Fitting ARP rod bolts
Discussion
Just bought a set of these for my M3 but I'm unsure whether to use them or just replace for standard bolts.
The issue I have is that in order to ensure the bolts have stretched enough you really need to be fitting them on a bench where there's room to measure the bolt stretch. As I can't do that (engine in the car) I'm a bit worried about using them.
I also read the following post on M5board which makes alot of sense:
The issue I have is that in order to ensure the bolts have stretched enough you really need to be fitting them on a bench where there's room to measure the bolt stretch. As I can't do that (engine in the car) I'm a bit worried about using them.
I also read the following post on M5board which makes alot of sense:
M5board said:
Ok, time for a long, summary post!
First, a BIG thanks everybody for your inputs, suggestions and advice. Without these, I would not have thought about the technicalities regarding rod bolts.
I have finally decided to use the standard, oem BMW rod bolts and I've ordered them.
This was after discussing with Chris Jurkiewicz from VAC Motorsports and a couple of the guys at ARP with regards to the tightening method, clamping force and the resultant slight change in shape of the big end after properly tightening. This needed to be checked and corrected if necessary. This is where the re-useable ARP bolts come in handy, to fully tighten, check and then re-assemble.
With the standard oem bolts, the big end is most likely to remain the correct shape using the BMW tightening procedure. Using the ARP bolts might be different in some way, so I've erred on the side of caution and stuck to what I (and everyone else) know will work.
For reference, the BMW tightening procedure for the oem bolts (new ones) is:
First stage (application torque) to get them into position properly - 5Nm
Second stage (Jointing torque) to get them all to a reliable starting point for the final stage - 20Nm
Third Stage (final - Torque angle) to preload the bolt to the correct amount - tighten a further 65degrees.
I've worked out that the final 65degree tightening will stretch the bolt by 0.2257mm. This does no include the 20Nm jointing torque, so in fact the oem bolts are stretched quite a bit. This is why they are 'use once' bolts, they are stretched a little past their initial yield point (they will come out longer that when they went in).
The ARP recommendation for their bolt that would fit:
Use their Ultra Torque Fastener Assembly Lubricant.
Then Either:
Stretch the bolts to 6.5thou to 7thou (0.17 to 0.18mm)
or
torque them to 50ft-lbs (67.8Nm)
I've calculated that to achieve the 0.17 to 0.18mm stretch would require a rotation of 50degrees.
This got me thinking that the resultant clamping force could be different, which might affect the roundness of the big end.
I would really have liked to have used the uprated bolts, but without stripping down the engine and possibly additional machining work, I thought it best to stick with what I know will work. I don't really want to strip the entire engine at the moment, funds would not allow it.
If I were to use stronger, aftermarket con-rods, I'd certainly be using uprated bolts with them.
As for now, I won't be increasing the engine's power by any significant amount (headers would be nice one day, but that's about it) and I won't be using the car on the track so the oem bolts should really be sufficient.
If anyone else is thinking of using ARP bolts, I hope this information will be of use to you.
I would also like to thank Chris at VAC for taking time to read my emails and to answer them properly and honestly. He would have been able to send me 16 of the correct bolts, to arrive within the week, should I have decided to go with them.
Although several companies in the UK had the 12bolt kit, none of them were willing to split a kit and I would have either had to wait until their next stock order (weeks) or pay a silly postage for fast delivery from the US.
Obtaining them from VAC would have been faster and cheaper!
Finally, thank you to the guys at ARP who also took the time to look up part numbers, consider my emails and answer them accordingly, it is excellent customer support, even though technically, I'm not a customer.
Thanks all and all the best.
So am I better off just fitting standard bolts along with BMWs tightening method?First, a BIG thanks everybody for your inputs, suggestions and advice. Without these, I would not have thought about the technicalities regarding rod bolts.
I have finally decided to use the standard, oem BMW rod bolts and I've ordered them.
This was after discussing with Chris Jurkiewicz from VAC Motorsports and a couple of the guys at ARP with regards to the tightening method, clamping force and the resultant slight change in shape of the big end after properly tightening. This needed to be checked and corrected if necessary. This is where the re-useable ARP bolts come in handy, to fully tighten, check and then re-assemble.
With the standard oem bolts, the big end is most likely to remain the correct shape using the BMW tightening procedure. Using the ARP bolts might be different in some way, so I've erred on the side of caution and stuck to what I (and everyone else) know will work.
For reference, the BMW tightening procedure for the oem bolts (new ones) is:
First stage (application torque) to get them into position properly - 5Nm
Second stage (Jointing torque) to get them all to a reliable starting point for the final stage - 20Nm
Third Stage (final - Torque angle) to preload the bolt to the correct amount - tighten a further 65degrees.
I've worked out that the final 65degree tightening will stretch the bolt by 0.2257mm. This does no include the 20Nm jointing torque, so in fact the oem bolts are stretched quite a bit. This is why they are 'use once' bolts, they are stretched a little past their initial yield point (they will come out longer that when they went in).
The ARP recommendation for their bolt that would fit:
Use their Ultra Torque Fastener Assembly Lubricant.
Then Either:
Stretch the bolts to 6.5thou to 7thou (0.17 to 0.18mm)
or
torque them to 50ft-lbs (67.8Nm)
I've calculated that to achieve the 0.17 to 0.18mm stretch would require a rotation of 50degrees.
This got me thinking that the resultant clamping force could be different, which might affect the roundness of the big end.
I would really have liked to have used the uprated bolts, but without stripping down the engine and possibly additional machining work, I thought it best to stick with what I know will work. I don't really want to strip the entire engine at the moment, funds would not allow it.
If I were to use stronger, aftermarket con-rods, I'd certainly be using uprated bolts with them.
As for now, I won't be increasing the engine's power by any significant amount (headers would be nice one day, but that's about it) and I won't be using the car on the track so the oem bolts should really be sufficient.
If anyone else is thinking of using ARP bolts, I hope this information will be of use to you.
I would also like to thank Chris at VAC for taking time to read my emails and to answer them properly and honestly. He would have been able to send me 16 of the correct bolts, to arrive within the week, should I have decided to go with them.
Although several companies in the UK had the 12bolt kit, none of them were willing to split a kit and I would have either had to wait until their next stock order (weeks) or pay a silly postage for fast delivery from the US.
Obtaining them from VAC would have been faster and cheaper!
Finally, thank you to the guys at ARP who also took the time to look up part numbers, consider my emails and answer them accordingly, it is excellent customer support, even though technically, I'm not a customer.
Thanks all and all the best.
No not really, just that torqueing them alone doesn't reliably produce the same preload as measuring stretch due to friction etc, which is why ARP recommend the stretch method. So I'm worried that I'm more likely to have a failure with a incorrectly preloaded ARP than I am a correctly fitted OEM bolt. After all the standard bolts have done 145k and are fine so far.
les3002 said:
No not really, just that torqueing them alone doesn't reliably produce the same preload as measuring stretch due to friction etc, which is why ARP recommend the stretch method. So I'm worried that I'm more likely to have a failure with a incorrectly preloaded ARP than I am a correctly fitted OEM bolt. After all the standard bolts have done 145k and are fine so far.
And I bet BMW recommend servicing every 10k, using BMW parts, bla bla blaDo you follow that religiously ?
Given 80%+ of users will just follow the torque procedures and never have a problem. It's a fairly safe bet you'll be fine.
But as you've also already stated, they have done 145k already. So what on earth put it into your head you need to take things apart to fit uprated rod bolts ?
There is only one reason to fit "uprated" rod bolts, and that is to get a "more reliable" big end joint. Now, if your uprated bolts can't be fitted in such a manner as to ensure there preload is sufficient to avoid overload, well, what's the point??
(The std BMW bolts and procedure clearly work, and have millions of miles of durability under std operating conditions! If you are increasing rpm, then ok, maybe "Uprated" bolts might be useful, but some might be uprated rods/pistons etc ??)
Also, has anyone actually benchmarked the std BMW bolt for stress/strain?? For all you know, it might actually be "better" than the ARP one ;-)
(The std BMW bolts and procedure clearly work, and have millions of miles of durability under std operating conditions! If you are increasing rpm, then ok, maybe "Uprated" bolts might be useful, but some might be uprated rods/pistons etc ??)
Also, has anyone actually benchmarked the std BMW bolt for stress/strain?? For all you know, it might actually be "better" than the ARP one ;-)
stevieturbo said:
And I bet BMW recommend servicing every 10k, using BMW parts, bla bla bla
Do you follow that religiously ?
Given 80%+ of users will just follow the torque procedures and never have a problem. It's a fairly safe bet you'll be fine.
But as you've also already stated, they have done 145k already. So what on earth put it into your head you need to take things apart to fit uprated rod bolts ?
I do indeed follow BMWs servicing schedule and always use genuine parts unless there's a known 'better' product available.Do you follow that religiously ?
Given 80%+ of users will just follow the torque procedures and never have a problem. It's a fairly safe bet you'll be fine.
But as you've also already stated, they have done 145k already. So what on earth put it into your head you need to take things apart to fit uprated rod bolts ?
To be honest fitting the bolts was just a bit of preventative maintanence (as the car is going to be used on track) whilst the sump was off anyway. The S50B32 engine is well known for spinning big end shells due to rod bolt stretch and many people do fit ARPs, although from what I've read it's an even split for people who just torque up and people who measure stretch.
Max_Torque said:
There is only one reason to fit "uprated" rod bolts, and that is to get a "more reliable" big end joint. Now, if your uprated bolts can't be fitted in such a manner as to ensure there preload is sufficient to avoid overload, well, what's the point??
(The std BMW bolts and procedure clearly work, and have millions of miles of durability under std operating conditions! If you are increasing rpm, then ok, maybe "Uprated" bolts might be useful, but some might be uprated rods/pistons etc ??)
Also, has anyone actually benchmarked the std BMW bolt for stress/strain?? For all you know, it might actually be "better" than the ARP one ;-)
Well from what I've seen there has been numerous failures of standard BMW bolts but I've never heard of an ARP failure on the S50.(The std BMW bolts and procedure clearly work, and have millions of miles of durability under std operating conditions! If you are increasing rpm, then ok, maybe "Uprated" bolts might be useful, but some might be uprated rods/pistons etc ??)
Also, has anyone actually benchmarked the std BMW bolt for stress/strain?? For all you know, it might actually be "better" than the ARP one ;-)
The same way BMW do at the factory. Pre torque of 5nm followed by a further torque of 20nm and finally angle tighten to 65°. I can see what you're trying to get at though - that I'd have no way of knowing whether the BMW bolts had stretched enough.
Having seen people use the stretch method with ARPs and them needing 65lb/ft to stretch enough I'm wary of only tightening them to 50lb/ft.
Having seen people use the stretch method with ARPs and them needing 65lb/ft to stretch enough I'm wary of only tightening them to 50lb/ft.
les3002 said:
The same way BMW do at the factory. Pre torque of 5nm followed by a further torque of 20nm and finally angle tighten to 65°. I can see what you're trying to get at though - that I'd have no way of knowing whether the BMW bolts had stretched enough.
Having seen people use the stretch method with ARPs and them needing 65lb/ft to stretch enough I'm wary of only tightening them to 50lb/ft.
But then you could call into question the accuracy of their torque measurements too. Was their wrench accurate ? did they use the correct lube ? Correct amount of time in between stages. Was the temperature of the room and workpiece as it should be ?Having seen people use the stretch method with ARPs and them needing 65lb/ft to stretch enough I'm wary of only tightening them to 50lb/ft.
What size are the actual fasteners ?
Sounds like it's homework time!
Get some spare rods, and a few spare ARP and std fasteners!
With the spare rod on the table, measure rod bearing ovality / stretch / torque for a number of different tightening scenarios (including the std BMW procedure and all the way to failure if you have enough spares!) (ideally you also want to measure installed preload (because that is what you actually care about (you never want to get to zero preload under dynamic loadings!)) but that takes a piezo pressure sensor and Charge amplifier to to properly, so i'll assume you can't do this.
Then, fit the ARP bolts and repeat.
The ARP material characteristics are i think published (in terms of elastic / plastic limits etc)
So from stretch you should be able to calculate installed load
You will have to take a guess at the material used in the std rod bolts, (or measure it/get it measured).
However, if your testing shows no big end deformation right up the point the fastener yields, then you can just pick a sensible stretch (snug torque + angle for when installed) to aim for when you do the real build
Get some spare rods, and a few spare ARP and std fasteners!
With the spare rod on the table, measure rod bearing ovality / stretch / torque for a number of different tightening scenarios (including the std BMW procedure and all the way to failure if you have enough spares!) (ideally you also want to measure installed preload (because that is what you actually care about (you never want to get to zero preload under dynamic loadings!)) but that takes a piezo pressure sensor and Charge amplifier to to properly, so i'll assume you can't do this.
Then, fit the ARP bolts and repeat.
The ARP material characteristics are i think published (in terms of elastic / plastic limits etc)
So from stretch you should be able to calculate installed load
You will have to take a guess at the material used in the std rod bolts, (or measure it/get it measured).
However, if your testing shows no big end deformation right up the point the fastener yields, then you can just pick a sensible stretch (snug torque + angle for when installed) to aim for when you do the real build
Max_Torque said:
Sounds like it's homework time!
Get some spare rods, and a few spare ARP and std fasteners!
With the spare rod on the table, measure rod bearing ovality / stretch / torque for a number of different tightening scenarios (including the std BMW procedure and all the way to failure if you have enough spares!) (ideally you also want to measure installed preload (because that is what you actually care about (you never want to get to zero preload under dynamic loadings!)) but that takes a piezo pressure sensor and Charge amplifier to to properly, so i'll assume you can't do this.
Then, fit the ARP bolts and repeat.
The ARP material characteristics are i think published (in terms of elastic / plastic limits etc)
So from stretch you should be able to calculate installed load
You will have to take a guess at the material used in the std rod bolts, (or measure it/get it measured).
However, if your testing shows no big end deformation right up the point the fastener yields, then you can just pick a sensible stretch (snug torque + angle for when installed) to aim for when you do the real build
I'll translate.Get some spare rods, and a few spare ARP and std fasteners!
With the spare rod on the table, measure rod bearing ovality / stretch / torque for a number of different tightening scenarios (including the std BMW procedure and all the way to failure if you have enough spares!) (ideally you also want to measure installed preload (because that is what you actually care about (you never want to get to zero preload under dynamic loadings!)) but that takes a piezo pressure sensor and Charge amplifier to to properly, so i'll assume you can't do this.
Then, fit the ARP bolts and repeat.
The ARP material characteristics are i think published (in terms of elastic / plastic limits etc)
So from stretch you should be able to calculate installed load
You will have to take a guess at the material used in the std rod bolts, (or measure it/get it measured).
However, if your testing shows no big end deformation right up the point the fastener yields, then you can just pick a sensible stretch (snug torque + angle for when installed) to aim for when you do the real build
Torque as per ARP specs and instructions with a good quality torque wrench and never worry about it lol
stevieturbo said:
I'll translate.
Torque as per ARP specs and instructions with a good quality torque wrench and never worry about it lol
That's the plan!Torque as per ARP specs and instructions with a good quality torque wrench and never worry about it lol
I'd love to do that homework MaxTorque and if the engine was in bits on a bench I probably would.
As it is I'll be lying upside down attempting to change the bolts so I'll just torque them to spec and forget about it.
Fair enough!
The homework CAN be fun though!
(many years ago, as a fresh out of uni engineer, my first "assignment" on a real engine dev project was to characterise the cylinder head retainment on a new V6 engine. So i set up a test procedure to tighten each head bolt, into yield and to failure, whilst recording torque, compressive load, and angle. Luckily for me, being required to man the instrumentation mean't that i had to get someone else to wield the torque wrench. And of course, because you couldn't exactly tell when each bolt would fail, the poor guy pulling the wrench repeatidly went shooting off across the buildshop, usually ending up arse-over-tit when the fastener failed. it was funny the first time, and 32 bolts later, i was still laughing my arse off everytime he fell over......... ;-)
The homework CAN be fun though!
(many years ago, as a fresh out of uni engineer, my first "assignment" on a real engine dev project was to characterise the cylinder head retainment on a new V6 engine. So i set up a test procedure to tighten each head bolt, into yield and to failure, whilst recording torque, compressive load, and angle. Luckily for me, being required to man the instrumentation mean't that i had to get someone else to wield the torque wrench. And of course, because you couldn't exactly tell when each bolt would fail, the poor guy pulling the wrench repeatidly went shooting off across the buildshop, usually ending up arse-over-tit when the fastener failed. it was funny the first time, and 32 bolts later, i was still laughing my arse off everytime he fell over......... ;-)
If you use oem bolts use manufacturers procedure If you use ARP bolts use their reccomendations you don't honestly believe that factory measures stretch on b/e bolts on the proction line do you ,they will have quality preset non ad justable torque wrenches which will be monitored and changed regularly
There's a serious amount of overthinking of the problem going on in the M5board post. Yes it is ideal to use measured stretch to extract every tiny bit of the bolt's yield potential but unless you're building a Top Fuel drag engine where everything is on the bitter limit it's somewhat overkill. A half decent torque wrench and settings that have been tested for the lubricant used will get you far enough into the ball park for there to be no problem in use.
What most people don't realise is just how over-specced even standard rod bolts are in most engines. When I run the bolt loading versus bolt strength calcs at peak rpm I usually find that the total loading on the big end joint is about equal to the tensile strength of just one of the bolts i.e. a 100% safety margin. So why do bolts occasionally break? Mainly because of an unusual combination of factors that use up the safety margin. The very rare poorly formed thread with high friction, the very rare faulty bolt, friction variations due to cleanliness of assembly and lubricant coverage, torque or angle setting variances due to out of calibration equipment, over-revving, valve bounce which removes the cushioning effect of gas pressure acting on the piston crown at TDC.
If enough of these stack up a bolt can fail but usually when a big end breaks it isn't the bolt that failed even though it's almost certainly snapped once the event is over. I wouldn't trust most "enjinears" let alone Joe average car owner to be able to diagnose whether the bolt broke first or the bearing failed due to lack of lubrication or a shell spinning. Anecdotal claims of engines with "known" bolt problems are somewhat suspect IMO.
The critical thing IME is to torque a new bolt up several times to condition and burnish in the contact surfaces and reduce friction scatter. Friction levels can drop by 25% after several cycles. I used to torque every new bolt to 90% of its full value and loosen again three times with ordinary engine oil before cleaning everything again and doing the final fit with the chosen assembly lubricant. Rather like running in a new engine on dino oil rather than synthetic you want a not too slippery lubricant to let the threads burnish in and then use moly or ARP lube or whatever, and obviously a lower torque setting, on the final fit. If the lubricant is so slippery there is not much friction anyway then multiple torque-ups don't achieve much.
Most road engines use 8mm or 9mm bolts. The huge 10mm ones in the M3 seem like massive overkill. I don't know the exact piston and rod weights but putting in some educated guesses the total joint load at 8000 rpm on a 91mm stroke is about 10,000 lb. The 90% tensile yield strength of even a single 10 x 1.25mm grade 10.9 bolt (150,000 psi) is about 13,000 lbs. For a 220,000 psi ARP bolt it's about 19,000 lbs. Even at a proof load of 80% of the yield strength the ARP bolts are 3 times as strong as necessary for 8000 rpm and the OE bolts twice as strong as necessary.
You could cock the torque settings or friction levels of the ARP bolts up by a massive amount and still achieve sufficient clamping load to be as safe as houses. Most Chevy V8s with 100mm plus diameter pistons and massive strokes only use 3/8" bolts and will still go way past 8000 rpm.
Theoretically the M3 ought to be able to run to 14,000 rpm with ARP bolts correctly fitted. 12,000 rpm on the stock ones. I very much doubt if any big end failure on an M3 engine was a bolt problem.
Fit the ARP bolts as I suggest. Find something else to worry about.
What most people don't realise is just how over-specced even standard rod bolts are in most engines. When I run the bolt loading versus bolt strength calcs at peak rpm I usually find that the total loading on the big end joint is about equal to the tensile strength of just one of the bolts i.e. a 100% safety margin. So why do bolts occasionally break? Mainly because of an unusual combination of factors that use up the safety margin. The very rare poorly formed thread with high friction, the very rare faulty bolt, friction variations due to cleanliness of assembly and lubricant coverage, torque or angle setting variances due to out of calibration equipment, over-revving, valve bounce which removes the cushioning effect of gas pressure acting on the piston crown at TDC.
If enough of these stack up a bolt can fail but usually when a big end breaks it isn't the bolt that failed even though it's almost certainly snapped once the event is over. I wouldn't trust most "enjinears" let alone Joe average car owner to be able to diagnose whether the bolt broke first or the bearing failed due to lack of lubrication or a shell spinning. Anecdotal claims of engines with "known" bolt problems are somewhat suspect IMO.
The critical thing IME is to torque a new bolt up several times to condition and burnish in the contact surfaces and reduce friction scatter. Friction levels can drop by 25% after several cycles. I used to torque every new bolt to 90% of its full value and loosen again three times with ordinary engine oil before cleaning everything again and doing the final fit with the chosen assembly lubricant. Rather like running in a new engine on dino oil rather than synthetic you want a not too slippery lubricant to let the threads burnish in and then use moly or ARP lube or whatever, and obviously a lower torque setting, on the final fit. If the lubricant is so slippery there is not much friction anyway then multiple torque-ups don't achieve much.
Most road engines use 8mm or 9mm bolts. The huge 10mm ones in the M3 seem like massive overkill. I don't know the exact piston and rod weights but putting in some educated guesses the total joint load at 8000 rpm on a 91mm stroke is about 10,000 lb. The 90% tensile yield strength of even a single 10 x 1.25mm grade 10.9 bolt (150,000 psi) is about 13,000 lbs. For a 220,000 psi ARP bolt it's about 19,000 lbs. Even at a proof load of 80% of the yield strength the ARP bolts are 3 times as strong as necessary for 8000 rpm and the OE bolts twice as strong as necessary.
You could cock the torque settings or friction levels of the ARP bolts up by a massive amount and still achieve sufficient clamping load to be as safe as houses. Most Chevy V8s with 100mm plus diameter pistons and massive strokes only use 3/8" bolts and will still go way past 8000 rpm.
Theoretically the M3 ought to be able to run to 14,000 rpm with ARP bolts correctly fitted. 12,000 rpm on the stock ones. I very much doubt if any big end failure on an M3 engine was a bolt problem.
Fit the ARP bolts as I suggest. Find something else to worry about.
Excellent post Puma. I thought I was being a bit of a big girl about it, at least now I can just get on and get the thing back together, I'm sure I'll find something else to worry about.....
I wonder what the issue is then that causes so many spun big ends on the S50 if the bolts are more than up to the job?
I wonder what the issue is then that causes so many spun big ends on the S50 if the bolts are more than up to the job?
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff