Composite Panel Buckling Load 1

[midsidenode]

Can someone help me here? I may be having a brain fart. Please take a look at the enclosed figure involving buckling of a panel in pure compression with 4 simply supported sides. How is the enclosed Figure obtained? My plot of the allowable buckling load vs the aspect ratio looks nothing like this but this is indeed what a typical buckling load chart looks like. For example, for an aspect ratio of 1, the allowable buckling load is obviously .161 N/mm for this case. What am I doing wrong here? Just assume m=1.

 

[midsidenode]

In other words, when I plot an allowable buckling load vs the aspect ratio using the buckling load equation, I get a curve that looks like the enclosed curve. How in the he!! is this not correct?

 

[ESPcomposites]

For each a/b ratio, you need you solve for m=1,2,3...and take the minimum value. For lower a/b, m=1 is usually the critical value, but as a/b increases it will go to m=2 or greater. You should get a solution like the attached (for uni-axially loading). For shear loading, it is a smooth curve and does not have a "wave-like" result.

Brian

http://www.espcomposites.com

 

[midsidenode]

Brian,
With the equation that you provided, keep the dimension of the plate parallel to the load at a=.5 meters and vary the plate width from b=5 meters to b=.333 meters. Just keep m=1 half waves. Then plot Nx(crit) vs a/b and see that you get the same curve that is enclosed. Why doesn't this look like a typical buckling curve? In other words, as a/b approaches zero, Nxcrit should go to infinity. Does Nx(crit) need to be normalized as stress before it's converted to running load (N/m)?

v/r,
msn

 

[ESPcomposites]

Perhaps your order of operations is incorrect in Excel (it can sometimes be tricky and difficult to spot). Here is my sample problem. I did not double check it, but it looks as expected.

Brian

http://www.espcomposites.com

 

[midsidenode]

I guess that is an interesting thing. To get these curves correct, one must ensure that the plate width (b) is held constant and the plate length (a) is changed. Thanks Brian.

 

[ESPcomposites]

Yeah, while the form I posted can be condensed as you showed, it may be less intuitive if you do so (potentially causing some confusion as to how to generate the curve). Also, the form I posted is more typically presented in publications and more aligned with the solution for metals.