Bolt connection simulation

[jkelindberg]

Hi...

I don't understand how to use the bolt connections in simulations. I have tried to search for some documentation but come up empty-handed. The help button in the bolt connection window takes you to the "Simulation, Contacts" help page which does not contain any information about the bolt connections.

So that's enough whining for now, maybe anyone of you on the forum have had better luck with setting this up to work properly?

I made a very simple model; two blocks, a clearance hole in one and threads in the other. Then turning to simulation and add a M16 bolt connection with 60kN pretensioning. I don't want the block to "merge" so I add a separation contact condition on the connecting surfaces with 0.15 static friction coeff. Lock all DoF on one of the surfaces of one block and apply a slight pressure on one surface of the other block.

The first thing I notice is that the pre-check complains about a possible conflict between the separation contact and the bolt connection. So that's one clue right there that something is amiss...

The simulation will run anyway and you'd think if the pressure was low enough not to overcome the friction caused by the pretension of the bolt you would not see a displacement between the blocks. But in the simulation you will.

This is the displacement plot:

And here is a section showing the contact pressure due to the pretension:

Obviously, this is not an accurate simulation of real world bolt joint.

So, anyone got any ideas what I'm doing wrong?

Thanks!

//jl.

[whunter]

I've not tried this in F360, but in other software you'll have to do a non-linear analysis (really just steps of linear analyses).

First you'll do the pretension, then once everything has 'settled', you'll apply the pressure on the side as a separate and new load case to see if it slips or not. I suspect that if you do the analysis in one go, the static linear solver (assuming that's what is used) will apply the slight pressure and pretension simultaneously, so you'll get odd results since the friction doesn't get time to "build up".

Regardless of my guesswork above, the two surfaces shouldn't slip if there's enough pretension.

[jkelindberg]

Hi whunter thanks for your reply.

I can't imagine this was the design intent when developing a simulation tool specifically for bolt connections.

Your workaround may work, but using this approach would render the bolt connection tool irrelevant (since you would need to model the bolt in 3D (F360 will not allow you to make 1D beam elements, as far as I know), and it would also mean you have to approximate the contact pressure over an area instead of using the nodal values. And it may be OK for doing two blocks bolted together. But I want to make a simulation on a complete assembly and it would mean a lot of work...

I still believe there must be something I have missed, I don't believe Autodesk would push this to end users without a) documentation and b) having tested it thoroughly.

//jl.

[AndrewSears]

@jkelindberg, can you attach your model? 

Thanks,

Andy

[jkelindberg]

Absolutely Andy. File is attached.

[todd_alford]

@jkelindberg, Thanks for pointing out the incorrect link.  The connectors are mentioned in the how to section under constraints on this page in the help.  That said the information is not as in-depth as I would like it to be, thank you for pointing this out.  I am on the documentation team and will work to improve this area. 

Thanks,

Todd Alford

[AndrewSears]

@jkelindberg, I did not forget about you.  I am still collecting data about your setup, our bolt idealizations and creating comparison examples.  

Thanks,

Andy 

[j.hancock01]

Hi All,

Is there any update to this post as I think the answers would be most useful to other users (myself included)?

I've just run a similar trial to the one described, this has two 200x200x90mm thk plates with a 48mm diameter hole through them both. I've constrained the back face of one of them in all degrees-of-freedom and the contact face between the plates with a separation contact with a coefficient of friction set to 0.3. I applied a 45mm diameter bolt with a pre-load of 279477 through the holes in the plates using the Fusion simulation, constraint, bolted connection option. I then applied a vertical load on the outer plate of 40000N. See below a screenshot of the set up:

With a friction of 0.3, the pre-load should ensure the joint does not slip for a load less than 279477*0.3 = 83843N, so no sliding expected in this case.

With this set up there was a displacement of 2.7mm, so the joint slid:

I wondered if the default value of the "Friction stiffness for stick" was too low, so increased this to 4E5 N/mm (see below), assuming this would reduce displacements in the "friction locked" case to something like 40000/4E5 = 0.1mm

Indeed this did reduce the displacements as expected:

I then thought I'd apply a load that should cause the joint to slip, I chose 200000N, which should be over twice the value that can be resisted by friction. This lead to displacements of only 0.7mm, see below:

I wondered if the stiffness of the bolt was increasing the pre-load of the joint or doing something else, so took the bolt out of the analysis and replaced this with a two constant forces equal to the pre-load applied to the edges of the bolt holes:

I then cranked up the lateral load to see if the joint would slide under a lateral load of 600000N, which is twice the pre-load and over seven times the expected frictional resistance. I didn't actually expect it to solve as the outer block is only restrained by friction so should "fly off". However, the run did solve and gave a minimal displacement of 0.33mm.

Any thoughts on how I can get the friction in this type of contact to work correctly would be much appreciated.

Best wishes

Jon

[Anonymous]

I'm also really interested in the bolted connection (or constraint), so it would be great if any issues in this feature get resolved.

[Anonymous]

Any news here? I also get the same problem with the "pre-check complains about a possible conflict between the separation contact and the bolt connection. "

[tinkercadP2U5F]

also very interested in this