Newbie doing rebuild needs advice! Rod Bolts

[mmasse]

I am rebuilding my 993 engine. I am still waiting for the parts to come back from the machinist, and in the meantime I am reading as much as I can about the engine assembly process.

I have one question regarding the rod bolts. Following advice on several posts I decided to replace the stock rod bolts with ARP ones. In reading the assembly instructions for these bolts, I found out that the recommended method for installation is to torque them until a specific stretch of the bolt is achieved. Now, this is not possible in my case, given that I am doing a top end rebuild only and the cranck is still in the case, so i don't think I will be able to fit the measuring tool in there.

So, there is an alternative method mentioned on the instrcutions which is basically a torque specification. But, what concerns me a bit is that it recommends to torque the bolt to the specific value, then take the nut out and torque them again in order to secure proper sitting of the bolt head agains the rod. Now, my understanding was that these are strectch bolts and that once you torque them one time you can not re-torque them. am I wrong?

Please provide your advice on this.

Thanks

[Flieger]

Factory bolts are torque-to-yield. The ARP and other aftermarket bolts are made of stronger steel and operate within their .2% offset linear elastic region, with no yielding taking place. Therefore, you can reuse them. Running the nut on twice burnishes the threads giving a torque to tension relation closer to what they found in the lab when making up the specs.

[mmasse]

Thanks Flieger

[CaptainCalf]

Yep,you can torque the ARP conrod bolts over and over and still get great stretch readings...I know cuz I've done it

[AlfonsoR]

Also note that when the ARP fastener is tightened to its recommended stretch value, the ARP fasteners apply different stresses into the rod than compared to the OEM fasteners. That change in stresses causes the shape of the rod to change slightly. Most engine builders recommend "resizing the rods" which means taking the bore back to a "perfectly" round shape.

[boyt911sc]

Mario,

Since you could not do the stretch measurements for the rod bolts in your case, just follow the recommended torque recommended by the manufacturer. I did one engine last winter and another one last month using both stretch measurements and torque application of 50 ft-lb for the SC rods and the results were 'perfect'. All the stretch values (elongation) for the 24 rod bolts were within the recommended values of 0.0095"- 0.0100". You should be fine as long as you follow the guideline for lubricating the faster before torque application. BTW, have you received the P-9191 cam tool already?

Tony

[chris_seven]

Originally Posted by Flieger

Factory bolts are torque-to-yield. The ARP and other aftermarket bolts are made of stronger steel and operate within their .2% offset linear elastic region, with no yielding taking place. Therefore, you can reuse them. Running the nut on twice burnishes the threads giving a torque to tension relation closer to what they found in the lab when making up the specs.

I have been confused about 911 con rod bolts for some time although I think it is a good idea to replace the with ARP bolts.

The early manual (Volume 1) describes the con rod bolts as malleable and advises their replacement. It doesn't give any stretch characteristics and I am not sure that 'stretch' bolts were used in 1965.

The comments about malleable bolts disappeared in later versions of the Workshop Manual at least up to 89.

The 993 seems to be entirely different to the earlier cars and there are different types of rod being used.

I have cut and pasted the relevant explanation and I hope this helps rather than adds confusion.

[mmasse]

Originally Posted by boyt911sc

Mario,

Since you could not do the stretch measurements for the rod bolts in your case, just follow the recommended torque recommended by the manufacturer. I did one engine last winter and another one last month using both stretch measurements and torque application of 50 ft-lb for the SC rods and the results were 'perfect'. All the stretch values (elongation) for the 24 rod bolts were within the recommended values of 0.0095"- 0.0100". You should be fine as long as you follow the guideline for lubricating the faster before torque application. BTW, have you received the P-9191 cam tool already?

Tony

Tony, the tool arrived yesterday...THANKS!!! I called the machinist yesterday and the parts should be back by mid next week.


Thanks for your input on the rod bolts, as well.

Cheers

[mmasse]

One more question.on a different post of mine on Rennlist guru SteVe Weiner advised to check the bearing crush. I will give a call to the machinist today to see if he had considered this. I am relacing the rod bearings for new ones but should I expect any issues with this. As mentioned before I am not taking the full engine apart therefore I will not be able to measures the crank. Is it still possible to measure the bearing crush appropriately?

Thanks

[CaptainCalf]

Originally Posted by mmasse

One more question.on a different post of mine on Rennlist guru SteVe Weiner advised to check the bearing crush. I will give a call to the machinist today to see if he had considered this. I am relacing the rod bearings for new ones but should I expect any issues with this. As mentioned before I am not taking the full engine apart therefore I will not be able to measures the crank. Is it still possible to measure the bearing crush appropriately?

Thanks

Yep, you can use plastigauge

[KTL]

Be sure to use an appropriate lube for torquing the bolts. It does make a significant difference. For example, I was doing my Pauter rods and their recommendation is to use engine oil and torque to 50 ft-lbs. Stretch requirement for the Pauter spec'd ARP bolts is 0.0055" Well the 20W50 engine oil I used (Valvoline VR1 racing) would allow some stiction to occur approaching 50 ft-lbs. The bolt was completely lubed- under head, shank and threads. Also, at 50 ft-lbs the stretch was quite a bit low.

So I chose to use Permatex red sticky Ultra Slick assembly lube and that prevented the sticking/chatter of the bolts as I was torquing them. When comparing the torque vs. the stretch, I was getting 58 to 64 ft-lbs of torque when achieving 0.0055" of stretch.

My point is, since you're not measuring stretch, make sure to use an appropriate lube to torque the bolts. I'd recommend using ARP's special lubricant to ensure you get suitable clamping force.

[mmasse]

Originally Posted by CaptainCalf

Yep, you can use plastigauge

Could you explain how to do this with the crank on the case? Sorry, I have never used Plastigauge before

Thanks

[CaptainCalf]

Originally Posted by mmasse

Could you explain how to do this with the crank on the case? Sorry, I have never used Plastigauge before

Thanks

Just lay it across the crank and install rod caps & bolts and torque rod bolts then remove the bolts and caps/brgs and check the plasigauge by measuring it with the gauge that comes with it to make sure you're within spec.

There are a number of plastigauge colors/sizes so make sure you have a variety in case you need more then one color.

[AlfonsoR]

Originally Posted by boyt911sc

Mario,
All the stretch values (elongation) for the 24 rod bolts were within the recommended values of 0.0095"- 0.0100".

Tony

Tony, what torque wrench did you use?

Thanks.

[chris_seven]

Its interesting that when questions involving measurements of torque and stretch and bearing clearances arise there are always different methods discussed and when they agree all is well and when they disagree some confusion normally occurs.

The real property that we need to measure when we tighten a bolt is preload.

If we could directly measure load it would be easy but systems that allow this are generally impractical.

Torque is relatively easy but often inaccurate and stretch is relatively accurate but not always feasible.
On a con rod bolt where stretch can be measured it must be the preferred method as the vast majority of fasteners used obey Hooke’s Law in a reliable and repeatable manner.

The influence of variations in lubrication and thread stiction are eliminated completely and confidence that the correct preload has been achieved should be high.

To put it simply if the stretch is correct then the torque applied to achieve the stretch is irrelevant.

I would always lubricate threads and make sure that there was no thread damage but this is basic good practice. The absolute level of lubrication and thread stiction will have no impact on the preload if the bolt is correctly stretched.

A +/- 25% variation in preload would be a good guide for bolts that are torque tightened and this is due to variations in lubrication and other details that affect nut factors.

A variation in stretch from 0.0095"- 0.0100" would be a Preload variation of around 5%.

The same issues arise when comparing bearing clearances. Plastigauge is relatively easy and cheap method for measuring bearing clearance. It is also fundamentally inaccurate compared to bore and shaft measurements using good quality micrometers.

I would always measure bearing bore with a gauge that was set and then measured with the same outside micrometer that was used to measure the shaft as this eliminates the errors associated with using two different measuring devices.

If the Plastigage says its OK and the direct measurements contradict this, then don’t believe the Plastigage.

[AlfonsoR]

Originally Posted by chris_seven

A +/- 25% variation in preload would be a good guide for bolts that are torque tightened and this is due to variations in lubrication and other details that affect nut factors.

.

Are you quoting John Bickford? If so, I don't think his text includes enough data to make such a conclusion. Doesn't say how many tests, what type of fasteners, and what other factors might come into question. For example, grade of bolting, rolled or machined threads, type of torque wrench, prep and installation method, temperature, etc.

I believe that a well calibrated torque wrench, good quality fasteners, and a good installation/assembly method will get you much closer, maybe within 5-10%.

However, I do agree the measurement method is more precise. It has to be since it is a direct measurement as opposed to the indirect method that the torque wrench is. The measurement method should be used when ever practical. I'm just arguing the validity of the general statement that a torque wrench is accurate to +/- 25%.

[chris_seven]

NASA Reference Publication 1228 suggests a torque to preload accuracy of between +/-15% and +/-30%. These figures needed Control bolts, nuts, washers and a consistent finish, dimensions, hardness and lubrication to obtain these results.

AIAA Document 2001-36-51 - Reducing Bolt Load Variation by Angle of Twist suggests variations of +/-35% are common when bolts are only torque tightened.

It is data of this type that resulted in the above comments.

I have not read John Bickford's book.

I think that to acheive a preload variation of +/-5% with torque tightening is a bit optimistic although I am sure good practice and attention to detail must help.

NASA 1228 suggests a variation of +/-1% to +/-8% for preload variation when using stretch measurement.

I think this NASA Document is rather conservative and is looking at even more critical applications that engines and this is why I think +/-25% and +/-5% give a guide to the difference in precision between the two methods, which was my basic intention.

I don't think my comments concerned torque wrench accuracy but preload variability.

I have to say that I am not a great fan of torque wrench calibration methods but that's an entire dabate in itself.

Again NASA 1228 suggests a variability of between +/-3% to +/-15% for typical torque wrenches but I think +/-10% would be more typical.

For rod bolts stretch is easy but for head studs we have just bought a load cell and plan to make some measurments and ry to see how much variation we get with some new studs an some Kaynar stainless steel lock nuts.

[Walt Fricke]

Chris

What is your opinion of the angle method? It would seem that if you can get the initial conditions to a known point, only the inaccuracy of that point ought to affect the result?

On a stud, if one had a way of measuring the 3 dimensional (really only one of the three) location of the exposed tip, might not measuring how much it had apparently lengthened (measurement of its axis from point above it shorter) be a way of approaching stretch?

[chris_seven]

Walt,

I beleive that the angle method generally reduces the variability which is probably why Porsche adpoted this for many of the bolt tightening methods used in the later Workshop Manuals.

I would still expect around +/- 15% as a guide figure.

I certainlt tighten head studs in this way. I think the figure in the late

The problem with trying to measure the end of a stud is that it is not a free body in the same way as a bolt. When the load increases the stud will stretch and the head and barrel will compress. It could be quite tricky to decide actual stretch of the stud.

If you measure end to end and know the modulus of the material the compression in the cap doesn't matter.

I have seen large diameter studs gun drilled in the centre so the strect can be measured using an inductive transducer or a depth gauge but this is difficult and obviously head studs are too small.

There are Ultrasound systems available but I think they are expensive but relatively accurate and perhaps it is possible to hire them?

Boltight Echometer

[Walt Fricke]

Yes, in industry where things are critical there are ways. One engineer said that at big power plants they heated the bolts to X degrees, inserted and spun on the nut, and allowed contraction of cooling to give the needed preload/stretch. Highly accurate with bolts and nuts you needed machines to lift, apparently. Just not practical (of perhaps even possible given the low masses) for the likes of us.