Pocket Penetrometer in Split Spoon Sample 1

[garrettk]

There is a debate raging in my office about the use of a pocket penetrometer in fine grained samples retrieved from a standard split spoon sampler.

I say that this is absolutely wrong as the split spoon sample is disturbed and will not give you "accurate" results. Correlations have been developed that are readily available (NAVFAC for example) that provide strength relationships to blow count data.

Others in my office say that this is an acceptable practice because "it gives you one more number".

If I am correct, in that this practice is wrong, could anyone here point me towards a paper that backs me up? I think this is the only way to convince them.

If I'm wrong, please explain why.

Thanks!

 

[fattdad]

I think correlations between fine-grained soil strength and SPT N-value are just as problematic as correlations between fine-grained soil strength and pocket penetrometer values. Whatever problems you may feel with one, you can argue with the other. For sensitive marine clays that you are familiar with (i.e., where you already have TXC or consol data), you should already realize that the SPT N-value correlations are incorrect and the pocket penetrometer values are much more "telling". This is the case with the blue gumbo where I practice.

I like the pocket penetrometer for fine-grained SPT samples. I'd also continue to report N-values too.

In summary: Report both data. Use judgement and experience to evaluate the results. Select lab testing to confirm your assumptions.

f-d

¡papá gordo ain’t no madre flaca!

 

[GeoPaveTraffic]

I generally agree with f-d.

Our standard is to do pocket pen's in fine grained soils and report the values along with the blow counts.

My personal opinion is that in clays and plastic silts that the pocket pen is more accurate than blow counts.

 

[DRC1]

You may want to suggest 15 and 37 are two more numbers they could use. Pocket pens do give a rough approximation of insitu strength, and we have all used them to check retrived samples. If the clay is fairly stiff and not sensitive, it will provide an approximate strength.
The problem I see is that you are using SPT spoons to determine the strength of the soil. Strength of fine grained soils should be determined from samples obtained by pushing brass shelby tubes. Driving spoons can remold the clay raise pore pressuresand result in strength predictions that may be very in accurate in either direction. I think Fang, Bowles and ASTM warn against this practice. If you don't have good material to test, what good is the test? Testing the end of a shelby tube is common, although I think labs prefer to use small vanes - but I am not trying to start a debate on vanes vs. pocket pens.

 

[BigH]

DRC1 - you really use brass shelbies? Mmm

I sure hope that the strengths are not being based solely on SPTs and pocket pens (or torvanes), but I have no problem with doing pocket pens on the split spoons - because I will also have (hopefully) other data that will give me an idea of the sensitivity of the clayey material, some correlations of field vane or triaxial vs the pen (or torvane). If the clayey material is not sensitive, then the pocket pens (or torvanes - you can guess I like the latter over the former) will give you a lower bound and will be useful - especially in helping to identify anomolous softer (or stiffer) zones - other things being basically equal. Our company always did miniature lab vanes in the end of thin-walled tubes - peak and remoulded.

One can always argue that unless a piston sampler is used, any thin-walled tube (or shelby for the purist) will be somewhat disturbed - along with if you push too far or have to tap because of the material stiffness . . . and then, we must also remember the size of the tube makes a difference too.

In the end, it comes down to judgment anyway - of course judgment based on the best available strength data is preferable over more suspect data. And, in the end, it is usually the consolidation properties of the fine-grained soils that govern anyway (except in slope stability and some piling situations).

 

[oldestguy]

One more vote for doing the pocket pens, but recognize the limitations.

In my experience, when comparing unconfined compressive strengths with pocket pens, the correlation drops off pretty fast above 1.5 t/s.f.

Also one has to recognize the disturbance part of a split spoon sample, being highest along the sides. For example, split the sample lengthwise and let it dry. You can see the effect of side shearing easily. Poke the ends, not the sides.

However, in addition to the P-P, I'd rather take the time to do an added simple test. It does not require a Shelby tube or brass tubes, since you are likely to get split spoon samples anyhow.

Trim the ends of the jar sample and do a quick unconfined compressive strength test on a Remac valve spring tester, or equivalent. If you don't have that, a bathroom scale set under a drill press with flat foot in the chuck works OK also. This is why the places I worked in the past used olive jars to get a suitable sample height. A lot of these simple tests can be run at low cost. Rather than closing your eyes to the benefits and also the limitations, take advantage of every thing that you can, within reason and then >>>

Then use your head as to how you use the data, per BigH.

By the way, in some 54 yrs this old guy has handled hundreds (more likely thousands) of split spoon samples and I have yet to see that you can get a good correlation of Qu to N, so I'd not use that NAVFAC relationship with any confidence.

On that score also, I once tested side by side two different drill rigs and gear. The pin guided weight with cat-head and one wrap (bet many here never saw one ) got half the blows of a center pipe guided weight. Now take the mechanical weights these days and who knows which one is really the best. So that means N can vary all over the map also. Now I bet Mr. Gow is rolling over in his grave. Do a Google search for him and good stuff shows up.

 

[emmgjld]

Oldestguy sounds like my father.

Get local correlations (best to take the time and do them yourself) and then use as much data as possible. I have used data from 4 types of SPT hammers and have personally obtained that data (I ran the equipment). I am AMAZED at how much confidence is expressed in the test.

Just about the time you are feeling confident, the sky will drop.

 

[DRC1]

Big H -
I remember Brass Shelby Tubes - We used to have trained dinasuars push them down with their tails. Yes I meant thin wall tubes which I refer to as shelbies. Like saying jello for gellitian.
I agree with everything you have said, and even tough somewhat disturbed, a prperly sampled thin wall tube will be better than an SPT sample.
I would also like to echo that N value corralations should not be used to compute strength.

 

[Bluefoot]

It may be more of an engineering practice problem. An engineer should always ask himself the question "How much do I believe in the number I got" before blindly rely on the correlation. Most conventional geotechnical tools are not invented as an absolute measure, but a comparative measure which means you should not completely throw that old thing away (ironically, the old things are often better than the new fancy stuff) but need come down to the base and try to establish your local database for different conditions/soils.