ARP Rod Bolts data

[addictionMS]

I thought I would post some data from my crank assembly.

The crank was magna fluxed, center drilled and micro polished, std/std, using ARP rod bolts, rods were re-sized with the ARP bolts

So here is the data not sure how much value it has, but here it is anyway

I measured each bolt before and after the stretch, I used a stretch gauge to measure lentgh, goal was 0.0120-0.0115 inches, so all measurements are in inches.

I considered myself done when the gauge said I have between 0.0110 and 0.0120 stretch. So the stretch in the list is measured by calipers .

To respond to the concern in this ARP stretch failure thread there was no way I would have not noticed that the bolt stretched too easily, it took considerably more effort the get to the stretch than the alternative torque method of 45 ft/lbs

the data

before/after/stretch

2.795 / 2.803 / 0.008
2.792 / 2.802 / 0.010
2.793 / 2.803 / 0.010
2.794 / 2.802 / 0.008
2.794 / 2.805 / 0.011
2.786 / 2.798 / 0.012
2.790 / 2.801 / 0.011
2.800 / 2.813 / 0.013
2.786 / 2.797 / 0.011
2.791 / 2.801 / 0.010
2.795 / 2.806 / 0.011
2.792 / 2.803 / 0.011

[cstreit]

Good stuff Jim.

Can you estimate what torque you ended up with? (on average)

[addictionMS]

not really, I would guess it was 45 ft lbs plus about a 45-60 degree additional turn....not too surprising the last 0.001" took a lot of torque to get, I mean a lot, not much movement inthe nut, but a lot of force.

I guess my walk away is that the calipers are not sufficient for the job, too much slop, the "C" style might be better, but the money spent in the stretch gauge was worth it,

Jim

[Wayne 962]

Good work. This is really the proper way to assemble these rods. However, the torque method works very well too (haven't heard of any problems).

-Wayne

[911pcars]

Thanks for posting this.

Did you use ARP's thread lube?

Sherwood

[addictionMS]

yes I used the ARP thread lube.

JIm

[mpdevelopment]

The problem with torque is friction and the accuracy of the torque wrench. The friction is greatest the first time the bolt is torqued, after a few times the theads and surface of the nut-rod interface will become burnished and impregnated with the moly lube, reducing the friction and thus the torque. Also if the torque wrench is out of calibration, errors can result. I rememember an incident at a shop I once worked at. One of the mechanics was retorqueing the 5-8 cylinder head on a 308 Qv Ferrari when one of the studs pulled apart. This resulted in a mad scramble to have everbodies torque wrench tested and recalibrated. The wrenchs were all fine but the Ferrari stud material was not.

[addictionMS]

a little more on the how to, at least my how to....

I measured the bolts for initial length, and set the stretch gauge to 0 on one of the bolts, for me it was always the one on the side of the rod with the stamped numbers, and then measured the offset from 0 for the other bolt, usually something like +3 or -4 in 0.001".

Then I put the rod on the crank torqued it to 45 ftlbs on both bolts, measured the stretch to be sure I was getting some (typically 0.050-0.070") and to set a base line. then one at a time I tightened a bolt to get the desired 0.0120-0.0115" stretch, and repeated on the other bolt.

On a few I backed off the nut and redid the process to convince myself I did it right, allowed because they are ARP bolts.

I think once I got the hang of it it was pretty quick, a trick learned was the process, and to put a little pressure on the top of the gauge (gauge had a what was an extensiion of the tip which contected bolt sticking out the top of the gauge, so pressure ensured a solid seating of the tip on the bolt). I also wiggled the gauge a little to be sure I was seated giving me the least amount of stretch measured possible, I did this for the inital measurement and the final one.

Jim

[ChrisBennet]

I've done the bolt stretch method with Raceware and ARP rod bolts and both had insufficient stretch at their recommended torque setting.
-Chris

[addictionMS]

good to hear, puts my mind at ease a little more...

Jim

[911pcars]

"I've done the bolt stretch method with Raceware and ARP rod bolts and both had insufficient stretch at their recommended torque setting."

There are fewer variables measuring fastener "tightness" using the stretch method. The only variable would be the accuracy of the stretch gauge (micrometer, vernier caliper, etc.). There are more variables using a torque wrench: the accuracy of the wrench, torque loss from socket extension-to-socket-to-fastener, type of thread lube and tightening technique.

Sherwood

[vichang4]

I understand that the stretch method is very accurate if the bolt is not defective. But how about the bolt that stretched past the required amount with only 1/2 of the estimated torque. The bolt was probably defective, and if only the stretch method was used the faulty bolt would not have been found.

I suggest that the stretch method should be used, but the final torque or a minimum torque should also be measured. The torque necessary to stretch the bolt should be close to the recommended torque.

[addictionMS]

if you read my observations above, you will realize it is very unlikely to have reached the stretch without realizing there is a problem, it takes a LOT more torque that the spec torque to get to the stretch.

Really there is no way you will get a defective bolt without knowing it.

Jim

[snowman]

The ammount of clamping force produced by a bolt is NOT measured in by the ammount of torque reqired to tighten the bolt.

The fancy SPS bolts in my Carillo rods were tightened the first time using a stretch guage. I leave the stretch guage attached to the bolt while tightening it. I use a home made "crows foot" box end wrench and a torque wrench just for s and g's. I have torn down and reassembled the engine 4 times and noted the ammount of torque required to achieve the desired stretch, in my case 0.006". The first time required 22 ft lbs the last time only 14 ft lbs for the same stretch, AND the same clampiing force. In other words the SAME clamping force or stretch required 57% more torque initially than after 4 cycles. The ammount of torque also varied about 2 ft lb (14%) between fasteners. Had I used just torque I may very well have ruined the fastener and not known it until the bolt broke in a race. (Don't let the low numbers throw you, these are very special 1/4" dia bolts and hold far more than the stock rod bolts ever could have)

The other down side of using torque is in order to provide an adequate safty factor and insure no one ever over tightens a bolt, the torque is set so low as to give up a whole lot of the fasteners capability. Using stretch it is not necessary to give up anything. This is the likely explanation to what Chris has observerd in his post above. Whats that mean? Well those fancy ARP bolts MIGHT only hold 100,000lbs instead of 220,000 lbs and arn't as good as a stock 180000 lb Porsche stretch bolt. Why, because the bolts weren't tightened enough to hold as well as they could. Why? because they were tightened to a lesser torque, and not an optimum stretch.
(numbers are for comparason only NOT exact statements of fact)
PS I wouldn't trust any caliper to be accurate enough to make this measurement. Micrometers are good, stretch guage, because you can watch it while you tighten, better.

[mpdevelopment]

Those SPS MP35n Carr bolts are the finest rod bolts in the world. Was that 1/4" bolt in a motorcycle application? You can use 5/16 Carr bolts in a 911 custom rod rather than the 3/8" normally used to reduce weight at the big end.

[snowman]

The 1/4 inch bolts are used in Carillo rods for Porsche 912's. Problems hitting the camshaft with bigger ones.

[mpdevelopment]

Jack What RPM do you run the 912 engine at?

[snowman]

Try to shift before 8200 RPM. Power peaks at 7800.

[350HP930]

Since I had a set of 930 rods with ARP bolts that I just resized sitting around I thought I would measure them before disassembling them to add some extra data to this discussion.

The ARP documentation says that the stretch should be 0.012 +-0.0002.

While my rods were only on the second torquing almost all the bolts were stretched between 10 and 11 thousandths.

Its certainly enough bolt stretch to hold them together well but agrees with an above post that the torque method can be about 10% off of the stretch method, on the safe side of course.

[addictionMS]

350HP930,

great additional data, it would seem that there is some wisdom in the ARP instructions to torque the bolts three times if you are not measuring stretch...

Jim