I tested (3 in all?) return back to their original length, as best as I could measure. 0.01mm type of accuracy.

Stretch that I measured:
20Nm-90 degree stretch of bolt oiled sparingly: .0096" (.245mm)
55Nm oil dumped on: .0083' (.21mm)

The original verbus 12.9 bolts returned to their dimensions when pushed to 55Nm and seemed ok at 20Nm+90, though I should have re-measured them. The only bolt that didn't return was the one that yielded under aggressive torquing to 75Nm with some oil and it was easy to feel. I then took that same bolt and went 15Nm+90+90 and it did not snap, but elongated permanently by .4mm (.016"). This bolt loosened and torqued to 55Nm would see it permanently elongate even further, so I guess once it goes plastic it keeps elongating every time you re-torque it up until it snaps? Is this expected at that point? (I'm more of a electrical kind of guy)

So, I don't know exactly. I still plan on going 20Nm+90 at this moment, but it's a tough one that shouldn't have been. Is there enough info to go with the +90? Like I said, the nuts that came off the car were on pretty tight to very tight and felt alot like the angle torque method in terms of difficulty removing the nuts, but I guess the oil on the threads had long passed their lubrication point, so who really knows.

Phil

[Walt Fricke]

Someone - ARP? - says that if you use torque, you should burnish the fasteners by tightening and loosening them several times. I guess the idea is that after you do that you have mostly taken the friction variables out of the process, and the specified torque will more closely achieve the needed elongation.

The angle method, since it is based on the pitch of the threads, looks like it could avoid having friction produce a result of less than needed stretch (i.e., clamping force), and without the extra steps of burnishing.

I was helping a guy once do his crank with my stretch gauge and his aftermarket bolts. Took way more torque than I thought would be needed to get the stretch. We called the bolt manufacturer. Did you lubricate, he asked? No, which is why we were struggling. But we had gotten the stretch.

[ahh911]

Yeah, friction for sure using the torque method. The starting point for the 20Nm is even around 10 degrees off depending on lube or no lube. I guess it is 55Nm and the Dempsey book is in error? I assume that's what his update is trying to say? And Bentley? They image the carrera 3.2 bolts in the front of the book, so them too? Seems weird. Have you ever heard of a nut coming loose with the 55Nm oiled method?
Phil

[Walt Fricke]

The only rod problem I had was on a 2.3 (66x85) race motor. At the time there were no super bolts available (or I didn't know of them), and I routinely ran it up to 8,000-8,200 RPM. A bolt broke. Not good. The nut was still attached to its end of the bolt. The bolt showed the characteristic necked down shape at the point of fracture. I chalked this up to exceeding the strength of the stock bolts, but perhaps that nut backed off a hair and it was all down hill from there. It doesn't have to back off much to cause trouble, and it is progressive.

For a street motor, I'd expect that tightening to 50Nm several times to burnish, and then reoiling, torquing to 50, leaving it and going on to do the same to the next, would do the trick. Or go to 55.

One of the guys who tears down lots of motors might have a qualitative idea on how often he finds an obviously not fully tight nut on a rod. I tend to use an impact wrench for the removal process, so hard to tell with that.

Here is the TSB on the 3.2 rod bolt change. I don't see that this applies to the 3.0s, which use the larger 10mm bolt. My Holiday Inn Express degree in metalurgy is inadequate make a guess as to what change in materials would lead to a different method of torquing/turning X degrees. If the steel were of a higher than normal 12.9 grade, it would have a higher yield point so could be stretched more safely.

[ahh911]

Thanks Walt, it is tricky, I was going to the 20Nm+90 method, but the final torque is very close to the yield point of the bolt. I've read about your bolt failure as well on other posts, I assume you mean that you torqued using the well oiled 50Nm method and the nut still held on.

I too can't see how the 12.9 marked 3.2 carrera bolt or the 00 (what does 00 even mean?) marked bolts can possibly relate to the 3.0 and earlier cars. The 12.9 on the 3.2 is clearly a yield fastener considering the two 90 degree turns and the 00 marked bolt is who know what. The early 3.0 bolts were never re-labeled (always were 12.9 marked) and there was never as far as I've seen a spec change in any Porsche spec book so You'd think that since the bolt is the same material as always, then the original 50-55Nm is the correct value. The rods are sitting there with assembly lube in zip loc bags ready to go. There are so many conflicting but credible points of view on this.

Phil

[HaroldMHedge]

I would not trust the 911SC Bentley Service Manual for the rod bolt torque Stage 1 20 Nm (14 ft-lb), Stage 2 90 +/- 2 degrees. They copied the Carrera manual and did a poor job of updating the information. You are correct to question that data Phil.

My copy of the 82-83 Technical Specifications lists two different torques for the conrod nuts (page 57).

911 SC M10 x 1.25 50-55 Nm

911 Turbo M10 x 1.25
1st step (initial torque) 20 Nm
2nd step (final torque ) 90 +/- 2 degrees torque angle

(Sorry I can't seem to post a picture of this PP software thinks my file is too large.)

[HaroldMHedge]

Here we go. Needed to change the file to a JPG file.

[ahh911]

Thanks for that Harold. Here is the same type document published in 1987. Also a screen shot of the rod part numbers for the 83 911SC and 911 Turbo from Porsche PET. The Turbo part number is different and is the same as the Carrera 3.2, i.e. 9mmx1.25 and therefore would not follow the same torquing as the SC.


KTL goes into detail as well on this post:
911sc oem 10mm rod bolts and nuts question

Here he tracks down the last known torque spec from Porsche for the 10mmSC bolts that I've seen published in 1989. If someone has a later screen shot showing a different method, everyone in the future going through this same process would be very grateful. Also, how much does it really matter if we use either method? Surely there must be a few hundred who have angle torqued these older 10mm bolts and several thousand who have used the 50Nm torque value in north america. How much does it really matter if we approach or slightly enter the yield point of the bolt? Chris seven wrote pages on this stuff, I need to find it again.

Conrod Bolt "re-use" question
Chris Seven wrote some interesting stuff.


Phil

[Walt Fricke]

Phil
Now that you have read what Chris Seven (RIP) had to say, you will not learn anything which is more useful about this subject. You will have noted that the experienced shop guys agree with his engineering analysis, even if they differ on business methods. Both methods work. Pick one. Put the motor together. It isn't like Porsche has divine inspiration, or Papal infallibility, or there is some obscure Porsche Torah wisdom out there to guide you if only you can uncover it from some archive.

It is certainly a shame that you had to resize new bolts, but that is a different issue than what values to use tightening them. If you use the angle method - or the stretch gauge method, you can follow up with what someone (KTL?) does: set the torque wrench to 50Nm, and check each nut after you are done with the angles. If somehow there was a bad bolt in the batch, and you couldn't feel it when you got out there close to 90*, this should catch it.

You could take a page out of the CV bolt literature: stretched bolts are said to relax a bit shortly after stretching (a bolt engineer said that, so there is credibility there). So wait an hour to check torque if you are wont to take every potentially useful precaution.

Bolts which are manufactured to a standard (like DIN) are batch tested, and if the test bolts meet spec, or a specified percentage of them do, the batch is called good. Does this prevent there ever being one sub-standard bolt in a tested batch? No, I suppose, but surely experience guides the manufacturers, and they can identify how defects can occur, like dies or whatnot getting worn. This system seems to work well enough, as opposed to X-raying every nut and bolt you use, and doing your own non-destructive proof test on every one of them just to be sure.

[ed mayo]

Frankly I think the bigger issue is counterfeit bolts, even from the dealer, we only trust U.S. made aftermarket bolts. That may be the reason for your bolt diameter problem. This didn't seem to be a problem 30-40 years ago.

[Harpo]

I would also trust bolts from Germany