8 replies to this topic

[MatsNorway]

I have been doing som calculations on threads for an welded and machined assembly. And the book states that μ (friction) when dry can be between 0,18 to 0,35..

 

I then did my headscratching and got all the formulas down, did the numbers, found and fixed some errors and summed it up. And i still managed to get some very big and not likely numbers.

 

And after a hunch i check my other book containing a much more comprehensible table with friction for all the different materials and instances. And that book states steel on steel friction when dry to be 0,1...

 

Is there something i have missed? I also believe there is even better tables with the surface finish coming into play. But anyway, the difference between these two books is massive.

 

I once had a online calculator linked to me by someone here. But i lost it. So im hoping something like that comes up again.

 

 

If your really interested here is the info.

 

M24x3

Force is 15000N

calculating without lubrication

 

Head friction/plane surface not included (im doing the other part in the assembly)

 

 

 

[saudoso]

Most likely it dropped from 0.18/0.35 to 0.1 after we lost PII to China

[Greg Locock]

http://www.roymech.c...co_of_frict.htm

 

there's a list for various threads/materials

 

Needless to say the correct answer is to get your torque wrench and load cell and do the experiment, if it is that critical.

[MatsNorway]

I thought i deleted the thread the same day after work...

 

Anyway, my math all checks out it seems. Had a minus one place (losening bolt) where it should have been + (tightening). I was suspecting it too, otherwise the formulas was all correct it seems. I just managed to punch something wrong into the calc repeatedly leaving my doubting my choice of formulas or thinking i had missed something else.

 

Done right i got a nice low number as we hoped for. Going to keep 0.35 as it is conservative and the total force was only 120N/mm2

 

Im still confused about the total lack of writing about the needed height/amount of threads.. I need to check more tomorrow, it has to be in that book somewhere.

Edited by MatsNorway, 23 April 2015 - 19:38.

[desmo]

Isn't directly measuring the fastener's stretch preferable to a torque spec anyway?

[Lee Nicolle]

Isn't directly measuring the fastener's stretch preferable to a torque spec anyway?

In theory yes. Many pro engine builders use stretch on conrod bolts though you have to spot face both ends. Screwing into parent material makes stretch a very iffy thing. 

I have had various specs from various suppliers re tensions. Some recomend standard  tensions with simple oiled threads, some recomend use moly lube and less tension, other standard tensions.

Using a stud locker [Loctite etc] does alter tensions too. Nominally a lubricant but not as good as oil to me. For bolts/ studs that work loose it is generally a good idea. I do not use loctite on any internal bolt, but flywheel bolts, accesory brackets and the like it is generally a good idea.

 

With tensions too, use a lubricant under the bolt heads, lots of friction there too. And if using washers [hard ground]  lube both sides of the washer. Clamping will be better.

[bigleagueslider]

Mu for clean dry metal-metal contact can vary greatly depending on surface finish or whether it's static or sliding conditions. With threaded fasteners, if you are using wrenching torque to tighten the bolt/nut, then it is common practice to first measure any prevailing torque in the threads. And then subtract that prevailing torque from the applied torque.

 

Due to the large difference between static and sliding friction coefficients (ie. 0.30 vs 0.15) you will see a difference between the amount of wrenching torque it takes to initially loosen a bolt/nut vs the final wrenching torque used to install it.

[MatsNorway]

The challenge here is to find the material needed as that is lacking in the original documents. Normally you would just go for 42CrNiMo (whatever it is called) and call it a day. But the design is welded so even 8.8 like qualities are unsuited i was told. So i had to go a bit more down in detail. When i have done my work a Civ ing at materials is supposed to take over and spesify the needed materials, welding info++ for the design.

[bigleagueslider]

The challenge here is to find the material needed as that is lacking in the original documents. Normally you would just go for 42CrNiMo (whatever it is called) and call it a day. But the design is welded so even 8.8 like qualities are unsuited i was told. So i had to go a bit more down in detail. When i have done my work a Civ ing at materials is supposed to take over and spesify the needed materials, welding info++ for the design.

42CrNiMo is the same as AISI 4340 alloy steel. This material can be heat treated to fairly high tensile strength (>220ksi) and is used for highly stressed structural components, but it is not commonly used as a threaded fastener material.

 

You also mentioned something about the design being "welded". Can you provide more details about this particular fastener application?

 

If you want a good technical refernce for threaded fastener analysis as it is approached in the aeropsace industry, get a copy of MIL-HDBK-60.