ever take a torch to a piece of steel, heat it and bend it? i'm sure that an ARP rod bolt will respond just the same as a stock bolt in that environment. i'll bet the bolt(s) in the disaster pic failed from excessive temperature after the engine lost oil pressure and was then run until something popped. definitely a situation where you can't blame the fastener. i doubt the bolt broke first, and the engine continued to be run with a thrown rod long enough to blue the rest of the rods as badly as shown.

Is this true?

If you go to a Miata thread on repairing the crank bolt fiasco ( http://www.miataforum.com/ubb/Archives/Archive-000001/HTML/20000913-5-004321.html ) and go down to a post by SOFTOP he says:

>>>" Disclaimer: I work for Loctite."<<<

>>>" The threadlocker on the bolt is also very important. Torque wrenches work on a theory called torque/tension relationship. Basically this theory determines that by applying a certain amount of twisting force to the fastener, a predictable amount of clamp load will be generated. This is only true if friction can be controlled. The quoted torque specification was valid for a NEW fastener in "as received" condition. This means that it was coated with a substance with lubricity equal to that of 30W oil. Less lubricity than this and the resultant clampload will be too low resulting in loosening. Higher lubricity will lead to more clampload which leads to thread failure and fastener stretch. The Loctite threadlockers mentioned have a lubricity controlling agent which provides lubricity in the correct range. Beware - this is a patented property of Loctite threadlockers and cannot be found in any other brand. It will also lock the bolt and prevent it from loosening." <<<<<<

I'm reading these posts with great interest. I would seem once the new fastener is cleaned of the " as received " lubricant and locktite then used, the locktite's lubricity would equal that of the as recieved. Providing the new fastener was spec'd with a 30W oil for torque values. It sure would be nice to see something from Locktite showing these lubricant correlations for their various products.

Hi guys,
Torque is just a means to get to the desired end result, which is a given clamp load in the rod bolt, within a designed for specific stress (pounds/sq in) in the rod bolt.
Porsche allows this stress to exceed the yield strenght of the material (which is why they dont allow re-use of the rod bolts). The alloy steel of Porche's bolts have to be fairly ductile (good elongation) to do this, which generally means lower (tensile) or ultimate strenght.
ARP etc use steel alloys and heat treats to achieve a higher strenght, (but less ductility) with the material yield strength and tensile strenght fairly close together. In this way when properly tensioned, a higher clamp force can be achieved to offset bolt stretching due to over-reving. (dealt with in previuos correspondence)
From an Engineering perspective these bolts are generally tensioned (or preloaded) to about 80% of the yield strenght of the bolt material.
Torque (irrespective of lubricant) is a poor way to accurately preload or tension the bolt, because the coefficient of friction will vary a great deal even during the torquing. (Think about it. If the coefficient of friction varies only 25%, rod bolt preload will be effected 25%, and result in bolt stresses either way too high or way too low).
The best way to achieve correct bolt preload is to measure the lenght of the bolt (bolt head to end with a micrometer after just snugging down the nut, and then rotating (torquing) the nut (in increments) untill the rod bolt has (elastically) streched the correct increment (about 3 thousands inch as I remember for my last 911 build). For personal experience, the rotation torque needed to rotate the bolt one flat at a time, between each micrometer mearurement, far exceeds any recommended "straight" torque number.
The relationship engineering formula between stress and elongation is very well known and is very linear within the elastic limit of the material (delta L = FL/EA). E is the modulus of elasticity of steel (30 million psi). Another method commonly used for say a blind rod bolt, is "torque -turn", where after an light initial seating torque, the bolt head is rotated a specific number of degrees to achieve the desired for elongation (elastic preload).
Because the bolt thread has a specific number of threads per inch, the desired elongation (say 3 thousandths of an inch) due to the specific rotation (degrees) allows prediction of specific bolt elongation and thus preload stress. or clamping force.
Rod bolts when properly tensioned should be seen as extremely stiff springs. and as long as they are not yielded (stresses past material yield point) will retain their preload. Obiously the higher the preload (ARP etc versus Porsche) the higher the "allowed" revs before rod bolt permanent streching and subsequent problems.
Rod bolts(or nuts) will not back out when preload is maintained.
If streched past yield point, the (clamp) joint will fail, the rod bearing will fail, and short term fatique will break the bolt. Finally the rod will break and scatter pieces out of a new and expensie hole in the case.
Loctite will not help. It will only confuse the ability to achieve correct preload. Loctite is aenarobic,and will start to set up as the joint is being put together (the degree of which sort of depends on the assembler's sequence of assembling rod on the crankshaft and then torquing rod bolts. As such it could resist the efficiency of torque being trabnslated into bolt stretch (preload), which would cause the joint to become loose at lower revs.
I have tried to address this interesting rod bolt contraversy from an unemotional Engineering perspective. Hope it helps.
Martin.

I'm sure that this issue can be debated until the cows come home. I did do a lot of research on the properties of Loctite, etc. when I wrote the book, but I don't have any of that in front of me right now (it's in a large box somewhere in the garage), so I can't speak specifically on the lubricating properties of Loctite prior to its curing.

All guesses aside, what a really counts is the stretch of the rod bolt. I don't have the time right now, but the best way to put this issue to rest would be to:

- get a crank and a rod
- get some factory rod bolts
- clean the heck out of the bolts
- apply the oil to one rod bolt and torque (I haven't seen in the factory manuals where it specs the type of oil? - In my early 911 manual, it just says "oil threads and mating surfaces")
- apply the Loctite to one bolt and torque
- apply nothing to another rod bolth and torque
- Measure all three with a stretch gauge and compare

That would give you the definitive answer...

-Wayne

This is becoming as much fun as materials testing lab in college. Let's get a Tinnius/Olsen machine and check the yield strength of the bolts while we're at it. Only kidding, Wayne that sounds like the way to go, a comparison using the various methods.

Mr. garibaldi, based on your reputation it is obvious you have knowledge and experience. What you seem not to realize is that you are not alone here my friend. There are many mechanical engineers on this board. I have over two decades of experience in mechanical reliability improvement on industrial rotating equipment.
I appreciate your exchange of information and opinions as long as I can manage to ignore your holier than thou attitude. It isn't appropriate or justified.

Loctite is a band-aid as you said. It has it's uses, in some cases it is a help and others a hindrance. In all cases its use is only beneficial to cover up a design or assembly deficiency.

I personally don't believe it would be much help at all on a rod bolt that is tensioned beyond yield as in an over rev condition.

But, torquing is not an exact science. The torquing factor for lubricants is an average, an estimate, and an educated guess. Saying that torquing with straight weight motor oil is more accurate than when using loctite is just not supportable.

I have done testing on over 15 lubricating products in an attempt to determine an accurate torque factor when using those particular products. The intention was to develop a chart that would compensate torque values for different thread lubricants.

The repeatability was not accurate enough to qualify as statistically accurate.
The products I tested varied from STP to ISO 32 mineral oil (turbo 32), from nickel-eaze to lubriplate anti-seize. I even tested loc-tite 242, 262, and 272.
Most products were mil spec and were used as directed by the manufacturer.
We tested products based on mineral oil, paraffin oil, napthetic oil, copper, zinc, lead, and magnesium on several grades and diameters of fasteners.
The repeatability just wasn't there. The conclusion was that the friction of the contact surfaces in the threads and on the mating faces of the nuts and or washers was unpredictable when an accuracy of over 80% was required. That was on new fasteners, the repeatability worsened on fasteners that were used previously or weathered.

The idea of using a torque factor for each lubricating product was basically scrapped.
We now calculate yield based on the manufacturer's testing and certifications, and then calculate the desired stretch on the fasteners.
The desired stretch is achieved and documented by one of the following techniques:
Measuring overall length before and after torquing with an outside micrometer or similar tool.

Measuring overall length of the fastener before and after torquing with ultrasound. This is very accurate especially on longer studs and bolts but is expensive.

Calculating the number of flats of the nut required to properly stretch the fastener (not as accurate as the other methods but more accurate than a torque wrench).

I consider using a torque wrench to tension a fastener just one step above using a pneumatic impact gun.

For most fasteners it wouldn't present a reliability problem due to the high safety factor designed in.
I the case of the rod bolts in a 3.2 liter engine the safety factor is not as high.
If someone were really concerned with the reliability of the engine I would recommend measuring the actual stretch of the bolt, not the twisting force of the nut.

Which is why in some applications special torque washers (or rather clamp load washers) are uset to achieve the correct clamping load. These washers deform in a specific way when the specified load has been achieved. Needless to say, they too can only be torqed once.

I am no expert in this field, but find this discussion very interesting.

If you go back to the first post on this thread, I said that loctie is a bandaid. I also said that the issue was with the rod bolt, and I said that the solution is to engineer at the source- THE ROD BOLT.

All that the subsequent postings by many of you have done is just reconfirm that fact in varying degrees of elaboration.

I said it, I am not apologizing. I am no expert, nor am I better than any one of you in here or out there. Its a shame that my comments were misconstued by some of you as me being arrogant. Many of you seem to liken it to me walking into your party, going up to the host, kicking him in the nuts and then going over and grabbing myself a drink. Sorry that it may not have been as politcally correct as many of you would have preferred.

That isn't the issue here nor the focus. All I did was disagree with the recommendation of using Loctite on a rod bolt in this instance. I never brought up the book, it was repeatedly referenced to, that is why is came up I knew of no book. Look closely, all I began by saying was not to use the Loctite. I made no mention of anyone, nor their work, I didnt even mention anything about myself. Its irrelevant, but alot of you are offended by this, so what would you like me to do? Leave, keep my mouth shut? Take what anyone says witha grain of salt including me, and lighten up, its only a damn rod bolt. It is what it is.

If you are offended by another view, then dont read it. Use what you want to in your engine. And I have no secrets to hide, I am not sure where that one came from. I am more than happy to share the things I know, while it is not much by any stretch of the imagination, but perhaps it could help.

And I appreciate all of you engineers and others that responded and elaborated on what I stated in the beginning in a more condensed verse.

Welcome aboard - your initiation is complete!

:)

Your knowledge and opinions are indeed welcome here.

-Wayne

Absolutely what Wayne said....

I once p!ssed in the laundry tub at a party 'cause the toilet was full... The host didn't like me after that either :)

To summarize...
 

So, my conclusion is loc-tite will do no harm, and may even be equivalent to 30W oil for lubricating properties (according to loc-tite anyway). In additions, we all believe that the bolt stretch method is far superior to using even the best torque wrench and lubrication available, as far as producing accurate clamping force....

So where do I find the stretch spec for a stock 911 rod bolt? I have never seen this spec is ANY Porsche reference book or in this forum. Racewere has a stretch spec for their bolts, but none for OEM!

Here is a nice picture of a stretch gauge for anyone who does not know what I'm talking about. You torque the nut with the guage on the bolt.

http://forums.pelicanparts.com/uploa...1078361319.jpg

The other problem about Loctite is that if you apply it on too much, it will escape to the mating surfaces between the nut and the rod thus will act like a thin soft washer after it has cured.

Stop. READTHE FIRST POST. Take care of the engineering at the source. THE ROD BOLT.

Agreed. Particularly on the 3.2/3.3/3.6 engines - do not use the factory bolts - use aftermarket ones. (page 62).

-Wayne

Ok, time to through in my two bits...

If you're using a stretch-gauge like the one shown, it doesn't really matter if you use Loctite or not. This is because, as others have mentioned, it's the bolt preload that really matters when it comes to the bolt functioning properly. The load has everything to do with the nut coming loose, and the fatigue life of the jointed assembly.

However, if you are not using a strech gauge, I would argue that you are doing more harm then good. The torque-stretch relationship of fasteners is greatly influenced by friction (both between the bolt/nut and the nut/con-rod surface), and by using a lubricant other than what was envisioned by the Connecting Cod Fastener Design and Release Engineer (someone translate that to German!), you are affecting the bolt preload. Under-loaded and you're at risk of the nut coming loose and/or fatiguing the bolt. Over-loaed and you're just looking at fatigue. Either way, it's best to use what's recommended.

If you happen to use Loctite, I still don't see how it would prevent a more catostrophic failure from happening. Say the nut does start to loosen and the Loctite prevents it from coming all the way off. If you look at the stress the bolt experiences during the up-down cycle if there is no preload (i.e. the case where a nut is coming loose), it is rather enormous and I imagine if would yield in no time.

OK, let's get nasty:
So Sammy, who's your ghostwriter????
Just kiddin. We missed you at Willow Springs this last Wednesday. Is your car back together?

We saw a rod cap peeking through the 2.4 case of Scuffy. My theory is that lots of detonation due to a lean running condition we had all day hammered the hell out of the rod bearings. We could hear some (bearing?) noise when coming into the pits from the second to last run. We thought is was the tranny or a CV joint. During the next session the engine grenaded while Jack was driving the car. Apparently, the detonation hammered the crap out of the rod bolt. Here is a picture of the culprit.

http://forums.pelicanparts.com/uploa...1078437426.jpg

Ingo

UM, that stretch gauge looks like a dial gauge hooked up to a "c" shaped holder. Am i correct in assuming so. Where can you get that holder and as already stated, where does one find the specs on stretch for the rod bolts?

Ingo, I'm signing copies in the lobby;)

Car is back together and running at conservative boost levels for now until I have time to tune the CIS for higher levels. Goes like a skinned cat compared to before.

I'm still stuck at work, my next day off will prolly be in April.

Re: To summarize...
 

That is roughly what I concluded as well for the Engine Book. There are people out there in this world who will argue up and down about the "right and wrong" ways of doing things. In this case, the torque measurement is based on 35+ year engineering and documentation. There are many things in tne Engine Book that I altered or updated from the original factory manuals to incorporate and emphasize progress that has been made since then.

The comment about a torque wrench being one step up from an impact wrench - funny but true. I know from my days working on Satellites at Hughes that torque specs indeed are very subjective. In reality, if a design is engineered properly, then it will have enough built-in margin to compensate for the variations that exist in the tightening process. These variations include material deformations in both the nut and screw, temperature variations, humidity, lubrication, surface prep - the list goes on.

The bottomline is that rebuilders have been using Loctite on the rod bolts for many rebuilds without any problems. Many rebuilders have not been using the Loctite too, and have had many successful rebuilds. In my mind, it's relatively minor issue in the grand scheme of things.

-Wayne