Submitter Response for 2023-C-09

Impact of Directionality on Response Spectra for Seismic Bridge Design

From: [email: Richard.pratt@alaska.gov]

Comments:

FHWA Evaluation of C-09 Reviewed By: Derek Soden/HIBS-10 Comments: The statement doesn't address how the RotD50/100 directionality interplays with the 100%/30% directional load split used in current design specifications. Is the 55% underestimation mentioned in the problem statement with respect to demand calculated using the specified directional split? While still supportive of this statement, my job title needs to be updated in the statement. We would recommend prioritizing C-07 over this project. December 17, 2021

We offer this response:

This topic does not so much relate to the 100% / 30% RSA combination as to the manner in which the two-component orthogonal ground motions are combined into a singular spectra. Historically, the two orthogonal ground motions were combined using the geometric mean method [for example, (x-part * y-part)^0.5] but more recently the "ROTDnn" method (a vector sum) has been used. In the ROTDnn method, the vector sum of the cos*x-part + sin*y-part are combined over a suite of angles between zero and 180 degrees in small (say 5 degree) increments. The resulting spectra accelerations are organized from lowest (ROTD00) to highest (ROTD100) values. The mean value (ROTD50) is commonly used in lieu of the previous geometric mean combination. The question is, what ROTDnn should be used that best predicts the displacement demand of "real world" multiple degree of freedom structures? This project intends to address that issue and provide design specification language to clarify the appropriate ROTDnn or correction factor when using ROTD50 to best predict seismic displacement demands.

Review Date:

December 17, 2021