Civil Engineering Association

(01-31-2009, 09:50 PM)juice Wrote: [ -> ]To ETABS newbies, always check your "GLOBAL FORCE BALANCE RELATIVE ERRORS" in Last Analysis Run Log.

Don't get too excited at once on the output. ;)

Dear Juice,
The errors are in percentages,but what are the critical percentages?
This is what I got for a 12-story building(wind load and seismic loads aren't defined yet).
LOAD FX FY FZ MX MY MZ
DEAD 0.003611 0.102726 2.10E-11 0.538303 0.002214 0.056797
LIVE 0.014544 0.097379 2.34E-11 0.568369 0.027720 0.059392
SNOW .000000 .000000 .000000 .000000 .000000 .000000
TRIBUTAR 0.019676 0.118920 2.09E-11 0.614789 0.064894 0.075489


There are high percentages as you can see,so?

Dear civilphongvan [ Post: #102 ]
ETABS use 'terminology' center of mass & center of rigidity.
In 'lateral load resistant system', you must add 'eccentricity'
Eccentricity simply can be said as distance from center of mass to center of rigidity.
For simple & low rise building 5% additional eccentricity is enough for design.
As additional, rigid diaphragm, when you assign it, it have meaning all the element of structures that you select to be added with 'rigid diaphragm' ( could be slab, beam, column & wall : you see at building plan then select then add diaphragm ) will move all together when lateral load applied to your structure.
To set your own calculation of lateral load you can use 'static user load', in this option you can set your loads + the eccentricity
Thanks
Paladin

Dear deadlord (post#111)
according to ATKINS ETABS Analysis Manual, if you see 8.1.2 "Global Force Balance" it states thats "These errors are calculated bases on the relative difference in applied external loads and the base reactions for each load case. The global force balance relative errors should be very small in order of 10e-10 and less. It should be noted that this is a necessary but not sufficient condition to ensure a sound modelling and analysis"

1-these errors should be of small magnitude 10e10 & smaller
2-this check must be satisfied for sound modelling PLUS ALL OTHER CHECKS.

Dear Coolestbliss(post# 113),
So looking at the numbers I've put this means that I have lots of errors.
Does the slab meshing affect these errors?I use the slab meshing of Ram Concept,but
Yesterday I noticed Ram Concept's mesh is poor compared to Adapt builder mesh,and had to
Correct it manually because there were very small elements.

Can the unorganized replication of floors increase this error?
"It should be noted that this is a necessary but not sufficient condition to ensure a sound modelling and analysis" with much respect ,I don't like this kind of sentences,the errors either produce analysis errors or not?so how come now he says that a sound model can be made without checking everything to be right?

Regards,
Husam.

I never used RAM Concept or Adapt Builder.
Slab meshing is an art and you will understand it with experience.
What do you mean by unorganized replication of floors? I can make nothing of it.

NECESSARY BUT NOT SUFFICIENT MEANS...

For a sound ETABS model, there are plenty of checks to be satisfied and one of the checks is this GLOBAL ERRORS check. Do not think that satisfying only this check will be the "Only" SUFFICIENT CONDITION but rather it is a NECESSARY CONDITION along with many other conditions.

Your errors depends on the model geometry etc. so it is difficult to tell you that an eror og magnitude 0.0003 or 0.0004 is significant or not? but i think they are not passably well

Dear coolestbliss,
I found a huge error in the model,once I corrected it the global errors at at least (X)E10^(-10).

Could you list the checks that we should do?or recommend any document that mentions them.
Regards,

I prepared some tips for creating a default.edb file for ETABS, please download it here

also check this topic for checking ETABS model, tips and tricks

Dear BennyK,
It is not a simple software to learn in 2 days..Please see this CIVILEA topic