ARP Rod Bolt Torque Values
 

Sorry but I checked the archieves with no luck.

Does anyone know the torque values for 9mm ARP rod bolts for a 3.3 turbo? I can not measure stretch, so I need torque.

Thanks

With the ARP lube, it's 35 ft/lbs, (they recommend tighten/loosen three times).

-Eric
90 C2

I just put these on my rods and used a stretch gauge. They took a fair amount more than 35lbs to reach proper stretch. I am not sure how much stretch gets retained in the bolt after it is loosened prior to retorqueing, but I the bolts do have some elasticity.

But eric is right, that is the spec ARP uses if you do the torque method.

Cheers

Thank you

If you can...
 

I would consider torquing up a rod or two in a vise measuring the stretch to get 'torque values' for when you put them on the crank. Almost best of both worlds.

tadd

Don,

My 930 ARP rod bolts were torqued using both a torque wrench and a stretch gauge. We considered them to be "torqued" when the prescribed stretch value was reached -- the torque wrench was just for reference. When they reached the proper stretch value, the torque wrench read 41 ft-lb. I wouldn't say that you should necessarily torque yours to 41 ft-lb absent the input from a stretch gauge, but you can add this to your mix of incoming information. It seems to confirm what Jeff said above.

Rob

Rob, that is what I was trying to say but couldn't see to!! When playing around with the gauge I noted that it yoy put up to say 15-18lb of torque on the bolt and then undid it, it would return to its origianl length. My machine shop, which has used 1000's of ARP rod bolts for different applications, also confirms that the two methods don't yeild the same torque on the nut.

All that said, I *think* the majority of people use the torque method and not a stretch gauge and I have never heard of a problem from it.

Cheers

Dont you use an dial caliper to measure stretch?

I found the threaded end of the ARP rod bolt was not ground flat, so I couldn't get a reliable reading with a micrometer (I had a micrometer, but not stretch gauge). Maybe a micrometer, or caliper with a point to fit in the center "divot" on the bolt........

-Eric

I used the ARP stretch gauge that incorporates a dial indicator, here it is.....

http://forums.pelicanparts.com/uploa...1145386716.jpg

http://forums.pelicanparts.com/uploa...1145386728.jpg

Cheers

Note that that gauge is calibrated in .0005" increments unlike most other cheaper guages that are in 0.001" accuracy. Its works great and I had the same results as Jeff where the torque to get the proper stretch is definitely higher than what Arp says.

I bought the gauge at the local auto parts center that has a performance shop.

I seem to remember the 8mm bolts for my 993 were .017" stretch...would have to look up my notes to confirm.

Cheers,

Mike

torque vs stretch
 

The extra torque is put into over coming the thread friction. As soon as you stop turning the bolt, that 'needed extra' just disappears. Angular measurments can be much better, but since there is uncertiantily in the starting torque... Best is a stretch measurment.

That said, I have had KICK ASS results with a torque wrench that uses a piezo element and cacluates the rate of force change. That sucker will nail the very, very begining of the plastic range every time for tension applications regardless of the bolt type, material type or size.

tadd

I not sure if I am saying the same thing here or not, but it is my understanding when you tighten a fastener what you are really trying to achieve is a specific tension. Stretch measures it very accurately, torque puts friction into the equation. That said, it would seem to me that the manufacturer of a given fastener takes that into account when they provide a torque spec.

Cheers

Jeff,

You're exactly correct. Because torque is an indirect (but convenient) way to measure bolt tension, manufacturer's torque specs are designed to acheive that desired tension, and have already accounted for frictional losses and other uncertainties inherent in the torquing process. But those uncertainties can be significant, so it was not surprising (to me) to see that the torque spec for these ARP bolts (35 ft-lb) was 17% lower than the torque needed to achieve the specified stretch (in my case). The good news is that there's usually a pretty wide range of acceptable torque levels, and I haven't heard of any rod bolt failures that resulted from assembly when they were torqued with a calibrated torque wrench to 35 ft-lb.

Rob

sort of...
 

Catca:
Your insight is spot on, measuring the stretch of the bolt is a direct, accruate way of determining the installed preload on the bolt. That preload is the basis of the friction between the threads that keeps the bolt from 'coming undone'. You want to add enough tension that the bolt won't come undone in operation, but not enough to put the bolt material in its plastic range where it perminately deformes. You want to hit the sweetspot where the bolt is elastic (can stretch and deform) from operational loads and not unloosen, but not over do it and stretch into the plastic region where it does not 'snap back'. Typically this is about 60% of its ultimate yield strength (where it breaks).

Unfortunally, when you install using a 'torque wrench' that installation friction from rotating the threads of the bolt in the hole can vary widely thanks to the metal types, metal finish (surface roughness, platings), lube (dry,oil, moly), and bearing area. So the typical torque tables are only a guide, not gospel. Until the derivative torque wrenches which actually measures the rate of change in applied torque, so the wrench knows when you start into the plastic range, the only accurate way to get the correct preload was to measure stretch directly. For blind tension applications (where you can't get to both sides of the bolt) they came up with these way cool washers that have a thin plastic coating that changes color to let you know when you have the correct preload.

This is also why 'lock washers' are an abomination. Anything that is soft enought to deform while tightening will be soft enough to deform under load. When that happens all the wonderful friction holding your bolt in the hole evaporates and the bolt can unloosen in that fraction of a second. Thread locking compounds avoid this althogether.

Finally, it is always better to overtighten than undertighten from a fatigue standpoint, hence the 'sort of' in the sig line...

tadd

I'm pretty sure this is for the first use only, not successive uses...

Wanted to add that you should really use a torque/stretch method. Obviously stretch is the most accurate, HOWEVER I have seen a few bad ARP rod bolts that stretched to the proper value but had VERY little torque ont hem because the bolt was faulty. Would have resulted in a dropped rod.

That said, use stretch but verify that your torque value used is close to their recommended value for torque method.

Good point Chris. That, in essence, is what I did.

Cheers

Nope - I checked with ARP tech support yestreday and you can re-use as long as the threads are in perfect shape and you should keep the bolts paired with the nuts. And always use the moly lube for rod bolts.

Don,

I was speaking of the "torque/untorque three times" procedure. You only do this the first time you torque new bolts. When you re-use them later you don't need to torque/untorque them 3 times...

...

As a side note, the ARP stretch gauge I have will not be used all that often. If someone wants to use it and pay the shipping, they are welcome to it. A paypal deposit (refundable on return) is all it would take! Let me know if this is of interest to anyone.

Cheers

Nice offer Jeff!

Got it - sorry.

Stretch gauge
 

Whoops - meant to send a PM
PM Sent Jeff!

AM

This is the result of my search for ARP rod bolts. I am at this moment installing the rods on the crank of my 3.0 SC. The ARP paper says stretch 0.0115-0.012". I have only a micrometer ( I have dial indicators, but am not interested in reinventing the wheel here). After three (3) go arounds as per ARP at 45 ft-lb, with the lube, the measured stretch in each of the two bolts I just did is only 0.008". I did measure them after the second go round, and the length of the bolt had increased by 0.005" from the non tensioned out-of-the-box length, so they are indeed stretching.

Tried calling ARP, but they're gone for the day. Next best thing is to ask you guys (maybe better, actually).

For reference, I just last week did my 2.7 using Raceware bolts, torqued per Raceware instructions. After torquing them, I found the Raceware bolts to be of the prescribed length according to the micrometer. So, I am a bit puzzled by the results of this ARP tightening sequence.

I am open to suggestions.

Edit:
It just occurred to me that the rods have been resized, and quite possibly were resized with the ARP bolts installed. So, it is possible that the bolts may have had a few thousandths stretch before I ever go to this point.


Thanks!
Pat

ARP says you can re-use them multiple times.

1. Beg, borrow, or steal an actual stretch gauge.

2. If assembling on the bench, just use the gauge.

3. If assembling in the case, use torque-angle method:
3.a. Put rod in soft-jaw vise and torque both nuts to 5 foot pounds.
3.b. Use a thread marker, liquid paper, to mark the nut/bolt
3.c. Turn bolt until proper stretch level achieved.
3.d. Measure angle and then use 5 foot pounds plus that angle when installing on crank.

This comes out so consistently it's almost always spot on. The reason is because at 5 foot pounds, the differences in friction coefficients have barely begun to come into play - it's just snug. When you look at how quickly the required torque to keep turning ramps up with each successive degree of turn, it's easy to understand why this can result in widely-varying torques for the proper stretch when you look at the entire set of 12 nut/bolt/combinations.

The ARP site USED to say 3 times on/off with torque. NOW it says 5 times on/off. Hmmm, wonder why? Because while this method may be "good enough" in most cases it's obviously flawed. The whole idea of on/off is to burnish the nut/rod surfaces to try to get those friction coefficients more consistent.

Torque-angle method gets around that by not letting the coefficents come into play in the first place. This is precisely why Porsche changed the head stud nut procedure to this method on the 3.6 at some point. Not sure what other engines they specify it for now - but engineering-wise it makes a heck of a lot more sense.

Chance
Thanks very much for the response. The methodology you describe makes excellent sense. However, in my case, there is pretty much no possibility that I will have a stretch gauge at my disposal.
I did speak to the ARP tech support people today, and discussed the metallurgy of the bolts (an 86xx alloy, which I recall as being a wartime-developed alloy substitute for the 4130/4340 series chrome molybdenum steels), what it takes to exceed the yield and get the bolts into the plastic range, and the realistic differences between stretch method and torque method. Evidently ARP has suggested five torquing of the nut/bolt to allow the lubricant which they supplied to pretty much equalize the frictional components of the torque process, exactly as you describe.

What I have arrived at with four successive torquing of the bolts is a very consistent stretch of 0.009" per bolt , measured with a micrometer. Intuitively, I would think that another 4 ft-lb or so would put me at just about the prescribed stretch. If I leave them alone as they are, I figure I'm losing a very small fraction of the clamping power of the nut/bolt assembly. If I try to honk them down a bit more, I risk getting into the plastic range (permanent deformation). With an engine which will rarely see 7000 RPM, I will leave them as is for now. Knowing a bit regarding the alloy they are made from, I feel better about them not letting loose on a straightaway somewhere.

Of course, this becomes academic if a stretch gauge falls into my lap over the weekend. Far cry from the days of putting Chevy head bolts in with the farmer method..."tighten them until they squeak, and then go another 1/2 turn".:p

Pat

My guage has been all over Canada now, I think it wants a holiday the USA, so if anybody still needs to use one, drop me and email and I can get it sent out to you once I get a deposit!

Cheers

This method should apply to all ARP bolts assuming:

1. thread pitch is the same as for the 3.6
2. desired stretch is .012"

Make sure rod bolts are fully seated in the rod cap - you can do this on the bench by assembling rod and cranking down a bit on the nuts. Once you're in the case, torque each nut to 5 foot pounds. Turn each another 135 degrees. Results in .012" stretch.

FYI

P.S. - Anyone who is able - please verify this on the bench - then this should become standard procedure for in-case rod assembly.

Is 5 foot pounds enough to reliably seat the bolt and squeeze the two halves? Or would a first pass at higher, and then redoing it be better?

-Eric
90 C2

Good question - it's kind of a trade-off between having enough snugness vs. getting into too much friction. Just lube up your bolts and 5 should be enough. Also, and this is important, make sure your bolts have bottomed out in the rod caps first - that definitely takes more grunt to get those seated. I'll update my previous post to reflect that...

I just had a nice long discussion with the engineering staff at ARP. My concern is having insufficient tension on the rod bolts...this has been obviously gnawing at me for a few days now. Being a former machinist and mechanical engineer, I do not feel quite right about the 0.008 stretch instead of 0.0117 prescribed by ARP. Without going into too much detail, I have considered at what point will the reciprocating load on the rod overcome the tension in the bolt, and start the rod cap to wiggling around on the rod? If I take a chance and try to guess how much more I can torque these bolts without a gauge, I could quite easily get into the plastic range of the bolt, and at $35 bucks a bolt, I can't afford too many of those mistakes. I don't really need to blow up this engine due to my own negligence or stupidity.

So, I am heeding the advice of this thread, taking the guesswork out of the picture, and procuring a stretch gauge. And, I'll probably take the 2.7 case apart (that I just put together) and redo those rod bolts also. Thanks also to ARP for taking the time to review this stuff with me.

Pat

Wise decision - there are so many things you can mess up with these things as a weekend warrior - anything you do "above and beyond" is going to make you feel MUCH better afterwards. Otherwise these things will gnaw away at you later because you're just not 100% sure it's right.

i have torque my new rod bolts 35ftlbs 3 times in the case

should i torque them more?

i read here you say better more than less torque

I just had my case opened. The ARP bolts went back to the unsprung length when I took it apart (I had written down the measurements). I put them together again a week or so ago, with a stretch gauge, and rechecked the length. I basically ended up at 0.011" stretch.

spjuvern, you really need to either use a stretch gauge, or a micrometer to do this correctly. If you under stretch them, you are leaving a lot of clamping force on the table. If you overstretch them, they are junk. I think if you folllow ARP's instructions EXPLICITLY, you should be OK.

I just went through this yesterday and my opinion is that the 35 ft.lbs. tighten and loosen 3 times is a bunch of b.s.. (part # ARP 6005). If you do that I'm measuring about .0085-.009" of stretch when you hit 35 ft. lbs on the final torque. I found it was hitting upwards of 45-50 ft. lbs. on my wrench to hit the .0115" value and get into the correct range. So my opinion is that the 35 lbs is a bit low and it sounds like that's the consensus. What makes a lot more sense to me as someone else stated is to go to X torque and then add some Y degree of rotation much like the Porsche procedure. Its sort of annoying to buy $225 worth of bolts and then find out you need to chase down the special gauge. Oh well...

Keep in mind too that not everyone's torque wrench in calibrated exactly the same, but I think 10-15 ft. lbs more is a sizeable jump over 35.

I also think while its good to be as thorough as patkeefe is, you can drive yourself insane worrying about rod bolts; after all you bought the ARPs anyway to get a stronger bolt than factory. I believe if you get close to the stretch range you're in good shape. I would probably worry a bit more if I was entering LeMans with it...

And I will add that I think Pat's statement above is dead on IMO (for whatever that's worth :) ) and also why I think the 35 ft lbs. is a bit too low.

Stress Strain
 

The ARP rod bolts are similar to the curve in the following link:

http://darkwing.uoregon.edu/~struct/courseware/461/461_lectures/461_lecture24/461_lecture24.html

Once you get past the "top of the straight line hill", you are in the plastic range, which means permanant deformation. An ARP rod bolt is essentially junked if you pass this spot, and further torquing (stretching) will result in "necking" of the fastener. The limit is, as ARP says, around 0.0115" stretch, which does not correlate to torque for many reasons (there is a lot of good info here on ths board regarding this).

Engines cost, what, $7000 to build? Buy a micrometer for $50 and check the stretch. Very cheap insurance.

http://forums.pelicanparts.com/uploa...1236603214.gif

Pat

We just did some ARP bolts a little while ago... from my failing memory, I recall that to get to the proper stretch we were 50-55 foot pounds on my uncalibrated wrench ... at 35 foot pounds, I think we were at about 60% of the spec stretch.

JB

The limit is higher depending on the ARP bolt you have- they have about 5 different ones for Porsche so anyone reading this should refer to the ARP site where part numbers and stretch limits are noted. If you measure .0124" and the upper spec happens to be .0122", I wouldn't sweat it. I would imagine ARP leaves enough on the table that the bolt won't explode your engine.