ACCEPTED SOLUTION
Accepted Solutions
OK, now the worksheet really is attached and a question:
Why don't you use the built-in functions "eigenvals", "eigenvecs" or "eigenvec" ?
You may note that symbolic and numeric evaluation return differently normalized vectors.
I guess that nx and ny should be dimensionless. But it sure is a good habit to add the unit Pa to the zero-results. For your example this is not really necessary, but on other occasions it may be mandatory to add units to zero values, too, because Prime does not use static unit checks (which is a good thing anyway even though it can be a bit bothersome on some occasions).
When you use symbolic evaluations and assign the result to a variable its a good idea to avoid numeric inline evaluations. Avoiding this and telling the symbolic (which unfortunately does not know anything about units and treats them as unknown variables) that the used units (in your case only MPa) are positive reals does the trick and you get the resukt for ny which you probably expect.
As the picture shows you may also create a function that reflects the relationship between ny and ny.
Worksheet in format P5 attached.
OK, now the worksheet really is attached and a question:
Why don't you use the built-in functions "eigenvals", "eigenvecs" or "eigenvec" ?
You may note that symbolic and numeric evaluation return differently normalized vectors.
