Silt or clay?

Dear fellow permies and natural building enthusiasts,

I started building with cob and sourced some clay from a local brick factory. I added sand from the river and hemp fiber and it works quite well.
Next, I would like to use the material I have on site.

One thing I don't really understand is how I can differentiate between silt and clay. I'm expecting more silt than clay, as we are located on the lower slopes of a mountain range.
Any tips?

PS: I haven't really figured out how to actually do a good search on the forum, so excuse me, if this is already answered in 50 other threads.

The classic method that a lot of us use is the settling test. Mix a sample of the soil with water in a glass jar, put on a lid, shake vigorously for a couple of minutes, then set it down. The sand fraction will fall to the bottom immediately, followed in short order by the silt. After ten minutes, some of the clay will have begun settling, but it can take a day or longer for the water to clear. You will see bands through the glass and have a good view of the particle sizes, and this will give you a pretty good idea of the proportions.

There's also the worm test: take a handful of the moist soil and roll it between your palms until it forms a worm-shaped cylinder. High clay content will allow you to make smaller diameter worms that continue to hold together, while lots of sand can make it impossible to even get one started. Silty soils will be sort of in the middle, but the worms will be fragile. A good clay worm can be picked up by one end and swung back and forth.

However, for cob and plaster mixes, what we're really interested in is shrinkage and strength (compressive and tensile). The easiest way to evaluate this is to take some soil samples, mix with water until you've got something that can be shaped by hand (mud pie consistency). Pack the mud into some equal-sized forms - I like to use sections of 100 mm PVC pipe and level off the tops. Set them somewhere to dry, and then do a visual assessment of the pucks. The more clay is in the mix, the more it tends to shrink and crack. High amounts of silt or sand can sometimes crack as they dry, but will have little or no shrinkage.

Finally, do a shatter test. Drop each sample from 1 m (about waist height) and see how much breakage you get. A strong mix might only lose chips from the edges of the puck. Sandy mixes will fall apart. Silt-dominated soil will land somewhere in the middle. A "goldilocks blend" will have low shrinkage and withstand impact, and this normally happens with clay fractions of 10-20%.

Test pucks have become my go-to because the soils in this region are dominated by very fine-grained silts that are often mistaken for clays because of their consistency.

Hi there, Benjamin,

My first thought on reading your query was, "Clay's sticky and cold; silt's just sandy mud," but I double-checked on the internet to make sure.  Apparently one definition of silt is that its particles are 'between clay and sand in size'.  So, it's more claylike than sand, but sandier than clay.

Hope that helps!

-- Jojo

Hi Jojo,
thanks for the answer. I read that too, but I'm not sure what the consequences of that on cob are.
I'll make some "test muffins" with the material and report back.

Hi Benjamin,

Right -- I forget about practicalities, which are the whole point!  I'm thinking clay will take longer to dry, and will stand a higher chance of cracking, but you might struggle to form a solid mass with silt.  I'm interested to see how your test cakes turn out...

Jojo Cameron wrote:...you might struggle to form a solid mass with silt.

I don't think you would make a usable cob with a high-silt subsoil. Most silts are slippery in texture and don't stick together as they dry. You can get away with some silt but you typically want a sand-clay mixture. Sand provides the compressive strength (much like ballast/aggregate in concrete) and clay sticks everything together, as the clay particles have a (negative) electrical charge and are attracted to one another and other particles in the soil.

As suggested, the simplest way to test whether your soil is suitable for cob is to test it. Either in little muffins or as bricks. Play around with the test pieces and see how strong they are. If you find something is almost working, try adding a little more sand or clay and retesting - or dig a hole in a slightly different spot.

Thank you all for the input. I’ll make some test cakes when I’m back at home and share the results.
The settling test showed almost no sand, but doesn’t let me distinguish between silt and clay.

The sediment in that jar looks like silt to me, because there's still water in it: clay is heavy enough to push all the water out of itself and form a unit.

Benjamin, unfortunately that doesn't look like great cob soil. As Jojo says, it looks like it's predominantly silt. You might find a layer of clay appears after a few days, once the water completely clears, but it will be a small constituent of your soil.

Are there any other spots on your site that you could dig? If you look for changes in what grows there, that might indicate a difference in soil makeup. Also, wetter areas often have higher clay content in the soil as it doesn't drain as freely as sand or silt.

Benjamin Dinkel wrote:Thank you all for the input. I’ll make some test cakes when I’m back at home and share the results.
The settling test showed almost no sand, but doesn’t let me distinguish between silt and clay.

This is my area of expertise, so I'll weigh in here.

I agree with Luke, this looks like poor cob soil. Also Luke makes the crucial point about why you need clay. It is the glue that binds together the ingredients of any soil-based construction method. Without clay, there is no strength.

Let's get some specifications on what we are dealing with. Soil has 7 parts, here's how they work in soil based building materials:

Air - mostly squeezed out by compression, obviously adds no strength
Water - also not wanted in the final building structure, but needed to make a workable mixture, hydrating the clays to make them slippery
Organic matter - mostly undesirable, excepted added fibers like straw improve tensile strength
Big rocks - gravel and larger rocks are bad for soil-based building materials
Sand - 2mm to 0.05mm - mostly made of quartz or other resistant minerals, adds compressive strength, zero tensile strength.
Silt - 0.05mm to 0.002mm - basically smaller particles of quartz, so fills in the spaces between sand particles, but behaves the same way mechanically
Clay - smaller than .002mm - here is the crucial ingredient...

Clay is vital, like concrete needs Portland cement, or mortar needs lime. It is not just a smaller size quartz grain, it has very different chemical and mechanical properties. Clay is made of tiny plates with ionic bonds dangling from their flat surfaces. Water is strongly attracted to these ionic charges, so well hydrated clay acts almost like a viscous liquid - the water lubricates the plates allowing them to slide past each other easily.

A small particle of quartz, on the other hand, has almost no chemical affinity for water, so a soil composed mostly of quartz particles  - no matter the size - behaves very differently than a clay soil. Think of a sand castle at the beach. Wet sand holds together a little bit due to surface tension - the water is a weak temporary glue - but as soon as it dries out, it crumbles back to loose sand. The same happens for any size quartz grain. A pile of rocks has no strength, a pile gravel - no strength, a pile of sand - same. A pile of silt - almost the same, but it's actually hard to find pure silt, there's almost always some clay in every silt, even if its just individual clay particles sticking to the silt particles. Now you've got some glue action, and silty soils have texture and strength.

The real magic happens when clay dries out. The slippery plates of hydrated clay lock together when the water leaves. Some of those dangling ions bond with their neighbor on adjacent plates. The plates mostly line up parallel, like stacked plates, but overlapping in every direction. These ionic bonds are pretty strong, so drying clay converts from a thick slippery liquid to a rigid solid. Pottery is almost pure clay. Once the water content drops enough, its a weak solid. Take away all the water, its quite strong. This is reversible. Add water, unfired pottery becomes soft, add more it becomes a liquid. Firing pottery removes ALL the water and then melts the individual plates together. This is irreversible, water can't get back in between the plates.

So how do we make buildings out of soil? Clay is the glue - its binds ingredients together and provides almost all of the tensile strength - but just like concrete and lime-mortar, we need other ingredients. Portland cement by itself sets up to make a strong building material but it cracks as it sets because it shrinks. So we add sand and gravel to minimize the shrinkage to the distance between adjacent sand particles instead of the whole block. Same with soil building. Clay is the glue binding together the other ingredients. Unlike cement-concrete, there is much less chemical bonding between the glue and the matrix, so just like unfired pottery, soil buildings need comprehensive protection from water. And pure clay cracks as it dries because it shrinks a LOT.

So soil building needs clay, some other filler material to add compressive strength and minimize shrinkage, and some fibers are nice to increase the tensile and flexural strength.

The soil settling test in the jar looks to me like mostly sand. I can see individual grains almost to the top of the sample. If you can see it, its sand. Please confirm by examining a bit of soil in the palm of your hand. Add water, smush it around with your finger, sand grains should be evident visually and by feel.

Silt is just the same material as the sand, just too small to see individual grains. Tested by tasting it. Dig into the top layers of your soil sample jar after pouring off the water and letting it dry out a bit. The very top layer will be clay (if there is enough to form a distinct layer). Clay feels slippery and smooth in your mouth. Dig deeper, find the silt, it feels gritty on your teeth. The silt grains are strong quartz particles, not slippery clay particles.

Ok, now wash out your mouth. Yes, geologists are disgusting and nasty. That's why we drink a lot of whiskey - gotta wash out that dirty mouth and kill the germs.

Now for the practical application.

If you've got a soil that's mostly sand, it will not be suitable for cob. At all. It will just fall apart as soon as it dries, like a sand castle. But its not hopeless, you just need to keep looking for clay rich soil to add to it. Dig deeper, dig in low spots near water, but not immediately adjacent (floods drop sand next to a stream channel, silt and clay farther out in the floodplain). Clay is created by weathering of certain kinds of rocks. If you find the right rocks, you'll find clay nearby.

Another tactic for sandy soil is to remove some of the sand. Put your soil in a barrel or bucket, mix thoroughly with water, then put a hose down to the bottom with a slow trickle of water. At the right flow rate, sand will remain in the barrel, silt and clay will overflow. Catch the overflow, let it settle out. Hopefully this fractionation will leave you with silt and clay in a useable ratio, just add a little more of the sandy soil. You can try two-stage fractionation - sand in the first bucket, silt settling out of a second, much larger container, clay flowing out with the water. Clay will not settle out if there is any movement at all. Clay needs very still water and many days or weeks to settle.

I hope you can use this to make progress on finding a source for cob. Let us know the results of further testing.

One more thing:

I just remembered a far easier way to extract the clay from soil. If there is any clay at all, it will settle last after mixing with water. That's what happens in the jar test. Just scale that up. Get some large tubs, pans, barrels, buckets... almost any container will do. Fill it part way with soil and plenty of water, mix it up, let it settle for a few weeks. Gently scoop out the water off the top without disturbing the top layers. Then let it dry completely in the sun. If there is a clay layer,  it will shrink, separate from the silt layer below, and curl up into plates or flakes. These are easy to remove. The dried silt below will also have some clay in it, so it may also crack into blocks, but not curl up. So it's still useful, and you can adjust the proportions as needed, discarding the sand or silt that exceeds the quantity needed to achieve the optimal ration of sand/silt/clay.

Thank you so much for your input and your very  thorough explanation. When I come back I’ll run some tests anyways and maybe look at some other parts of my property.

I found some promising material that was shoveled out of irrigation channels. It was hard enough to be picked up in a shard and had drying cracks. I soaked it and made test pucks/muffins with different "clay"/sand mixtures.
Unfortunately it seems way too silty. I was able to crumble the pucks with my bare hands, even the high "clay" ones.

The mixes you see in the photo are (sand:clay-ratio) top row, left to right:

0:1 (just the mud)

1:1

2:1

4:1

mid

6:1

10:1

1:0 (just the sand)

The search continues!

This will tell you more than you probably want to know, but google 'Soil Texture Triangle'

also, visit here:

https://www.bing.com/videos/riverview/relatedvideo?&q=Soil+Texturing&&mid=9854BE4AED0F59D52C979854BE4AED0F59D52C97&&FORM=VRDGAR

Yes, the mason jar test works. It is the equivalent of what I used in my soil classes 36 years ago at Purdue University.  You will know your clay content vs silt in two days. (Sand as well)

Thank you mark. I will try adding some detergent and draw the lines for sand, silt and clay. I didn’t know about the exact times before. I also didn’t know about importance of sifting. Helpful material!

Look up Andy Ward on youtube and watch some of his videos on finding wild clay, it might help you find clay-heavy soils in your area. Note that his goal is to make pottery out of it, which requires a higher clay to sand ratio than cob/adobe/etc. From everything I've read, silt is a neutral to negative contribution to the mix (my soil doesn't have a high silt percentage, so I can't tell you about that from any personal experience).

At a glance, your test ratios look like too wide of a range to me. And almost all of them are majority sand, if I understand your notation right. Try using the highest-clay soil you can find locally, and test a range more like 10% sand to 50% sand. Probably also include fiber in your test mixes (you didn't mention whether you did this or not, but in your photo I don't see any obvious fiber.) My understanding is that it's somewhat rare to find a soil that won't crack without added fiber, even if the clay percentage is close to ideal.

both Phil and B Beeson are on the money.
I  my experience, e soil is always found insitu in the ground, never in a channel or a pond.
if you remove the topsoil, the next layer s usually clay.
If you do the soil tests at different locations on your property, you will find a suitable naturally occurring mix.