ARP rod bolt question?

[irobertson]

I have begun reassembling my 3.2.

I installed the rods on the crank this evening.
I'm using ARP rod bolts and I measured the bolt stretch as recommended, and I also kept track of the torque while I was bringing the bolts to the correct stretched length.
In lieu of stretch measurement, ARP states 40 foot pounds.
I'm surprised at the variance in the torque required to get the correct stretch.
I used the supplied lube and I measured with a micrometer.
Curious if anyone else has experienced this?
Are these numbers out of whack?

Here are the results:
Torque Stretch
42 0.0100
38 0.0102
50 0.0104
48 0.0100
45 0.0099
52 0.0101
41 0.0103
40 0.0105
50 0.0099
45 0.0101
43 0.0104
44 0.0101

[winders]

Now you see why measuring stretch is superior to measuring torque....

Scott

[chris_seven]

+/- 30% would not be unexpected on commercial fasteners such as Grade 12.9 Cap Heads.

If you measure preload it a load cell and a carefully designed fixture the results are even more frightening.

You can reduce the scatter by 'burnishing' the individual nut/bolt prior to final torque tightening but with stretch as the ruling measurement there should be no need to carr out this operation.

The only question I have is why you have no measurements below 40 lbsft? This would be statistically unusual.

Have you skewed your results by always tightening to the minimum suggested torque ?

[irobertson]

Originally Posted by chris_seven

Have you skewed your results by always tightening to the minimum suggested torque ?

Quite possibly, as I started by taking them all to 20 ft-lbs, then went to 38 on all and did my first stretch measurement.
Typically they came in between 0.0060 and 0.0075 at 38.
With the exception of the one bolt that hit the mark at only 38 ft-lbs. I'm now wondering if that bolt is weak and should be replaced.
However, on the ones that are high, like 50 & 52, I backed off a quarter turn and re-torqued to the amount shown to get adequate stretch.

This was the third tightening of the bolts.
1st was when the rod ends were resized by the machine shop.
I measured them before I loosened them, then again after they were undone.
2nd was when I was measuring my bearings with a dial gauge.
I took them to the recommended stretch at that time. Highest torque required at that time was 47
Then measured them again loose to ensure they had no permanent deformation.
3rd was going on the crank last night.

[chris_seven]

I know it is difficult but I would ignore the torque results and put faith in the stretch numbers.

The measurement with the lowest uncertainty must be stretch and trying to square the circle of reducing the scatter of torque to preload to a 'straight line' with good correlation is, IMHO, the road to a nervous breakdown.

[irobertson]

Indeed.
I was thinking about this more and found it funny that I was concerned with one bolt being torqued 2 ft-lbs under the recommended torque, yet not worried about the three are 10 ft-lbs over recommended.

[ZOA NOM]

Definitely rely on the stretch numbers. I'm rebuilding my motor for the third time due to a failed circlip on #6 wrist pin, and I've never had a problem on the fat end of the rods. Other rebuilds were due to oil starvation causing #6 rod bearing to eat itself, and the first rebuild was just to install the ARP's and clean things up.