ARP Rod bolts

MrPerkles · MrPerkles's Garage

Hello I have a set of ARP rod bolts to fit but dont own or have access to an ARP stretch gauge,is there any other safe method of fitting them IE using a large micrometer ?

304065 · 09-27-2009, 04:24 AM

I don't see any reason why not, assuming you have (or can practice to develop) enough skill with the micrometer to generate repeatable measurements. Micrometers have a "feel" for when they are properly in tension on the part-- the very flat surface finish of the anvils will actually create surface tension against the part and "stick" the anvil to it.

You can practice this by using a micrometer standard. A "standard" is a piece of steel of known length, say 75mm, which also has flat ends. This is inserted between the anvils and the mic is tightened down using the ratcheting feature of the mic-- setting it at exactly 75mm. Then you use the little wrench that comes with the mic to zero it out. Anyway you can practice repeating the 75mm measurement over and over and over until you can do it with your eyes closed. A light oil on the standard and the anvils will help with the feel. Wipe off the instrument with iso alcohol when you are finished and put some light oil on the anvils and standard faces or it WILL rust and throw off your measurements.

There are inexpensive digital mics available, they switch between metric and decimal-inch, whichever you prefer.

My ARP bolts had a little divot in the end, seems to me that a micrometer with a ball bearing anvil attachment could fit repeatably into the divot, then you could use the conventional anvil on the other end. Rather than buying a ball micrometer (which has a ball end on the anvil permanently) you can buy accessory anvil tips (ball, flat, point etc) for not too much money. The key is that you have to subtract the length of the tips and they must absolutely fit the same way every time.

So it's not impossible and mic's are used in aerospace frequently to determine preload.

It's also a good idea to measure the new length of the bolts. That way if one deviates significantly from the others, this is a red flag that there may be a manufacturing defect. Not likely, but still possible and easy to check. Find a way of marking all 12 bolts, then create a log of the unstretched length. That way when you tear down later, you can tell if one or more has permanently deformed. If I remember right ARP says that if a fastener has stretched permanently more than 0.001" it's considered junk.

Anyway, the procedure would be to put the bolts in, turn the nuts finger tight, take the measurement and then do your first torque to a reasonable torque level, say 25 ft-pounds. Before the pundits jump on me here, I'm not suggesting you use the torque method, what I am saying is you want to creep up on the correct stretch, so just randomly reefing on a breaker bar could cause you to overshoot the correct preload. Make a stretch measurement, write it down, then subtract the "new" length to get a sense of how far you have to go.

Good luck! Ask me the time I give you instructions on how to build a swiss watch. . .

Flat6pac · 09-27-2009, 10:17 AM

I believe as you read about them, the stretch will be there as you apply the torque through 3 different assembly. Take them through the 3 sequences and then torque to spec of the bolts and the stretch will be there.
The torque is more important than the stretch.
Bruce

Jim Sims · 10-02-2009, 03:48 PM

"The torque is more important than the stretch."

No, stretch is a much better and direct method of determining correct bolt or stud tension. The amount of the applied torque that manifests as tension is very dependent on highly variable thread and bearing (bearing at the nut/bolt/washer contact face) friction. Friction depends on fastener/thread materials including plating, surface finish, thread clearances and lubrication. Compounding this uncertainty there is also the variability of the torque wrench output/reading.

Flat6pac · 10-02-2009, 07:20 PM

Well, then you are saying that torque is of no value because of too many variables. If this were a true statement there would be no torque values given and there would be no reason for ARP to include thread lube with specific torque requirements, there would be only tighten and measure specs..but I dont see any other spec in my spec books on length.
Bruce

Jim Sims · 10-02-2009, 08:31 PM

I not saying the torque method is of no value, I'm saying the stretch or extension measurement method is superior. A nominal fastener situation in which the tension achieved is only measured by the torque applied has to be designed to work within a broader range of actual tension which in practice means the lowest likely value of tension and hence a lower joint performance. If the friction is higher than assumed the fastener will be under tensioned; if the friction is lower than assumed the fastener may fail from excessive combined torsional and axial stress during tightening. The way ARP solves this dilemma (and there are other strategies) is to more tightly control friction. ARP supplies both extension (stretch) and torque values for their fastener hardware. The torque values provided are very specific to ARP hardware (down to the actual part number), using their specified lubricant and five tightening cycles to smooth down the asperities on the friction surfaces - these torque values were likely confirmed (calibrated) by stretch measurements or load cells in the fastener grip. Visit the ARP website for more information and read what they say regarding the merits of the two tensioning methods.

88-diamondblue · 10-02-2009, 09:16 PM

When I assembled my rods the torque required to get the correct stretch was 15-20lbs more than the ARP torque value. They had been torqued/measured 3 times before final installation and ARP lube applied to the threads. I did use a micrometer to measure the stretch on the rod bolts. Just takes longer to accomplish as the stretch gauge I had did not stay tight enough for me to use accurately.

MrPerkles · MrPerkles's Garage

All good stuff ,I think Mike Bainbridge will have the final say on this for me.Even though hes a long way from me if it means taking the crank up to him I will.

lucittm · lucittm's Garage

This is a good debate but it misses one point. The stretch method is better than the torque method only if the specific fastener has been designed to stretch a predetermined amount. This has to do with high quality and very pure metalurgy which can only be controlled by the manufacturer (ARP in this case). Also the reduced diameter of the fastener is one of the key factors in the stretch measurement. Most fastener manufacturers do not have the QC or even the ability to mass produce fasteners +/- .001" in length, so they use torque for a tightening specification.

ARP knows that when their rod bolts have stretched to the designed spec, they are at the peak of both strength and clamping force. Too little stretch and the clamping force is weaker; too much stretch and the bolt is closer to its yield limit under severe operational conditions. The guys at ARP will be glad to explain this over the phone and their on-line catalog does a good job at giving some technical details.
ARP | Catalog

I love this stuff.

Mark

jacksd3 · 10-03-2009, 01:04 PM

+1 on using the stretch method. Using torque I was way off. You gotta get a stretch gauge because the problem with using a micrometer is that you won't be able to engage the dimples on either end of the bolt.

Steve@Rennsport · 10-03-2009, 01:55 PM

Mark is spot on,..............

Fastener design determines what method is best for proper clamping: basic torque, torque-angle, or stretch.
I've found that ARP rod bolts reach their specified lengths around 50-55 lbs-ft so its critically important to use their procedures for reliability.

AlfonsoR · 12-29-2010, 10:11 AM

doing research...

spuggy · 12-29-2010, 05:05 PM

Agree with Mark/Steve - and the ARP catalog I read before my set was fitted very strongly recommended stretch as opposed to torque (or angular rotation, another method).

One member (I forget who) posted long ago that he noted that ARP bolts only reached some fraction (I think 80-85% ?) of specified stretch in his tests when torqued to the rated values.

Only downside is that you can't apparently get the ARP stretch gauge (a micrometer anyway) into the crankcase, so can only use it to assemble the rods to the crank before mating the case halves.

I saw a suggestion to workaround that once - to use a digital caliper to get a reading on the rod bolt, re-calibrate to zero, fiddle with it to get it out (thus losing the setting) and close them up to read the length (as a negative from the set point).

Only downside of that approach would seem to be less accuracy from a caliper than a mike (heck, the spec I'm reading is 1/100 of a millimeter acceptable range on the stretch), and any issues with getting consistent measurements from the end of the fastener without appropriate ends.

AlfonsoR · 12-29-2010, 07:35 PM

I agree on the measurement method, much more accurate than the torque translation method.

I haven't bought any rod bolts so I don't know the stretch spec for sure, but I read in one thread the spec was 11.5-12 mils (1 mil = 0.001"). From what I understand, the accuracy of most dial calipers (0-6 inch) is +/- 1 mil. I don't think I would want to be using a dial caliper if that's the case.

Plus, not only are you fighting the case to get the tool in there, the micrometer should provide more consistent, repeatable readings.

spuggy · 12-30-2010, 08:52 AM

Quote:

Originally Posted by AlfonsoR

I haven't bought any rod bolts so I don't know the stretch spec for sure, but I read in one thread the spec was 11.5-12 mils (1 mil = 0.001"). From what I understand, the accuracy of most dial calipers (0-6 inch) is +/- 1 mil. I don't think I would want to be using a dial caliper if that's the case.

Plus, not only are you fighting the case to get the tool in there, the micrometer should provide more consistent, repeatable readings.

For a 930 (and presumably the 3.2, which has the same crank, rods & bolts), ARP spec their 204-6005 rod bolts and give the required stretch over each bolt's individual "at-rest" length as .25-.27mm. Which I make 0.0098" to 0.0106".

So you're looking for ~10 thou stretch, ideally just slightly on the fat side of 1/100". Getting repeatable readings within a couple of tenths of a thou with a good vernier mike can be tough enough, so yeah, a caliper doesn't seem like the ideal tool for that job.

burgermeister · burgermeister's Garage

i bought a JEGS stretch gage to use for this. It was totally worthless. Impossible to keep on the bolt even before starting to torque it up, and repeatability of measurement was less that with my digital calipers. The calipers were actually pretty repeatable, which is what I ended up using. I found their torque spec got me to 80 - 95% of required stretch every time.

I used to work in a lab where one department did torque-tension testing. Basically, a fastener is drilled with a small hole into which a strain gaged bolt is threaded. Then, the torque vs. bolt stretch is measured as the bolted joint is tightened with representative tooling. This is how torque specs for important fasteners were developed. The variation of torque to tension is pretty large ... but the goal is always to get the desired clamp load. Torque is just the easiest way to measure it.