Basement Foundation design using Piled Ground Beam

[JohnnySm]

Hi Guys,

I have a few questions regarding foundation design for a basement support structure.

Please, first review the attached to familiarise yourself with the geometry/scope of my questions.

The scheme, at this point is to create a basement retaining wall box structure with slab over and suspended slab (using the ground floor under as permanent form work) for the basement floor. All loading from the above slab and superstructure over is to be distributed into the perimeter and spine RC walls onto underlying ground beams supported by a collection of equally spaced piles. NOTE, we have chosen to technically model the basement slab as a suspended floor not as a raft foundation as there is no sporadic column arrangement ( it seems the more value engineered method as a piled raft slab will require substantial reinforcement/depth to redistribute the loading, we can discuss this also).

My main question regards how to go about modelling/designing the ground beam element. The usual way i would consider is to assume a simply supported continuous beam using the pile positions as supports. However this method does not incorporate any of the additional effects from varying stiffness due to the sub grade modulus from the pile.

Do i just assume rigid stiffness and increase the required rebar for a conservative estimate on the ground beam or do i need to do vigorous analysis and model the supports as springs? (FYI im still new and haven't actually done spring support analysis) blame my university. any matierial on this would be splendid.

FYI, i have done research on winkler foundation and elastic foundation design. It seems very complicated. is this really necessary?

Also the FEA model for the superstructure is done on Tekla Structural Designer. Not sure if there's a way of inputting the values for pile stiffness K value.

Sorry its a long one!

Cheers guys!

 

[JohnnySm]

see attached for the other sketch.

 

[bhiggins]

If the piles are bearing on solid rock, it seems correct to neglect subgrade modulus as long as you meet allowable bearing capacity and the beam spans between piles. If the piles are bearing on a compressible material then it makes sense to model the subgrade modulus.

I'm not a geotechnical engineer, so in either case, I would contact the project geotechnical engineer for his recommendation and go with that.

 

[BAretired]

Unless the piles were unusually stiff relative to the grade beam, I would ignore the pile stiffness and analyze the grade beam as if it were bearing on hinged supports.

BA

 

[JohnnySm]

BA Thanks. So we are expecting RC concrete ground beam with CFA RC piles, therefore i should model the beams as a simply supported continuous member( treating the beams as suspended)?

Just out of curiosity in what scenario should i be considering pile/subgrade stiffness/modulus?

Thanks.

 

[jdgengineer]

Personally, I would ignore the pile stiffness when you assume that the piles are significantly stiffer than the gradebeams and design the piles as pin supports. I believe this is what BAretired meant as well.

For small projects it's our experience it's typical to assume the piles are pin supports. This makes the calculations significantly more simple and is generally sufficiently accurate.

It's our preference to receive pike stiffness from the geotoech and to model the stiffness directly in our model. We use RISA Foundation for these designs and it's pretty simple to include.

It's been our experience that when we have a pile supported slab that we neglect the subgrade modulus. Generally, you are using piles because the geotech does not want to support the foundation directly on the subgrade. With a rigid pier support, unless your gradebeams are very flexible or your subgrade is very stiff it shouldn't take very much load anyways. I'd check with the geotech to see what is appropriate for your project.

Often times the geotech is doing the pile analysis in LPile and you are supplying them with your reactions. In these instances they may be more receptive to give you pile stiffness.

On smaller jobs where this doesn't happen, we've found the geotech to be reluctant to provide this stiffness. If they do give it, ideally I'd like to see it bounded with lower and upper expectations. As the piles are less rigid you will introduce more demand into the gradebeam. As the piles are more rigid you will introduce more demand into the pile. Bounding it should cover worst case of these scenarios.

In reality, the soil profile may vary across the site and the stiffness values may not be very accurate due to unknown soil characteristics. Just understand the limitations and design accordingly.

It sounds like your project is quite substantial. You should be working closely with the geotech to determine appropriate modeling and foundation approaches.

 

[jdgengineer]

Looking at your attachments, I think your retaining wall will act as a very deep beam capable of distributing load to multiple piles. If your loading is uniform it won't matter but if you have concentrated loads you may want to try and model pike stiffness to take advantage of the load sharing between piles.

 

[BAretired]

For grade beam design, it is conservative to ignore pile stiffness except at the outer ends of the end spans where it may be prudent to include some additional top reinforcement in accordance with engineering judgment.

Usually, the steel cage in a concrete pile is positioned with less precision than is the case with a concrete column, so it may be inadvisable to rely on piles to provide significant moment resistance. The piling crew are working in muddy conditions and cannot be expected to achieve the same workmanship as the general contractor. Sometimes, the top of pile must be kept low because of site conditions. Frequently, piles or pile caps are doweled into grade beams with nominal dowels which are incapable of taking significant moment.

In general, I would not rely on piles to resist calculated moment for a variety of reasons, some of which I mentioned above but there may be exceptions; can't think of any offhand. Offset piles are a special case which need to be carefully considered and properly reinforced but fortunately, they are relatively rare on most job sites.

BA

 

[JohnnySm]

Thanks guys,

The concrete retaining structure does support additional columns loads from above, so concentrated loads are applied but then distributed down the walls. In any case and for construction simplicity we are going to allow for simply supported single span ground beams supported on hinge joints which will provide the platform for the retaining walls over.. This is in reference to you BA as i assume your above comments refers to my mention of continuous beam analysis, i.e i should model single span to avoid hogging moments that the pile might not be able to cater for?

I will liaise with the piling engineer and see what he says, i haven't had to ask this question before so he might be reluctant, and if in the case that ignoring pile stiffness is conservative it should not make too much difference anyway? BA could you expand on your reasoning for additional top reinforcement in the corners?

Cheers

 

[BAretired]

The concrete retaining structure does support additional columns loads from above, so concentrated loads are applied but then distributed down the walls. In any case and for construction simplicity we are going to allow for simply supported single span ground beams supported on hinge joints which will provide the platform for the retaining walls over.. This is in reference to you BA as i assume your above comments refers to my mention of continuous beam analysis, i.e i should model single span to avoid hogging moments that the pile might not be able to cater for?

I would consider each beam as continuous resting on piles having zero stiffness. Hogging (negative) moments would not be shared by the piles but would still exist in the beams. Using simple spans may be okay except where there are continuous walls bearing on them but I see no advantage to using simple spans anywhere.

I will liaise with the piling engineer and see what he says, i haven't had to ask this question before so he might be reluctant, and if in the case that ignoring pile stiffness is conservative it should not make too much difference anyway? BA could you expand on your reasoning for additional top reinforcement in the corners?

The end piles will take moment whether you like it or not, the magnitude depending on the amount of reinforcement continuous between pile and beam. This will result in negative moments in the end spans which should be reinforced to limit cracking.

BA