Response Spectrum Analysis_Queries

I am performing Response Spectrum Analysis in Staad Pro 2007. I have some queries regarding obtaining base shear and designing the members for seismic forces.

To start with, I have defined various gravity loads in load case No 1 to 5. In load case 6, I have defined applicable masses (in terms of loads) in all three directions X, Y and Z. The spectra in X direction is defined as follows:

Combination Method: CQC

Spectrum Type: Acceleration

Spectrum Table: Plugged in site specific spectra as defined in our seismic report.

Interpolation: Liner

Damping: 0.05 (5%)

Scale: 9.8 (as the values of acceleration provided in table are in terms of %g)

Direction: X: 0.385. Please see below how I defined this factor.

Query No. 1:

I believe the response modification factor for the type of frame (R ) and Importance Factor (I) are not explicit inputs to Staad Spectrum Analysis; but these are to be defined in the form of factors X, Y and Z in the input stream. Please clarify.

In my case, my structure in X direction is Ordinary Steel Concentrically Braced Frames. Therefore, R = 3.25 as per ASCE 7-05, Table 12.2-1. I value for Occupancy category III in my case is 1.25 as per ASCE 7-05, Table 11.5-1.

As per ASCE 7-05, Cl. 12.9.2, the base shear should be divided by R/I, which means the base shear in X direction should be multiplied by 1 / (R/I) = 0.385 as per ASCE 7-05. I, therefore, assume that the factor in direction X for Spectrum Analysis in Staad should be X=0.385. Please confirm that this is correct. If not, please advise what factor for X should be used in Staad.

Similarly my Z direction, being Special Steel Moment Frame, R = 8, as per ASCE 7-05, Table 12.2-1. I value for Occupancy category III is 1.25 as per ASCE 7-05, Table 11.5-1. Hence base shear in Z direction should be multiplied by 1 / (R/I) = 0.156. I, therefore, assume that in Staad Spectrum Analysis the factor Z should be 0.156. Please confirm whether or not this assumption is correct; and if not, please advise the correct value to be used.

 

Query No. 2:

This relates to the load combination. As suggested in Staad technical manual, the response loads will be absolute load. So, we need to consider two cases, Gravity + Seismic and Gravity - Seismic. Please clarify.

In my model, load case 7 is Spectra in Z Direction.

As per ASCE 7-05, Cl. 12.5.3 (a), we need to combine two orthogonal seismic forces for seismic category D. (100% seismic X +or - 30 % seismic Z and vice versa). How to do this in response load case as I get combined seismic force effect in a single load case only?

Query No. 3:

My third question is regarding distribution of seismic base shear in terms of vertical and horizontal distribution? How does Staad take care of it? In ASCE 7-05, this is calculated based on a function of the Effective Seismic Weight at each floor and the height above base. Please refer to equation 12.8-11 & 12.8-12 in ASCE 7-05.

Query No. 4:

My last question is regarding design of various members in Staad for this response spectra load. How should the load combinations be defined, and how the design parameters should be defined?

Thanks & Regards

Himali

Parents
  • Himali,

    Please see responses keyed in to your queries:

     1. Yes, you have the right idea here.  The only refinement you may need to make is based on ASCE 7-05 Cl. 12.9.4, which says in effect, that the base shear from the response spectra analysis shall not be less than 85% of the corresponding Equivalent Lateral Force Procedure base shear.  If the RS base shear in a particular direction turns out to be less than 85% of the corresponding ELFP base shear in the same direction, then you would also need to incorporate the scale factor into the direction factors you cited.

     2. The nature of a RS analysis and the statistical methods that are used to combine the responses of the various modes make it impossible to accurately carry algebraic signs through to the results.  For this reason, the best we can do is create an envelope by generating load combinations that are based on the positive RS results and load combinations that are based on the negative Rs results, and assume that we have effectively "bracketed" the actual design forces.

    Regarding the second part of your item 2, the method that we demonstrate in the Dynamics and Seismic Analysis class is to manually create Repeat Loads that cover all possible permutations of 100% of the seismic force in one direction with a concurrent 30% of the seismic force in a perpendicular direction.  We also demonstrate how to incorporate eccentricity into the permutations, if it is part of your design considerations.

    3. In a STAAD.Pro response spectrum analysis, only the inherent torsion (as per Cl. 12.8.4.1) is incorporated into the analysis.   There is no convenient way to incorporate the code requirement of 5% accidental eccentricity (Cl. 12.8.4.2), however, I have also heard conflicting opinions as to whether it is even necessary to do this.

     Regarding the vertical distribution on forces, the Clauses that you cite appear in section 12.8.3.  This section generally pertains to ELFP, and I do not see any explicit reference back to this section from section 12.9 (response spectra analysis).  It is my understanding that the "vertical distribution" of forces in a RS analysis is inherently considered based on the mass model, and that there is no explicit distribution step required.

    4. When performing member design based on the RS loads, I recommend using Repeat Loads to form "envelopes" that consider both the positive and the negative forms of the RS forces.

    The proper use of design parameters requires engineering judgement.  The online help documentation provides a section named "Design Parameters" that gives definitions for each parameter.  Additional information is available in the online Example problems, the Verification problems, and in the Steel Design and Concrete Design courses.

    Regards,

    Chris

     

     



Reply
  • Himali,

    Please see responses keyed in to your queries:

     1. Yes, you have the right idea here.  The only refinement you may need to make is based on ASCE 7-05 Cl. 12.9.4, which says in effect, that the base shear from the response spectra analysis shall not be less than 85% of the corresponding Equivalent Lateral Force Procedure base shear.  If the RS base shear in a particular direction turns out to be less than 85% of the corresponding ELFP base shear in the same direction, then you would also need to incorporate the scale factor into the direction factors you cited.

     2. The nature of a RS analysis and the statistical methods that are used to combine the responses of the various modes make it impossible to accurately carry algebraic signs through to the results.  For this reason, the best we can do is create an envelope by generating load combinations that are based on the positive RS results and load combinations that are based on the negative Rs results, and assume that we have effectively "bracketed" the actual design forces.

    Regarding the second part of your item 2, the method that we demonstrate in the Dynamics and Seismic Analysis class is to manually create Repeat Loads that cover all possible permutations of 100% of the seismic force in one direction with a concurrent 30% of the seismic force in a perpendicular direction.  We also demonstrate how to incorporate eccentricity into the permutations, if it is part of your design considerations.

    3. In a STAAD.Pro response spectrum analysis, only the inherent torsion (as per Cl. 12.8.4.1) is incorporated into the analysis.   There is no convenient way to incorporate the code requirement of 5% accidental eccentricity (Cl. 12.8.4.2), however, I have also heard conflicting opinions as to whether it is even necessary to do this.

     Regarding the vertical distribution on forces, the Clauses that you cite appear in section 12.8.3.  This section generally pertains to ELFP, and I do not see any explicit reference back to this section from section 12.9 (response spectra analysis).  It is my understanding that the "vertical distribution" of forces in a RS analysis is inherently considered based on the mass model, and that there is no explicit distribution step required.

    4. When performing member design based on the RS loads, I recommend using Repeat Loads to form "envelopes" that consider both the positive and the negative forms of the RS forces.

    The proper use of design parameters requires engineering judgement.  The online help documentation provides a section named "Design Parameters" that gives definitions for each parameter.  Additional information is available in the online Example problems, the Verification problems, and in the Steel Design and Concrete Design courses.

    Regards,

    Chris

     

     



Children
  • Hi Chris

    Thanks a lot for the value adding input. It indeed help me a lot to go further.

    However I am still not able to understand the explanation using Repeat Load command in second part of reply no. 2. The Repeat Load specification says Repeat Load should not be used with Response Spectrum or Time History Analysis.

    Could you please elaborate?

    Best Regards

    Himali

  • Hi Himali,

    You are absolutely correct!  I should not have suggested the use of REPEAT LOAD with reference to Dynamic Loads.  I fell into the trap of working down through my list of considerations for an ELFP analysis.  Sorry about that.

    You will need to create your load combinations and permutations of 100% and 30% using the LOAD COMB command.

    Thanks for catching that for me!

    Chris



  • Hello Chris

    Thanks for the info.

    Say for example, I have a load case 6 as Seismic in X direction and 7 as Seismic Z. Can I combine them along with other gravity loads as 100% X + 30% Z?

    You did mention about eccentricity into permutations. I am not clear about that as well.

    Thanks & Regards

    Himali

     

     

  • Hi Himali,

    Yes, the syntax of the LOAD COMB command looks like this:

    LOAD COMB #

    LOAD CASE X (space) FACTOR FOR LOAD CASE X (space) LOAD CASE Y (space) FACTOR FOR LOAD CASE Y...(space) LOAD CASE n (space) FACTOR FOR LOAD CASE n.

    So, if Load Case 1 is Dead Load and Load Case 2 is Live Load and Load Case 6 is Seismic in the X direction and Load Case 7 is Seismic Z, then the combinations (WITHOUT orthogonal effects) might look something like this:

    LOAD COMB 3
    1 1.0
    LOAD COMB 4
    1 1.0 2 1.0
    LOAD COMB 5
    1 1.0 2 0.75
    LOAD COMB 6
    1 1.0 6 0.7
    LOAD COMB 7
    1 1.0 7 0.7
    LOAD COMB 8
    1 1.0 6 -0.7
    LOAD COMB 9
    1 1.0 7 -0.7
    ...

    LOAD COMB 17
    1 0.6 6 -0.7
    LOAD COMB 18
    1 0.6 7 -0.7

    So to incorporate orthogonal effects, you could create a series of LOAD COMB commands to combine gravity loads with all the possible permutations of 100% seismic in one direction and 30% seismic in the other direction.  A sample might look like this:

    LOAD COMB 17A
    1 0.6 6 -0.7 7 0.21 

    LOAD COMB 17B
    1 0.6 6 -0.7 7 -0.21

    LOAD COMB 18A
    1 0.6 6 0.21 7 -0.7

    LOAD COMB 18B
    1 0.6 6 -0.21 7 -0.7

    Hope this helps!

    Regards,

    Chris



  • Hi Chris

    Thanks a lot for the input.

    I guess these combinations should capture all possible scenarios.

    Best Regards

    Himali