Align or Mate on Helix Surface

[barrykingwill]

I am trying to mate the surfaces of an extruded thread on a bolt and nut, such that the helix surfaces are mated.

I need this as as ti import in into a FEA program, and perform an analysis on tightening the nut, using surface to surface contact, such that I can measure the tension, shear vs torque applied.

I have modelled the true thread of both parts and have them in an Assembly, but cannot find a way to mate the surfaces such that the nut is in the groove of the bolt.

See attachment

 

[HaroldL]

There is no "Helical" mate, I don't thing even SolidWorks can do that. What you might try is to make the threads as a pattern of circular cuts, v-groove with the thread profile then linear patterned. Then you may be able to use a Mate from the surface of the groove of the nut to the groove on the bolt. Not exactly "real world perfect" but it may be close enough to get you an idea of the tension, etc.

H.
v12.0

 

[barrykingwill]

Thanks H

But circular grooves will not result in tension come analysis time.

I want to run a non linear surface to surface contact (with friction) such that a prescribed rotation on the nut, will produce the resultant tension and torsional shear in the bolt. It needs to be helical for that to work, but the nut also needs to be perfectly seated on the bolt at the outset, else the surface to surface contact will not work in the FEA side.

Another way, might be to cut the thread extrusions in such a way that the nut located at a specific place, in terms of height and rotation matches the cut extrusion of the bolt. ie Extrude them in their final positions and identically......and then just anchor both and not mate anything.

The problem with this I found was the lead in/lead out of the thread....but might have to have another stab at it if no choice.

Was hoping it would helically mate.

Thanks

 

[DavidJ]

I've tackled this slightly differently in the past.

I modelled a 'bolt' with a head - aligned underside face of head to be BELOW mating surface by an amount equal expected stretch in bolt - defined joint between the non-head end of bolt and the hole it is in as welded contact. Tell FEA package that there is an interference fit between bolt head and mating surface.

Not as sophisticated as what you are trying to do, but easy to model and gives reasonable results. Different FEA packages treat contacts very differently - some have tools to help, others need very significnat workround (e.g. using false temperature differences to expand or contract the bolt).

 

[Hop]

Is it realistic to expect a simple solid modeling program like Alibre to have the chutzpah to also model the mating of screw threads? Heck, it can’t even model mating involute gear teeth!

As we all know, a screw thread is just an inclined plane wrapped around a cylinder. To expect two of these to mate perfectly over their length of engagement is naïve at best. So, given that there will be an interference fit between the bolt and the nut under tensile load in the bolt, the compressive forces at the interface will vary greatly depending on pitch variation, “bumpiness” of the profile, and friction between the two surfaces. A decent FEA program should include a “tool box” that allows the user to specify this geometry explicitly. I would not expect Alibre, or any other solid modeling program, to specify exactly how two mating threaded parts actually mated. In my opinion, that should be the job of the FEA software package’s mesh generator.

Plus, in Alibre, there is no way provided to specify the lead stack-up tolerance on helical mating threads, which will surely affect the tension and torque on the bolt. I am purchasing some ceramic ¼-20 threaded rods for use as high temperature electrical insulators in a vacuum arc system. The rod manufacturer cautions me not to expect more than five to ten turns of thread engagement before the lead stack-up gets large enough to crack the rod. Should I expect my Alibre model to include this factoid?

That’s just my 2¢. :| I wish someone would offer an inexpensive version of the COMSOL Multiphysics FEA modeling software. Anyone at Sourceforge listening? There is a free Linux/Unix package, SALOME, available for downloading here: http://salome-platform.org/ Unfortunately, I don’t do Linux yet, although the new Ubuntu release is tempting.

For more clues on free FEA modeling read the attached PDF file.

 

[DavidJ]

Picking up on Hop's comments - the closer you look at a problem, the more complex you realise the reality is (also tends to be more variable, and less certain).

Very often the best you can hope for is a reasonable approximation that gives a sensible value for the stress (or whatever you are looking at). You might be able to carry out several analyses to attempt to define reasonable upper/lower limits. You still have to remember that no matter how sophisticated, it is only an approximation (that's what 'design factors' or 'safety factors' are to allow for).

Using tools like FEA might allow a better handle on stress than a simpler approach - allowing some reduction of design factor. Sometimes FEA might be used to study specific aspects of a mechanism to help reveal which factors are most important, and which aspects are less critcial.

It sounds like the OP is trying something which is very ambitious even assuming 'perfect' threads and uniform coefficient of friction - the real case of course will be much more complex.

Even today many such problems are often handled by established 'rules of thumb', or by building a test rig and measuring what happens (e.g. you can get a reasonable direct measure of tension by measuring bolt extension - often using ultrasound).

 

[barrykingwill]

I have not yet looked much further on this as I have been tied up with other projects.....but at the end of the day it is just Maths (or Math, depending on where you come from), and yes there would be some assumptions like linear coefficient of friction, but those are parameters you can play around with once you have stable solution.

A lot more depends on how well the different FEA packages handle surface to surface contact with friction (non linear, dynamic)......more so that just getting the geometry right.

Will keep u posted if I get to wrap it up. Have other stuff on the go too, which is taking a bit of priority at the moment.

Thanks All