Correlation between SPT and Clay Properties 2

[jjsim]

Hi

Anyone has the Stroud (1974) and Stroud and Butler (1975) papers on the correlation between SPT and undrained shear strength and volume of compressibility?

Stroud (1974) The standard penetration test in insensitive clays and soft rocks
Stroud and Butler (1975) The standard penetration test and the engineering properties of glacial materials

Any idea how reliable are these correlations? Any other correlations that you normally use in your design? I am trying to estimate the pile group settlement and ultimate capacity. All I have right now is the SPT-N data.

Thanks.

 

[EireChch]

Jjsim,

See attached technical note which questions the use of SPTs in clays.

Although it raises questions, correlation of N values to undrained shear strength is widely used in the UK. I have even seen engineers use N values to determine undrained shear strength over shear vanes that directly measure undrained shear strength in the same borehole....

 

[LRJ]

EireChch, I share your frustration with the continued use of SPTs in industry. It seems to be one of those 'we've always done it this way' issues; older (senior) engineers teaching younger (junior) engineers to specify SPT tests, and so the cycle continues.

Anyone who has been following developments in site investigation techniques over the last 30 or so years knows that there are more reliable in situ tests than the SPT; the SPT is arguably the least reliable in situ test. Cost is obviously an important factor, but once you take into account the time taken to do a load of SPTs, other techniques (e.g. CPTs) are actually quite cost effective. If the cost of unreliable data on a design is also considered (e.g. in terms of additional conservatisms needed in the design due to the reduced reliability of the input data), then SPTs just aren't cost effective.

 

[fattdad]

jamming a pocket penetrometer into the SPT is more reliable then relying on the N-values. The pocket penetrometer should not be used in design, however. Just have to jam tubes and run lab tests - or use a CPT/DMT.

I go both ways.

f-d

ípapß gordo ainÆt no madre flaca!

 

[dik]

Eirechch:

Can you attach the file?

thanks, Dik

 

[EireChch]

I always seem to have trouble uploading files. Attached now Dik.

LRJ - Yes I would have to agree with you. A colleague of mine once told me that those were the most expensive words in business "we've always done it that way". SPTs have do have an advantage of being able to penetrate most soils. We use to have problems with CPTs when we had shallow gravel layers, i think they would call refusal as soon as they hit a qc of 30MPa . When you have to start pre drilling the cost goes up. Doing SPTs every 1.5m was the norm too where i worked. I asked one day for SPTs every 1m and the driller looked at me in shock. Again, "they had always done SPTs every 1.5m"

f-d - We regularly used N values for detail design. Although it was residential house foundations so consequences were not high risk.

 

[dik]

Thanks very much... Dik

 

[oldestguy]

What's wrong about running some unconfined compressive strength tests on split spoon samples? I'd much rather use that than any N values for clay.

 

[LRJ]

A UCS test on a disturbed clay sample? The sample disturbance on clay would make the test non-representative of the intact condition, but so would the UCS test since soil is affected by radial confinement; that's what distinguishes it from rock. It might produce a conservative result, but again it would not be reliable.

 

[oldestguy]

LRJ gosh what a shock. In over 50 years of examining and testing all kinds of soil samples and comparing quick sample tests to the least disturbed other samples from same site on thousands of jobs it's too late to change. Also budgets sometimes control how fancy you can be and also the degree of risk involved. Never been sued related to recommendations in reports. Many happy clients.

 

[Okiryu]

See attached extracts from Kulhawy and Mayne Manual. Note also the words of caution for correlating Su with SPT N-values.

 

[dik]

You can see from the one chart that a given N value cuts across a huge range... consistent with everyone's comment about an N value being nearly useless...

Dik

 

[fattdad]

Hi Eirechch, Thanks for the link to the Reid and Taylor paper. I've saved it for my virtual library!

UCS on an SPT? Sure, that'll return some index value, that may be useful in correlation? I liken that to a more difficult process then using a pocket penetrometer. You also get unit charges for an UCS, which you would not get if the drilling geologist logged the holes with a pocket penetrometer in hand.

I don't have any context on residential geotechnical engineering - well other than aftermath lawsuits. . .

f-d

ípapß gordo ainÆt no madre flaca!