Mat footing settlement 4

[Gus14]

I have an irregular mat footing similar to the attached sketch with light and heavily loaded columns. The lightly loaded columns are around the elevator opening.

For sketch,

https://files.engineering.com/getfi...5-7fc6-4b57-b781-2b5178e997ea&file=Review.pdf

I modeled it in CSI safe and checked the bearing capacity, punching shear, and flexural reinforcement. The soil reaction varies as in the following image:

https://files.engineering.com/getfi...45e8-bf43-35b06a293eb6&file=soil_reaction.PNG

I am worried about the footing differential settlement, and I can't find tutorials on how to model it properly. Do you have any rules of thumb or references to help with settlement calculations?

 

[phamENG]

Talk with the project geotechnical engineer. They should be able to help you with a sub grade modulus, which is essentially a per square foot spring stiffness. They can also take a look at your pressures and let you know how much of a concern settlement really is.

 

[Boiler106]

while youre waiting on the geotech, make sure youre not using the subgrade modulus they gave you for the slab which is based on much lighter loads. depending on where you are, your modulus will be much, much less than the slab subgrade modulus.

 

[JLNJ]

Maybe you can de-couple the two sections of the mat so that your bearing stresses are more uniform and that narrow section of the mat doesn't have to work so hard. It will also make your settlement calcs more simple.

 

[driftLimiter]

make sure you clearly understand what is meant by differential settlement. The soil subgrade modulus takes into account the elasticity of the soil in compacted state. The settlement can occur overtime due to different causes and is not necessarily related to subgrade modulus (by my understanding).

The spring constant provided in subgrade modulus is for immediately deflections under loads.

 

[Gus14]

Thank you, everyone, for replying.
The Geotech engineer told me to assume that soil subgrade modulus equals 120 x q allowable. He did not give me any other value for the subgrade modulus. He also did not address my concerns with the differential settlement stating that the soil sample only represents itself. Moving on

The spring constant provided in subgrade modulus is for immediately deflections under loads.

So I can only use this subgrade modulus (120x350 kn/m.s) to calculate immediate settlement, while the long-term needs a more advanced study. CSI safe deflection under service loads shows a 7 mm differential settlement along the entire footing, with only 4 mm varying along the elevator columns. Not sure how much I can trust these results.

de-couple the two sections of the mat

Do you mean making them into two separate footings?
My concern is that the footing with a heavier load soil pressure bulb will spread below the one with lighter loads causing the same settlement and complicating construction.

After some research, I feel since the soil is dense sand without any presence of groundwater. The long-term settlement should not be excessive. Do you agree? However, I can't help but think about if the soil was clay and groundwater was present. But maybe that's a day for a more experienced geotechnical engineer. I wish soil testing firms were more sensible.

Do you recommend any reference to start learning foundation settlement with some software applications? At least to know what is reasonable and what is not.

 

[SlideRuleEra]

1) The Geotech engineer... did not address my concerns with the differential settlement...

I wish soil testing firms were more sensible.
​

2) Not sure how much I can trust these (software) results.

1) Not surprising, the problem is one that lies in gray area that combines Structural and Geotechnical.

2) You should trust the software result, output is "screaming" that something is wrong, and there is. The mat is not even close to uniformly loading the soil:

The mat is actually a combined footing, with five point loads... but has not been sized/shaped with this in mind.

Combined footing design is complicated with one of the loads being 87% of the total load. The other four loads total 13% of the total.

A satisfactory combined footing (uniform pressure distribution to soil) will be much larger and it's footprint will be well outside the limits of the current mat. I would consider putting the 2000+ kN load on piling.

 

[Gus14]

Thank you SlideRuleEra for replying,

You should trust the software result

I was referring to the settlement results calculated by the software, and not the soil reaction.

will be much larger and its footprint will be well outside the limits of the current mat.

Yeah, that's the problem. I can't really stretch it enough to ensure uniform pressure distribution, so I thought about the differential settlement. That's why I avoid placing any columns near the elevator columns because I know their footing will be oversized to encase the elevator machine with barely any soil reaction.

But don't you think this problem applies to every building even if the footings are placed 3 meters apart since all columns will be connected with ground beams and roofs with the elevator footing slightly settling and the other footing settling more?

 

[HTURKAK]

- This is not a mat/raft foundation but an irregular combined ftg..

- The prediction of differential settlement requires geotech study .. You cannot estimate with deflection of Winkler springs..

- I will suggest you to design two separate spread ftg . as suggested by JLNJ (Structural) and a BPS for his post..







Tim was so learned that he could name a
horse in nine languages: so ignorant that he bought a cow to ride on.
(BENJAMIN FRANKLIN )

 

[Gus14]

I will suggest you to design two separate spread ftg . as suggested by JLNJ (Structural) and a BPS for his post..

Thank you HTURKAK for replying, it's easy to separate them, but,

1. I don't really believe this reduces differential settlement since the light footing will just tilt towards the heavier one as it settles more.

2. It will complicate the formwork, and the contractor (I have seen this before while working in supervision) will likely join them and not provide any top reinforcement.

 

[SlideRuleEra]

1) I was referring to the settlement results calculated by the software, and not the soil reaction.

2) That's why I avoid placing any columns near the elevator columns...

3) But don't you think this problem applies to every building even if the footings are placed 3 meters apart since all columns will be connected with ground beams and roofs with the elevator footing slightly settling and the other footing settling more?

1) Differential settlement is the problem and that is what the software output "tells' me, subgrade modulus is the least of the issues. Differential settlement will remain no matter what the subgrade modulus value.

2) Significant column loads (compared to 2000+ kN) near the elevator would actually help resolve the problem. Then a reasonably sized/shaped combined footing may be designed. The problem with the current plan is the the loads (2000 kN and 4 each, smaller loads) are both grossly different and spaced far apart (compared to the desired foundation footprint dimensions).

3) I do not agree, where I worked (poor soils), problems of this type are avoided by using foundations appropriate for size of loads... footings for "light" loads, piling and properly design mats for "heavy" loads.

Individual footings will solve the differential settlement problem for that one column... but this brings up another issue. Footings for the 70 and 100+ kN loads need to be "small" for proper soil contact pressure (kN/M²). The footing for the 2000+ kN load will have to be "large" for proper soil contact pressure. Soil contact pressure for each of the footings needs to be approximately equal, otherwise, differential settlement (of one footing compared to the others) remains a problem. This can be done, but getting it all to fit within the current mat's footprint will be a real challenge... there is likely not enough area to make it happen.

What you want to do (five-load combined footing) is a geometry problem, and size/shape of the current mat footprint do not give enough space to solve it. IMHO, piling are the solution to fit the footprint.

 

[Gus14]

Thank you, SlideRuleEra, for sharing your experience. This discussion was very informative to me.

 

[SlideRuleEra]

Gus14 - You are welcome.
Best Wishes.

P.S. You mentioned that you had avoided placing columns (load) near the elevator.
If it's ok with you, in the software model shift (subtract) 420 kN from the heavily loaded column and add 105 kN to each of the 4 existing elevator columns. See what this does to soil contact pressure distribution for the entire footing.

 

[HouseBoy]

I would think that to combat differential settlement, you'd want the bearing pressure to be relatively uniform throughout the area of the footing For starters....can you find the location of the "center of action" for the dead loads and then proportion a footing for which the geometric center is very close to that location?

Might want to compare the location of Dead load center to Dead + Live load center.
Dead loads seem to be the more critical (since they are always there) AND they are somewhat larger so... I'd guess "center of footing area" would be closer that location.