Modelling Issues/consideration in ETABS

[Badar (BAZ)]

I want to comment on some modelling issues in ETABS. Though some of these things are discussed elsewhere in the forum, I hope to extract some more useful conclusions.

First thing is related to modelling the bending stiffness of flexural members, for strength level loads, that is representative of their condition near failure.

The ACI code specifies the modifier of 0.35 on gross moment of inertia to represent its condition at yielding.

Some people say that the factor should be multiplied by 2 to represent the stiffness of T-beam. This approach would be justified if you are not taking into the account the out of plan bending stiffness of slab.

But, ETABS does include the out of plane bending stiffness if you have modelled the slab by using shell elements. So, a factor of 0.7 would overestimate the stiffness of your structure in this case, and will lead to under-design.

If one has used the modifier of 0.35 in ETABS for beams in beam-slab floor system, then what value should be adopted for slab? It should not be 0.25, as this value has been specified for flat plates and flat sab floor system.

If one is using some value of modifier for out of plane bending stiffness on shells, then the share of the bending moment in beams will be reduced accordingly. This approach is correct if one will be providing the reinforcement in column strips of slab. But, if you are providing reinforcement in slab in the direction perpendicular to supports only, i.e. beams, as is the general practice in Pakistan, then you are under-estimating the flexural demand in beams.

Now, there is also a question of factors to be used while deciding the amount of reinforcement required in beams, columns and shear walls.

If you are using factors 0.35 for beams and shear walls, and 0.7 for columns, then you are finding out the demand in members at the point of yielding, and this conforms to the code. But, this also means that the structure might experience unacceptable cracks widths. So, if you are using 0.35 for calculating the demand at strength-level forces, then you should also perform crack-control-check at service-level loads by using the factor of 1.

If you are calculating the strength-level demand with a modifier of 1 for all structural members, after you have decided the location and the number of shear walls with modifier of 0.35, then you are overestimating seismic forces, as you are underestimating the time-period. But, the structural performance will improve.

[ILYAS]

very useful information

[Ayesha]

Remarkable observations.  Kudos.

[UmarMakhzumi]

Class A stuff!

Sir, I think you should write a SEFP Consistent Design Article by combining this topic and the previously posted one by you. 

 

 

[Badar (BAZ)]

Thanks. It can be done.

[Waqar Saleem]

Thought provoking article, however there are a few queries which need further consideration:

(1)     The reinforcement is always being provided in slabs for negative moment both along the beam and perpendicular to beam. Due to availability of sophisticated software’s such as ETABs and SAFE, this reinforcement is provided based on compatible behavior of slab and beams. Moreover, concrete shall crack in tension (may be at service level gravity loads) and reaches/crosses the rebar yield strength (at ultimate strength considering DBE earthquake level) both in beam and slabs simultaneously as slab load is going to beam/columns. Therefore, it is realistic to assume same stiffness modifier of 0.35 for both slab and beams in case of two-way beam slab flooring system. Thus, underestimation of demands in beams is not understandable.

(2)    Crack control width may be checked at ultimate load, if it is OK at strength level, it will be automatically OK at service level. However, if you provide the cover and reinforcement spacing/diameter as per serviceability limit of code, then crack control width may not be an issue at the service level and ultimate level loads for given size of cross section and spans.

(3)    The concrete shall crack when its tensile strength exhausted and this will happen in most of the cases even well below the service level gravity loads, therefore it is hard to justify the modification factor of 1 especially for flexural members such as beams.

(4)    This is not clear how structural performance will improve if we design with modification factors of 1.0. The earthquake resistant design philosophy relies on ductility i.e.: plastic hinges are allowed at the base of the wall and columns and at the end of beams. The plastic hinges are not allowed in the columns. The modification factor of 1.0 will increase flexural strength of beams much more than the seismic flexural demand. This will not allow the beam to yield/development of plastic hinges in case of design level earthquake and there is pretty much chance of shifting the plastic hinge mechanism to the columns ends in several stories converting them to soft stories, drift concentration and failure.So, designing beams for modification factors of 1.0 in all members contradicts the earthquake resistant design philosophy and one need to reconsider the “R” factor for these cases to (a) avoid development of plastic hinges in columns (b) forces foundation.

(Article is written with a fruitful discussion with my Senior Colleague Dr.Munir Ahmed,)