PEMB foundation uplift

[jeffhed]

I designed a foundation for a PEMB. I want to know peoples methods for designing the footings for uplift. I have typically designed the footing to weigh as much as the net uplift load (0.6D+(W or 0.7E)). The calculated net uplift force has a factor of safety of 1/0.6 = 1.67 in it from the reduction of the dead load. I received a plan check comment that states that with my calculations there is only a factor of safety of 1. But this is a factor of safety of 1 in regards to the net uplift force. Should we really be increasing the footing weight to maintain a factor of safety = 1.5 over the net uplift force? To me this is more like a factor of safety of 1.67*1.5 = 2.51. Not to mention we know what the actual weight of the footing will be. This is a commercial building and the city inspectors are very good in requiring construction matches the plans, especially foundations. Am I wrong? Do others design PEMB footings that weigh 1.5 times the net uplift force?

Thanks in advance for your replies

 

[jeffhed]

JAE,
I am confused by your last post. Didnt you state earlier that the F.S. = 1.5 is no longer required since it is taken care of in the load combination? That you would expect the dead weight of the footing, slab, etc to be equal to the net uplift calculated from 0.6D+W? If so, I am being more conservative than that by reducing the footing weight by 10% and comparing to the net uplift from 0.6D+W. Maybe I am misunderstanding one of your posts.

 

[jeffhed]

Maybe to clarify. For this case where a steel building has a net uplift force. I am arguing the following to anchor the building for uplilft: 0.6D+W <=(Footing wt + slab wt).

 

[JAE]

jeffhed - no, you are not correct in somehow assuming that the footing is not a dead load.

You are checking the footing and column against overturning. Your column has dead loads coming down on the footing and the footing itself is a dead load too.

The check is against UPLIFT of the column/footing assembly.
At least that is what your original post said:

I want to know peoples methods for designing the footings for uplift

So you sum up ALL your dead loads - column dead load reaction, footing weight, column pedestal weight, any tributary slab on top, any earth load above the footing pad, etc. That is your total dead load D.

Then you take your wind uplift on the footing, W

The sum - 0.6D + W should not result in a "net" uplift.

In other words 0.6D > W

 

[jonathanwilkins]

I read a publication by metal building manufacturer's association recommending footing design be done using 70% of the wind loads. They did, however, recommend designing anchor bolts w/ full design load.

This struck me as very practical, however, in violation of the code. I've used it in medium spans where the footings where otherwise enormous. I have a hard time visualizing a 6x6x2 mass of concrete flying throught the air w/ the building intact. In small buildings I haven't worried about it. I don't design anything that's really large.

(I interperet the code to read 0.6D+W as the correct load combination for getting your required concete weight).

 

[BAretired]

I'm not sure why there is so much confusion about this issue. Whether you use Limit States Design, 0.9D = 1.5W or Allowable Stress Design 0.6D = W, it amounts to the same thing.

In Limit States Design, the 0.9 factor accounts for overestimating dead load. If 150 pcf is used in the calculations as the density of concrete, it is probably an overestimate. Concrete weighs between 140 and 145 pcf. The 0.9 factor accounts for such overestimating of dead load. The factor of safety against overturning is 1.5.

There have been a number of questionable arguments presented elsewhere on this forum suggesting that safety factors in the codes are combined in such a way as to provide an overly conservative design. In my view, those arguments are fallacious. The codes that I know provide a reasonable and proper way of providing an adequate level of safety to the users or occupants of the structures governed by those codes.

BA

 

[WillisV]

BAretired - the confusion comes in when someone (read code official or other reviewer) who is not familiar with the built-in 1.5+ safety factor of the 0.6D+W load combination requires you to satisfy a 1.5 safety factor IN ADDITION to the safety factor provide by the load combination. e.g. 0.6D=1.5W which is, of course, wrong, but ingrained in some who are looking for that magic 1.5 number to appear in plane numerics and not be buried in a load combo.

 

[dcarr82775]

I agree with BA, the code is perfectly clear on this issue, there is no confusion.

If a plan reviewed gives me a comment in which I know he is not following the code (i.e. WillisV's example) then I will show him the error of his ways because I know he will not find a code provision that supports his argument. If that fails I go to his boss who hopefully is smarter. The code official always comes around when I can show they are wrong.

BA, I have always understood we use 150pcf to account for the heavier weight of steel than concrete. I have never bothered to do a calc to see if that is realistic or not.

 

[jeffhed]

JAE,
My argument is not that the weight of the footing and slab is not dead load. It is. In the interest of getting the most economical design, my argument was that the dead load of the footing is predictable. Also, the steel building base plate and anchor bolts are designed to transfer the uplift load calculated from 0.6D+W of the STRUCTURE at the base plate. So I was wondering if the structure was designed to transfer the uplift force calculated from 0.6D+W of the structure (Anchor bolt design force), why increase the weight of the footing above and beyond the anchor bolt design force by multiplying footing weight by 0.6. If i should multiply the footing weight by 0.6 and have even more enormous footings than normal, that is fine. I only want to do the best economical design but also I want to do it correctly. I posted this question specifically to determine if I am crazy or not. It appears I may be crazy.

 

[slickdeals]

Also, the steel building base plate and anchor bolts are designed to transfer the uplift load calculated from 0.6D+W of the STRUCTURE at the base plate.

I think it is incorrect to provide base plate and anchor bolts to resist 0.6* superstructure weight. You should be consistent in your design assumptions. By doing the above, you have essentially created a weak link in your load path.

 

[jeffhed]

slickdeals,
I made my statement regarding which load combination would result in the largest uplift force and anchor bolt tension force. Wouldnt 0.6D+W result in the largest uplift force?

 

[a2mfk]

I'm confused by that statement too. May I offer a summary of this entire discussion, since its been plaguing me my entire career here in Florida and the SE US:

1. Letter of the codes at the moment, all dead loads are put into D, including slab and footing weights, for stability analysis:
0.6D + W

2. Same goes with anchor bolt design, minus the weight of the footings and slab of course.

3. Many SEs (I know, not all BA ) on this board and other see the need for ASCE/IBC to perhaps consider a separate, less conservative factor for slab and footing weights, that may be something like:

0.6D + W + (0.7->0.9) (footing and slab weights)

And those footings and slabs are engineered for uplift.

4. The flip side of #3 is that we have now reduced the OVERALL safety factor by reducing the DL reduction factor of the footing. Frankly, I think we would not be having this discussion if the equation were something like:

1.5/1.67W + D

D= all dead loads accurately calculated, up to you the engineer which to reduce slightly like roof MEP, ceilings, etc.

That we clearly see the factor of safety is for the unknowns of the wind and to provide an overall FS.

Is this the way you see it BA??

 

[jeffhed]

a2mfk
Thank you for doing what I could not, summarizing this thing into a more understandable format. We in our office agree that a different factor should apply to footing and slab weights as it is a much more reliable estimate than the dead load of the building. However there is no current code allowance for this. We further argue that if the anchor bolts and base plate are designed for a design uplift force of 10000 lbs (Resulting from 0.6D+W of the structure), why should the footing be designed for 10000/0.6 = 16667 lbs? Seems like a lot of extra concrete to me just for the hell of it when the anchor bolts and base plate have not been designed to lift 16667 lbs.

 

[BAretired]

WillisV,

As you have said, the use of 0.6D in combination with 1.5W is simply wrong. We are fortunate in our area as the authority having jurisdiction does not review structural calculations, so it is not an issue I have experienced.

dcarr82775,

I have heard that too (150 pcf includes steel). I have never bothered to check it either, but it is certainly an approximation, particularly for lightly reinforced mass concrete as found in foundations which are being sized by weight alone.

jeffhed,

So I was wondering if the structure was designed to transfer the uplift force calculated from 0.6D+W of the structure (Anchor bolt design force), why increase the weight of the footing above and beyond the anchor bolt design force by multiplying footing weight by 0.6.

Using ASD, you do not design the anchor bolts to yield at the net uplift. You use their allowable stress. Thus the anchor bolts are capable of carrying 50 to 60 percent more at yield.

By the same token, if the total weight of foundation just balanced the uplift, there would be a factor of safety of 1.0 against overturning or pullout. That is clearly inadequate. The 0.6D term simply ensures a similar factor of safety as used in the anchor bolt design.

Think about it.


BA

 

[BAretired]

a2mfk,

Sorry, I missed your question due to my slow typing (and perhaps thinking). I'm not quite sure why you feel that a separate, less conservative value should be applied to the dead load of concrete slabs and footings.

In Canada, we use Limit States Design for stability calculations. We have a slightly different combination than you do. Ours is (0.9 or 1.25)D + 1.4W. When dead load contributes to the effect being considered, use 1.25. When dead load resists the effect, use 0.9. The 0.9 factor cannot reasonably be increased beyond that. When converted to ASD, 0.9 becomes 0.9/1.4 = 0.643.

jeffhed,

We further argue that if the anchor bolts and base plate are designed for a design uplift force of 10000 lbs (Resulting from 0.6D+W of the structure), why should the footing be designed for 10000/0.6 = 16667 lbs? Seems like a lot of extra concrete to me just for the hell of it when the anchor bolts and base plate have not been designed to lift 16667 lbs.

If the anchor bolts are designed using ASD for 10000#, they will yield at approximately 16000#. That is considered failure.

If the foundation weighs 10000#, failure will occur when the uplift is 10000#, long before the anchor bolts fail.

BA

 

[jeffhed]

BA,
I know what you say follows the letter of the code. I also know that multiplying the weight of the footing by 0.6 follows the letter or the code. I am just questioning it because in my opinion it does not seem reasonable to add 6667 lbs (1.5 yds) of concrete per footing when the footing weight and slab can be accurately estimated. As I have stated earlier today, it appears that I will need to multiply my footing weight by 0.6, like it or not.

 

[BAretired]

jeffhed,

Your post was just 2 minutes after mine, so you probably didn't have time to read it. Think about what I said. You are not adding 6667 lbs...that is where you are mistaken.

Certainly the footing weight can be accurately estimated. So can the yield strength of the bolts.

BA

 

[JAE]

jeffhed - you keep saying that the footing dead load can be "accurately estimated". That isn't the point.

The 0.6D+W combination provides about a 1.5 safety factor against uplift due to all sorts of variabilities such as:

1. Footings might be built smaller than actually designed.
2. Footings might be shallower than designed.
3. Wind loads might / WILL vary considerably
4. The wind tributary area on the building could be altered in the future.

I'm sure there are others. You have to realize that the 0.6 factor on D in this case is not there to just deal with the variability on the value of D. In most combinations the load factor is intended to directly correlate with the load type's variability.

In this particular combination, the 0.6 factor isn't just for dead load variability.

 

[a2mfk]

BA " I'm not quite sure why you feel that a separate, less conservative value should be applied to the dead load of concrete slabs and footings."

It stems from my opinion that 0.6 (FS=1.67) is an unreasonably high factor of safety for the uplift of a footing during a wind event (lets neglect overturning for this argument). I do not know if there is statistical data or testing that backs up this opinion, but I am just trying to get the ball rolling here that I would like to see a formal study done by people much smarter than me

I have been heavily involved in forensic engineering investigations since the hurricanes that hit Florida in 2004, and made several trips to the greater Biloxi, Mississippi area after Katrina. I do not recall either first hand observing or seeing in pictures, etc. of any uplift failures of a footing. I mean a footing being pulled out of the ground and relocated or at least being the root cause of a structural failure.

Now having read a lot of your posts and respecting your opinion, I think you could easily argue that is because load path has not been maintained so that these footings were ever even tested. I assure you most footings and slabs that were wiped clean of their wood framed structures were never tested, the roof cladding failed one piece at a time, the diaphragm failed, and then the walls easily failed. I saw so many bare slabs and foundations, it was really upsetting and sad to look at... At least I got to meet a lot of the homeowners who got the hell out of dodge.

Forgetting residential, with PEMB and single story retail box, you often get very large uplift loads resisted by a concentrically loaded pad footing. It seems hard to imagine that this column would be able to pull a footing out of the ground through the slab, even if undersized, because so many other things would fail first, ie, roof decking and then we have no uplift loads.

Ideally if everything were designed and constructed properly, the max uplift would see its way to the foundation, and then we'd have a bigger need for a larger factor of safety. You would probably argue that this means we need to improve the load path quality in design and construction so this happens, and I would agree. However, I think that a 0.7->0.9 reduction factor is a more realistic representation of how things actually react under real wind load max events.

But, like I said, I would like to see much smarter people discuss and research this topic. As of now I think we are wasting a lot of concrete to meet code...

 

[a2mfk]

I get JAE's points above too, good counter arguments. Its a reason why I have a pretty high range on my proposed reduction factor...

On a related side note, I did once see a DOWNWARD pressure failure of a PEMB moment frame. BUCKLED THE FRAME but left the cladding intact. A house about 100 feet away with a shingle roof did not have a scratch on it!

It was a tornado in Daytona Beach, FL.. These things bounce around a lot in Florida and will cause all kinds of weird damage and then leave other structures untouched, very localized.

 

[BAretired]

a2mfk,

The uplift resistance of a foundation can include more than its own dead load. If there is soil above it, the weight of soil is part of the dead load. If there is a grade slab above that, the tributary weight of slab becomes part of the foundation dead load.

In the case of a pile, the dead load is usually a small part of its resistance to uplift, the main portion being carried by skin friction.

I agree that the foundation will not feel the full uplift of the wind if the roof blows off or other elements fail first but surely that cannot be the basis for design. Every structural element must be designed on the assumption that the rest of the building remains intact.

The OP in this thread and some of his co-workers seem to be under the misapprehension that the 0.6D term provides more resistance to uplift than the anchor bolts can carry. That is simply not true.

BA