Direct Measurement of soil natural density

[cmorgh]

Hi to all,
I appreciate any comments on insitu methods regarding the direct measurement of soil natural density in deep depths, where no field density test- say by sand bottle- is practical.
Please note that I do not mean indirect estimation by correlation of SPT or CPT values.
Regards

 

[cmorgh]

Hi again,
Please exclude the down-the-hole or cross hole tests.

 

[GeoPaveTraffic]

If you don't want to use ANY correlations, then the only way to get a density is to get a sample, measure, weigh, and dry it. That is the only way to get desnity without using a correlation of some sort.

Therefore, in you case I would suggest taking 3" diameter Shelby tube samples at the locations where you want/need densities.

 

[BigH]

As a caveat to GeoPaveTraffic - some materials swell on release of overburden pressure - so the density you seek, depending on material, would not be the density you measure.

 

[dgillette]

Let me temper GeoPaveTraffic's suggestion with the caution that 3" Shelbys may work well, or be be a wasted effort, depending on the soils involved. If the material is plastic, or is partially saturated and has enough fines for capillarity to hold it together, it may work. A non-plastic saturated material is likely to be trouble, as is anything with more than a trace of gravel. In such cases, you may need to go to an Osterberg or other piston sampler, larger diameter dry coring, etc. Loose, saturated, non-plastic materials are the worst because they want to densify at the least suggestion of disturbance, and you end up with free water in the top of the tube. In extreme cases, where the material is loose, saturated SP, and the density needed to be known very precisely, people have even frozen the ground with liquid nitrogen and cored the frozen soil. (I'm not making this up. As I recall it was for a power plant, maybe nuclear, in Japan.)

On the other hand, if you have plastic clay that you know to saturated, you can get a pretty decent density from water content and Sp Gr.

Why do you need to know the density more precisely than you can get from correlations? Also, what materials are you dealing with?

 

[GeoPaveTraffic]

Good points both BigH and dgillette. I was assuming, for no good reason, that the material was a near normally consolidated plastic silt or clay with little sand or gravel.

 

[cmorgh]

Thanks to all for the comments,
what about the measurements (this method also needs correlation) by gamma ray i.e. the nuclear method?
Is there any possibility to perform the test- and regarding the available apparatus- in deep depths?

Regards
cmorgh

 

[PKWATCGS]

Hi,
Recently, we have had a similar problem as yours at a site comprising uncontrolled, sandy clay fill with some gravel to a depth of over 10m. The fill is currently unsaturated and we wanted to measure the density and density ratio accurately for assessment of potential hydroconsolidation when the water table rises in the future. We tried everything ranging from PMT, SPT and CPT correlated with in-situ density testing in deep (6m) test pits, tube sampling, and down-hole gamma probing.

I am sorry to give you the bad news that none of the above worked. When the material is not saturated, PMT, SPT and CPT resistances are sensitive to moisture content variations, as well as the amount of gravel in the soil matrix. Density from tube samples is easily affected by sampling disturbance as indicated by others. Gamma probing gives wet density only, and is sensitive to uniformity of the hole drilled (i.e. sensitive to size and shape of the hole as may be affected by localised collapse and/or enlargement, and probably also the presence of gravel particles). If you want to try the gamma probe you will at least need to lower the probe down several times at different orientations and also take caliper readings.

So far we have not had any success so I would be pleased to hear from anyone who has a "silver bullet" for this problem.

 

[dgillette]

No silver bullet, whether literal, figurative, or Coors Light. I would be inclined to try dry coring - largest available diameter. We've had good luck (or at least we THINK we have) with 6" coring with acrylic liners, working through hollow-stem augers w/o fluid.

 

[BigH]

dgillette had the hammer on the nail - you could always, and expensively, ground freeze. Then core, using chilled water. The core itself could then be measured for density. The only caveat is if the freezing causes ice lenses to form (i.e., available source of water) - but for sandy soils or clayey soils where the process is quick, this could be done. Again, expensive, but . . .

 

[dgillette]

Maybe you should forget about the density measurements and analysis, and call these guys to thump on the fill:

http://www.rapidimpact.ca/index.htm

Might well work enough to prevent significant collapse on wetting if the fill isn't too wet at present. While on their website, be sure to watch the video of blast densification.

 

[cmorgh]

Hi again,
If using nuclear device, is there the posibility to lower the probe down to say 25 to 30 m? (regarding the technology of the test and related devices)
The texture I am encountered with is rather non-cohesive. Hence, no traditional sampling would be useful.
Regards

 

[jdonville]

cmorgh,

I would think that the proximity of the sidewalls of the hole, if close enough, might throw off a nuc. Besides, nucs are not the tool of choice for measuring insitu density of granular material, especially if there are relatively large openwork voids.

Jeff


Jeffrey T. Donville, PE
TTL Associates, Inc.

http://www.ttlassoc.com

The views or opinions expressed by me are my own and do not necessarily reflect the views or opinions of my employer.

 

[dgillette]

Let me back up a little bit, PKWATCGS. I've never dealt with a situation quite like that, so this is new to me. How do you analyze the settlement from wetting? We've dealt with collapsing/hydrocompacting silts in canal foundations, expecially in the Four Corners area of the southwestern US, but we had much thinner layers, which were shallow and easily accessible, and more important, of known thickness. It was pretty easy for us to go from oedometer collapse tests to a prediction of settlement, and we were able to deal with it by simply ponding water before final trimming and placement of the concrete lining.

Your problem looks more complex. Any collapse tests you could do on samples would be at a "micro" scale (2-4 cm). The gross-scale collapse (of presumably heterogeneous fill) would be a function of the nature of the looser portions and their overall percentage in the 10 m. Can you get reasonably undisturbed samples of the looser material for the oedometer tests? If you measure in-place densities with nuke or something, that would give you density on a different scale, perhaps up to a meter, and you would have to relate it to some reference density that indicates collapse potential. The difference in density between collapsing and noncollapsing materials could be pretty subtle and might be masked by small amounts of gravel in the fill, which would raise the unit weight of the whole soil without affecting the fine-grained matrix where the collapse takes place.

Am I missing something, or is it really that difficult? In part I'm asking out of idle curiosity, but I could run into the same problem on one of our projects since we often run into dry fills in the southwest.

How about an in-situ test with injection wells?

Thanks,
DRG