Calculating bending moments from element stress

[rn14]

Hello,
I have recently been playing with Calculix a bit and have found it pretty useful. I have generated a model of an annular section that is quite thick (66"OD 24" ID 13" T) and was hoping to compare the radial and tangential moments to those calculated using Roark's equations (I wanted to see if the thickness throws roark's way off). I have been trying to find any resources that may point me in that direction. Is my best option simply to use sigma=My/I? I'm also a bit confused with which stresses should be used. The model has the ring center at 0,0,0 with z as the axis of thickness. Load is applied as point loads at the nodes at the ID. I could take a slice along the x axis and look at sigmaxx but it seems like I should be able to look at principal stress instead. Thanks.

 

[rb1957]

"Load is applied as point loads at the nodes at the ID." ... pressure or shear (Fz) load ?
it looks like you're applying radial loads near the top surface ??

how constrained ??
it looks like you're restraining the lower edge ??

you're plotting stress xx ... remember stress yy and zz affect the result. This also suggest that Mc/I probably isn't going to get you very far.



Quando Omni Flunkus Moritati

 

[rn14]

The point loads are applied in the -z direction at the nodes on the top of the ID and the lower OD is restrained from translation in x,y, and z.
The plot was provided more or less to show what I was modeling but I also plotted stress xx because I was considering pulling out stresses from a cut through the xz plane to calculate try and calculate a moment. It definitely would only be an approximation but I haven't been able to come up with much else. In particular I need moments so I can compare to Roark's and also because the real object is reinforced concrete and I would like to size rebar to compare to other design methods.

 

[rb1957]

unless full restaint in all directions is real, i think you're over constraining the model. if you're only applying z loads, then i think a better conraint scheme would be z at each node, + x and y at one end of a diameter, and x or y at the other end (think of the rigid body motions you're trying to restrain).

if you want to see really "sensible" stresses, I'd consider using a cyclindrical co-ord system (to give you radial and tangential stresses). looking at principal stresses maybe be "better".

Quando Omni Flunkus Moritati