Boussinesq & AASHTO 2007 - off by a factor of 2?

[ARS97]

I'm designed a concrete "well cellar" for a gas company. This well cellar is a four-sided box with no top or bottom and is buried it's entire height (top flush with surface). The inside plan dimensions are 8' x 6', and the height is 8'. I'm trying to determine the induced stress on the face of one of the walls from a nearby vehicle load (HS20). I wanted to investigate this pressure using 2 methods:
1) A simplified method using additional backfill height (AASHTO)
2) A more exact method using a Boussinesq estimate

*DISCLAIMER - I'm a structural guy, and I'm not terribly familiar with AASHTO

Method #1
I have a copy of AASHTO LRFD Bridge Spec, 4th edition, 2007. From the best I can tell, I think an additional 4' of backfill would cover me (Table 3.11.6.4-1). This supposedly covers the "design truck" that has a 32k axle (HS20).

Method 2
I have 2 references for the induced stress from a point load: (1) AASHTO (3.11.6.2) and (2) Geotechnical Engineering, Principles & Practices, First Edition, Coduto. I've generated a basic spreadsheet that calculates the pressures/forces for each method. Here's my issue - the AASHTO method is giving me pressures/forces that are exactly TWICE the amount from the textbook. I've double-checked the equations, double-checked the inputs, and now I'm starting to believe that there is a either a misprint in one of the books, or there's a mysterious factor of 2 added in AASHTO for some reason. I was actually able to get in touch with Donald Coduto, but he didn't offer much insight as of yet. I'm trying desperately to track down some errata for both sources, but I'm not finding anything.

For you geotech gurus - attached is a PDF that shows both methods. Can anybody explain why the AASHTO method is generating results that are double the general Boussinesq method from the textbook?

 

[PEinc]

There are numerous discussions in the ET forums on this exact subject. If you search, you will find them.

http://www.PeirceEngineering.com

 

[ARS97]

Well, I'll try again, but I didn't find much initially. I have seen a few mentions of this mysterious "2" factor, and it appears to be a bit more involved than just a typo. I THINK it has something to do with the difference between the 1/2 space theory of Boussinesq and the at-rest assumption of AASHTO. Am I on the right track?

 

[RFreund]

This has been discussed many times and you really should put in some leg work. But this is a soft spot of mine....

It seems as though:

Rigid (i.e. restrained at top, basement wall) = 2
Semirigid = 1.5 (i.e. cantilevered concrete walls)
Flexible = 1.0 (i.e. cantilevered sheet pile walls (no tie backs), MSE walls (although I feel that there is an argument for even less on those walls)).

For your structure I would say 2.0 as it would be similar to a basement wall (non-yielding).

The civiltech manual gives a good discussion on this.

http://civiltech.com/downloads/sh_manu.pdf

Codes may dictate otherwise, i.e. when dealing with Railroads.

EIT

http://www.HowToEngineer.com

 

[Doctormo]

I think Bowles has the best discussion on this.

 

[ARS97]

Doctormo - I will check it out.....thank you....

 

[fattdad]

bousinesq equations solve for changes in the horizontal stress assuming that there is soil all around the point of interest. So, that means that there is some earth pressure that reacts against the calculated load. For a wall that's not true, there's just air. So, the design consideration is to double the Bousinesq values.

The surcharge fill equations typically relate the change in load in the active zone of the wall. The math is just different.

f-d

¡papá gordo ain’t no madre flaca!