ETABS- regarding Analysis Processing time

[madhavaperur]

Dear All,

I am analysing a 28 story concrete building on ETABS v 9.7.4. Building model has shell elements (slab with rigid diaphragm and shear walls) and frame elements (beams and columns). Following are the details from LOG file.

E L E M E N T F O R M A T I O N 19:34:06

NUMBER OF JOINT ELEMENTS FORMED = 41000
NUMBER OF SPRING ELEMENTS FORMED = 0
NUMBER OF FRAME ELEMENTS FORMED = 282898
NUMBER OF SHELL ELEMENTS FORMED = 86605
NUMBER OF LINK ELEMENTS FORMED = 557
NUMBER OF NONLINEAR DEFORMATION LOADS = 0
NUMBER OF CONSTRAINTS FORMED = 485

REDUCTION OF CONSTRAINTS AND RESTRAINTS:
NUMBER OF
CONSTRAINT MASTER DOF BEFORE REDUCTION = 1473
COUPLED CONSTRAINT/RESTRAINT MASTER DOF = 380
CONSTRAINT MASTER DOF AFTER REDUCTION = 1093

E Q U A T I O N S O L U T I O N 20:06:55

TOTAL NUMBER OF EQUILIBRIUM EQUATIONS = 462033
APPROXIMATE "EFFECTIVE" BAND WIDTH = -4328

NUMBER OF EQUATION STORAGE BLOCKS = 273
MAXIMUM BLOCK SIZE (8-BYTE TERMS) = 8388608
SIZE OF STIFFNESS FILE(S) (BYTES) = 1.079 GB

NUMBER OF EQUATIONS TO SOLVE = 462033
NUMBER OF STATIC LOAD CASES = 13
NUMBER OF ACCELERATION LOADS = 6
NUMBER OF NONLINEAR DEFORMATION LOADS = 0

To analyse this building my computer is taking 14 hours. I think this is too large.

My computer OS is Windows 7.0 with 8GB physical RAM and Intel Core i7 CPU (3.4 GHz).

Please suggest based on your experience if this type of run-time is justifiable. if not, what could be possible reasons for such slow processing.

Thanks for your time and efforts.

 

[johnbridge231]

Yes the solution time is too large. Did you check the generated mesh of the sheel elements?
After the solution, do the results appear proper?
The negative value of the effective bandwidth is a little weird...

Analysis and Design of arbitrary cross sections
Reinforcement design to all major codes
Moment Curvature analysis

http://www.engissol.com/cross-section-analysis-design.html

 

[madhavaperur]

Hi John,

thanks for quick response. I have used Auto mesh for walls and slab with 1.8m maximum mesh size.

Please let me know the reason for negative bandwidth. what does this term mean and where should I look for possible errors in model.

Results appear to be fine.

Thanks

 

[MrHershey]

1.8m (5.9ft) mesh seems awfully small for a whole building. I'm guessing you can go with a much rougher mesh and get similar results.

I typically don't ever let ETABS automesh walls unless out of plane loading is a concern. I'll try to limit horizontal mesh spacing of wall elements so that width:height is less than 1:1. Usually will do one or two elements per story. May need more if there are openings in the wall.

Generally speaking, where a real refined mesh will really help you is for out-of-plane design of slender elements (slabs, out-of-plane walls). A fine mesh won't yield as much of an increase in accuracy for elements looking at in-plane forces (walls, diaphragms) because the changes in curvature aren't as dramatic and reversals aren't as common. You could look at maintaining the fine mesh (or even going finer) where you're anticipating elements to be at higher moments (like at the base of your tall walls for the 28 foot building) and less fine elsewhere (like at the top). A 1.8m by 1.8m mesh seems pretty coarse but when you consider your building is probably 80m+ tall, that means for a single shear wall you're looking at 40+ by whatever elements. Adds up quick.