Need a book? Engineering books recommendations...

# RE: Factor of Safety against Uplift

• To: <seaint(--nospam--at)seaint.org>
• Subject: RE: Factor of Safety against Uplift
• From: "John MacLean" <john_maclean(--nospam--at)pomeroy.ca>
• Date: Mon, 12 Nov 2001 20:10:46 -0800

```William Sherman wrote:

<Why do you consider limit states design superior to allowable stress design
for overturning? Personally, I have never heard of using limit states load
factors in stability calculations (overturning, uplift, sliding, etc). In
Chapter 18 of IBC 2000, it states to use allowable stress design load
combinations with formulas in that Chapter and in Section 1610.2 it
recommends a factor of safety of 1.5 against overturning, sliding, and water
uplift of retaining walls. Presumably this safety factor is for service
loads as it would be excessive to apply it to factored loads. (This section
should be clarified that it applies to service loads, and it also should not
apply to seismic design conditions.) IMHO limit states should only be
applied to "strength" design (material stresses) and not to "stability"
design.>

William:

Sorry I'm late in replying, I've missed the list for a couple of days.

With respect to your last sentence, it seems fundamental to me that a
structure should be stable under factored loads. But what does it prove if
and the camel's back breaks. That's where the FS comes in. But it's just a
patch to work around the fact that the allowable stress method doesn't deal
with things out at failure loads where we're most interested in them.

I'm not familiar with the IBC and only marginally familiar with UBC 97 but I
note that UBC 97 in article 1611.6 has a similar clause, i.e.

"Retaining walls shall be designed to resist .....overturning by at least
1.5 times the overturning moment, using allowable stress design loads."

They're basically telling you to use a load factor of 1.5 on the soil
pressure and 1.0 on the dead loads and then check stability. That's pretty
much the same, in concept, as the LRFD method. At least as far as I can
tell.

Especially confusing if you consider that in section 1612.2 the code gives
an LRFD load case of 0.9D ± (1.0E or W) + 1.6H (I added the 1.6 H in as per
article 16.12.2.2 I think I'm reading the code right here but I'm not 100%
sure). If wind and earthquake are zero you have a load factor of 1.6 on the
soil pressure and 0.9 on the dead load which is more severe than the
requirements in 1611.6.

In LRFD I check stability for overturning by applying the factored loads and
then figuring out where the resultant of the soil pressure will be (e = S M
/ S V). If it's under the footing I'm okay. If the soil resultant is outside
the footing I redesign because the structure is going to overturn. You could
also do sum of the overturning moments <= sum of the righting moments. i.e.

I like limit states better for design for stability against overturning for
two reasons. First it makes more sense. With LRFD you're basically saying
what's the worst thing that can happen here? Lateral soil pressure could be
60% more than predicted, soil / concrete density could be 10% less than
predicted. Okay, let's look at what happens if that's the case. If you don't
like 60% and 10% then change them accordingly. At least you'll have an idea
why you're changing them. What does a factor of safety of 2 or 1.5 against
overturning mean? Are they just numbers that get dreamed up. What's the
rationale behind them?

Second, I can figure out what the soil pressure under the toe of the footing
is so I can design the reinforcing in the toe of the footing. How do you do
that with allowable stress design?

John MacLean

******* ****** ******* ******** ******* ******* ******* ***
*   Read list FAQ at: http://www.seaint.org/list_FAQ.asp
*
*   This email was sent to you via Structural Engineers
*   Association of Southern California (SEAOSC) server. To
*   subscribe (no fee) or UnSubscribe, please go to:
*
*   http://www.seaint.org/sealist1.asp
*
*   Questions to seaint-ad(--nospam--at)seaint.org. Remember, any email you
*   send to the list is public domain and may be re-posted