Torque Values - Wet or Dry?

[Ferrino]

Dumb question: are the torque values in the factory manuals assuming some level of thread lubrication or are they dry? If lubed, what are we talking here - a wee bit of oil or a smear of anti-seize grease or...

[winders]

If the Factory manual does not specify a lubricant for assembly, it means dry. If the Factory manual specifies a lubricant for assembly, it means when lubricated.

[Ferrino]

Thanks.

[T77911S]

i would do it wet.

you want the torque to be how much force is pulling on the stud, not the resistance of the nut turning on the stud.

this will be an on going debate i am sure.

[RedCoupe]

Technique is also important. No stopping and starting when you're tightening, but tightening smoothly and steadily up to the stated torque value.

[chris_seven]

This topic does need a debate. - I has probably been covered before but its worth a re-visit

I would agree with the previous comments that unless stated otherwise the Manual assumes dry torque figures.

When the manual advises lubrication try to do exactly as stated.

If you use a 'dry' torque figure and also lubricate you run the risk of ovetightening the fastener.

This overtightening can be between 30 and 50% depending on the lubricant.

Some of the ARP thread lubes suggest a 50% reduction in torque as an example.

[Quicksilver]

Originally Posted by T77911S

i would do it wet.

you want the torque to be how much force is pulling on the stud, not the resistance of the nut turning on the stud.

this will be an on going debate i am sure.

Which ignores the fact that the only measurement they gave you was torque. That torque was based on specific conditions which specifically included the resistance of turning the nut and to ignore them and say, "always wet" is inviting damage.

The correct answer is has been given: If the factory torque spec doesn't mention lubrication then install it dry. If it mentions a lube then use it.

To get the best working knowledge of fasteners it should almost be a requirement to get Carroll Smith's books:
- Prepare to Win
- Tune to Win
- Engineer to Win

These are car prep books done in language where you can understand and absorb information.

(Favorite tongue in cheek fastener quote from a friend just back from A&P training in the Air Guard: "I used 'military torque specs'. I tightened it until I felt it get a bit easier then backed it off an eighth of a turn.")

[Ferrino]

Originally Posted by chris_seven

If you use a 'dry' torque figure and also lubricate you run the risk of ovetightening the fastener.

That's exactly the condition that I am concerned about.

[Danimal16]

Torque values are dry unless stated otherwise.

[chris_seven]

It can all get quite tricky and often the answers aren't great.

Dry Torque would generally assume a 'clean' dry thread. Any contamination of the thread will affect the 'tightness' and so will any damage.

Brand new fasteners can also be a pain as threads are not always perfectly formed and it used to be common practice to 'burnish' threads on new nuts and bolts by tightening them a few times before final torque.

For anything remotely critical I would tend to look in the later 911 Manuals and use the Torque + Angle figures that replace the simple torque figures in the early manuals as this should result in a more consistent preload which must be a good thing.

Trying to do better with simple torque control methods is not really feasible.

It has been well established that about 90% of the torque applied doesn't preload the bolt and this has led to developments of 'Nut Factors' commonly used in the Aero Industry.

All of the factors which affect the load actually applied should have been taken into account when the manual was produced should mean we don't have to worry about these issues.

We should, however, strive to use consistent working practices in terms of thread cleanliness, freedom from damage and a strategy to deal with new threads.

I would only clean threads with a good quality lint free cloth and wouldn't de-grease chemically.

Cleaning under heads of bolts that turn is also important as either oil or contamination in this area can have a significant impact on tightness.

If you can use stretch gauges on a specific application then this would always be my preferred approach as it eliminates all dodgy variables.

[reddogmotrsprts]

I used to use anti-seize with a torque wrench until I read that certain anti-seize compounds could result in an over-torque of roughly 50%. Since then, anti-seize has a dramatically reduced role in my garage.

I've taken to replacing a lot of fasteners with stainless versions, especially on things like exhaust hangers and such -- a huge win here in the northeast. While I'd like to use stainless (or higher grade steel in some cases) more frequently, I don't know how the stretch compares between these materials and stock, so I've resisted using these in any critical role.

I've enjoyed a bunch of Carroll Smith's books, including "Screw to Win", but it's been awhile since I crammed it. Time to review, clearly! It really is a fastenating book and I strongly recommend it -- especially the early material which reviews the chemical and physical properties of metals.

[winders]

reddog,

What fasteners are you replacing with stainless and are you making sure that you are using a grade of stainless equivalent to what you are replacing?

[reddogmotrsprts]

Originally Posted by winders

What fasteners are you replacing with stainless and are you making sure that you are using a grade of stainless equivalent to what you are replacing?

That's my whole point; I'm replacing stuff where the grade doesn't matter -- think trim and stuff I've fabricated where I had to go off-book -- but I'd like to have the confidence to replace stuff where it does. I'm generally using 304 or 316.

[Danimal16]

Originally Posted by reddogmotrsprts

I used to use anti-seize with a torque wrench until I read that certain anti-seize compounds could result in an over-torque of roughly 50%. Since then, anti-seize has a dramatically reduced role in my garage.

I've taken to replacing a lot of fasteners with stainless versions, especially on things like exhaust hangers and such -- a huge win here in the northeast. While I'd like to use stainless (or higher grade steel in some cases) more frequently, I don't know how the stretch compares between these materials and stock, so I've resisted using these in any critical role.

I've enjoyed a bunch of Carroll Smith's books, including "Screw to Win", but it's been awhile since I crammed it. Time to review, clearly! It really is a fastenating book and I strongly recommend it -- especially the early material which reviews the chemical and physical properties of metals.

The only problem with stainless steel is they are susceptible to brittle failures. Of course it depends on the alloy. Nothing wrong with anti-seize, as long as you understand its impact on the torque value.

[Danimal16]

Originally Posted by reddogmotrsprts

That's my whole point; I'm replacing stuff where the grade doesn't matter -- think trim and stuff I've fabricated where I had to go off-book -- but I'd like to have the confidence to replace stuff where it does. I'm generally using 304 or 316.

Yep, I wouldn't shy away from those applications. No heat cycles there.

[reddogmotrsprts]

Dan, I apologize if I misinterpreted your comment as sarcasm. Reading it again, it sure seems that I read you wrong.

I just went back to the book to answer some of these questions, but unfortunately, "Screw to Win" makes only one recommendation with regard to stainless steel fasteners -- don't use them. Smith discards all stainless fasteners as worthless due to their reduced ultimate tensile strength versus standard steel.

When researching the yield strength of stainless, I found that the tensile strength of 304 is about 15% less than that of grade 8.8 steel, while 316 is virtually identical. This means that 316 can be swapped for 8.8 (and 304 substituted as an upgrade for lower-grade steel fasteners, such as the sort I found in an RX-7's suspension over the weekend).

With that said, let us consider this dangerous experiment conducted this evening to replace some bolts. Realize that this is an academic exercise, so if you're not interested, feel free to ignore:

I need to replace some non-metric motor mount bolts (thanks, Instant-G) with some M12 bolts in order to run WEVO mounts. For reference, the torque spec for stock motor mount bolts (grade 8.8, I believe, I don't actually have one) is 65 ft-lbs.

Burnishing a 304 fastener and torquing to 40 ft-lbs, which is the torque spec indicated by WEVO, cold-welded the bolt to the nut. I next used anti-seize and torqued to 40 ft-lbs. -- perhaps putting me above the elastic limit of the fastener, let alone the recommendation from WEVO.

This may be a textbook scenario in which the hardness of stainless could be dangerous. In fact, I found recommendations to use stainless only in simple clamping and never in mechanically-loaded applications. The physics of which forces are applied to this bolt are way beyond my ken.

So the question is, what will cause a failure of the bolt first?

Also, Dan raised the subject of heat. What role might heat play? I can't find anything specific to stainless steel here, while on the other hand, I found suggestions to use stainless in exhaust manifold fasteners.

Thanks for your thoughts.

[chris_seven]

Stainless Steel and fasteners is not really that complex.

There are several commonly available grades and these are typically sold as A2, A2-70 and A4-80.
They are equivalent to what you would expect from a normal High Tensile Fastener.

A2 and A2-70 will always be stronger than commercial bolts which tend to be around Grade 4 and their use would be reasonable in most non-critical circumstances.

A4-80 is very similar to a Grade 8.8 and could be considered a replacement.

The other issue, as you have discovered is the ability of stainless steel to 'gall' when being tightened.

This means that it important to use an anti-seize compound and this will impact on the torque versus preload of the fastener and hence the behaviour of the joint.

The figures provided by the Workshop Manual deal with most of the variables but if you switch to Stainless the 'nut factor' or K value will change and you will have less certainty that all is well.


The 'stretch' of individual fasteners will change little as the difference in Young's Modulus between Austenitic Stainless Steel and a typical HT Bolt steel is around 2.5% - the stainless having a slightly lower modulus.

The stretch being measured would normally be 80% of yield which is around 0.15% elongation so 2.5% of this figure is really difficult to measure in a normal environment.

Most stainless steels are fairly heat resistant compared to conventional steels ate least at the temperatures that cars operate but the expansion is generally higher so this may cause some loss of preload depending again on the joint.

If the performance of the joint is critical then I wouldn't tend to replace fastener type.

It is generally unwise to consider fasteners in isolation to the parts they are holding together.

[reddogmotrsprts]

Thanks for your comments; they jive with my reading and I would only add (for others considering substitution) that titanium has a similar galling issue.

You can probably tell that my experiment was more of an attempt to provoke discussion than a scientific exercise, but still I'm curious if there are best practices for torquing fasteners 'wet'. I've read that (dry) zinc-plated steel offers relatively inconsistent friction, so I wonder if wet torquing with angle measurement couldn't provide a more consistent result. Is there a better method? I expect that I can confirm this with experimentation on a case-by-case basis, but can accurate assumptions be made? And if accurate, how precise are they?

We might not need this extra consistency since we often just need to land in a sweet spot of elastic tension, but as we go up in grade, the size of the sweet spot shrinks. Again, an academic question that might help inform a mental image of how this works...

[82 SC]

Make sure you get your torque wrench calibrated. Each Wrench has a tolerance that it can do. SO 100 ft pounds of torque could be + or - a certain percentage. With a crappy wrench you could actually get 95 lbs or 105 lbs and be in tolerance with that wrench. Also the accuracy will very poor at the lower end of the scale. Don't use a wrench in the first 20% or so of its scale. As in you have a 100 lb pound wrench it will be very inaccurate at the lower settings as much as 30% or more. Best to check the specs on your wrench. This reason is why you buy a lower tolerance wrench to do it at the correct levels. Just because your Hazet wench has measurements from 10 ft lbs to 250 ft pounds you should not use it at 10 ft pounds. Have not had my coffee this morning so my percentages might be off but you can get the jest I hope. OH I work in a calibrations lab

[chris_seven]

As I suggested in an earlier post many 911 Torque values use torque and angle as a tightening strategy in later editions of the Workshop manual.

Plating of all types has a significant impact on nut factors.

Galling of Titanium varies greatly with the grade and I have found that Grade 5 isn't as bad a A4-80 Stainless Steel.

The impact of galling can be reduced by using dissimilar metals but you will always come back to the fact that making changes to fasteners can introduce uncertainty of preload.

The best way to look at it is to realise that if you consistently lubricate 10 fasteners and torque tighten them as carefully as possible and measure preload against an accurate loadcell you will see a variation of around 30%.

This would be a good result and assumes calibrated, high quality torque wrenches and skilled technicians.

If you changed to torque and angle this may drop to 20% - a big improvement.

If you now start to mess with plating and lubrication methods you just have a huge potential for error.


If you use CAD plated aircraft bolts that are very controlled you will have better consistency as the bolts will be batch tested prior to release but commercial fasteners are just all over the place.

Fortunately not many bolts on a car are critical so we generally get away with relatively poor practice.

The best way forward would be to buy an Ultrasonic Stretch Gauge but I think they are still $3-5000 each.