mat foundations sliding resistance: Adhesion or Friction? 1

[Francisco A.]

Good morning Engineers:
As you know, Table 1, page 63 of NAVFAC 7.2 presents friction factors and adhesion values for different interface materials. If you have a shallow foundation sitting on pure cohesive soil (in this case lean clay with fines >85%) do you provide a friction factor (say 0.30) or an adhesion value (say 500 psf)for the structural engineer to size the mat against sliding?
I am clear that if the soil is purely frictional (e.g. clean sand) a friction value is the correct recommendation, but in the case of a pure cohesive soil I am not sure, as the table provide friction values for clays. I guess that if you want to provide a friction factor for a clay soil you need to estimate the effective friction angle (phi prime) of the clay in terms of its plasticity.
Furthermore, the values in this table are called ULTIMATE. Do you reduce them applying a factor or safety, or do you inform the structural engineer that the values provided are ultimate and he needs to use appropriate factor of safety. What would be that appropriate factor of safety. Thanks for you time and input!

 

[EireChch]

Sliding resistance is comes from the sum of frictional component (phi based) and a cohesional/adhesional (drained or undrained cohesion based)

In CLAY you have to consider both drained and undrained conditions.

Work out your drained an undrained parameters:
=for drained phi'= somewhere between 20deg to maybe 30 deg for CLAY and c'= somewhere between 5kPa to 100kPa or more
=for undrained phi=0 and c[sub]u[/sub]= 20kPa to 300kPa

In the above slide,
Fs= factor of safety
Rn=Vertical force from soil and wall
tan delta[sub]base[/sub]= interface friction angle of base material
C'= drained cohesion
B= Base width
Rs = driving horizontal force

For drained conditions, you work our vertical force and interface friction angle. This is your resisting force. Some people also use drained cohesion (c') and work out cohesional component. I have in the passed when stuck, its good practice to not use it as its questionable if you actually get full cohesional resistnace. Eurocode 7 also states that it shouldnt be used.

For undrained conditions, phi is 0, for all you have left is undrained cohesion (c[sub]u[/sub]) times foundation width.

Generally one case will govern but its good practice to check both.

Factor of safety of 1.5 is typical

Disclaimer the above parameters are very approximate.

 

[Francisco A.]

EireChch,
Thanks a lot for your prompt response, very interesting. I have only one remaining question: In the case of a temporary lateral load (e.g. caused by wind or earthquake) should we only consider undrained (short-term) parameters (i.e. undrained cohesion)? in my case the lateral load will be temporary (foundation for shock absorber to stop swinging of a crane). Thanks again for your help.
Francisco.

 

[MotorCity]

Maybe I am in the minority here, but I have never relied on cohesion to resist sliding (or maybe I have if it was baked into the frictional resistance stated in a geotech report). I have typically relied on friction resistance by multiplying the gravity load by a friction coefficient, say 0.3 or 0.4, depending on the subbase material.

In my mind, cohesion is the internal force that holds the clay particles together, thus making it "cohesive", as opposed to granular materials that lack this internal force and simply disperse into individual grains. If that's the case, how does this internal cohesive force translate into an external force that resists lateral movement at the interface between clay and the bottom surface of a foundation?

 

[dik]

...and maybe some passive soil resistance offered by the side of the raft?

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[Mccoy]

EireChch,
Thanks a lot for your prompt response, very interesting. I have only one remaining question: In the case of a temporary lateral load (e.g. caused by wind or earthquake) should we only consider undrained (short-term) parameters (i.e. undrained cohesion)? in my case the lateral load will be temporary (foundation for shock absorber to stop swinging of a crane). Thanks again for your help.
Francisco.

Those loads act in a very short time span so the undrained condition should be used. Please note that, in analogy to the drained condition, probably we should not use peak Cu (also defined Su), but rather the remolded Cu, sometimes called Cu,r. the remolded quantity is often calculated by an alpha multiplier.

In drained conditions, usually phi = phi_cv (= delta if rough concrete) and c' is taken as zero.

There might be some cases where Cu in sliding may be maxed, but it may be risky to use peak Cu in lack of technically adequate motivation.

 

[aeoliantexan]

One other issue to consider is, "what might happen to the subgrade during construction?". The general answer is, "nothing good". Possibilities include; (a) the excavation is made using a backhoe bucket with teeth that plow the soil, and the operator levels the subgrade by sweeping sideways with the bucket, leaving several inches of loose soil, (b)Rain saturates the soil, which gets turned to mud while the steel is placed, and (c)The foreman places 2 inches of loose "cushion sand" based on his experience building sidewalks. Result is cohesion is gone and friction is reduced. Shear keys might be a good idea.