Response Spectrum Analysis (RSA) - Concept & Use

Career Coaching for Great Bridge Engineers!
To sign up to receive a free 4 Practices E-Book and early bird offers for our upcoming course on Context-Sensitive Conceptual Bridge Design and others, click here: www.spannovation.ca/school
This short video gives a succinct overview of the concept of a response spectrum and contains a very basic discussion about how to use it. The emphasis is on the acceleration response spectrum and its meaning.
My partner Raj Singh and I (Saqib Khan) have over forty years of combined experience in bridge, seismic, and erection engineering, which has included a wide variety of bridges (short creek spans to long river crossings and everything in between) on traditional bid-build and alternative delivery p3/design-build projects

[@RheeJaeHyuk]

Years of training summed up with 10minutes lecture

[@mr.d3916]

very good explanation. thanks

[@mostafaismail43]

Excellent teacher

[@allinengineering8168]

best explanation

[@000khalil000]

best explanation!!

[@MuhammadAli-en1js]

very good explaination

[@babulchandra8760]

excellent

[@ziaulhasan6720]

thanks

[@leonart8958]

That was a great explanation, well done.
I also want to ask, what is your take on multiple - support response spectrum (MSRS)? why some code provisions such as AASHTO doesn't consider this type of analysis? If I am not mistaken, the conventional RSA is based on the assumption that the bridge excites under uniform ground motion, however for long span bridges this may not be the case. MSRS, which is based on random vibration theory, consider this variation in ground motion (wave passage, incoherency, variation of local soil condition) which may amplify or deamplify the response of the bridge.
Do you also have lectures or references that demonstrate the use of this type of analysis(MSRS) in practical engineering?
Thank you.

[@spannovationbridgeandseism9421]

Leon: You make an excellent point here! I am not sure why the MSRS method has not found widespread usage. To the best of my knowledge, it has not been adopted by any code at this point. You are absolutely correct that the traditional RSA assumes uniform excitation at all supports and cannot capture the effects of spatial variability leading to multiple-support excitation effects. Conventional RSA using the envelope spectral values can be unconservative and would miss local demand concentrations at different support conditions. So far, I have not seen any softwares incorporating MSRS or code provisions that can help a practicing engineer adopt the MSRS methodology.
Generally speaking, multiple-support input is of more interest in longer bridges, whereby the design is checked using the Non-Linear Time History (NTHA) analysis incorporating variable time history inputs at different support locations. The variable input NTHA would provide the most "Accurate" results and the MSRS is therefore not needed for such cases. Interestingly enough, I am carrying out some research in collaboration with University of British Columbia on this topic. We are going to come up with some design guidelines for practicing engineers to design shorter bridges appropriately when the geotechnical engineer provides us with multiple response spectra for a given bridge site. However, it is not based on the MSRS methodology; the idea is to use NTHA assuming variable time history input corresponding to different site classifications and generate trends for bridges with 2 to 7 spans. Based on statistical analysis from the data obtained through extensive NTHA, we will try to propose some amplification/de-amplification factors, etc. to modify the response spectrum to be used. The design engineer should then be able to use the standard RSA technique and inelastic static pushover analysis (ISPA). The design can then be carried out using the standard analysis and design techniques as enshrined in the various codes of practice. That is the idea anyways but we will see what form the recommendations eventually take.
For MSRS references, please start your search with Dr. Der Kiureghian and you will start finding papers through various internet resources. Once we have got some credible results from our research, I will try to do a whiteboard video about this topic. Thanks for leaving an excellent comment and inviting discussion on this topic. Please stay engaged with our community!

[@civilideas1925]

👍🏻

[@hansenm6910]

Hi sir,
Can the matched response spectrum(of a certain time history ground motion record) be used as an alternative seismic input instead of the design response spectrum(target) from the code?

[@spannovationbridgeandseism9421]

One could do a site-specific response analysis (it's usually carried out by geotechnical engineers). For this, an outcrop ground motion is input at an appropriate level below the ground surface and wave propagation analysis is done to obtain the time history signature at or close to the surface. A design spectrum can then be generated corresponding to the mean or mean plus minus some standard deviation of the response spectra of a number of time histories.

[@hansenm6910]

@@spannovationbridgeandseism9421 thank you sir,
i'm just wondering, what if i have this ground motion record, then i did a spectral matching of its corresponding RS to the design RS of code (target). Since I already obtained the matched RS, instead of doing a time history analysis using the RS-compatible time history, I will used the corresponding RS of this time history, assigned it at each support, and then perform the conventional response spectrum analysis (but in this case i have different response spectrum at each support of the bridge). do you think it is an acceptable/appropriate procedure?

[@presxi3472]

1. what is a response spectrum? 2. why do we need to determine and what is the main purpose of response spectrum? 3. what is damping ratio and how it is differ from the critical damping? 4. What is natural frequency? hope these basics are clearly define.. thanks

[@spannovationbridgeandseism9421]

Thanks for your questions. We suggest that you keep track of our upcoming webinars/courses as these will be addressed in one of those.

[@bluebloo5574]

a 9 minus lecture better than a 50 minus lecture given by my professors.

[@spannovationbridgeandseism9421]

Thanks!

[@noorhanalkhatib4341]

Please doctor what about the point in the spectrum when T=0 sec? We can't use the equation A=u*w2 in this case right?

[@spannovationbridgeandseism9421]

The Sa value at T = 0 sec corresponds to the peak ground acceleration (PGA).

[@noorhanalkhatib4341]

@@spannovationbridgeandseism9421 thank you very much