Punching Shear for a Transfer Slab 1

[MHconc123]

I have a fairly deep transfer slab carrying column point loads.
As expected shear is the critical issue in most areas.
When determining the shear on checked shear perimeters is it still fine to assume the load acts uniformly around the full perimeter. There are situations where the columns over are primarily to one side of the supporting column.
The moments are taken into account etc but I am trying to convince myself that this would still be applicable. On the one hand it feels like the column load and hence shear stresses would be concentrated on 1 face of the supporting column but on the other hand, to fail in opunching shear the concrete slab on the otherside of the column would have to come into play first.
Any thoughts or similar experiences would be greatly appreciated

 

[rapt]

MHconc123,

The "beam shear" check that you also have to carry out will cover the fact that the point load is on one side only.

Generally, punching shear calculations are independent of the locations of the loads on the slab. Because the load is on one side only, you should get much larger column moments than you would for loads distributed on both sides which will significasntly affect your punchinng shear calculations.

 

[BAretired]

The load does not necessarily act uniformly around the full perimeter. Shear failure can occur on one side before the opposite side has been exposed to significant shear stress.

BA

 

[slickdeals]

See page 318-200 of ACI 318-08. The figure in the commentary shows what you are referring to.

We are Virginia Tech
Go HOKIES

 

[ron9876]

I agree with rapt. I have had people argue with me that the EFM applies to everything but I don't think so.

As I have mentioned on this site before always be careful with transfer beams/slabs because fixing either your mistake or the contractors mistake is very unnerving.

 

[rapt]

Ron,

How does EFM (I presume you mean Equivalent Frame Method) come into this?

 

[BAretired]

I thought he meant Finite Element Method but got his letters crossed. This suggests we should stop using acronyms when we are not certain that everyone is on the same wavelength.

BA

 

[ron9876]

I did mean equivalent frame method. There are people that would argue that you can use the full bay width to resist applied point loads from columns and therefore only punching shear would govern. I don't agree with that assumption at all.

 

[rapt]

Ron,

There were a couple of PT software people (whose names I will not mention, but they are regarded as "experts" in PT in USA) expousing that theory. I have had a lot of arguments with them over the years. Then, those same people do not recognise column/middle strips in PT flat slabs.

It really depends where the point loads are when using EFM. If they are in the middle strip, then their efefcts are felt through both the middle and column strips. If they are in the column strip, they should only be included in the column strip.

 

[hokie66]

MHConc123,
Was your question answered to your satisfaction? The answers are all correct, but the one by slickdeals shows it in pictorial form, if you have the ACI code. In summary, you can't just assume the shear stress around the column is uniform, you have to consider the real situation.

 

[rapt]

Hokie66.

The diagram on page 200 of ACI for punching shear is showing the distribution of shear stress due to the combination of an axial force V and a Moment Transferred to the column, M. It makes no other allowance for the distribution of the axial forces on the slab.

The Force V is the total reaction on the column.

The Moment M will vary depending on the loading condition and it is the only variable in the code method that alllows for non symmetric loading and its effects on the punching shear stresses around the column.

If there is a very large point load on the span on one side of the column and none on the other side, there will be a much larger moment transfer to the column resulting in a larger variation in shear stress around the column.

 

[hokie66]

rapt,
I understand that. Did something I say lead you to think otherwise? I was just trying to help MHConc123 understand your earlier comments and those by others.

 

[csd72]

This is a very good example of where awareness of the inherent assumptions is crucial.

As a designer you should be checking ALL possible shear perimeters.

With a simply supported slab it is clear that the critical shear is near the support, with a uniformly loaded flat slab it is also clear that the critical punching is just near the column.

In your case it is not so obvious what the critical case is so you should be checking ALL possible failure modes and only discounting those modes that you are confident are not critical.

You may find that strut and ties analysis is more approriate when the top and bottom columns are close together.

 

[MHconc123]

Thank you for all you input.

csd72 I am not 100% sure what you are referring to.
You would always need to check all shear perimeters outwards from the column face until the passes.
When determining the critical 'faces' of the shear then this is usually obvious pretty quickly based on the bending moments and the length of the column face/shear perimeter length. I agree that the calculations need to be done to a point that the critical face can be confirmed however.

I feel that in my situation that Rapt's response is most relevant. Whilst most of the point loads are fairly evenly distruibuted, the moments obviously increase significantly where they are not and hence have a signifianct impact on the calculated shear stresses.
Beam shear will obviously also be checked but so far it does not appear to govern in any areas.

 

[BAretired]

Beam shear is one possibility and should be checked.

Punching shear needs to be checked with a non uniform distribution of shear stress around the critical section to account for the presence of bending moment.

BA