Eccentrically Loaded Spread Footing Design Question 4

[ilikeconcrete]

I am designing a spread footing for a column that is part of a moment frame and am assuming that the column/footing joint is rigidly connected to reduce the deflections of my frame.

Say my service-level gravity force is 300kips and the associated moment is 150 ft-kips.

When checking bearing capacity, I can calculate my eccentricity to be e = M/P = 150/300 = 0.5ft and determine my bearing stress distribution.

However, I can't find a reference that comments on whether my factored bearing pressure distribution used to design for shear and flexure should be calculated using e = 0.5ft (from service-level forces) or if I calculate a different eccentricity associated with my factored loads such that e = Mu/Pu.

Can anyone point me to a published reference that I can include with a submitted calculation book?

Thanks.

 

[UcfSE]

I understand the points made and have debated with myself before. Either way, ask yourself, is it any more reasonable or accurate to use a factored e and factored loads in an equation that probably was not meant for ultimate load conditions? Do you violate any assumptions by doing so?

 

[ron9876]

I have always struggled with this also. In my opinion you have to use service loads to check service allowable soil stresses. But if you use ultimate loads to design the footing you probably will get a different soil pressure distribution which isn't correct. I use service loads, calc soil stresses and apply a psuedo load factor to the soil stresses to design footing. This also isn't strictly correct but you can use a multiplier that you know is at or above the LRFD stresses.

 

[structuresguy]

I never said anything about using factored loads to check bearing capacity. But to design the footing, you have to. If the factored loads change the e, well so be it. I don't think it would make that big a difference in the reinforcing of the footing to worry about it. We have all wasted more money (in man hours) talking about it, then it would have cost to add an extra piece of rebar or two.

My point is, there are many things we need to worry about. But a little extra reinforcing in a footing is not one of them. Many engineers analyze things to death which are relatively inconsequential to the overall performance of a structure.

 

[kslee1000]

One old practical method to design the footing with moment (assume concrete footing rest on soil):

1. Obtain service level soil stresses.
2. Multiply the service load stresses by a factor of 1.6, or 1.65 if LL>>DL, (since at the old time the DL factor was 1.4 & LL factor was 1.7)
3. Design the concrete footing use the factored stresses.

This method is acceptable for eccentricity/moment caused mainly by the transient loads (wind/earthquake). For permenent eccentrical loadings, you will need to find other solutions, best is to eliminate the eccentricity at the first place (using piles, ties, offset footing pad...).

 

[ron9876]

kslee1000 this is what I mean. If you use factored moments to calculate loads on piles in pilecaps for instance you may or may not get a tension load in end piles which will require tension reinforcing in the pile and top reinforcing in the pilecap. You very well may not get acceptable results when you use factored loads to determine pressure distribution under footings. This is especially true with moments from lateral loads such as shearwall pilecaps.

 

[kslee1000]

Add'l thinking on the method I described previously:

By doing so, you are not necessarily design the concrete footing to withstand the whole factored loads (which is at the limit state), but you are sure that your footing will not fail before noticeable cracks suddenly developed on the walls/floors due to excessive progressive/permenent settlement, since the footing was sized against the allowable (elasticsteady state) soil pressure.

The unit load factor also ensure the footing has adequate strength to resist pressure above the presumptive static soil pressure at the split second of peak reversible loadings.

 

[Lion06]

ilikeconcrete-
I would make a very large bet that if your e for service loads is 2.5' from centerline of a 6' square footing (the resultant is acting 6" from the edge of the footing) that your allowable soil stresses will rarely work.
This is not a hard discussion and I'm not seeing a case for using anything other than the factored loads. the factored loads are what they are. The e just comes along for the ride.

 

[ilikeconcrete]

StructuralEIT-

Trust me, it happens. Design a spread footing for a shear wall.

 

[Lion06]

I have designed footings for shearwalls, but the OP was stated as a spread footing for a lateral column.

 

[theonlynamenottaken]

Hmmm, very interesting. I read all of the above and ran a handful of scenarios with varying DL/LL and Moment/Axial ratios through my 'Footing with Moment' spreadsheet.

The two methods:
1) Determine soil bearing pressure with Service loads and pseudo-factor those pressures up to determine Factored level internal forces in the footing.
2) Determine the internal footing forces by using Factored level column forces, and the associated Factored level soil pressures.

None of the cases I ran showed more than a 10% difference in the critical internal footing forces. What was interesting was that the Overturning Safety Factor was up to 20% less when using method 2 (for higher Moment/Axial ratios). Of course there is some small error because my spreadsheet does not use any factoring for calculating footing self-weight.

I most agree with something kslee1000 touched on. The 10% discrepancy tends to indicate that the method selection does not substantially impact footing internal forces. I would focus my efforts on forcing the failure mode to behavior that warns the inhabitants of impending failure. Its scary to think that an overload situation would either break the footing or overstress the soil - because if that happens and the 'toe' moves... overturning stability disappears quicker than an Architect after receiving an invoice.

 

[Gumpmaster]

Use service level loads and design the footing with ASD from ACI 350. It's conservative and you only have to analyze the footing once.

 

[ron9876]

If you are designing a shearwall footing for mostly overturning moment and limited applied dead load where the footing weight has to be sized to provide the needed dead load to resist overturning(0.6D+W)and limit soil stresses you will find a substantial difference if you do the same calc with factored loads.
I don't think that designing this condition with applied factored loads is acceptable. It frustrates me that the building codes don't address things like this.

 

[asixth]

I may be bringing this post back from the dead but generally I design the bearing capacity of the footing for service loads, making sure the allowable bearing capacity of the soil is not exceeded.

When designing the base for flexural steel, I then assume that the full allowable bearing pressure is developed over then entire base, convert to ultimate loads (i.e. multiply by 1.35) then design the flexural steel based on this loading.

 

[oneintheeye]

i've been thinking about this on weekend and its not quite as simple as people are suggesting for some cases. Imagine the loading puts servicability outside middle third, you have a base area in tension. then if you factor your loads for RC design the e increases further, therby theoretically increaseing the area of base in tension and reducing your bearing area. You then have a possible condition where you may be able to design less or more steel depending on which e you use. i.e. tension steel in top or bottom over different zones.

 

[hokie66]

I agree with Gumpmaster and/or asixth. Why make this so complicated?

 

[KBVT]

If you are finding that factoring your loads gives you opposing effects than your service loads, than I would investigate the load combinations with one or more loads not acting as stated in ASCE-7.

 

[miecz]

It seems to me that I've had cases where using factored loads to calculate the moments resulted in dramatically higher moments than calculating service load moments, and then factoring those moments. It's been a long time, but I think using factored loads changed the bearing profile dramatically. It may have had to do with the eccentricity being within the middle third for service loads, but not for factored loads. In any case, I concluded that the intent of the code was to provide a factored strength derived from service loads. So, for the concrete design, I calculate moments and shears for service loads and then factor those results by 1.6.

 

[DRC1]

The concrete is designed on factored Loads. The Footing is sized based on allowable loads. "e" does change for factored vs. unfactored loads. Footing stabilty and bearing capacity are still more akin to allowable stress. For step by step see Bowles "Foundation Analysis and Design"