Summary evaluating the learning-and-experience journey made in this initial FEA endeavour

Here we are, looking back at 47 days of progress to 15June2016 - about 1~1/2 months.
The first two weeks (14days) of it while working full-time as a welder-fabricator in a steel fabrications company...

Moving beyond here - which I have done already - there's use of automeshers. I've already used a specific-purpose geometry-and-meshing tool to model a cylindrical pressure vessel with a nozzle.
That's beyond the scope of this initial programme; so we close it off and summarise...

All these FEA simulations were manually meshed.
For which the stages are

For the learning value of building an efficient model.

Also for the pragmatic reason that the free "preview" version of the FEA program I downloaded and used has a restriction of a very limiting maximum node-count. Which concentrated the mind on efficient expression of the physical situation as a FEA model. A good discipline in the learning-exercise phase.

Note that the "restricted maximum number of nodes" has often prevented me from determining whether the solution seems to be fully "converged".
So repeating the caveat about don't take any of these simulations in this programme of work as a "proof".

What's been achieved?

I'll let you be the judge of that for yourself.

For me - it looks good. For example - I did not expect the initial programme would include modelling parts of the bridge I worked on in 2015 - particularly the "orthotropic bridge deck" simulation. Where as an engineer I was very much involved in advising and guiding on good weld designs. Having even this level of modelling ability then would have enabled me to even more deeply engage with the designers.

There's been a goodly range of line-element structures (eg the trusses), shell-element structures (eg the orthotropic bridge deck) and area-element / volume-element simulations (eg the weld-shape and crack-tip simulations)

Long-standing curiosities have been satisfied - eg the "Stress profile visualisation in a centrally-loaded cylinder".

With a general scientific background, I found the Finite Element Analysis computer-program readily usable.
Sometimes deducing how you use the program, or deducing what the program is doing, required stepping back to very simple models which enabled particular aspects of building a model to be examined. Then stepping forward to resume the full model, on resolving the point being examined.

Already I know that my provisional objective has been more than achieved. As a welding engineer, I wanted a means to express ideas and explain my specialist welding insights into what might be advantageous design ideas in the form of FEA models.
I benefit from the limitation of responsibility that such models would have as a way of conveying ideas. Where the structural designer engineers would critically evaluate the ideas for correctness and usefulness.
In that "ring-fenced" limited context, of a welding specialist contributing to a structural project - I'm already there :-)

What lies ahead?

The theory of Finite Element Analysis beckons. The mathematics, engineering and materials science (which to some extent I already know, with a Materials / Metallurgy background).
I've already had a look at a book on FEA which I got those weeks ago. Will it prove enticing? Will the theory become a nice place my mind chooses to dive back into where it last left-off, as a relaxational activity?
Maybe, maybe-not...
That I like what FEA does gives the potential endeavour a sweeter set of associated sentiments.

New projects seem likely to be much more resource-intensive, use CAD (Computer-Aided Design) tools to generate complex detailed drawn-objects, then powerful expensive FEA programs with "automeshers" which work on the CAD drawn objects.

Whatever... Happy days :-)



(R. Smith, 16Jun2016)