A client requested a time-history analysis of an existing flare header.

The client requires that friction losses along the header shall be neglected.

The system is very flexible and we are having a hard time defining the cutoff frequency for the analysis.

Different model locations are excited by very different modes. As the cutoff frequency (above 60 Hz) increases, the DLF at locations excited by lower frequency modes seems to become too conservative (specified cutoff frequency too far to the right of the optimal cutoff frequency).

Question:
Is it possible to perform a Time-History analysis of large systems with very dissimilar resonant frequencies?

Best Regards,
Eduardo

I don't know what you mean by "too conservative". As you increase the number of modes included in the analysis, the dynamic component becomes more accurate. A "more conservative" analysis would have a large DLF applied to the "missing" mass (the mass not included in the modal sum).