Please be nice… 1

[sgk77]

Hi All.
I’m a structural engineer seeking some advice please. I need to do a high level model of a slab on an elastic foundation…apologies as I appreciate this is a structural engineer in a geotechnical forum talking about elastic models…..so to unwritten rules broken already. I’m trying to see if these is any kind of formal correlation that can be made between the modulus of sub grade reaction and the strain /deformation moduli Ev1 and Ev2. I can find anything so am assuming if there is it’s probably in a far more intelligent environment than Google search…any help gratefully appreciated!

 

[EireChch]

If you know your load and modulus then you can calculate your deflection. modulus of subgrade reaction (k) is load / deflection.

 

[sgk77]

Hi - I understand that the k value is just load/deflection - but I don't have access to the original plate testing results (only the Ev1 and Ev2 values). As I understood it (perhaps incorrectly), the k value is taken from a single (non-cyclical) plate test, whereas the Ev1 / Ev2 values come from cyclical (loading/unloading) tests...so wasn't sure whether they were strictly the same thing, and couldn't find anything that mentioned both k and Ev2 in the same guidance (aside from some esoteric research papers where it wasn't clear if there was a consistent or practically usable relationship). Sorry if being dense - but are you saying that I can simply use the Ev2 moduli as equivalent to k? I've seen this done - but wasn't convinced this was correct. Thanks!

 

[EireChch]

In absence of anything else I would calculate as shown on the attached. I would take Ev1 as E in the bottom equation.

 

[sgk77]

Brilliant - thanks very much. If you ever need a favour on anything (well) above ground let me know

 

[EireChch]

No worries. Ev1 and Ev2 are from repetitive Plate load tests. I have actually never done one myself.

They are different to unrepetitive (or normal as I would call them) PLTs, the main difference being that each load cycle is only held for 1 min.

Regardless of this, K from repetitive is still taken as load/1.25, which is load at 1.25mm. K and E are related through Poisson's ratio etc, so I think its a reasonable assumption to take EV1 as E.

Interested to hear others thoughts. . .

 

[sgk77]

I'm glad you said 'normal'...I thought it was just me. Can't say too much, but am working on a big site with a bigger crane and people get very concerned about settlement under its rail points...so the plate tests here quite rightly were the cyclical type...the issue I have is that when further plate tests are required for much simpler things like short term temporary works, the engineers concerned now think everything must be cyclical... KISS has gone out the window! Thanks again.

 

[ATSE]

Plate load tests generally give k values much higher than what is realistic for building foundation settlement. Sometimes by a factor of 4 or more - depends on the depth of influence ("lower zone") and time-dependent behavior, which are not captured.
Plate load tests are better for pavement design, not foundation design.

 

[EireChch]

Interesting ASTE, I have always found PLTs to be good for testing engineered fill. Typically get E' in the range of 20-50MPa for granular sandy fill which is what I would expect.

 

[ATSE]

EireChch:
The plate load test is typically 1.0 ft x 1.0 ft.
You can't possible get any meaningful lower zone engagement with that tiny footprint that applies to larger bearing areas.
Also, a short term loading test gives you no information for consolidation for plastic soils.
Plate load tests are ideal for short term wheel loads.

 

[EireChch]

Yes I wouldnt use them for direct foundation design. Like using those correlations to extrapolate it to a larger foundation size.

But I think its very good for assessing stiffness of engineered fill (granular), to use is a settlement calc