I cant understand that if springs play soil's role , in soil structure models, so why do we impose ground surface motion as input motion and don’t use bedrock motion?
Sahand,
When implementing the substructure approach for SSI you should be very careful what exact ground motion you are using. For brevity I'll divide the ground motions in two types:
1. W.r.t. to the depth (if we assume horizontal infinite layers) you may use the surface motion, bedrock motion, outcrop motion or the motion at some depth. If your foundation is very shallow (let's say less than 1-1.5 m) you may use the surface motion (or rock outcrop if your building is on one) without modification. If your foundation reaches the bedrock, you should use the bedrock motion. In all other cases (maybe 90% of the time) you need to use the ground motion at the depth of your foundation. The best way to do it is with deconvolution analyses: you have the surface ground motion and the soil layers with their dynamic properties and you calculate the motion at the depth of interest. In this way you consider both reduced intensities and shifted frequency content of the earthquake-induced vibrations. You can do this with SHAKE, EERA, DeeoSoil, NEERA etc. A simplified method is to reduce only the Spectral acceleration using the Rd-value (as they do in liquefaction analysis with Seed & Idriss method).
2. W.r.t. the "type" of the motion you may either use the free-field (no influence by structural stiffness) of the effective foundation ground motion. The EFGM includes the effects of incoherence on seismic wave propagation: base slab averaging, and embeddment. Both are influenced by both foundation size and stiffness. FEMA-440 gives a very good insight on the recommended procedures for calculating effective foundation ground motions.
Best regards,
Stoyan Andreev
- Badges
- Science topic
- Similar topics
- Agricultural Science
- Agronomy