Stretch Gauge

jamesjedi · 11-10-2011, 02:15 PM

What are the acceptable brands for a stretch gauge? It will be used for connecting rod bolts. I imagine that many may be inaccurate.

Thanks for any help.

Walt Fricke · 11-10-2011, 04:11 PM

I think quite the opposite. The stretch gauge is really just a fixture to hold a dial indicator. As long as your dial indicator is reasonably accurate, you will be fine. JEGS, Summit Racing, Tavia? Some like a sharp point on the indicator, the better to nestle into the hole on the end of the bolt.

I like to measure the bolts and record that, so I use a micrometer setting piece (carefully cut steel cylinder, usually comes with micrometer sets) to set the indicator and measure. But for just stretch you can zero on the untensioned bolt, and then quit torquing when you get the specified stretch. I can't imagine even the cheapest dial indicator being off enough to matter over a few thousandths of an inch range.

jamesjedi · 11-10-2011, 04:20 PM

Thanks. Good to know, as the results could be catastrophic if they were off by a large amount.

304065 · 11-10-2011, 07:38 PM

Quote:

Originally Posted by Walt Fricke

I think quite the opposite. The stretch gauge is really just a fixture to hold a dial indicator. As long as your dial indicator is reasonably accurate, you will be fine. JEGS, Summit Racing, Tavia? Some like a sharp point on the indicator, the better to nestle into the hole on the end of the bolt.

I like to measure the bolts and record that, so I use a micrometer setting piece (carefully cut steel cylinder, usually comes with micrometer sets) to set the indicator and measure. But for just stretch you can zero on the untensioned bolt, and then quit torquing when you get the specified stretch. I can't imagine even the cheapest dial indicator being off enough to matter over a few thousandths of an inch range.

A micrometer "standard." Comes with a couple of wrenches that look like an '80s fan belt wrench- flat steel with a hook and pin. You put the standard in the jaw, tighten down the anvil with the ratchet and look at the measurement. It's important to use the same gaging pressure when you calibrate as when you measure, so I like to tighten down with three clicks and that's it. Anyway, you tighten down on the standard, then use the wrenches to adjust the micrometer scale so it reads zero-- and then your mic reads precisely what the standard's measurement is. Takes about 60 seconds and should be done anytime the mic is dropped, changes temperature or is out of your possession.

Walt Fricke · 11-10-2011, 10:10 PM

John is the resident metrologist. I use the "standard" which came with my micrometer set to set my dial indicater in the stretch gauge so it reads actual length, so I can record that. This is not as accurate as setting a micrometer, because you are measuring using pointed ends, not flat ones, so it is easy to measure a little long if you aren't careful. If I did this a lot I suppose I'd just leave a dial indicator, once set to measure actual rod bolt lengths, in the stretch gauge in a padded box, so I wouldn't have to go through all this.

And maybe buy a second stretch gauge for the stretching.

Or I could use a micrometer to do the unstretched measuring directly. But rod bolts have nice little dimples in their ends which make using the stretch gauges so quick because you aren't worried about what you are measuring from and to.

I'm now thinking that once I have done all the initial length measuring, I should just zero the indicator for each bolt come stretch time, and record the stretch. Should speed things up not having to do addition in my head, no constant referring to what I have written down as I am tightening the bolts. I can do the addition later (in fact, have my spread sheet do it) in preparation for that far away time when I change the rod bearings and need to know if I can reuse each bolt.

But if I do this once every two years I am doing a lot, so inefficient (as long as accurate) methods aren't a big deal.

304065 · 04:05 AM

The ARP stretch gage is like $150. I bought an inexpensive one which I would never do again.

The trouble with these "stretch gages" is that they hold themselves on the rod bolt using the gaging pressure from the dial indicator. That's controlled by a wheezy little spring inside the indicator that's supposed to press the tip against the workpiece without distorting it (think Rockwell hardness tester).

The problem is that the gaging pressure's not enough to really hold it on there while you use your third and fourth hands to wrestle the torque wrench up to 50 ft/lb.

All right, so you have to hold it on the bolt. I did this and took a measurement. Then squeezed it on a little harder, and watched the indicator needle move! I could bend the beam of the stretch gage using the variable pressure of my hand! So depending on how much coffee I've had that morning, the bolt is either stretched to length or slack! Ridiculous. A more repeatable and reliable method is needed and that's where the micrometers come in.

Also, Walt points out a giant deficiency with these gages, the fact that there's no good, repeatable surface to measure to on the bolt. I mean, there's a dimple in the head, but the other end is just flat. So you can use a stick-on point on a flat micrometer anvil on one end, and try to get the flat anvil on the other end of the mic square with the rod bolt. Then hold it there while you're doing the stretch.

Or you could use something like this:

Or if you are really gold-plating it, a used one of these- this is the same idea as the stretch gage, with an order of magnitude more precision (and you have to set it up with gage blocks because it does relative measurement, not absolute)

But that is going overboard.

You know what would be cheap, simple and repeatable?

a) take all the bolts out of the box and measure them. File and polish the tips flat so they are all the same. I'm not talking about putting them on a bench grinder, blowing the temper and making stress risers, I mean a couple passes with some fine abrasive.

b) fabricate a c-shaped piece of steel bar stock that exactly fits around the bolt head and tip plus the max elongation per the instructions. This is a "snap gage" in the book, the "no-go" part of it.

c) fab another one with the design stretch built in. This is the "go" gauge.

Install and stretch until (c) just fits.
Upon disassembly, check the bolts with (b). If they are permanently deformed and longer than the (b) dimension, garbage.

This would be the kind of thing that could look like the letter H, you could make a go side and no-go side and stamp it in, spray the thing with cosmoline and hang it on the wall of the shop for onlookers to wonder about. You could even use it at the track.

Limitless possibilities in the world of metrology-- all designed to take the operator out of the equation.

jamesjedi · 11-11-2011, 03:35 PM

Thanks everyone. I will have to read this a few more times for it to properly sink in. It sounds like there are variables regardless of the instrument.

It appears to be a good idea to have a "standard" for calibration.

Thanks again.

Walt Fricke · 11-11-2011, 06:54 PM

Raceware and ARP rod bolts are dimpled on each end, which helps with using the stretch gauge.

I hold my stretch gauge lightly while I read it. Then I remove it and turn the nut a bit more. Then I apply the gauge again and read it. Of course, as you approach the stretch you want (which is a range, not an exact figure) each application of the wrench is a little less, so as not to overshoot. Once within the proper band of readings, stop and move on to the next.

It might be nice to watch the gauge as you torque, sort of like setting cam timing. But hardly necessary. And not possible with a socket wrench.

This seems to have worked well enough. Measuring a second time when all were tightened gave same readings. Bolts haven't let loose. When looking to reuse bolts, re-measuring gave believable results (some within spec, a few not).

To remind myself about the double dimples I looked at a used set of Racewares, each of which has a label I applied with its stretch. A couple, as I recall, were elongated just a little over the limit, which is why they did not go back into the race motor. Which reminded me that after I last rebuilt my stock motored SC, I had used new stock rod bolts. I should have used these Racewares. Even stretched just a little into the yield zone ought still to leave them stronger than stock bolts.

304065 · 11-12-2011, 11:45 AM

Quote:

Originally Posted by Walt Fricke

Raceware and ARP rod bolts are dimpled on each end, which helps with using the stretch gauge.

I hold my stretch gauge lightly while I read it. Then I remove it and turn the nut a bit more. Then I apply the gauge again and read it. Of course, as you approach the stretch you want (which is a range, not an exact figure) each application of the wrench is a little less, so as not to overshoot. Once within the proper band of readings, stop and move on to the next.

It might be nice to watch the gauge as you torque, sort of like setting cam timing. But hardly necessary. And not possible with a socket wrench.

This seems to have worked well enough. Measuring a second time when all were tightened gave same readings. Bolts haven't let loose. When looking to reuse bolts, re-measuring gave believable results (some within spec, a few not).

To remind myself about the double dimples I looked at a used set of Racewares, each of which has a label I applied with its stretch. A couple, as I recall, were elongated just a little over the limit, which is why they did not go back into the race motor. Which reminded me that after I last rebuilt my stock motored SC, I had used new stock rod bolts. I should have used these Racewares. Even stretched just a little into the yield zone ought still to leave them stronger than stock bolts.

Walt,

How do you zero the gage? I guess you zero it out on the particular bolt with some preload to take up the slack in the rack, then when you take it off and put it back on again, it returns to zero?

Walt Fricke · 11-12-2011, 04:23 PM

The bolts are about 2.75" long. 've been using a suitable standard, 2" I think, to set the dial indicator so it reads 1.0000. A fair amount of futzing to be sure that that is the smallest reading I can achieve (to deal with not being sure standard's ends are exactly normal to the indicator's axis).

Then I measure the bolt. While the dimples mean what I read won't be the true overall length of the bolt, it does mean that as long as I am using the same end tips I should be able to reset the indicator in the future to the same place with the standard, and to get the same bolt measurement.

Someone sent me a spread sheet program, so I can enter the relaxed length, and the sheet will add the min/max stretch and give me where I want to be in overall length. Armed with the sheet, I stretch in increments until I am within spec. I think I have usually started off with the recommended torque, and done my first measurement from that point.

What this discussion has done for me is to think that after taking a first measurement (and a micrometer might work fine for that), what I might do for each bolt thereafter is just to zero the dial of the indicator on each relaxed bolt, and just look for the number of thousandths of stretch I am getting. No need to remember 2.7645 + 0.004 = 2.7649, looking back and forth to my listing of lenghs.

I have always assumed that one of these two methods is what everyone uses.

TimT · 11-12-2011, 04:45 PM

I think you guys are making to much out of this...

I have measured elongation on bolts up to 2" dia on some suspension bridges in the NYC area... the process is the same

The bolt is the reference....zero the gauge on the bolt... tension it up.... note delta from your zero mark... wash rinse repeat..

You don't have the stretch gauge on the bolt while you are tensioning it..

the stress/strain curve is pretty accommodating on the up slope

304065 · 11-13-2011, 05:35 AM

Quote:

Originally Posted by TimT

I think you guys are making to much out of this...

I have measured elongation on bolts up to 2" dia on some suspension bridges in the NYC area... the process is the same

The bolt is the reference....zero the gauge on the bolt... tension it up.... note delta from your zero mark... wash rinse repeat..

You don't have the stretch gauge on the bolt while you are tensioning it..

the stress/strain curve is pretty accommodating on the up slope

Tim,

Every time my ez-pass clicks around town I am thankful for guys like you who measure the stress/strain of the monster bolts on our bridges! What kind of a micrometer did you to use to measure those?

TimT · 11-13-2011, 01:07 PM

What we used is just a larger version of the ARP style gauge, just a machined aluminum frame, and a dial indicator. The bolt manufacturer supplied the gauge.

In really large dia bolts with long grips, such as those found in moveable bridges, sometimes the bolt is tensioned using a hydraulic jack and jacking frame.. The jack is pressurized to an amount that equals the design tension, then the nut is snugged by hand .. pressure is then released from the jacks....

On the Marine Parkway Bridge the lift trunnions were reinforced by post tensioning..

Those measurement were taken using ultrasound and strain gauges .. as the shafts were to long to manually measure elongation..

Bolting Solutions - Multi-Jackbolt Nuts, Bolts, Studs & Custom tensioners - Superbolt, Inc.

[/hijack]