Why is model over stress due to Harmonic loading in AutoPIPE's reports?
Example:
The analysis shows that is fails severely under a harmonic load in segment B. The load is small you would think at 0.05in resultant displacement, but very high frequency.Applying the same displacement statically produces a significantly lower result. Can you explain why this is? Why is it that a very high frequency vibration can create such a high stress. I know they have gaps in dynamic cases that will be removed, but this isn't going to make any different at this location.
The following two settings for the analysis are contributing to high response for harmonic analysis: 1. No damping defined for the harmonic loading function: providing a damping of say 5% lowers the response2. Cut-off frequency under Edit model option was set to 33: Providing the cut-off frequency of 2,400 Hz seems to lower the response3. Change of direction due to bends results in high moment values caused by the high pipe internal forces Changing 1 and 2 lowers the response by some amount. For such a high frequency, the accelerations of the system must be quite high even for small amount of displacements (see accelerations report). This may potentially be caused a high response. You can reduce the response even further by better discretization of mass. Currently there is a near zero length element near bend B01. If the bend at that location is set to short radius (increasing the length for the near zero length element), you will see further drop in the response (I set the mass points per span to 9). Generating an equivalent time history loading for the harmonic seems to generate similar results. So the results seems to be correct although huge. Use the attached excel sheet to generate Time History loading. The frequency is quite high which will result in a very small time step.
The following two settings for the analysis are contributing to high response for harmonic analysis: 1. No damping defined for the harmonic loading function: providing a damping of say 5% lowers the response2. Cut-off frequency under Edit model option was set to 33: Providing the cut-off frequency of 2,400 Hz seems to lower the response3. Change of direction due to bends results in high moment values caused by the high pipe internal forces Changing 1 and 2 lowers the response by some amount. For such a high frequency, the accelerations of the system must be quite high even for small amount of displacements (see accelerations report). This may potentially be caused a high response.
You can reduce the response even further by better discretization of mass. Currently there is a near zero length element near bend B01. If the bend at that location is set to short radius (increasing the length for the near zero length element), you will see further drop in the response (I set the mass points per span to 9).
Generating an equivalent time history loading for the harmonic seems to generate similar results. So the results seems to be correct although huge. Use the attached excel sheet to generate Time History loading. The frequency is quite high which will result in a very small time step.
"Code Compliance" sub-report using Results> Output Report
Bentley AutoPIPE