Beta method - pile capacity 1

[ONENGINEER]

I am employing the beta method to calculate the ultimate load capacity of relatively long piles (35-40m). The soils encountered are predominantly clayey sandy silts, as well as sand. The Standard Penetration Tests (SPTs) conducted in the field often exceed 30 or are reported as refusal before reaching 24 inches of penetration. I would appreciate expert insights on the following points:

1. I am utilizing the Canadian Foundation Engineering Manual (CFEM), which provides the following beta values for cast-in-place piles:

- Clay: 0.25-0.32 (uncertainty exists regarding whether this is for driven or bored piles)
- Silt: 0.2-0.3
- Loose sand: 0.2-0.3
- Medium sand: 0.3-0.5
- Dense sand: 0.4-0.6

Question: Is there a more detailed reference for estimating beta values for silty clayey sand or sandy clayey silt?

2. The N values I employ are uncorrected as reported by the driller. The categorization into loose, medium, and dense conditions may remain the same regardless of whether one uses N, Ncorr, or N(60).

Question: In principle, should one use raw N values, Ncorr, or N(60) regardless of the above?

3. I have noted in other discussions that the Tomlinson or Fellenius method of design was recommended by BigH.

Question: Tomlinson advocates for a constant friction below a critical depth, whereas Fellenius considers this a fallacy. Do these two methods converge in practice?

Thank you for your assistance.

 

[geotechguy1]

Most building codes still accept / follow either Tomlinson or simply calculating skin friction and end bearing and combining them.

I have worked in some areas where (essentially arbitrarily) it was the norm to design only with skin friction or only with end bearing.

Personally I think the Fellenius unified pile design method is closer to correct than the normal code method or Tomlinson. However I'm not sure unified pile design is fully adopted anywhere. In many cases the average state of practice vs state of research differ so much that they are opposite; to go off on a tangent, if you follow unified pile design or similar for a pile in liquefiable soil, the vertical load in the pile decreases when liquefaction occurs above the neutral plane, whereas according to the code method / method that's adopted in practice, the load in the pile increases.

R.e. critical depth - I believe personally that fellenius is right and codes and old guys in corner offices are wrong. I always found the Fellenius explanation from the 94 paper somewhat confusing but to me it seems like an artefact of the fact that above the neutral plane load from the soil is being transferred into the pile and below the neutral plane load is being transferred out of the pile into the soil. So if you use load cells embedded in a pile zeroed prior to a static load test, removing the load that's already in the pile distorts the result and creates a false impression that a critical depth exists.

There are lots of other references for beta factors, don't know how different they are (eg. FHWA bored and CFA manuals; NAVFAC DM7.2 which gives clay 0.2 - 0.25, silt 0.25-0.35, sand 0.35-0.5).

'Silty clayey sand' 'sandy clayey silt' - I guess to me, both classifications seem like a misnomer. Or at least as a junior I remember having multiple logs unceremoniously dumped on my office floor covered in red ink and x's for daring to use such a classification. A soil thats 20-35% silt, 20-35% clay, and 30-60% sand, (or 20-35% sand, 20-35% clay, and 30-60% silt as in your second one) - both to me seem like they are fine grained and should be classified as clay, or as silt (if you believe in Our Lord the Casagrande Chart).

 

[ONENGINEER]

Thank you geotechguy1 for providing references on beta. The term I used like sandy clayey silt was meant to be either a clayey silt or sandy silt. My apology if sounded like a classification name. The point was if there are references for beta values that are not for purely silt or purely clay materials and may take the PI into consideration.
I have often worked with Tomlinson. I would like to know more about the Felonious unified pile design method in a simple way.

 

[geotechguy1]

Do you have FE software? An easyish way to understand unified pile design is to model piles in FE. Even if you only some basic grasp of how FE models work (stiffer areas attract more of the load) it should be really obvious that the conventional pile design methods are a bit lacking and not really theoretically or physically sound. The old ways work - but mainly due to the many layers of conservativism in deriving loads, assessing capacity, and assessing displacements.

Otherwise, fellenius red book:

https://www.fellenius.net/papers.html

 

[ONENGINEER]

geotechguy1- what is CFA manual? I could not readily find it. Thanks.

 

[geotechguy1]

https://www.fhwa.dot.gov/engineering/geotech/library_listing.cfm

It's Geotechnical Engineering Circular No. 8, you should be able to get a PDF from that page. They do have a bunch of stuff about β factors in there. The other one is GEC. No. 10 which is the drilled shaft LRFD design and construction manual.

 

[geotechguy1]

The beta value is just adhesion at the pile-soil interface times tan phi if I remember correctly. So it varies with phi and that adhesion.

 

[ONENGINEER]

I realize the beta=tan(phi)*tan(delta) but decided not to use it to avoid correlate the SPT and phi values.
The following formulations:
beta=1.2-0.2SQRT(Z)
beta=1.5-0.244SQRT(Z)
are given in GEC 8 and am comparing with the CFEM beta values. I am trying to understand the differences. Seems FHWA formulations incorporate the concept of critical depth within themselves but not CFEM. Is that correct? None of the references take into account the pile materials being wood or concrete etc. It would be more convincing if at least two methods yield similar results.

 

[geotechguy1]

It could also be that they are trying to account for high confining stress suppressing dilation (i.e. you move towards a normally consolidated or critical state phi and away from a peak phi at low confining stress with increasing confining stress and thus potentially a lower beta if it's the same material)