foundation stiffness 1

[lsuengut]

I am designing a spread footing for a large fan. The vendor has specified a required foundation stiffness of 1,000,000 lb/in. I have never designed a foundation with a stiffness requirement. Any guidance would be appreciated.

 

[frv]

I would assume the subgrade would be an order of magnitude (or more) more flexible than the foundation itself. I think you have to get the geotech to provide you with a stiffness estimate, then limit the pressure on the subgrade is order to limit the deflections to what is allowable.

 

[ishvaaag]

Well, either you can with your ideal setup, or you can't. Since it seems to ask for a stiffness in a spring constant basis, usually a modulus of subgrade reactio multiplied by an area, see if the geotech data adapted for the type of soil and spread of the footing or mat delivers the asked stiffness. This must not be difficult since in the range of what footings of 2 meter side can deliver in sound soils, much easier for good and very good soils.

If not, you may try to deliver the stiffness by using a piled footing or mat. You normally will be then in a very bad soil, that you may need to bridge through the piles to some tip resistance. Piles working by tip resistance to rock or stiffer soil are better to the stiffness purpose, because floating piles in a very bad soil will go with the soil under the load, at least far more than those firmly set at their tips.

If you can't use but floating piles, try to ascertain what the force-deformation for the piles will be from your geotech, specifying the pertinent previous and on works tests. Also, if the asked stiffness is also wanted in tensile action (the footing separating from the soil), piles are inmediately of order, and then you also shall specify the required tests to get the geotech "warrant" to the required reliability.

 

[WARose]

Something is wrong with that number. Most soils I am use to dealing with have a sub grade modulus of 100 to 250 lbs/in^3. I don’t know if you could provide that level of stiffness with a concrete block sitting on rock. I’d question that.

 

[structSU10]

It does look high but if you had a sufficiently rigid 7'X7' footing, loaded uniformly, on 150 pci soil you would achieve this value. If you then take the 1000000 lbs and divide it over that area, you get 20.4 ksf. You would likely design to a much smaller bearing value, so it seems reasonable to achieve such stiffness. If you add in unbalanced loads, you would probably find a larger size is needed.

I would model it in FEA, with my footing supported by appropriate springs. If i get less than 1" of displacement aynwhere, or 20 KSF (again very high) then i would consider this requirement met.

This does seem like an odd way of specifying, yet there are likely other factors, maybe dynamic or something, that led them to this number.

I'm no geotech though, so there might be something i am overlooking, its just my 2 cents.

 

[WARose]

It does look high but if you had a sufficiently rigid 7'X7' footing, loaded uniformly, on 150 pci soil you would achieve this value.

If you looked at it as a whole: yes. But if this thing is [for example] supported on piles that socketed into rock, the flexural stiffness of the supporting mat may not provide that level of stiffness at certain locations. It may be important to ask the fan manufacturer about that.

Another thing to consider here is: with just about any foundation for a fan I've done, I've been forced to do a dynamic analysis for the unbalanced load. He may want to ask the manufacturer if this needs to be considered here and (if so) what the unbalanced forces are. (With a lot of those guys, you can't get that info; therefore you are forced to assume [in the case of a fan] an eccentricity for the rotating mass.) And if you have to do a dynamic analysis, it will then be necessary to get dynamic spring constants from the geotech.

 

[ishvaaag]

Just as a comment, I have found in practice the less conservative appraisals of the modulus of subgrade reaction -and formulas to adapt it to some size- the more consistent with the actual behaviour of the soil, i.e., the actual response. Other thing is you ask from soils characteristic values, those that to be attained may overpredict average settlement by one order of magnitude. As interested in the actual behaviour, as the traditional practice in foundations has been, I normally would stay away of such paths, but sometime maybe it simply won't be commendable to so proceed.

In any case, in the simple way of tackling the problem, just a matter of juggling the data and formulae.