1
$\begingroup$

Is there any well-established benchmark to test the validity of a finite element analysis, i.e. FEA? For example, a 3D model with specific loading and boundary conditions. Along with the expected results for stress or strain.

Scientific journal papers could use such a benchmark for validation. Just wondering if there is such a standard benchmark.

UPDATE 1

Maybe one option is a simple 3D structure with a known structural response like this one taken from that reference:

Simple 3D beam

UPDATE 2

As @TomášLétal pointed out, probably a set of verification tests are needed to cover a range of aspects.

UPDATE 3

According to comments and posts, these are the options:

  • Simple 3D structure with known structural response.
  • Doing a modal analysis, as they apparently do in the automotive industry.
  • Compare against real experimental results.
  • ANSYS or ABAQUS Verification Manual.
  • ...

UPDATE 4

There are simply-supported structures along with their hand calculation analysis here: https://github.com/calculix/CalculiX-Examples/tree/master/NonLinear/Sections

They could possibly be verified benchmarks.

Improve this question
$\endgroup$
2
  • 1
    $\begingroup$ You cannot check FEA as a whole in one test. Usual way would be checking some aspects against known analytical solutions in simple cases. Check out ANSYS Verification Manual. $\endgroup$
    – Tomáš Létal
    Jun 29 at 16:13
  • $\begingroup$ @TomášLétal Thanks, it makes sense :) I'm more of an ABAQUS or CCX guy. I'm going to check if they have any Verification Manual. $\endgroup$
    – Megidd
    Jun 29 at 16:17

2 Answers 2

Reset to default
3
$\begingroup$

Yes, often against real experimental results.

We did a lab where we used theory to predict the bending under load of an aluminium beam (1m long, 7cm thick and 25cm deep), comapring the results to the values obtained from actually testing such a specimen on a specialised test rig (bought for the University by a large company so we could do experimental testing for them... students doing final year projects are a lot cheaper than employing people with phd's :) ).

Many papers do that type of comparison and especially so to prove the validity of a proposed model before using it to predict the behavior in a particular situation.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Can you share a few samples? Those might be references for me. $\endgroup$
    – Megidd
    Jun 29 at 16:09
2
$\begingroup$

Static deflection is a tough thing to measure in the real world, as you often have to model the test rig. Your example of a block on 8 simple supports is a good example of where the test rig would interfere - you haven't got rollers under the supports so you are going to see some odd effects, and you'll also get odd effects from the point loads.

In the automotive world we usually correlate an FEA model by doing a modal analysis.

Improve this answer
$\endgroup$
1
  • $\begingroup$ I didn't consider the option of a modal analysis before. Thanks. $\endgroup$
    – Megidd
    Jun 30 at 6:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.