Apparent earth pressure envelope

[jdonville]

All,

Everyone here is, I trust, familiar with Peck's apparent earth pressure envelopes for braced cuts. These seem excessively punitive (in terms of the lateral forces required) for braced cuts up to say, 18 feet or so in depth, using a single level of lateral restraint (strut, tieback, whatever).

Is there any research or papers that suggest alternate apparent earth pressure envelopes that provide a middle ground between the unbraced condition (generally triangular earth pressure distribution) and the Peck criteria?

Jeff

 

[PEinc]

Look at FHWA's Geotechnical Engineering Circular No. 4 or PADOT's on-line Design Manual 4 (DM-4). They both show the latest research on earth pressure diagrams for tiedback walls. Originally FHWA and PADOT specified a triangular load diagram for single-tier walls. However, tieback contractors usually designed single-tier walls using trapezoidal diagrams. The newer research now calls for a non-symetrical trapezoial pressure diagram for both single- and multi-tier tiedback walls. The shapes of the trapezoids depend on the tieback elevations. Becareful if you refer to PADOT's DM-4. They have soil weights in tcf when most other methods are using pcf. You will get some funny answers if you are not careful with all units

The non-symetrical trapezoidal diagrams are a pain in the .... because there are no computer programs that will automatically generate the earth presuure and/or shape. Therefore, there is more hand calculation now. Also, if you move the location of the tiebacks, you have to recalculate the pressure and the shape. FHWA fixed something that wasn't broken.

 

[jdonville]

PEInc,

Whoops! I hadn't looked closely enough at GEC4.

As for implementation, I recently contacted CivilTech Software about including Tschebotarioff pressure envelopes in additon to Peck, and I suspect that they would be pleased to implement the revised guidance in the next version of their design software, provided that I could offer some backup.

I will have to download a copy of PADOT DM-4 for comparison.

Many thanks for pointing this out, PEInc. I'd give you a purple star or three, if that was an option

Regards,

Jeff

 

[PEinc]

I don't know anyone who is using the Tschebotarioff pressure distribution. Most people I run into use T&P, FHWA, or Schnabel earth pressure envelopes. CivilTech allows you to vary the proportions of the trapezoid or use rectangular diagrams. You can also manually enter any shape you want. Too bad we can't yet automatically enter the FHWA diagram from CivilTech's epres program.

 

[PEinc]

P.S. What version of CivilTech are you using? I'm up to Ver. 8.8i now. Any version below that has significant problems. Also, thanks for the "stars."

 

[fattdad]

What makes the "Peck" envelope "punative" is the presence of more load closer to the surface. If you integrate the entire envelope the total load is just 30 percent greater than an active earth pressure diagram, which for a braced cut that includes horizontal supports is just about where the at-rest earth pressure load would be - actually the at rest earth pressure may be more like 50 percent greater.

Just some observations. . .

f-d

¡papá gordo ain’t no madre flaca!

 

[PEinc]

But that's what the research shows for tiedback walls. The pre-loading of a tieback anchor forces a higher pressure distribution near the tiebacks. Triangular earth pressure diagrams are not appropriate for tiedback walls unless the wall has a single tier of tiebacks and, in my opinion, the tiebacks are located high up on the wall, which is common with waterfront bulkheads, and where you also often have looser fill soils behind the sheet piling.

 

[fattdad]

Ah, PEinc, I don't disagree with you at all. I just point out that the rectangular/trapazoidal forces are only 30 percent different from the tranangular Rankine active earth pressures.

f-d

¡papá gordo ain’t no madre flaca!

 

[jdonville]

PEInc,

They upgraded mine from 8.5a to 8.8i when I complained about the old version not printing a filename on the results sheet, so I'm good there.

Jeff

 

[PEinc]

CivilTech is pretty good. I've pointed out to them many times program glitches and improvements. They are very responsive.

 

[PEinc]

fattdad,

You wrote, "...the rectangular/trapazoidal forces are only 30 percent different from the trinangular Rankine active earth pressures." What is your point?

Long term monitoring of tieback has shown that rarely does a tieback load change significantly from its lock-off load. Most often, the tieback load is designed using trapezoidal, active pressures and the lock-off loads are usually about 70% x DL. In addition, Coulomb earth pressure coefficients are frequently used instead of Rankine.

 

[fattdad]

Tough crowd. . . . Anybody that wants to use Coulomb (or log spiral) rather than Rankine when the third alternate is to use "Peck's" rectangular drawing is entitled to, I guess. . . . I'm just responing to the initial post (I thought. . . . ).

Here's my point: Let's say

1) That Ka is equal to 1/3.
2) That a braced excavation is 15 ft high.
3) There is no water table to worry about.
4) Moist unit weight is 120 pcf

The Rankine earth pressure (total) would be 4.5 kips.
The Peck "rectangle" would be 5.85 kips.
5.85/4.5=1.3!!!!!!!!!!!!

¡papá gordo ain’t no madre flaca!

 

[PEinc]

fattdad,

No offense intended. I agree with you. It is common to calculate the total triangular earth pressure and then increase it by 25% or 30% prior to rearranging it into a trapezoid. I'm more comfortable knowing that the tieback force or brace force is more than the theoretical load. After all, the tieback or brace is what's holding up the wall. One of the more common (although relatively rare) reasons for tiedback or braced wall failure is the use of triangular pressure diagrams instead of trapezoidal or rectangular diagrams.

T&P developed pressure diagrams that give the designer the maximum anticipated brace load, not the average load, not the exact load. Their diagram takes into account that there may be some random brace somewhere in the wall that may have a higher load than the others. An interesting thing to consider is that, most likely, when T&P made their Chicago subway support wall brace load monitor readings, the support walls also were subjected to various traffic and construction surcharges. Therefore, their pressure diagrams may already account for normal surcharge loads. However, today, we add additional load beyond T&P's diagrams to account for these surcharges. Comments?

 

[fattdad]

For a cantilevered retaining wall, the surcharge is characterized as a rectangle (I know that everybody knows this). What I don't really know is how the industry would relate this to a braced excavation, ala the T&P rectangle/trapazoid? Would we just make the trapazoid wider by taking the "active" rectangle and multiplying it by 1.3? (I guess that's what I'd do, pending further insight.)

f-d

¡papá gordo ain’t no madre flaca!

 

[PEinc]

For cantilevered, tiedback, or braced walls, an area surcharge can be handled as p = q x Ka (rectangular) or it can be handled with a Boussinesq equation where the horizontal magnitude of the surcharge load will vary according to the depth, footing elevation, distance from wall, and wall rigidity. Strip loads, point loads, and individual footing loads can also be handled with the Boussinesq equations.

There are many different ways to apply surcharge loads to a retaining wall. They all usually work unless the designer does something stupid or unreasonable. However, the method you use for applying the surcharge often depends on the reviewer or owner agency. For example, DOT's may require that surcharges be applied in a way that is different from a method dictated by a railroad. Knowing various ways to apply a surcharge load and still get a safe and economical wall design is all part of the game or art of geotechnical engineering.

Surcharge loads are not usually factored up unless you are using load factor design or LRFD design. If the surcharge load is not applied immediately at the back face of a wall, then the horizontal magnitude of the surcharge pressure should be zero at the top of the wall for cases where the ground behind the wall is flat and level or sloping downward.

Boussinesq equations, I believe, were developed assuming a rigid wall. Railroads usually insist on calculating surcharge pressures assuming a rigid wall even though many walls, especially sheeting walls, are not rigid. This is very conservative, especially for cantilevered sheeting walls. CivilTech and ProSheet are two programs that let you assume less than a rigid wall when calculating surcharge pressures. This is done by applying Ka or another factor somewhere between Ka and 1.0 to the Boussinesq lateral pressure for a rigid wall. This seems reasonable to me.

 

[fattdad]

PEinc - Check on all points. When I suggested a "factor" of 1.3 I was relating this to the stress difference between the Rankine triangle and the T&P rectangle/trapazoid and also considering an areal load. I would agree with your Bousinesq equations and would use them also for varying loads.

On the matter of Boussinesq's original equations, I want to check my notes. I seem to think that the equations for horizontal loads are applicable for an infinant space and there may be a need to "adjust" the calculated load if the "space" is missing (i.e., when there is a retaining wall). Too late to check right now. . . .

f-d

¡papá gordo ain’t no madre flaca!

 

[fattdad]

Postscript (referring to my Earth Pressure notes, J. Michael Duncan, 1985):

1) Bousinnesq solutions are based on, elastic, homogeneous, isotropic stresses for a given element.

2) They are not a function of soil modulus or poisson's ratio

3) In an infinant soil, for every "real load", there is an "image load", such that displacement does not occur.

4) Considering that a real load adjacent a wall cannot have the companion "image" load, the actual load on a wall is 1.5 to 2.0 times the calculated "free field" (i.e., Bousinessq) load (these are from field measurements by Spangler in the 1930s).

Summary: If you are using Bousinessq to calculate horizontal loads related to point or line loads acting adjacent to a retaining wall, you will understate the horizontal load. The design load should be increased by a factor of 1.5 to 2.0.

f-d

¡papá gordo ain’t no madre flaca!

 

[DRC1]

Question - I have looked up PADOT-DM4. It looks very interesting. I was unaware that there was research changing the aperent pressure envelopes for single level tied back walls. I have used triangular for single tie backs. It is somewhat conservative as the tie back genrally generates some passive load into te soil due to the lock off load vs. depth at lock off ( I generally lock off at 100%). But as far as I knew it was always pretty much thought that the apparent pressue envelopes did not kick in untill you had at least 2 levels of bracing. The other thing I did not understand in the DM-4 was the cohesionless envelope. The cohesionless envelope was traditionaly rectangular with pa = .65 ka gama H^2
but DM-4 is trapizoidal and arperently twice that value at the maximum. How did that come about? Another thing I noticed is that the commentary states that free draining material is typically placed behind the lagging, so drainage is ussually not a concern. Typically the native material is distubed as little as possible, so no free draining material is placed behind the lagging.
Did I miss something major somewhere?

 

[PEinc]

The FHWA research fairly recently showed that the trapezoidal pressure diagram is appropriate of single tier walls. That is why PADOT shows the unsymetrical trapezoid. The trapezoid gives higher tie loads than the triangular diagram. Specialty design-build tieback wall contractors have been using trapezoidal diagrams for a single tier wall for many years. Sometimes the rectagular diagram is used. Except for waterfront bulkheads and walls where the single tier tieback is near the top of the wall, triangulare pressure diagrams are rarely used or recommended.

Be careful using the equations from PADOT's DM-4. The units are tricky. The unit weight of dirt is supposed to be in tons per cubic feet but the answer is in PLF. Verify that all units are proper for the equations.

Free draining material is not placed behind the lagging despite what PADOT says. For a temporary wall, there is no drainage required. There should not be any water for a soldier beam and lagging wall. For a permanent wall, geocomposite drains (chimney drains) are usually attached to the front of the lagging before the permanent concrete face is poured or shotcreted. If a precast facing is attached to the front of the soldier beams, the space between the lagging and the back of the precast is filled with crushed stone for drainage. TiedbacK soldier beam walls should NEVER use precast lagging panels between the beams.

 

[Panars]

PEinc-

What is the problem with using precast lagging panels between the soldier beams? Would it be acceptable to use the precast lagging if the wall is a cantilever one (not tied back)?