silts vs. clays 1

[howardoark]

I was going over in my head how to explain something to non-geotechnical engineers when I realized I didn't understand it myself (embarrassing). Non-plastic silts are, imho, the worst construction material, weak, erodible, subject to collapse - assuming you won't use peat as a construction material. That got me to wondering what the physical explanation was for why some silts are plastic (i.e., why they hold onto water). Which brings me to my question, if silts are plastic, why aren't they classified as clays? Kaolinite, a clay, for example, can classify as either a low plasticity or high plasticity silt (generally plotting just below the A line). Terzaghi & Peck divide silts into inorganic (rock flour, non-plastic) and organic (plastic), but don't say why organic silts are plastic. Lambe and Whitman divide clays from silts at 0.002 mm (No. 200 sieve is 0.074 mm) which is unsatisfying as it tells you nothing about the material behavior. I have a feeling that an MH soil with a LL of 90 and a PI of 40 probably isn't dilatent (does anyone know?). Should we really classify low and high plasticity silts as another variety of clay and only classify non-plastic fines as silt?

Thanks

 

[geomane]

I was reading through Lambe and Whitman this week.

Clay particles are smaller than silt sized particles. Clay is a colloid and has a higher specific surface surface. Therefore clay has higher electrical forces between particles, because the magnitude of the electrical charge is directly related to surface area.

 

[Ron]

Yes..clays and silts are similar, but very different in why they are plastic. Clay particles, as jmcc3265 noted, have a higher specific surface than silts. This is because the particle sizes are smaller and are usually flatter. Silt particles are rounded and larger than clay particles.

Clay particles are attracted to each other by relatively high Van der Waals forces....they higher the force, the greater the plasticity. The flatter the particle, the higher the forces.

Silts are plastic usually for other reasons. Since they are round, Van der Waals forces have less influence but the medium in which they sit has a big influence. Most often that medium for silts is some organic "soup". That soup is the sticky part and the silt particles are entrapped in it providing complementary filler for the soup, thus making it "thicker" and stickier.

One clear depiction of this premise is doing a hydrometer test on the fine particles of soil. For most clays, a deflocculant (such as sodium hexametaphosphate) is necessary to "break" the forces. For silts, not so much because the "soup" is easily diluted by water.

 

[GeoEnvGuy]

I work a lot in tailings which is essentially non-plastic silt some sand ML aka slimes coming out of the discharge pipe. Contrary to Ron they are angular. Using tailings beaches we try to get it down to sand some silt SM. To compact SM the proctor for SM on this material is still crap coming back at 10% optimal MC, which is not achievable in the field, which means the mould is holding water, displayed in Holtz and Kovacs book as a poor construction material.

The behaviour of silt is the worst in my opinion. Completing a dilatency test aka seismic screening, tapping a ball of material in your hand, comes back as rapid, both clays and sand have slower reactions. Clays have the previously stated van Der waal forces of attraction. Sands have higher frictional strength like sanding wood with 40grit compared to 200grit.

From the research I looked at, when the non plastic silt becomes 25-40%, sand is floating in a sea of silt and the frictional strength is reduced. On the low side 5-10% clay fraction 0.002 is needed to get plastic behaviour out of a silt. On the matrix suction perspective I would assume silt is actually better than sand.

I have also seen natural silt deposits and they react the same to the dilatency test and plot as ML. Comparing tailings to natural silt not much difference in behaviour or strength.

A practical method to describe silt vs clay behaviour is from the dilatency test. Any material with a PI above 30 probably won't have much dilatency which is the medium plasticity CI or MI boundary 30-50 in casagrandes chart. Beyond the bottom left corner of the chart you are distinguishing how good the attractive forces are due to the alignment of layers. Or how much attractive clay to frictional silt portion is in the material.


The first stage of site investigation is desktop and it informs the engineer of the anticipated subsurface conditions. By precluding the site investigation the design engineer cannot accept any responsibility for providing a safe and economical design.

 

[maxim22]

This is a topic I'm interested in at the moment on the subject of mixed soil types and where the dominant behaviour comes from.

It helps me to scale up the different particles sizes and introduce the different mineral types that grains are derived from and the depositional environment a soil was produced.

If the clay particle is scaled to say a 25 mm disc, in proportion silt might be a big beach ball, and a sand particle the size of a car. The main engineering properties are determined by the finest particles if enough are present to overcome the coarse fraction on a case-by-case basis.

Different rock minerals weather/decay to grains plate shapes or blocks. Grains that are round may be shaped like that due to transport erosion over geological time, angular grains may have been less travelled or younger. The smaller clay plates don't erode off each other in the same way due to the interparticle forces which are not there for block shape silt/sands.

Glacial deposits often comprise variable layers and mingled fine soil clay minerals and coarse silicates. More stable deposition environments have less variation in particle size.

Chalk rock samples can be crushed to extreme and disintegrated in lab preparation stage to plates size to get a plasticity result in the standard test.

 

[fattdad]

it has to do with moisture sensitivity. Something about the inverted, "Vee" of the proctor?

I don't really relate much of this to mineralogy, albeit important!

If I crush up a bunch of quartzite and reduce it to flour, it'll be silt and it'll be non-plastic. Mineralogically, it'll be quartz. If I then add 10% by weight bentonite, ASTM will likely classify it as, "Fat Clay." But it only has 10 percent clay minerals!

No, in ASTM it's all about behavior! In USDA it's all about grain size. They are for different goals!

f-d

ípapß gordo ainÆt no madre flaca!

 

[geotechengineer1]

To be honest I've never found a soil classification system that I liked or fully agreed with. Most soil classification systems would work great if nature was nice and provided us with soils that are 100% sand, or 2% clay, 1% sand, 97% silt. Instead we get 30-30-30-10 or 35-25-25-15. The confused graduate engineer or technician is sent to the field with an easy to follow USCS flow chart or derivative, and finds he cannot make heads or tails of anything, eventually getting yelled at/red penned enough by the guy in the corner office that he starts producing roughly the same 'classifications' that the guy in the corner office produced when he was a graduate getting yelled at. After 6 months to a year the graduate finally gets promoted to writing reports where he never has to worry about logging again.

Also, what about the minearology? Apparently some types of 'clay' will result in a sample plotting below the A-line if they exist in high enough portion of the clay fraction with no montmorllionoite in the sample - should those soils be called as 'Clay' or 'Silt'?

What even is a 'Plastic Silt'? I've had several engineers insist a soil is a 'Plastic Silt' only to run a hydro and an Atterberg and find out tha tthe plasticity is because of the clay content

 

[maxim22]

The purpose of the classification is for understanding material engineering properties, and the possible variations.

I'm UK based engineering geologist, we have lots of mixed soils.

Logging and classification are linked, but not the same thing. Engineering behaviour of a soil is often dominated by the fines content and it's mineralogy.

My experience is that loggers rarely undertake simple field tests on soils (disintegration in water, dilatancy, rolling threads etc) then may be surprised by the lab outcome. I have some samples logged on site as dry fine sand which I added water to in the office and got a plastic thread rolled out easy. The lab results will confirm PSD & limits when I get them.

A plastic silt in the UK would be particles more than 75% > 2 um < 63 um but more than 20% < 2 um and the 20 % was a sensitive clay mineral. This imaginary soil would behave like a clay in the hand. If the 20% was kaolinite, it might not be plastic but the PSD would look the same. Due to the science of clays.

 

[geotechengineer1]

An 'elastic silt'according to the uscs is just a clay that plots below the A line on that chart Cassagrande drew while drunk or in a temporary lapse of judgement that an entire industry is doomed to follow until the end of time, at least in USA/Canada/New Zealand as far as I can tell. I'm not sure about the UK practice. The American/Canadian/NZ systems are based on behavior in theory but in practice it's a mess.

It's the clay- or certain percentages of certain types/combinations of clay mineralsthat makes it elastic or plastic, why not just call it clay?

Or perhaps the words clay and silt are just distracting from what's important- is the soil plastic or non plastic (if non plastic, is it just dry - I've also seen loads of plastic soil logged as sand or silt, or "clayey silt" (why does the term clayey have to exist...), is it dilatant, is it sensitive, etc.

Anyway I think I agree with the OP -"plastic silt" as a classification shouldn't exist (neither should clayey silt) and an entire industry shouldn't be bound to some line on a graph drawn by a dead dude a century ago.

 

[maxim22]

I think classifications exist for civil engineers to apply understanding/process to soil layers on large projects for generalisations. They help with interpretation of big data, but don't lose sight of any facts.

I don't tend to use classifications and look at the hand specimen engineering behaviour and the lab data.

Materials can be defined by the characteristic properties and keep in mind the variations. Soils can be more variable than other materials, so people want to generalise.

Perhaps we don't need classifications, but I think clayey silt is valid in a description for example to tell the engineer that the material permeability will be lower than a pure silt.

 

[Blairsy]

OP - The article I linked to below pretty much sums up what the others have said.

https://fl-nzgs-media.s3.amazonaws.com/uploads/2017/11/NZGS_Symposium_20_Hind1.pdf