semi-rigid diaphragm in ETABS

[vetchengineer]

Hi everybody,

In ETABS 2018, I am working on a building model. When I define the diaphragm as "rigid diaphragm" there is no warning but when I define the diaphragm as "semi-rigid diaphragm" "ill-condition and unstable warning" occurs.

Please kindly inform me why this warning occurs and what should I do?

Thanks & Regards

 

[JLNJ]

Is your building model unstable? That’s what the model seems to be telling you.

 

[jayrod12]

Or it's telling you that without a stiff enough diaphragm, your lateral system is unstable. Maybe you've got a huge distance that you need your diaphragm to move the load to get to the resisting walls/bracing. If a semi-rigid diaphragm isn't stiff enough, then technically your model is unstable.

 

[vetchengineer]

Thank you for your replies but, Then if so, I mean if the model is unstable, what should I do?
What should I check?

I have got 2 possibilities, in my mind:
1- Numerical warnings I see while analysing with standart solver (6-7 digits warnings I have)
2- I have huge soil loading from one side of building.

Do you think these 2 items cause the “unstable warning” ?

 

[jayrod12]

You have to fix number 1 first.

Number 2 is just the situation. It's not like you can magically wish away the soil loading. It is there. Fix number 1, it may result in the issue going away.

If not, you may need more lateral elements added to the structure, i.e. braces or moment frames or shear walls, or you may need a significantly stiffer diaphragm to get the loads to move to where they need to go.

Don't skimp on your lateral system if you have a significant imbalances soil loading. That loading is real, and permanent. Make sure you've got plenty of strength.

 

[EZBuilding]

With a semi-rigid diaphragm, is the model stable if you don't have P-Delta active? Is the model Stable if you don't have any applied wind loads?

Review any moment releases in your frame elements. Make sure you don't have moment releases at pinned joints. Do you have any connectivity issues between frame elements - frame joints which are just off each other? Do you have one or two way slab modeling? One-way slab behavior is tricky and has more issues with stability. Shell behavior can cover up some of those issues instead.

 

[JLNJ]

First, you need to understand what makes structures stable or unstable. Then you need to understand the difference between rigid and flexible diaphragms, why you would use one or the other, and how ETABS handles modeling of diaphragms. Work that information in with your intent - what do you want your structure to do? A good engineer creates a model which behaves per the design intent. You don't simply connect nodes with elements and then just sit back and see what ETABS spits out.

If the structure relies on a rigid diaphragm to be stable and you don't want or can't have a rigid diaphragm, then you need to decide how you want the structure to work. That's what engineering is.

 

[milkshakelake]

It could be something as innocuous as a meshing error. Try applying a super stiff diaphragm defined as semi-rigid, like a 20000mm concrete slab or whatever you have with huge stiffness modifiers, and see if you have the same problem. If you still get an error, then it's a modeling problem, not a structural one. If the problem disappears, then I defer to the other commenters.

 

[dold]

I'd start with EZBuilding's point about member end releases. In my experience, instability warnings that occur when switching diaphragm types are mostly due to under constrained joints/nodes. It probably depends on the software and what settings you have for diaphragm out of plane stiffness. In RAM SS if you set 'neglect semirigid out of plane stiffness' (or whatever the setting is), joints may have zero rotational restraint depending on member end releases. I don't use ETABS but in RAM SS it will tell you which joint/node is unstable and in which degree of freedom. Will ETABS tell you this?

An easy way to check that is to save out a copy of the model and fix all member end releases. If that doesn't clear the error then something else is afoot. Or see if ETABS has a flag for neglecting diaph out of plane stiffness and play with that.

 

[Enhineyero]

I suspect that you have beams (or beams modelled as plates) pinned where they need to be rigid in order for the frame to be stable. Hence, when you select a rigid diaphragm the instability goes away. If you don't have beams then check the slab stiffness modifiers for out-of-plane direction, also make sure the beam nodes are connected to wall/column nodes.

If you have a maintenance subscription with CSI make the most out of it and give them a ring, they will help you with modelling issues.

 

[vetchengineer]

Hi everybody,

Thank you for your all replies. My problem has been solved.=)

I have reviewed all the "Numerical warnings I see while analysing with standart solver (6-7 digits warnings I have)" one by one.

Then my unstable warning has disappeared.



In fact when I compared the periods and base shear forces and weight of the buildings are similar in "unstable warning model" and "not unstable model".(for my model)
But as you know it can be very effective on the results.
Therefore; I think we should review all the numerical warnings for our all models with running standart solver. (as I do in old versions (ETABSv9)

Thanks& Regards