Why ncode rms stress result different from Ansys Random rms result

[hayka]

Hi. I am using ansys ncode design life v 13 and doing vibration fatigue analysis. I modeled a simple beam with tip mass and firstly run a harmonic analysis with unit load that is 1 g. The material of the beam is Al 7075 T651 and I enter 1.6% damping ratio for the results between the 5-500 Hz. I also save the db file for post-processing the results. After harmonic analysis is finished I import the file.rst file directly into designlife and run a fatigue analysis for the 0.025 g^2/Hz 5-500 Hz white noise. I get the most critical node with 1 sigma rms value and stress psd. Then for the same mathematical model in ansys I run a random vibration analysis and get the 1 sigma rms value and stress psd for the same node. 1 sigma result are different from each other and when I compare the stress psd values at each frequency and especially peaks at the natural frequencies are very different from each other. The difference for peaks is not because of the damping ratio as I mentioned above I use the same mathematical model hence the same damping ratios. I have to find the reason please help me!!!!!!

 

[izax1]

I don't know either of the codes you are using (Ansys and nCode??)but the posiblities that comes to my mind is:
Do the codes use the same element formulation?
Do you calculate the peaks at the same frequency?
Are the boundary conditions defined in the same way?
Is the damping defined the same way?

Do a Natural frequency calculation first and compare the mode shapes and frequencies. If the element formulation is similar, the natural frequencies and mode shapes are similar, the damping definition is similar then you also should get similar results.

 

[hayka]

All boundary conditions, material properties, and damping ratios are defined in Ansys and Stress transfer function is calculated by Ansys during harmonic analysis. In theory, Ncode must use this stress transfer function and calculate the stress psd. Using the same mathematical model stress psd is calculated by Ansys random vibration tool. Hence everything is same however results are different. I wonder that after transfer function is calculated by harmonic analysis stresses are multiplied by a calculated coefficient in Ncode Designlife but I do not how it is. In fact multiplying stresses by a coefficient is not correct because all stress concentration factors are taken into care by Ansys .Hence the calculated stresses by Ansys must be used by Ncode directly.