how can I relate SPT - N values to Youngs Modulus in sands 1

[lvkb]

Is there an empirical equation relating SPT-N values to Youngs Modulus E in sands? (preferably one that gives me E in kN/m2)

I am doing a pile foundation design where knowing E would be useful.

Thanks

 

[PSlem]

E=0.5(N+15) in MPa

 

[lvkb]

Thanks - do you have a reference for this at all?

 

[PSlem]

I got it here.

http://www.geoforum.com/knowledge/texts/broms/viewpage.asp?ID=147

However it is in Bowles' Fdn. Anal.& Dsn. Table 5-6
He gives several versions and reduces E by half if saturated, i.e. 250(N+15) below W.T. and 500(N+15) above in KPa

 

[Efty]

By Schultze and Menzenbach (1961)

E=2400+530*N±2100 (kPa) for silty Sands
E=3800+1050*N±9300 (kPa) for gravelly Sands

 

[Mccoy]

Some of such correlations I saw display such a scatter of data points that I wonder about their usefulness.

It's probably better to use the Burland & Burbridge method, which eliminates one source of error, getting settlements directly from NSPT without calculating E.

By the way, anyone has got a copy of the original paper?

I only saw redesigned graph, the original though has supposedly all datapoints in it, with soils (sands, gravels) differentiated.

 

[Efty]

Burland & Burbridge method refers mainly to shallow foundations, so I don't know whether it works well with pile problems.

A guide to get a rough estimation of E in sands is by multiplying N by 0.5 for silty sands and by 1.0 for gravelly sands (E in MPa). It is obvious that the silt content and the size of sand particles are significant. For example one should reduce the value 0.5 if there is a high percentage of silt or the sand is fine, whereas he should increase it if the sand is coarse.

 

[Mccoy]

Burland & Burbridge method refers mainly to shallow foundations, so I don't know whether it works well with pile problems.

Probably not, I missed the last part of the original question (piles).

I agree, sometimes it is better to simplify the correlations and reason in terms of plausible intervals rather than deterministic numbers.