Torsion in second mode? 3

[curiousinvite]

there is torsion in second mode , what might be the reason ?
what can be done to make translation in y axis second mode ?

 

[HTURKAK]

You may increase the torsional stiffness of the structure.

Consider to increase the dimensions of corner columns but keep always symmetry.

 

[Brad805]

What does the deformation look like in mode 2? That usually provides a clue as to why.

 

[curiousinvite]

Tried that there is still torsion in second mode , what can be done ? Do I have to provide shear wall ?

 

[JoshPlumSE]

Do you have stairwells or ramps modeled as part of the lateral force resisting system? Usually you try to keep these structurally separate from the main lateral force resisting systems.

 

[Brad805]

I assume all your columns, beam, and loads are symmetrical? I think JoshPlum has a good idea. It would be easy to create a model without the stair elements to see the effect. After that you can decide if it is important to include those in your model or simply apply the gravity loads you need to the model.

 

[WARose]

Just eyeballing it....it appears your center of mass and center of rigidity probably have good eccentricity to each other.

I agree with the others, I'd leave out the stairs (as they should have slip connections anyway) and see what happens. After that, I'd try changing the lateral resisting system so that more stiffness is on perimeter column lines.

 

[curiousinvite]

May be the stairs are acting as bracing causing torsion , but still Wouldn't it be more realistic to model with staircase than without one ?

 

[JoshPlumSE]

May be the stairs are acting as bracing causing torsion , but still Wouldn't it be more realistic to model with staircase than without one ?

How to handles stairs in any structure, or how to handle ramps in parking garages tend to be problematic topics in structural analysis. The issues (as I see them) are the following:
1) These inclined members tend to act like very rigid braces in the model and take a lot of the lateral forces.
2) These inclined members are not generally designed for the type of seismic ductility (or redundancy) required in the seismic design codes. Therefore, we usually need to figure out a way to prevent that load from getting into those inclined members.
3) In a lot of projects the Structural Engineer does NOT design the stairwell. Rather that is handled by a specialty engineer. Therefore, because of the issues above there need to be a lot of interaction between those two engineers if the stairs are going to be taking load from the structure. Alternatively, there needs to be some good details to ensure that the stairwell is isolated laterally from the rest of the structure.

 

[KootK]

May be the stairs are acting as bracing causing torsion , but still Wouldn't it be more realistic to model with staircase than without one ?

Yes, it would. In some high seismic juridictions, codes are evolving to encourage designers to either:

1) Model the effect as you have done or, more likely;

2) Detail movement joints at stair connections that facilitate interstory drift.

I suspect that part of your issue here stems from the particular combination of:

3) The stairs providing a stiff lateral bracing load path and;

4) Your primary lateral load resisting mechanism being relatively flexible moment frames.

 

[Brad805]

"Wouldn't it be more realistic to model with staircase than without one ?" I think it all depends on the stiffness of the stairs and how they interact with the structure. Say they are cast in place stairs, anchored to rigid slabs at all levels, then yes it is something you need to contemplate. If they are steel stairs with pinned connections, I would say many of us would be inclined to neglect them as being part of the SFRS. Regardless, I would rule our their involvement in your question with a quick model.

 

[dold]

We've been getting plan review comments lately regarding stairs. ASCE 7-16 requires the following:

That said, if I were to include the stair in the model, I would only model it as a plate/shell if it is truly a cast in place stair. Typical steel utility stairs are obviously way less stiff than a concrete plate and will attract less load. Of course it will be more stiff in the direction-of-walking (parallel to stringers) than it would be for transverse drift. We never have and likely never will include stairs in our global models. We just spec a slotted angle at the bottom stringer with a larger than normal hole and a big washer or something to that effect (we always design/detail stairs as part of our package).

The easier way of dealing with this if you really want to find member forces in the stair assembly would be to model the stair by itself (stringers and landing) in risa3d or ram elements or whatever, and apply an enforced displacement at the upper floor corresponding to amplified story drift. I wouldn't be surprised to see some decent forces, especially when including overstrength, for a ~3" displacement (0.02*12ft).

Whatever the case, I would absolutely neglect them as any part of your LFRS. I.e., run one model without stairs to design your building. Then add the stair back in later if you're interested in looking at the forces.

But to your original question, OP, I'd remove the stairs and rerun the model. If the torsion stuff goes away then you have your answer.

 

[curiousinvite]

yes they are cast in place stairs so I was inclined to include them in my model
I removed the stairs but still there is torsion in second mode
Brad805
Regarding steel stairs with pinned connections , I haven't come across one if you have a picture can u post one
And what if i provide corbels and rest stairs on it?

 

[HTURKAK]

curiousinvite (Civil/Environmental)(OP),

I do not know the applicable code in your region. But in general, it is not desirable that the first or second mode is a torsional mode.
Apparently , this is a low rise structure with four storey.

IMO, you have two options;

- Increase the torsional stiffness with ( increasing the size of corner columns ,increase the stiffness of perimeter frames, or adding shear walls at perimeter frames keeping symmetry ....)

- If you think that the dimensions of columns , beams are reasonable , you may design the SFRS for torsional effects. Check your structure for extreme torsional irregularity and for redundancy clause. ( For SDC E or F, Extreme Torsional Irregularity is not allowed and there is a penalty clause ; if SDC is D and having extreme torsional irregularity redundancy factor shall equal 1.3.)..

 

[JoshPlumSE]

Can you post the mass participation ratios for the model without stairs? I'm just curious to see how much the translational modes have changed (in either mass participation or period).

 

[JLNJ]

Your aspect ratio appears to be north of 2:1. Maybe your stiffness in the long direction is so much greater than the short direction that the torsional mode hits before the long-direction mode activates. I suppose this would tell you that your transverse stiffness is too small.

 

[curiousinvite]

I have attached my model file for reference