whit3rd writes:

On Sunday, November 15, 2020 at 2:14:47 PM UTC-8, Richard Smith wrote:
Is "polish and coat with a hard brittle varnish" the answer to this
question?

I've made a "square-cornered U" of RHS100x50by8thk (4"x2"by5/16"thk.).
45degree mitres full-pen'd for the two corners.
...
I'd like to put a hydraulic cylinder / hydraulic jack toward the open
end of the parallel sections forming the "U", and see what happens
when it's increasingly loaded.

Precautionary note - one assumes it would be a good idea to lash the
hydraulic cylinder with rope to constrain it if it were elastically
ejected from its original position...

If things go well and forces pass those necessary to cause yielding -
how would one reveal where that plastic yielding is happening?
I've heard of polish surfaces smooth and apply a hard brittle
lacquer...

I'd think, in testing-machine fashion, you'd want to step the pressure (good
hydraulic gage required for that) while watching a machinist's dial gage on the
deflection. You know what direction it moves... no need for a two-D grid,
and buckling would require 3-D before-and-afters.
Do multiple pressure trials, relieving pressure and looking for permanent deflection.

I'm not sure I could get full penetration on those internal corners, but maybe cut-a-slot/insert-a-wedge
in two or three places could make a U with continuous inner metal, all the welds being under some
amount of compression. You'd want to hot-work the bends, of course.

Thanks for responses. Things moved on a long way quickly.

This is the write up about what I was then in the early stages of...
http://www.weldsmith.co.uk/tech/stru...lyse_test.html

This one gets testing conditions directed at the test weld
http://www.weldsmith.co.uk/tech/stru...beam_test.html
The Lu..ders Bands spall the mill scale - "witnesses" the deformation
penetrated around the weld.
Found what I wanted...
Rich S