Best technique for incorporating screws into a structure for a basic stress analysis

[SwimBikeRun4342]

Hi Everyone:

I'm planning on doing a relatively basic stress analysis on some structural components. These components are going to be quickly used for both experiments and demonstration under heavy loading and impact conditions, so accuracy is very important.

One region in the structure concerning me is the areas with screws. At a local level (ie, the screw and it's surrounding region), how does one go about simulating the strength of the screws embedded in a structure, on Abaqus? At a global level, I plan to just construct ideal bolts (by imposing point-based constraint) onto the structure, in place of the actual screws - is this generally okay?

I worry that my global analysis (described above) is not sufficient. Hence I want to do a local analysis of the actual screw embedded in the structure. Is this common on Abaqus? It seems that constructing a contact model with a screw (geometrically, pretty complex) would be totally inefficient (taking a long time to both model and the run the simulation).

Are there any common methods for this "local" screw analysis I'm describing. Is it generally assumed that a rigorous screw/structure simulation is not worth the time? If so, are there any short-hand, approximate techniques for modelling screws that are common practice? Do you have any suggestions or references?


In summary, my questions are:
1) What are common Abaqus FEA procedures to simulate the strength of structures with many screws? Are the screws literally modeled in the simulation (ie, deformable bodies with complex geometry inbedded into a separate deformable body), or are they replaced with ideal constraints (and thus not included in the model)? The latter seems MUCH more practical.
2) Is there any clear and basic literature for FEA modeling of screws in a structure? I would think this is an extremely common design problem.
3) What assumptions can be used when modeling structures with many screws? This is more of a structural engineering question rather than a Abaqus question, but I'll go ahead and ask it. Is my "global" method of replacing the screws with ideal point-based Abaqus constraints accurate enough?


Thanks!

 

[sobe8]

Bolted joints (or "screws") can be modeled with Abaqus connector elements (basically springs that connect the things being fastened together) with stiffnesses equivalent to the actual joint. There is a lot of literature out there on how to calculate these stiffnesses. After you run your analysis, you can see what sort of forces this connector element sees to make sure the fastener you choose is strong enough for that joint.