Transfer Slab Design 3

[damo74]

Hi all. I'm designing a transfer slab that is supporting 2 storey blockwork dwellings. The transfer slab is 400mm tk. and is supported by r.c walls and columns in a basement car park. I'm using finite element analysis software to design it.

I have taken the worst case loading over the complete structure. This is 1.4DL + 1.6IL (as per BS8110 Part 1). However, I'm wondering whether I should be looking at pattern loading. My DL is 14.2 kN/m2 including self weight. My IL is 3.0kN/m2.

Due to the layout of structure and unsymmetrical supports I would like to take full loading over the entire structure, and ignore pattern loading. The code does suggest that I can do this if certain conditions are met (Clause 3.5.2.3). My slab does qualify for these conditions, but at the end, it suggests that all the resulting support moments be reduced by 20%, with a consequential increase in span moments. This would prove very difficult to achieve as it is the FEA software that I have used.

Has anybody any experience in using this clause of the code? If not, does anybody have any logical suggestions that would allow me to ignore pattern loading?

 

[FalsePrecision]

Since the live load is so small, compared to the dead load, then why is pattern loading such a problem?
If you absolutely need to avoid pattern loading, then perhaps increase your live load by a conservative percentage, and state so in the calculations. (Hoping the plans checker will concur)

 

[damo74]

Thanks FalsePrecision. I agree with you about the difference in the dead to the imposed loading. I was just wondering if anybody had any guidelines to go by.

 

[JAE]

From a US perspective (which probably doesn't directly apply to your situation), most US codes require pattern loading in floors if you have continuity in your framing.
If you ignore the pattern loading, you are ignoring a true-life condition. That said, as FalsePrecision mentioned, the actual effect of the pattern loading may not be that much...but you should make a judgement on that only if you know the true effect.

As far as redistributing moments in a finite element model, one thought would be - don't redistribute the moments, rather, calculate your final design top and bottom reinforcing that is required by the actual FEA model and then re-distribute the percentage of the areas of steel...they are linearly related.

 

[rapt]

Are you designing for the actual positions of the walls above and their loads as line loads or distributing it as a UDL?

If they are distributed as UDL's I would definitely allow for patterning of both the LL and the extra dead load from above the slab. Extra dead loads are only treated as such if they are of known value and location, otherwise they are live loads.

If you are allowing for the real load positions and values I would not worry about patterning it.

How are you calculating your deflections using FEM?

 

[damo74]

Rapt,

Dead loads of all walls are in exact, real locations, inserted as line loads. The deflections are calculated using the finite software. The software gives me the total non-linear deflections.

Thanks for the good advice.

 

[pba]

When I did something similar, I reasoned that the 20% reduction in support (hogging) moment could safely be ignored as this would lead to a conservative design. The corresponding increase of span moment is going to be less than 20% so you could conservatively create a load case 1.2x(1.4Gk+1.6Qk) for the design of the bottom rebar. - It worked for me...

NOW - 3.0kN/m2 is 1.5kN/m2 on two floors - What about roof loads? - usually 0.6kN/m2 snow plus 0.25kN/m2 access. If there are access ways to the buildings there may also be higher loads associated with these.

It is also common to find that domestic buildings are modified during their life with resulting changes in loading. Even if the client is adamant that this will not occur, I'd try to ensure some spare capacity to cover this.

 

[rapt]

damo74

By non-linear do you mean they include for cracking as well as creep and shrinkage deflection effects?

 

[damo74]

rapt

Yeah. It is the total deflection.

 

[HungryHorace]

damo74,

FEA is right on the borders of BS8110 guidance and therefore some engineering judgement is required. For FEA on something like a transfer deck, I do one of two things:

1) Draw up a checker-board on the supports, and use these as the basis of alternate loadcases. Where irregular bays hinder this, I use judgement at each location to input the worst case scenario.

2) Perform a risk assessment on the likelihood of pattern loading being a critical factor. In some locations or loading conditions, this may not be likely. In addition, for a transfer slab I imagine that you are already using load takedowns which are on the conservative side. This is likely to have more effect than pattern loading on the deck itself.

I would be interested to see how other UK engineers handle this area, as there are several grey areas between FEA and BS8110.