static analysis, constraints

[G.Kowalski_CBKK]

Hello everyone, 🙂

I have two sheet elements connected by a bolted joint, what connection should i use? in static analysis

Sliding / No separation?
or
Separation / Non-sliding?

Second question and my request.
Could someone write to the following bindings (static analysis), typical use.

for example:

bound = welded elements

Sliding / No Separation = 

Separation / No Sliding =

Shrink Fit / Sliding = 

Shrink Fit / No Sliding = 

Spring = 

Thank you for your answer.

Greetings.

[swalton · Accepted solution]

Have you used the calculation tool in the Bolted Connection app? 

If you want to use the built-in Inventor FEA, try this: https://forums.autodesk.com/t5/inventor-forum/fea-on-bolted-structures/m-p/2799892/highlight/true#M3...

If you don't have confidence in the approximation above, use Inventor Nastran.  See https://knowledge.autodesk.com/support/fusion-360/learn-explore/caas/sfdcarticles/sfdcarticles/How-b...

[Cris-Ideas]

You question is a "tip of an iceberg" and as simple as it sounds as difficult to answer it is.

What contact type you use depends on what you need to calculate and what results/parts are of interest and how much time you want to spend or waist on the analysis.

It all depends.

If you are hugely interested in joint itself you should use contact that is most realistic. But will result in longer or, depending on your particular model, very long solve times.

In general most general answer I can give to your question is:

Use the contact that allow movement of parts you expect to occur in practise. Most likely they can separate but there is rather low probability of sliding "use separation no sliding", if they can slide but definitely will not separate use "sliding no separation", if you really have no clue you can use "separation" but remember it  does not account for friction,

if you have no time and just want a general overview of what is going on use "bonded".

And be mindful of the "issues" with contact detection and " development" (changes during solve). So always examine deformation of your model.

So It all depends.

[G.Kowalski_CBKK]

Hello, thank you for your reply. Unfortunately I don't have Inventor Nastran. Greetings.

[G.Kowalski_CBKK]

Hello, you are right about the extended analysis time (I'm still waiting for the results 😉 ).
Thank you for your answer.
Greetings

[Cris-Ideas]

Show you model. Perhaps you have run on to other known problems.

[G.Kowalski_CBKK]

Hello,

Here is my model of stairs. The structure is attached to the wall on one side, and the other part of the structure rests on the floor.

I am interested in the pressure exerted by structures on the floor and the overall deformation of the structure, with a normal load of 100 kg / m2. I used S235JR steel. 

Thank you in advance four your help.

[Cris-Ideas]

What is your inventor version? If max 2021 can you post the model.

If you are only interested in values indicated above I would suggest to go with "bonded" contact for all.

If you would like to check local areas around this top bolts you need to use other type of contacts, depending on the type of joint you design. Also have in mind that separation contacts tend to produce errors if parts are not in contact initially as automatic closing of the haps does not work properly in these cases.

So it is accessary to model the structure with no clearances (this brings other problems) or add extra parts to mitigate this behaviour.

Also you did not mentioned what are your boundary conditions (supports). Please give some data over that.

If I was doing that I would most likely put full model (with steps) in to FEA and make some adaptations just for FEA purposes.

This would allow me to:

• apply loads as they are in real world
• quickly check key connections

and it would only come at a cost of only a little longer calculation times.

[swalton]

I usually try to start with simple FEAs, then add complexity. 

In fact, I'll often start with a 2D or 3D hand sketched Free Body Diagram.  The FBD helps me think about the expected loads, the expected reaction forces at the supports, and the type of supports and connections that best match the physics of my problem.

I might run a study with just bonded connections to see that I am getting the expected reaction forces.  Once the model matches the FBD, I'll add complexity.  It might be changing the supports at specific locations, or changing some bonded connections to seperation, or better defining a bolt.