We have an exciting announcement about badges coming in May 2025. Until then, we will temporarily stop issuing new badges for course completions and certifications. However, all completions will be recorded and fulfilled after May 2025.
General Mechanical

General Mechanical

Topics related to Mechanical Enterprise, Motion, Additive Print and more.

Why does the prestress from Static analysis does not add up to the stress in Random Vibration analys

    • Hui Liu
      Ansys Employee
      PSD stresses are statistical and reported as standard deviations from a zero mean (always positive numbers). These stresses take into account stiffening effects of a pre-stress analysis, but are themselves simply statistical deviations from a zero mean.nOn a component-by-component basis you can add or subtract PSD stress values to component values from your pre-stress to find the 'worst case' max/min stress that can occur in that component. This can be thought of as rigidly shifting the stress distribution reported by the PSD run away from its zero mean.nYou cannot simply use the Von Mises Equivalent Stress equation to calculate PSD equivalent stresses, as that algebraic equation is not mathematically suited for statistical values. If you want to Superpose Random Vibe Stress with Static Structural then there are a couple of points for your consideration on this issue:nThe default output from a Random Vibration analysis will be 1-sigma values (i.e., standard deviations about a zero mean). Therefore, we have a Gaussian distribution of stress at each node. If you have pre-stressed the modal analysis, then the effect of the static loading is accounted for in the stiffness that is used to determine the natural frequencies. This consequently filters down into the PSD analysis.nYou could add/subtract the 1-sigma component value (i.e., Normal Stress in X, Y, and Z) at each node to/from the static analysis to get a sense for the 1-sigma variation of the component stresses about that mean value. This would be the same as rigidly shifting the Gaussian distribution from zero mean to be centered about the static condition. There will not be a way to directly plot contours like this in ANSYS Mechanical. However, I can think of one possible workaround:nRight Click on the PSD result and select Export. This will give you a text file containing stress at each node. It will also contain the X, Y, and Z coordinates of the nodes if you have specified this on Tools > Options > Mechanical > Export > Include Node Location.nnRepeat this process for the Static resultnAdd and subtract appropriately to get the min/max values in ExcelnSave separate Text files with the Node locations and Min/Max valuesnRead in the Text Files using External Data in ANSYS Workbench. Make sure to specify the appropriate column data, especially Stress (and units!)nLink the External Data cell to the Setup of a duplicate Static StructuralnRight Click on the Imported Load folder in ANSYS Mechanical and specify 'Initial Stress'nSpecify 'Apply To' > Corner Nodes and 'Component' > Desired ComponentnANSYS Mechanical will then try to map the stress to the duplicated mesh. The result should be an Imported Initial Stress that really is the combination of your static and PSD results.n
    • Karthik Remella
      Administrator
      Thank you, for sharing this tip.nKarthikn
    • MTBXC
      Subscriber
      Recently, I did such an analysis -> Static + Prestress-> Modal-> Random Vibration. You have increased my knowledge of it. Thank younnMichaln
Viewing 2 reply threads
  • The topic ‘Why does the prestress from Static analysis does not add up to the stress in Random Vibration analys’ is closed to new replies.