Steel moment frame modeling with fiber sections

[YudongWang]

Hello all,

I'm modeling a MRF in Opensees with fiber sections. However, I have some problems with the period of the structure.
It is an 11 story MRF building, there are 2 MRF frames in each direction along the perimeter.
First, I modeled this building in ETABS in 3D with full story mass (double of mass in Opensees), it gave me T1 = 3.22 sec (N-S direction).
Then, I built a 2D model in Opensees for the MRF in N-S direction. I assigned half of the story mass in each story and put half of total gravity columns in Opensees model to consider P-delta effects. However, I got a super large period, where T1 = 4.58 sec.

Also I checked with a 2D ETABS model, same frame, but half of story mass, with no gravity components, it gave a period of T1 =4.9 sec (in this 2D model, it doesn't have gravitation beams and columns, if i built gravity components, the period will reduce a bit, may be closee to the Opensees 2D frame).

I found that if I assign 1/4 of the total building to the frame, the first 3 periods from Opensees are very close to the first 3 periods from ETABS in the direction of that frame (in ETABS, they are T1, T4, T7).
Could someone help me and tell me what should I do with this situation? I've been stuck by this problem for a long time.

Thank you very much.

[selimgunay]

Two questions and a comment:

1. Are you using a rigid diaphragm in the ETABS model?

2. Are you modeling only one of the NS frames in the 2D model?

3. It looks like you are having inconsistent periods from the 3D and 2D ETABS models as well. So the problem looks like it is not related to the Fiber sections in OpenSees, but it is due to a general modeling problem. It would be useful if you aim at achieving similar periods from the 3D and 2D ETABS models first.

[YudongWang]

Hi Selimgunay,

Thank you for your comment.
To answer your questions:
1. Yes, I used a rigid diaphragm in the ETABS 3D model.
2. Yes, I'm modeling only one of NS frames in 2D (there are 2 frames in NS in total).
However, in ETABS 2D model, i don't know how to model rigid diaphragm. I built the 2D model by deleting other members as well as floors, and assign additional mass to column joints (my mass source is from addtional mass only, the mass is half of 3D model), also I set the analysis to be in Y-Z plane only.

Thank you very much.

[selimgunay]

You don't need to assign a rigid diaphragm in the 2D model. Axial stiffness of the beams are generally high enough to represent the rigid diaphragm condition.

Why do you add additional mass to your joints while converting your model from 3D to 2D? That may be the reason of substantially increased periods.

[YudongWang]

Hi selimgunay,

I didn't add additional mass, the way I did is to assign additional mass, while in the mass source I set it to consider only the additional mass, so that I could make sure in 2D model the mass is exactly half of in 3D model.
(In this case the program will not consider self weigth, deal load, snow load in calculation of mass, it will only consider the mass I assigned to it).

[selimgunay]

Please check if the stiffness of your 2D model is really half of the stiffness of the 3D model. You can simply do that by applying lateral force to the top floor in the linear elastic range in both models.

[YudongWang]

I'll try. Reaaly appreciate your help an patience.

[ahmadbsr]

selimgunay wrote:
> You don't need to assign a rigid diaphragm in the 2D model. Axial stiffness
> of the beams are generally high enough to represent the rigid diaphragm
> condition.
>
> Why do you add additional mass to your joints while converting your model
> from 3D to 2D? That may be the reason of substantially increased periods.
for that work , we should use "equaldof " to see the effect of the rigid floors in our 2d models?

[ahmadbsr]

selimgunay wrote:
> Two questions and a comment:
>
> 1. Are you using a rigid diaphragm in the ETABS model?
>
> 2. Are you modeling only one of the NS frames in the 2D model?
>
> 3. It looks like you are having inconsistent periods from the 3D and 2D
> ETABS models as well. So the problem looks like it is not related to the
> Fiber sections in OpenSees, but it is due to a general modeling problem. It
> would be useful if you aim at achieving similar periods from the 3D and 2D
> ETABS models first.
for that work , we should use "equaldof " to see the effect of the rigid floors in our 2d models?