Steel tank wall - combined stresses - D100
Steel tank wall - combined stresses - D100
(OP)
For a steel water tank standpipe (h=60ft, D = 40ft, high seismic) with mechanical anchorage (J = 3). Code is AWWA D100-11.
Plate tension is in the left right direction (X axis) due to static and seismic hoop stresses.
Plate tension is in the up down direction (Y axis) due to overturning "bending" stresses.
The D100-11 code is silent (or appears so) about combined tension. For any other structure regardless of code, I would combine code-level (ASD) tension using "von Mises" criteria (deviatoric stress), and compare the resulting deviatoric demand with design shear strength = 0.577*Fy / Omega.
It appears like Equation 3-3 is for members, not plates. If I combine tensile stresses, I expect some strong pushback from my fellow engineers and reviewers, since I am not following the code (but instead following mechanics of materials).
This tanks meets the code if combined stresses are not considered.
Plate tension is in the left right direction (X axis) due to static and seismic hoop stresses.
Plate tension is in the up down direction (Y axis) due to overturning "bending" stresses.
The D100-11 code is silent (or appears so) about combined tension. For any other structure regardless of code, I would combine code-level (ASD) tension using "von Mises" criteria (deviatoric stress), and compare the resulting deviatoric demand with design shear strength = 0.577*Fy / Omega.
It appears like Equation 3-3 is for members, not plates. If I combine tensile stresses, I expect some strong pushback from my fellow engineers and reviewers, since I am not following the code (but instead following mechanics of materials).
This tanks meets the code if combined stresses are not considered.





RE: Steel tank wall - combined stresses - D100
To be honest, the vertical tension is not normally even checked, just the compression, and the allowable compression will be a fraction of the allowable tension. The only failures I've ever heard of relating to vertical stresses in a shell were all compressive failures, and that is the limiting factor in the code design.
If I'm visualizing it right, the seismic hoop stresses would be highest on the compression side of the shell, for that matter. The formulas calculate a single stress at a depth, but it's not actually uniform around the tank.
Note that there is never an effort to fully analyze all the stresses at a point. For example, at a circumferential seam where thicknesses differ, you get bending stresses which are not normally calculated or considered in the design. Ditto at the corner weld, where the bottom is presumed to be yield. You get residual stresses in the shell due to rolling, forming, and welding that are not calculated or considered in the design.