To add more info, Tom pointed out the MDS (Material Data Sheet) on Loctite 271, which says that no compensation is needed to achieve correct torque values:

http://tds.loctite.com/tds5/docs/271.pdf

-Wayne

1fastredsc,

As far as I know, factory rod bolts don't have the appropriate "dimples" for measuring stretch, nor does Porsche specify the required amount. That's why it's best to follow their instructions if using their rod bolts, which includes no Loctite.

I don't know any specific sources for the strech gauge other than the typical fastener suppliers, but I'm sure some searching will yield other sources.

I'm a bit surprised that, with all the engineers that frequent this board, no one has posted a stress-strain diagram, or explained what plastic deformation is. Here's a generic stress-strain diagram for a couple of ductile metals:

http://forums.pelicanparts.com/uploa...1078537456.jpg

Remember, yield bolt are not hardened, they are ductile. That's the main difference between factory rod bolts and aftermarket bolts, and why you tighten them differently.

I've labelled the yield point and the plastic deformation region. Stress is the tensile force divided by the cross-sectional area of the bolt. Essentially, that's the clamping force. Strain is the deformation of the bolt relative to it's original size. That's how much it stretches. What do we see? We see that as we tighten a bolt the clamping force increases rapidly until the bolt yields. As we continue to tighten the bolt, it stretches, but the clamping force increases far less than it did before the yield.

What does this mean for factory rod bolts? Once the bolt yields, we're on the flattest part of the curve. Errors in torque are minimized at that point. In other words, if we're 10 or 20% off in torque, it may only translate into a 5% change in clamping force.

Here's another one for steel:

http://forums.pelicanparts.com/uploa...1078535807.gif

This one's interesting because you can see how it work hardens. But even here it's obvious that relatively large changes in stretch result in small changes in clamping force.

Now, I'm not an engineer, I'm a chemist. So I can tell you that glues are polymeric in nature. Uncured glues like loctite are more like oligomers, which are simply short polymers that link together as they cure. Know what else happens to be an oligomer? Oil. Basic oil is identical to polyethylene except in length. The job of lubricant under contact wear (like what happens with your rings, but not your main bearing) it to interfere with the electrostatic interactions that cause friction. Any organic molecule can do it, big ones like oils do it better than others. Remember that gasoline lubricates your fuel pump? Tightening a bolt is a contact wear situation, pretty much any organic substance will prevent galling. Including loctite.

Summary: being a bit low in torque (assuming loctite doesn't lubricate as well as oil) is not that important when it comes to yield bolts. But since we're using loctite, which in a contact wear situation will behave similarly to oil, we really won't be that far off anyway.

Now's here's the point where I offend people. I take real exception to guys busting into this board professing knowledge that they obviously don't have, essentially calling us fools for using loctite on our rod bolts (I did too! And I will again!). We argue with facts here, not reputations. If you can't draw a stress-strain diagram, then you don't know a damn thing about stretch bolts.

Rondinone, your info kicks some major a$$, and i really enjoy your information (i'm a second year ME, and hoping to learn about what you just posted sometime in the future).But what i don't understand is that if these factory bolts, or any bolt made of any material have a certain amount of elasticity that will continuously apply pressure to the bottom of the rod bolt no matter what, then why use loctite? If it's been stressed beyond it's yield point to begin permanent deformation, than isn't the rod bolt already done for anyway?
Also with the previous statement reguarding that loctite is an oligomer and so is oil, what about the permatex red thread locker, isn't that basically the same as loctite thread locker?

And one more thing, is it possible that because of the cure time (based on the graph that wayne posted) that the loctite is begining to cure and therefore develop some of it's locking stregth before the nut has had a chance to fully clamp down with the rod bolt? I mean the average garage during a decent summer is let's say 25-30 degrees C, which according to that graph only takes approximately 5 min to develop 25% of it's stregth.

The reason to loctite on rod bolts is because they can stretch beyond where they were when they were originally tightened, yet still have significant tensile strength. So many people on this board have been acting like stretch bolts are the same as hardened bolds, which cannot undergo significant plastic deformation. So what Wayne is saying is that the bolt can loosen up if the engine is overreved, and loctite may save your engine. Personally I agree.

Plastic deformation is permanent (ductile), elastic deformation is not. Hardened bolts like ARP's only deform elastically, then they break.

This point has been made before:

http://forums.pelicanparts.com/showthread.php?s=&threadid=132255&highlight=stretc h

I would prefer to examine the stress/strain curve for the ARP and Raceware bolt material before making a blanket statement about no ductility beyond the yield point; some of the more exotic fastener alloys may surprise one. A SPS MP35N bolt still has around 10% elongation at the 290,000 psi ultimate tensile strength level.

Cheers, Jim Sims, P. E. (mechanical)

"Hardened bolts like ARP's only deform elastically, then they break."

Yes, I should know better. I was attempting to illustrate, and simplify, the difference between yield bolts and non-yield bolts. But since you're here, would you post the stress-strain curve for a few of the other alloys typically used in these applications? I think it would be insightful for all of us.

Maybe later this weekend; I'm home now away from my references. I happen to know the MP35N data off the top of my head as my team is using screws of said alloy in a piece of equipment we're building. Cheers, Jim

While not containing any stress-strain curves here is table showing the common alloys used by ARP for connecting rod bolts and head studs.

http://www.centuryperformance.com/fastenermaterials.asp#AERMET

I believe the ARP 3.5 alloy is basically MP35N, a multiphase alloy with 35% cobalt content.

Jim

Page 7 of the below PDF file has an MP35N stress-strain curve with engineering strain plotted. The drop in strength around 4 to 5% engineering strain is due to plastic necking but note the material hasn't ruptured. 2000 MPa is about 290,000 psi.

http://www.nims.go.jp/photocatalyst/MZMT/matsumoto.files/RevHPST.pdf

Jim

I beleive that garibaldi's first post is 100% right on, no exceptions. The rest is just hocus pokus, ie stuff by people who do not take the time to understand a problem and therefore are prone to beleive hockus pokus.

For the non engineering types lets put it this way- If locktite were good for a Porsche, why wouldn't the factory use or at least recommend it? They do NOT because it isn't useful, period.

A rod bolt is a streached spring. If that spring is streatched beyond its capability to streatch, it will fail, NOTHING, can save it at this point it will no longer clamp the parts togather. At the point where the streatch is exceeded it will simply allow both parts to bang against each other, that will cause certain destructin, within 100 cycles or so. 100 cycles is how long? esp at say 5000 RPM??? something like 1.2 seconds!!!

Think about it, if you have ever blown up an engine, how long was it when you heard the clank, clank vs the boom?

And that would be the same factory that:

- underdesigned the rod bolts on the 3.2
- designed the chain tensioners that failed
- used Dilavar studs
- used poor valve guides
- swapped out brand new Boxster engines when the block cracked

One thing I learned in school was to question the boundaries of the problem. Blindly following the Porsche factory without question (especially when the information is 35+ years old) is not the wisest path, in my opinion.

In this case (I'll say it again, Sam), I recommend the Loctite as a small preventative measure against the shortcomings of an underdesigned component. Based on information I've gathered, it is my opinion that the addition of Loctite has no known negative side effects from its usage. The only thing I can think of is the $4 you spend on the tube.

-Wayne

I disagree. I believe (similar to failure analysis tests we performed at Hughes) that the failure mode will be a latent failure. What does that mean? That means that an over-rev one day may significantly weaken a rod bolt to the point where the failure mode will not be discovered until several thousand miles (or several years) later. I do not believe that the rod bolts will fail catastrophically within seconds after and over-rev. The people whose cars (3.2/3.3/3.6) whom have self-destructed have most often mentioned it was not done when they were driving hard (I think Garabaldi mentioned that with respect to the pics that he posted).

This supports the latent failure theory. I believe that an over-rev may over-stretch the rod bolts and cause them to yield ever so slightly. This yielding means that the clamping force that holds the rod nut on is signifcantly reduced. Additionaly, over-revs may yield (damage) the rod bolt further, *or*, repeated vibration from the engine may cause the rod nut to come loose because it is no longer being held together tightly, since the rod bolt yielded. It is my theory that the Loctite will reduce the chances that the rod nut will fall off if the rod bolt yields and thus reduces the clamping force.

Another important thing to note is that the early rod bolts (pre 3.2/3.3/3.6) have no known defect that causes them to fall off. The reason why I recommend Loctite across the board is because there doesn't appear to be any large downside to using it. Suppose that your torque wrench is slightly out of cal, or there is a burr on the rod nuts. The Loctite is an added measure of protection on this critical joint.

Again, this is only my theory, and others do disagree with me.

-Wayne

Wayne,

But as I stated before, if the rod not is about to fall right off, the bolt will likely break anyway because of the lack of preload which _greatly_ increase the stress seen by the fastener. Dust off one of your old engineering books to see what I mean... ;-)

Also, Locitite _can_ do more harm than good if you're using the standard torque method because your affecting the friction and thusly the amount of strech (which is what you're really after in the first place).

The thread that won't die... ;-)

>But as I stated before, if the rod not is about to fall right off, the bolt >will likely break anyway because of the lack of preload which >_greatly_ increase the stress seen by the fastener. Dust off one of >your old engineering books to see what I mean... ;-)

The velocity of the piston can easily be plotted as a sin function. There are several places during the travel where the rod switches from pushing to pulling or vice-versa. You talk like an engineer, so it should be trivial for you to calculate the actual forces involved. I'm not an engineer, and even I could do it. So pull out some equations and back up what you're saying. Feel free to make some simplifying assumptions and disregard the cushioning effect of the oil.

http://www.eznut.com/btech.htm

See the above page of information about "the latent" fatigue failure Wayne is referring to and how reduced clamping force from one over rev, slightly stretching a bolt, can produce such a delayed failure. The figures and charts contained within the page demonstrate the behavior of a bolted joint that is overloaded and stretched with a reduction of clamping load (but note the clamping load doesn't go to zero). Fatigue loading is thereby increased and the joint life reduced but the joint doesn't fail immediately. The reported 911 engine rod failure while sitting at or leaving a traffic light is quite possible; the fatigue life of one of the rod bolts was likely reached even while at normal operating loads. Take the info on the website about torquing/tensioning methods with a grain of salt; they are "selling" a special nut that does fastener tensioning without the variability problems of thread friction. Recall that much of our civilization is put together with torque or turn of the nut methods and most threaded joints are quite satisfactory. Cheers, Jim

-If the yield, in your mind, is reached during an over rev to just slightly over stretch it, then technically there's nothing wrong with the design since it's been over reved.
-I thought it was part of the recommended maintenence to periodically change those tensioners?
-They've been changed and improved constantly over the years, and when i pulled mine out of my sc none were broken.
-Again, some that's been improved with better materials over the years which was stated in your book. It also seems like most of the guide issues occur on the 911's with high head temps compared to other models. (2.7 with thermo reactors)
-Well, it is a boxter.:D

The failures that have been seen have been detected in 3.2 and 3.6 engines, which have rev-limiters. Technically, I guess you may say that the rev-limiter was set too high for the structural limits of the rod bolt?

-Wayne

Thanks, Jim.