Modelling of Boundary elements incorporated walls.

[Muqtadir]

In Etabs, Can I model a wall with columns on it's ends and is it the same as the wall with boundary elements defined in section designer????
 

[WR1]

You dont need to separately model the columns for boundary elements. Read the wall design manual from ETABS help. It explains the design procedure.

[Muqtadir]

But If I do, will it behave like boundary elements?

[UmarMakhzumi]

19 hours ago, Muqtadir said:

But If I do, will it behave like boundary elements?

I have seen some people model 2 columns at wall ends but I don't agree with that approach.

The only way to find this out if the approach of modelling columns at the ends of wall is correct or not is to make two models.

The first model should contain a wall with boundary elements using section designer and the second model without boundary element and 2 columns at the end. Apply a 1 kN load at the top of wall. Calculate the stiffness for both cases. If stiffness is same, both approaches are the same. If not, then they aren't. Calculating stiffness is easy. You can apply see the displacement for each model and divide the 1kN by that displacement. Stiffness is the force required for unit displacement. 

Let us know, we are waiting.

Thanks.

 

[Muqtadir]

5 hours ago, UmarMakhzumi said:

Let us know, we are waiting.

So, I modeled both cases. 

Column Case = 0.00712"

Wall Case = 0.009072"

But I have a query. In the column case, Half the column overlaps the wall part. I believe the stiffness is more due to the extra part of column resisting on both the sides. What are your views in how to deal with this?

P.s: I created a case where I moved the column in such a way that the column doesn't overlap the wall but just touches it. Upon running the software The whole deflection was taken up by column and the column and the wall behaved independently. Whereas when their centre meet, case like in the photo, the column and wall behaved as a single unit. 

[WR1]

There is no question of stiffness in this case. Yes, stiffness will also be affected by the reinforcement but we ignore it. Only concrete dimensions are taken for stiffness.

Second, what is the purpose of providing boundary elements? This should answer your question.

They are not separate physical members, but just a concentration of reinforcement like columns. It has nothing to do with stiffness and modelling. Its a reinforcement detail for ductility.

[UmarMakhzumi]

 

8 hours ago, Rana said:

There is no question of stiffness in this case. Yes, stiffness will also be affected by the reinforcement but we ignore it. Only concrete dimensions are taken for stiffness.

The intent here is to evaluate and analyse the statement that why modelling columns at the end of shear wall in an incorrect approach. This approach is used widely in Pakistan by design consultants that are too lazy to model a proper boundary element where boundary elements are thicker than the wall and in some cases in all situations just to pass the models. One way to prove that this approach is incorrect  is to show that modelling columns is not as same as having boundary elements.  To prove that, you can check the stiffness in both cases to establish that both modelling approaches are not apples to apples and in fact by modelling columns at ends of walls you are doing stuff to pass the model that doesn't make sense.

9 hours ago, Muqtadir said:

So, I modeled both cases. 

Column Case = 0.00712"

Wall Case = 0.009072"

The numbers answer your original question asked in the first post that "Can I model a wall with columns on it's ends and is it the same as the wall with boundary elements defined in section designer????". The answer as you can see yourself is no. Modelling columns is not same as having boundary elements defined in section designer. For the case where you have columns at the end of shear walls, your stiffness is 140.5 kip/inch. For the case where you don't, your stiffness is its 110.3 kip/inch. Adding columns results in an increase of 27% stiffness.  

9 hours ago, Muqtadir said:

But I have a query. In the column case, Half the column overlaps the wall part. I believe the stiffness is more due to the extra part of column resisting on both the sides. What are your views in how to deal with this?

The increase in stiffness due to modelling columns is not justified. Stiffness comes out of E and I and what you should be getting is same as modelling a single wall where end regions are assigned as boundary in section designer. Generally FEA software work from node to node so I doubt if the software would consider visual overlaps.

 

9 hours ago, Muqtadir said:

P.s: I created a case where I moved the column in such a way that the column doesn't overlap the wall but just touches it. Upon running the software The whole deflection was taken up by column and the column and the wall behaved independently. Whereas when their centre meet, case like in the photo, the column and wall behaved as a single unit. 

Well that should be the expected results. If you move columns farther out, the are not connected to wall. 

To summarize, use section designer to define boundary elements. I hope this exercise provided you with some insight about the issue.

Thanks.

[Muqtadir]

On 9/14/2017 at 1:20 AM, UmarMakhzumi said:

To summarize, use section designer to define boundary elements. I hope this exercise provided you with some insight about the issue

It definitely did. Thank you.

[kkmcastillo]

Good day Sir Muqtadir!

Have you tried assigning Pier Labels to column section? This might consider the wall and columns (as boundary element) as one. Though, I havent tried this myself yet.

 

Thank you!