Sand internal angle of friction - Retaining Wall Design

[kellez]

So if i told you that the backfill material of a 10m retaining wall is compacted sand what would your first comment be?

1) water table is well below the wall foundation
2) rainwater will be drained but there will be cases where the sand will be damp not saturated but damp due to rainwater flowing through it.
3) what internal angle of friction do you suggest?

 

[EireChch]

How do you know it was compacted, have you seen / verified?

If there is no verification testing of the soil (shear box test) and I had not observed the placement I would be conservative and adopt 30 degrees. If you observed placement and have testing it could be up to 34-36 degrees.

Just use a saturated unit weight gamma = 19 instead of 18 for example. Personally I dont really consider saturated / unsaturated granular, I just pick a unit weight for the compacted material gamma = 18 is usually ok.

 

[geotechguy1]

Is the sand angular or rounded?

What is the gradation of the sand?

Was the sand dumped in and barely compacted?

Was the sand compacted in lifts to an appropriate specification?

What is the plasticity of the fines content, if any fines are present? According to our lord and master the USCS, 'Sand' could be anything from 100% uniformly graded material slightly with a slightly larger particle size than silt, or a material that is 51% coarse grained, 49% fine grained, and 50.1% of the coarse grained fraction is sand. It could also be 51% coarse grained, 100% of the coarse grained fraction is sand, and the remaining 49% is bentonite and technically the material would be a 'clayey sand' (although I think most practioners would just ignore the USCS and call that a clay but lately I've been working with some extremely pedantic engineers).

You're interested in the water table and rainwater...would that change the friction angle? Or would it change the effective stress and thus the shear strength?

 

[r13]

I would be conservative and adopt 30 degrees.

Lack of any information addressed by others, I think this the route you shall take, also use a conservative unit weight.

 

[kellez]

Good questions, didnt expect some of them.
The sand will be backfilled after the wall is build, but again no one can confirm if it will be perfectly graded right? and ofcourse you cannot wait for tests since the wall will be already build. Therefore the best choice is to go conservative.

 

[fattdad]

34

f-d

ípapß gordo ainÆt no madre flaca!

 

[geotechguy1]

If the wall is already built and someone is asking me for a friction angle my 'Possible Lawsuit', and 'someone is trying to trick me into signing off on something dubious' alarms would be going full blast. Fireman to the rescue situations are dangerous. I'd insist on drilling a hole behind the wall with in-situ testing and sampling or doing a CPT behind the wall. Especially since the wall we are talking about is 10m high.

And sure, you can get an idea of the gradation. By taking some samples, and putting it in some sieves. An experienced field technician/engineer/geologist should also be able to give you a pretty good idea just by looking at it. And if you can't look at it, test it, or see it, why would you give a friction angle to it? What will your defense be in court?

 

[fattdad]

Jam a cone in the backfill. Heck, do it three times!

f-d

ípapß gordo ainÆt no madre flaca!

 

[EireChch]

f-d, your answer to everything, "jam a cone in it" ha. Jokes - the CPT is a great tool and probably under utilized. Not to side track the thread but, how do you feel about parameters derived directly from CPT data?

For example phi and E'. I have always found that CPT parameters are always a lot higher than parameters from say SPT correlations, shear box tests, PSD correlations. I admit that SPT correlations, PSD correlations are likely conservative but even considering that the CPT derived parameters are always a lot higher. Especially if you are using Geologistmiki software (based on Robertsons CPT guide I believe). Lunne et al. Guide to Cone Penetration testing has correlations too, that are typically closer to the other methods. I typically adopt "design" parameters somewhere between the two (CPT derived parameters and other methods).

 

[geotechguy1]

I've been using CPET-IT alot recently and one downside of it is the manual and the software doesn't present the standard deviations / uncertainty for the correlation. It's a good tool but it's by no means perfect. Eg. Nkt could range from 2 to 20 but everyone takes 14 as gospel. There's a Kulhawy and Mayne manual that's pretty good for presenting the scatter plot used to derive the correlations

 

[geotechguy1]

https://www.geoengineer.org/storage/publication/20489/publication_file/2745/EL-6800.pdf

 

[fattdad]

Hi ErieCHCH,

Just being a bit glib, I guess. So many posts are asking about first principals, but are only providing indication of soil type - i.e., "What's the friction angle of sand?" No way to answer that and we (should) know that!

Too much to type in many instances!

f-d

ípapß gordo ainÆt no madre flaca!

 

[kellez]

If the wall is already built and someone is asking me for a friction angle my 'Possible Lawsuit', and 'someone is trying to trick me into signing off on something dubious' alarms would be going full blast.

The wall is not build, thats why i am asking this question. if the backfill soil is not there and not compacted, how can someone find out the angle of friction of a soil that is not there. Do you track down the source of the material to be used?

For the base soil that is already there and compacted, yes you can take samples and run a test.

 

[r13]

It is good to know your suppliers but unnecessary. This is the stage you need helps from geotechnical engineer/lab - take in-situ samples and specify backfill material.

 

[fattdad]

If you are designing and know of sand backfill in a pit or a local formation, you'd get a sample, run a Proctor, figure out the maximum dry density and prepare a sample for testing. Decide if you want the backfill placed at 95 percent Standard Proctor, or some other reference density. Then prepare a sample in the laboratory at that reference density and run a laboratory strength test. Direct shear or triaxial - you get to decide!

Now you have a friction angle to use in determining your active, passive and/or at-rest earth pressures. You'll also have the in-place density. You'll have the benefit of first principals to inform the design also!

There are certainly work-arounds! DM-7 has correlations for USC and friction angle. These are based on relative density. So, you may find a curve for SP, SM, SP-SM, etc. You'd have to guess at the relative density, but here's the cheat. Every 5 percent relative compaction equates to about 25 percent relative density. So, 80 percent relative compaction (likely modified proctor) is 0 percent relative density, 85 is 25 percent, 90 is 50 percent, etc.

Have fun and enjoy geotechnical engineering!

f-d

ípapß gordo ainÆt no madre flaca!

 

[geotechguy1]

Depending on the contractual arrangements, you could just specify sand / gravel backfill with certain criteria for gradation, fines content, etc, and a compaction specification. You have to be careful to spec something that's actually available in the local area. Then you can use some correlation like fattdad suggested and select a design friction angle.