RESPONSE SPECTRUM MODAL ANALYSIS

Question

Good morning, 
 I'm trying to perform response spectrum modal analysis with Response spectrum user define data option on a multistory concrete building. The porpoise is to introduce in the FE model, seismic actions according to Italian design code (NTC2018) or, at least, Eurocode. 
 The problem is to manage to introduce, for seismic calculation, also non structural loads (different from self weight) that are represented , for example, by member loads (distributed and/or concentrated) or nodal forces directly assigned to elements with the aim to make STAAD compute them as structural mass. 
 Is there a specified procedure to follow? 
 If consider the attached model, in the SEISM +X load case, I have defined: 
 - the self weight of the structure in X direction because I want to calculate seism effects in X direction; 
 - user defined response spectrum in X direction; 
 - distributed member load equal to 2 kN/m in X direction. Running the analysis, I have noticed that this load has not been considered in the calculation of structural mass and so in the evaluation of global seismic effects on the structure. Therefore, how can I define this load correctly in order to make it consider in seismic calculation? 
 
 Thank you, regards 
 Response_Spectrum_Analysis.std 
 Silvia

Sye · Accepted Answer

The member loads have been defined correctly. The only thing you need to do is to move the loads ahead of the spectrum data. A modified file is attached. There is a difference in base shears when the member loads are included vs when they are not, which is one of the ways to confirm that the member loads are being considered for the mass calculation.

Sye · Answer

Defining only spectrum for load case 5 is fine as the dynamic mass for all cases will be calculated based on the load input provided as part of the first dynamic load case. 
 You do not have to specify a master node. Rather leave it to the software to calculate that for you. When you do that it would report the center of mass in the analysis output file. 
 When you define the mass type reference load, you do not have to define the self weight and member loads as part of the spectrum load case. 
 When floor diaphragm is included, the torsion due to eccentricity between CM and CR is already accounted for in the analysis. 
 A modified file is attached for your reference 3733.Response_Spectrum_Analysis_rev.std

Sye · Answer

For a custom response spectrum, there is no automatic way to account for the accidental eccentricity. This is what you can do. Use the SET PRINT 17 command before the JOINT COORDINATES in your input command file and the analysis output file would then report the story forces at each level of the structure ( These are referred to as Floor Modal Base Actions in the analysis output file). These forces are reported for all the modes. To get the force at each story you may use a modal combination method like SRSS. This has to be done manually and once you have the shears at each story level, you can multiply those by the 5% eccentricity value at each floor to compute the accidental torsion. You can create a separate load case for the accidental torsion and use load combinations to combine that load with the seismic loading.

Sye · Answer

Yes you have to calculate the square root of summation of squares of the floor modal base actions to get the story forces at each level. There is no direct way to get the story forces reported in the output file for a spectrum loading. You can get that for equivalent static seismic load generations. The accidental torsion moments can then be combined with the seismic loads as indicated in my last post ( put those in a load case and then use load combination). The 5% eccentricity torsion moment that you are talking about is accidental torsion and this is different from the torsion that accounts for the eccentricity between the CM and CR ( inherent torsion). If you have a rigid ( or semirigid) diaphragm then you would have to account for both.

Sye · Answer

The 5% eccentricity is to account for the fact that center of mass may shift from it's assumed position on either side. This is different from the torsion that is generated due to eccentricity between the CM and CR. So it is not an either or.

Sye · Answer

When you define a rigid diaphragm, the software calculates the CM and creates a master node at that location. If there is a node already existing at that location, that node is considered as the master node and if there is no existing node, the software internally creates one. These nodes that you are seeing in the analysis output file are internal nodes and so you would not be able to locate these in the model. However you do not need to apply the moments at the master nodes because as the diaphragm is rigid, it would rotate as a whole and so the exact location of the moment is not going to have an effect. You may choose any node in the plane of the diaphragm. I would usually go with the one closest to the center but again, it really does not matter.

Sye · Answer

For a response spectrum analysis you cannot add the individual story shears to get the base shear. The reason for that is explained in the following wiki. The base shear is the one reported by the software. 
 https://communities.bentley.com/products/ram-staad/w/structural_analysis_and_design__wiki/29001/base-shear-does-not-match-the-summation-of-shears

Sye · Answer

CM information is calculated and printed only when you use FLOOR DIAPHRAGM option to model a rigid diaphragm. If you are not using a FLOOR DIAPHRAGM, you will have to calculate the center of mass yourself. 
 If you are using a response spectrum analysis and your model has FLOOR DIAPHRAGM defined, you would be able to account for both inherent and accidental torsion automatically by using the torsion option. If there are no rigid diaphragms defined you would have to manually calculate the accidental torsion and apply it to the different floors. You may create a separate load case for this torsion load and use load combination to combine it with the response spectrum load case.

RESPONSE SPECTRUM MODAL ANALYSIS

Top Replies