Decomposed Granite mixed with Clay 2

[subsearobot]

We are looking at simulating a 'worst case' scenario. We are testing a novel percussion coring system that is to take samples in Antarctica, down a 1km hole in the ice sheet. ( the 1 km hole will be melted)

since up to 4km of ice has overlaid the sediment in the past (geologic time), we assume that the till is overconsolidated.

We can not get glacial till locally (cost effectively, anyway). I found some refrences stating probable compositions of 10-40%clay, 10-20% silt, 20-40% sand and 10-20% granules and pebbles.

we are going to fill a steel tank 10' deep with compacted sediment and top with another few feet of water. Initially, we thought that using compacted decomposed granite (DG) would be a good worst case (hard); after consulting with clients, they would like to have clay mixed in also.

Question is, when we mix the clay into the DG, will that affect the cohesiveness of the DG?

Please, ask any questions to clarify- my expertise is certainly not in soils!

cheers,

 

[fattdad]

It's unfortunate you don't have materials to work with. Glacial till is VERY DENSE stuff! It can be difficult to drill for sure! Ultimately what you need is a mixture of grain sizes from "clay" to "gravel." Bear in mind that till is like ground-up rock. Not so much "decomposed rock" but rock that's been ground. The little "clay" sized particles are really clay-sized rocks. So, there is very little "plasticity" to the overall till material.

I'm no till expert. I used to be a field geologist in Seattle and we'd see it all the time. I would imagine if you tried to make "till" using real clay minerals, you may not get the same level of compaction. Compaction of clay requires more moisture content and results in soils with a higher void ratio. I'd bet a similar grain size sample (but with rock flour rather than clay minerals) would give you a lower void ratio, a higher density and (much) greater strength.

Don't think this has to do with the amount of gravel rocks that are obstructing the sampler. No, it's the overall density and the ability of the drill to move the stuff out of the way that affects drilling. So, you need to focus on the greatest density and the lowest void ratio. You could do density testing of various mixtures (ala ASTM D-698 or D-1557) to figure the mixture that gave you the best results.

Hope this helps.

f-d

¡papá gordo ain’t no madre flaca!

 

[Ron]

Re-creating a material from which you have no samples is a shot in the dark. It's difficult enough to re-create mixtures when you have samples....I wouldn't believe any of the data you would obtain without correlation with actual samples.

 

[dgillette]

This sounds like a fascinating project, certainly out of the ordinary for most of us.

Yes, the clay would affect the cohesiveness of the till, but I'm not sure that actually matters. With so much confining pressure, the cohesion just might not make much difference in the resistance to penetration of the sampler. 1 km of ice is way beyond "normal" soil mechanics. We rarely deal with anything more than about 2,000 kPa of overburden pressure (100 m of soil), maybe twice that under the largest dams. Your ice is something like 5 times that. (Does the density of the ice increase a lot with depth?)

Most of the shearing resistance of the soil comes from internal friction which is a function of the confining stress. I would worry about whether a tank test can replicate the confining stress and penetration resistance in the field. If you have any way of applying a large confining stress to the soil, that would make for a more realistic test. If you can't do that, make the stuff as dense as you possibly can, using the modified Proctor test ASTM D1557 to find the optimum water content. Then soak it and freeze it?

If you could somehow apply 10,000 kPa, the initial density of the soil may not matter much because the densest/stiffest of material would consolidate to a very low porosity, about the same as the loosest. See, for example, Vesic and Clough, "Behavior of Granular Materials Under High Stresses," Proceedings of the American Society of Civil Engineers, V 94, no. SM3, pp. 661-668, 1968. Some of that is also shown in Atkinson and Bransby's textbook "The Mechanics of Soils." (Had it out on my desk by coincidence.) By the time their confining stress got up to 10,000 kPa, the initially loose sand and the initially dense sand were both consolidated to about the same high density, quite probably accompanied by a lot of particle crushing.

Till can be darned near anything, depending on the source material. In the Great Lakes, it can be clay. Other places, it can include hard boulders, even tougher to deal with.

Bon chance!

 

[subsearobot]

dgillette,
thanks, great input!

we will be sampling in 'free' water. Locations will be close to the ice grounding zone, but where the ice is presently floating on the sea water. In geologic history, there has been great overburden, but presently, sea water (at about 1500 psi hydrostatic pressure).

does sediment have 'memory'- once compacted like that, will it retain its shearing resistance even though it is completely saturated?

thanks again!

 

[dgillette]

The strength and compression of the soil are governed by the effective stress, which is the difference between [the total overburden stress from the weight of overlying soil, ice, and water] and [the water pressure within the pores of the soil]. That difference is the stress that is actually borne by the soil skeleton. If the soil actually experienced anywhere near the full weight of the 4 km of ice, it would be tremendously dense and there would probably be a lot of horizontal pressure locked in and high strength even with the load removed. Therefore, expect very high friction on the sides of a sampler, as well as dense, dense, strong material at the tip, probably before the sampler is even 1 m into it.

Yes, it does have a memory of the previous overburden stress, especially if there is clay present. I had a site where a lacustrine silt had been consolidated under 50? m of volcanic rock. Where it was undisturbed, the standard penetration test sampler "refused" consistently.

On the other hand, if there was always a film of water below the ice or the ice was always "almost floating", the soil may not have experienced extremely high effective stress. If it was very dense, then got frozen and thawed without much effective overburden on it, the expansion from freezing could fluff it up a little (as it apparently did at the upper surface of my silt). However, I suspect that under several thousand kPa of pressure, the ice might not expand as much, and higher pressures tend to lower the freezing point. (Don't know whether a few thousand kPa would matter much to the freezing point, considering Antarctic cold. Somewhere on the internet, there must be temp-pressure-volume-phase charts for water.)

Am I correct that the main purpose of getting these pseudo samples verifying that the sampler would work OK? Is your client by any chance the USGS?

BTW - some DG contains clay from the breakdown of the feldspar. You might be able to find that without too much trouble in some places. Particle size could be critical unless the particles are soft enough for the percussion sampler to break up.

 

[cvg]

while it is not part of the original question, I would echo dgillette regarding particle size.

"granules and pebbles" do not adequately represent the vast range in size of glacial till. Decomposed granite does not really resemble till, even if you add clay to it. Till is not a well graded mixture, it is usually highly stratified mixtures of boulders, cobbles, gravel, sands, clays and silts.

You could (and probably will) find significant amounts of large cobbles and boulders depending on the specific location that you sample.

Under this high of pressure, is there a chance it has formed sedimentary rock such as conglomerate?

 

[fattdad]

confining pressure alone is not sufficient to convert loose materials to dense materials.

confining pressure alone is not enough to convert sand into sandstone.

I think of glacial till as well graded, in opposition to cvg's comment. Glacial till is frequently represented by many different particle sizes, which is the definition of well-graded (unless of course the till is "gap-graded", i.e., missing the mid-sizes).

To replicate the most dense of the most dense till materials, you need to focus on getting the highest density with the lowest void ratio. It's as simple as that. I don't think that'll happen with a highly plastic clay.

I would agree that the clays derived from the weathering of feldspars could render this decomposted granite questionable for your purpose.

I worked as a field geologist for 8 years before I got my masters in geotechnical engineering. I just have a feel for geology and am addressing your inquiry from that perspective.

Oh yeah, I also agree that till can be variable, so if you are trying to make a till that is mostly fine grained that'd be different from one that does have sand and gravel. That said, if it's hard to drill it's got a high density and a low void ratio.

Do you all use drilling mud to make the hole?

Sorry in advance if I'm too blunt.

f-d

¡papá gordo ain’t no madre flaca!

 

[dgillette]

"confining pressure alone is not sufficient to convert loose materials to dense materials."

For the extreme pressures being talked about here, it really can do that. Refer to the Vesic and Clough reference above, or to Atkinson and Bransby where the Vesic data are replotted in a little different form. They started with a very loose sand with void ratio ~1.0, RD = 20%. With 10,000 kPa confining, the void ratio was reduced to less than 0.6. At 40,000, it was about 0.4. They did the same with dense sand, RD = 88%. Its initial void ratio was 0.7. The loose sand reached 0.7 at about 8000 kPa, roughly equivalent to 400 m or so of soil (w/o pore pressure). Lee and Seed (1967) got similar results with a different clean sand. I would expect more change in void ratio with dirty materials. The highest embankment dams in the world are in the ballpark of 200 m, so these pressures are outside of "normal" soil mechanics.

 

[cvg]

well maybe well graded is not the correct term, however the deposition of till occurs in several ways, some of which can cause sorting. So you may have gap or skip grading, uniform grading or truely well graded. No way to predict.

According to wikipedia (which is of course the ultimate authority...)
"Till or glacial till is unsorted glacial sediment. Glacial drift is a general term for the coarsely graded and extremely heterogeneous sediments of glacial origin. Glacial till is that part of glacial drift which was deposited directly by the glacier. It may vary from clays to mixtures of clay, sand, gravel and boulders. Clay in till may form in spherical shapes called till balls. If a till ball rolls around in a stream, it may pick up rocks from the streambed and become covered by rocks; thence it is known as an armored till ball.

Till is deposited at the terminal moraine, along the lateral and medial moraines and in the ground moraine of a glacier. As a glacier melts, especially a continental glacier, large amounts of till are washed away and deposited as outwash in sandurs by the rivers flowing from the glacier and as varves in any proglacial lakes which may form."

 

[fattdad]

unsorted and well graded mean the same thing.

dgillette, interesting references. I studied under Clough, but haven't read this research.

This is fun to talk about!

f-d

¡papá gordo ain’t no madre flaca!

 

[oldestguy]

My question is why try to duplicate "glacial till" when there are plenty of places you can test the drilling system under a variety of conditions.

My experience with glacial till (that's some 56 years) is that it varies highly in gradations as well as density. The the densest is in the St. Lawrence River area. The contractor we worked with went broke trying to excavate it for the Seaway project.

I'd explore that alternative first.

 

[subsearobot]

We're doing some "simple" tests in the shop to prove out our design concepts. We are going to run some tests in the field (probably Lake Tahoe area), but as everyone has said, the conditions in the Antarctica are impossible to predict- no one has ever sampled where we are going to be.

Our client is a consortium of Universities that are running an impending Antarctic program. It's pretty amazing the scope of the project. We are providing a few instruments, another team provides a 10 million BTU hot-water drill, and the US polar services will provide logistics.

We are designing the core tube to withstand impacting a boulder; if it strikes one, we pull the machine up, and wait a bit before re-deploying. the Ice moves meters every day.

thanks everyone, your input is proving very informative, as I know very next to nothing about soil science.