What is Rigid response ? Details are in the Theory Reference, but the simple explanation is that we try to get a ‘max’ or ‘final’ response in Response Spectrum analyses through mode-combination methods.  The mode-combination methods try to account for modes acting together at once. When we have a ‘rigid’ response (high frequencies), the modes are thought to act together (in-phase), so they may be directly added (algebraic sum).  On the other hand, if modes may be out-of-phase with each other, we use other approaches (SRSS is the least conservative of these approaches). The Rigid response tries to separate ‘rigid’ and non-rigid (‘periodic’) responses separately.  To do this, we have two methods – Gupta and Lindley-Yow.  Gupta uses two frequencies – below f1, we don’t algebraically sum; above f2, we algebraically sum, and we vary in-between when f1 < f < f2.  Lindley-Yow uses a ratio instead. Once the 'rigid' response is calculated, it is then combined with the 'periodic' response via SRSS. If one wishes to include Rigid Response effect, this does change the mode combination completely (see Equations in Theory Reference cited above) by separating modes into 'rigid' and 'periodic' modes and combining them differently. When to use Rigid response ? To address whether rigid response is needed, one needs to answer this question: Does one want to separate 'periodic' and 'rigid' response for the dynamic behavior of the system? One needs to study the dynamic behavior of the system and understand when the modes are in phase and when do they transition to being out of phase. ANSYS Mechanical offer users multiple mode combination methods to determine the 'peak' response of the system (SRSS, CQC, ROSE, etc.).  Are these mode combination methods sufficient for the analysis being performed? If userse understand their system behavior and they understand that, at certain frequencies, the modes should behave in-phase - SRSS or other mode combination methods would not account for this.  Therefore, one wants a way to manually specify frequencies at which the modes are summed algebraically, and the Rigid Response method offers such an alternative to standard mode-combination methods. If one doesn't know at which frequencies his/her modes act in-phase, it's probably better not to use rigid response method.  If an analyst knows the cut-off at which frequencies their modes act in-phase, that is the frequency they input for Gupta or Lindley-Yow methods.