Buckling calculation/formula for parallel flanged channel

[AdamMech]

I am in need of help regarding 'Buckling' calculations/formula to determine the size of parallel flanged channel needed. I have been searching but i find different formulas and deciding which is right is hard.

The load/force exerting upon the channel is on the 'C' or 'U' face, a pressure vessel will have support legs consisting of these channels so i need help in calculating buckling. Also if the size of channel, for example, 100x75x8x8, the calculation if the force exerted on the 75mm length face (if the channel is comprised to make a skid with equipment on).

I know of Eulers formula for this calculation but need help from scratch really.

 

[CELinOttawa]

Sketch?

 

[AdamMech]

I am very new to this forum, how do I sketch? When I click on the pencil symbol it comes up with some text.

Help?

 

[rb1957]

draw something in word and attach the doc.

it sounds to me as though the channel is supporting the long side of the cylinder with one flange attached to the pressure vessel and the other to the ground ? so the web is in compression with possibly an offset load

Quando Omni Flunkus Moritati

 

[desertfox]

hi

I think I can see this without need of a sketch, both 75mm flanges are welded to a doubler plate or to the vessel itself yes?

If so take your vertical load and move it to the centroid of the channel, then add a moment to that centroid to account for moving the vertical load, calculate stresses due to moments and compressive load, also check for buckling.

 

[rb1957]

ok, i missed the flanges, i think df has the right orientation (though it could be that the 75mm could be on the ground side, doesn't change things much. mind you i won't've called these "legs" ... "skids" sounds better.

as opposed to df, i think you can straight-forwardly apply 1/2 the load to each web ... there's some off-set with the flange, but the load will drift toward the stiffer loadpath. if you want to limit bowing of the sides, add some web gussets into the middle of the U, yes?

Quando Omni Flunkus Moritati

 

[JStephen]

It sounded to me like two different problems, one with columns supporting a vertical tank, another with skids supporting a horizontal tank.

Anyway, draw the sketch any way you can, convert it to a pdf, and attach the pdf.

 

[rb1957]

heck, "print screen" with your phone/camera if you have to !

Quando Omni Flunkus Moritati

 

[AdamMech]

Thanks for your input people, attached is a PDF so this might give more clarification for what I am trying to explain and need.

 

[rb1957]

ok "legs" was right, and we were all hopelessly off track.

the legs are simple columns ? I think these are probably short columns; we'd use "Johnson-Euler" but maybe you mech's have your own approach.

legs-to-skid ... i get the gussets (trying to fix the end to react tank longitudinal loads). it looks like they are a separate weldment bolted onto the skid; looks complicated to me ... why not weld (or bolt) the gusset directly onto the skid and leg webs ?
some gussets could be symmetrical about the leg; if you moved that middle brace (on the base of the skid) then the middle leg could have a symmetric gusset too. if the loads are large, i'd brace the web of the skid (a couple of angles welded/bolted on, several ways to do this).

Quando Omni Flunkus Moritati

 

[JStephen]

In the US, I'd suggest the AISC structural steel codes. Don't forget lateral loads on the vertical members. It would help if your channels were regular channels with tabulated properties.

 

[AdamMech]

Thanks everyone for your input, I've found some formula's to use, especially Johnson-Euler solution regarding 'slenderness ratio' and which one to consider using based upon that equation.

 

[chicopee]

To support the flanges of the channels, I would weld stiffeners within the channels.