I've got a rather unusual case. I have an inline pump that has a steam turbine driver. Therefore, the centerline of the turbine is along the Y axis (assuming Y as vertical).

When I go into the NEMA module, the options for turbine centerline are only for X and Z. So, I went back to the model, flipped the special execution parameter to make Z the vertical axis, and reran.

However, when I go back to NEMA and read in the values from the Caesar run, it's still assigning the X, Y, and Z forces and moments based on Y being the up axis, even though the load report from Caesar shows the weight loads on the Z.

If I try checking the box in NEMA to indicate Z axis vertical, the the centerline cosine options change to X and Y instead of X and Z.

I ended up having to keep my Z as the centerline of the turbine, and manually entering the values from the restraint report with my Z axis vertical in order to make the analysis.

This would have been a non issue if you guys had just included Y as one of the centerline cosine options on the NEMA input. Although, I must say it is a bit disturbing that after changing Caesar to give Z as the vertical, the NEMA importer still assigned that value to the Y. I would definitely consider that one to be a bug.

Nema SM-23 explicitly addresses horizontal turbines. All the discussion and formulas in the code are for turbines with horizontal shafts. This is why the input screens for the Nema SM-23 analysis have only two (horizontal) direction cosines for the equipment centerline. Thus, if Y is vertical, the cosines to enter are X and Z; and if Z is vertical, the cosines to enter are X and Y. Currently, this module will not address a turbine with a vertical shaft (and there is no way to manipulate the software to accomplish this as you discovered).

Having the Y or Z axis vertical is merely a user preference in handling and entering data. Flipping the Y/Z-up option in Special Execution Parameters (of the Piping Input) also converts the delta coordinates of elements, any predefined displacements and forces/moments to the new system; it does not “change” or “flip” the model or any part of it in the space. This means that if any element of the model was vertical, it will still be vertical, unrelated to what system of coordinates is used.

This is a guess, but this is what I think happened in your situation. Once you flipped Z-up in your piping model, it converted the input and output to the new coordinate system. Any auxiliary analysis that are being done (like NEMA, APIs, WRCs, etc.) must match the same axis convention in order to be valid. However, you left the Z-up check box blank in order to be able to access the Z-coordinate for equipment centerline. These made CAESAR II believe that the loads it is about to transfer must be converted from a system with Z-up (your current model settings) to a system with Y-up (in the NEMA file) on the fly. So it did the conversion, and this now looks like it “was not changed”. If you click the Z-up check box button several times on and off on the NEMA SM-23 screen with loads defined, you will see the loads getting converted to the selected different coordinate system on the fly.

This is not a bug, this is how the program was intended to work.

As to adding a vertical cosine option to the NEMA SM23 module, we could do this, assuming there are specific guidelines on how (such as the special rules for vertical pumps in API-610).