orthogonal component

04-Future developments & requests
Post Reply
Posts: 1
Joined: 12 Apr 2014, 14:31

orthogonal component

Post by mimo »

When would be possible to get the orthogonal horizontal component of developed synthetic ground motion records by SeismoArtif????
Posts: 987
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA

Re: orthogonal component

Post by huffte »

Hi mimo. This is an excellent question. Given that the correlation between components of strong ground motion can take on a range of values - including 0 - one approach would be to generate artificial accelerograms for the same seismological parameters (magnitude, distance, site class, duration, time step) and use any two random pairs as horizontal components of the same motion. When target spectra are geometric mean based as opposed to maximum component based, then synthesizing two records independently to the same target seems valid since the geometric mean is thus preserved. Interestingly, the correlation of components should be checked (using Excel 'correl' function, for example) and the pair should be rejected if the correlation coefficient is greater than 0.3 to 0.5. I have done this and actually had the correlation end up being larger than 0.5, which is unrealistic according to research - see the reference below for further details on the correlation between components and best of luck to you.

Hadjian, A. H., On the Correlation of the Components of Strong Ground Motion - Part 2, Bulletin of the Seismological Society of America, volume 71, no. 4, pp. 1323-1331, August, 1981.
Tim Huff
Post Reply

Return to “04-Future developments & requests”