I assumed you meant that it didnt see "much" stress until its past its preload phase. But obvioulsy if the caps are separating, then the bolt is stretching more, so its being stressed higher, so the statement almost makes no sense at all to me.

I think a better statement in my mind would be that the bolts see hardly any stress/fatigue under optimal/normal separating loads from the cap, but if the loads increase and surpass the bolts rated tensile strength/stretch point, it sees exponential stress increases

??????Am I thinking funky, or am I just restating what he said maybe a different way?

 

My $.02

 

I don't understand why anyone is talking about TTY and torque-angle type bolts. Stretch is for high quality bolts, which is really what we are concerned about, right?
Bolts are like springs: you want to preload the spring a certain amount so it holds the caps on the rods, but you don't want to exceed Young's Modulus for the bolt material so that the spring won't return to its original equilibrium position (then the bolt has yielded and is no good). You also don't want the forces on the bolt due to the engine running to stretch the bolt past the yield point.

As far as what loads the bolt sees, the preload is the only load... until the engine starts running. Then when the crank pin pulls down on the rod cap, the bolts pull down on the rod, which pulls on the piston pin which pulls on the piston which goes down the bore. When all that happens, the bolts see the inertial and friction forces due to the piston and rod, in addition to the preload force. So you want

inertial force + friction force + preload force < yield force

and inertial force + friction force < preload force

so that the bolts won't yield and the caps never separate from the rods.


Right?

 

Sounds right to me.

Rich Krause

 

I should have known better than to try to give a short explanation in this forum, but I usually find I don't have the time to go through everything with all the questions that pop up.

Yes, many rod bolts are torqued to yield by OEs. The 5.7L Hemi is one, and so is another engine I'm working on now. The spec on the Hemi rod machining print even specifies .025-.127mm of permanent elongation.

I think if you talked to some of the bolt manufacturers out there they could give you many more examples.

 

Thanks for the info. So does that mean that (new) Hemi rod bolts are one-use? FWIW: .001-.005 in (approx) seems like a large tolerance for desired yield. Non-yield stretch tolerances are about 1/10th of that for the really finincky aftermarket suppliers.

Just curious: So why is it you post if you don't have time for the follow ups? Some of us might understand your detailed explanations and learn something which is part of what this forum is all about.

 

Although IMO this has been a purely academic debate... it has been somewhat interesting.

Does remind me of my days at Algor working with engineers. Never a dull moment.

-Mindgame

 

True, it is a very large tolerance. In reality we are typically in the .001-.003" range even with production style cycle times and equipment. Permanent elongation checks are just one of the means of auditing that the nut runner is still operating within spec. The real check on each bolt is the yield point control algorithm on the nut runner.

On the machine line the bolts are just tightened using torque + angle to a point approaching yield. On the engine assembly line they are pushed to yield. You are correct though. The bolts should not be reused if you're doing any work on the engine afterwards.

I know I shouldn't post if I'm not prepared to back everything up, especially in this forum. Sometimes I just see a topic that interests me and figure I'll drop in and post some info that seems to be lacking in the thread. I came from the aftermarket side of the business but moved to Detroit for just that kind of thing. The aftermarket has their mindset but there is so much to learn in this town working for an OE or supplier. The technology and resources are just in a different league.

By no means can I explain everything I come across, and this is one area where I'm fairly new. My real background is pistons and rings, but many times designing rods (and bolts) comes along with the rest of the power cylinder components.

Let me explain one thing though. The bolt will see a very small percentage of the additional load put on the joint from engine operation. If I wasn't clear on that before, I'm sorry. However, the percentage of the load the bolt sees until joint separation is VERY small. This allows a bolt to be installed into the yield area and still be a very robust joint. We are getting our rod bolts from a local source and are now switching to a torsional yield spec for rating the bolts instead of the tensile strength and hardness. This allows us to make sure the bolts are more consistent in delivering a certain preload. That rating strategy is one used by GM, BTW. The real purpose behind all of this is to get the most consistent preload possible in the cycle time allowed in mass production. We can't take minutes measuring the stretch of each bolt. Since you can, as a private engine builder, measuring stretch will deliver similar results to yield point control using our tools. Here's an example link:

http://www.a1technologies.com/pdf/te...ctionlink2.pdf

The hydrogen embrittlement issue is one difference the OEs take vs. the aftermarket for critical fasteners. ARP and A1 won't hesitate to harden a bolt above 45 HRc core. At this level the chance of hydrogen embrittlement is very high and the bolt is quite fragile. A bolt this fragile doesn't lend itself well to TTY because the point of no return is so narrow. However, it does allow you to have a very high tensile strength and thus the ability to greatly increase joint preload using the same packaging space as the OE bolts. Chrysler has banned bolts of that hardness (12.9 and above class) across the board, well almost. SRT doesn't always follow DCX's rules.

 

Good info. I especially like the link. Thanks for posting.

'Tis true that OE engines face a very different "load life" from engines modified from OE to produce multiples of the original hp and rpm high enough to produce multiples of tensile loads in the rods, but with generally the same size parts. Piston g's double in a 350 SBC from 5500 to about 7700.

Do you look at bending in the rod bolts when the big end of the rod "ovalizes" ?

 

So let me see if i'm understanding this.

You torque to the yield strength and you preload is established. Then you run the motor, and the bolt sees additional loading, albeit small because you have designed a 'hard joint'. The bolt plastically difforms (yields) more from the additional loading, but is still well below the ultimate tensile strength. This yielding reduces the clamping force, but not nearly to the point of seperation where bending and shear forces arise and would cause failure. The additional yielding has effectively strain hardened the bolt and also slightly reduced the preload on the bolt compared to the original TTY value before you run the motor. For the life of the fastener it operates elastically on the new strain hardened stress strain curve. Because you have the same max rpm for the life of the motor the loading will never exceed the yield point established from running it to max rpm the first time. Therefore it behaves elastically on the *new strain hardened* stress strain curve and no additional yielding can occur.

I don't think that your bolt science page adaquately demonstrated this, but thats the way I have resolved this in my mind based on my classes. A requirement for this type of design is that the UTS must be significantly higher than the elastic limit and .2% yield strength (which will be very close) where the initial TTY loading was stopped. If the UTS was too close to the .2% yield strength there would not be sufficient strength to allow the strain hardening to occur from the additional yielding when you run the motor to max rpm the first time.

Maybe that will make sense to someone else, or maybe they will point out where I am wrong. I think the key here is that noone has mentioned the strain hardening and the 'new elastic region' that it creates.

-brent

 

This is a good thread! Thanks for the info all, especially "94Bird" who offers a new perspective.

Rich Krause

 

Brent, I believe you hit the nail on the head. In searching for some more information to help last night I found another web site. Ajax Fasteners is from Australia but they have a very in depth paper on fasteners on their web site and it presents information in equation format to show what happens with bolted joints. Notice in particular the reference to how bolts react when put into yield on p. 10.

One of the things I noticed in this paper was their strong recommendation to not TTY a fastener that will see additional tensile loading during its lifetime. I thought that was kinda humorous given the situation.

http://www.ajaxfast.com.au/PDF/AISCPaperV05.pdf

Yes, we certainly do look at bending load in the fasteners too. As a matter of fact, more modern FEA techniques for connecting rods cycle the rod through a complete combustion cycle every 5 deg. or so to take into account whipping loads. Whipping loads are when the rod is not just in push-pull loading in line with the axis of the rod, but when you are for instance at 15 deg. ATDC, when the beam of the rod is put under loading that would induce bending. When we include those offset loads we can drop our required factor of safety down to 1.2 in certain cases instead of the previouly required 1.5 or better. That means we can get the mass of the rod down quite a bit, and place it just where we need it.

 

Is current Hemi designed for any manual trans applications? Auto only prevents overreving, but on a manual, downshifting can lead to very high overspeed which PCM obviously can't control.

Transaxle C5 Vette required a driveshaft "bumper ring" in M6 cars for this reason to keep the whipping d/s from hitting the torque tube at some rpm like 8000+ when the 4-1 "dumb$hit downshift" mode is engaged by the driver. This was an add-on after some "executive" performed the maneuver during a test drive, or so the story goes. Automatics don't have it.

What's the absolute design rpm limit you would use to figure rod loads?

Again, just curious.

 

There is currently a manual trans application for the 5.7L in trucks. I've never seen one on a lot, but I know we have a flywheel for such an application. Anyway, IIRC, the 5.7L has a peak HP speed of 5200 rpm and the overspeed is 5800 rpm. Generally you have a designed shift speed 200-300 rpm above peak HP and an overspeed about 300 rpm over the designed shift speed. Keep in mind the Hemi has an electronically controlled throttle body so things can be done to protect the engine if the driver does something a little stupid.

 

Not really on a manual trans. Over rev of the engine when it's powering the vehicle or free-revving is easy, of course, but mechanically overreving when the vehicle drives the engine due to driver error downshifting a manual was my point. That's not a PCM or throttle body controllable thing. Even Nextel Cup guys occasionally scatter one due to this on a road course.

With a heavy enough hand, you can probably get almost any syncronized manual into a gear, that when you engage the clutch drives the engine to revs WAY above your design. BTDT, but paused during clutch engagement when I noticed the driver's error.

My $.02