Rammed earth farm house - Soil Testing

Soil

I started looking for suitable soil to build the structure. I stumbled upon a possibly good soil sample. Look and feel of the soil was good. It had lightly rained couple of days before I found the soil. Holding a handful of soil and pressing it made a nice and compact soil ball - which also gave an indication of good soil. Soil had bright red colour indicating no organic content. FYI - soil with organic content is usually dark/black/brown and smells like decayed organic matter. I decided to take this soil sample for further testing.

I did these tests:

1. Mason jar or sedimentation test.
2. Cigar test.
3. Biscuit test.

Soil Test : Mason jar / sedimentation test

I decided to begin with Mason Jar test or sedimentation test. This is a simple test needing little time, effort and material. Take a mason jar with water tight lid. If you don't know what is mason jar - don't worry, use a 1-2 kg/litre sized cylindrical glass bottle/jar with wide mouth as shown in the picture. Don't use fancy shaped bottle because it becomes difficult to see the ratio of different contents after sedimentation. Take soil sample and remove large pebbles from it (don't sieve it though). Put this soil to 1/3rd level in the jar. Pour water up to 2/3rd level of the jar. Close the lid tightly and shake the jar vigorously for few minutes so that soil particles mix well with water. Keep the bottle on a stable surface and wait for 1 minute. Heavy particles like sand settle down within first minute. After that silt particles settle down - within first 30 minutes. Fine clay particles remain suspended further and settle down within first 24 hours. Anything that remains suspended in water is too fine clay particle - ignore that. We should see several clear layers (typically 1-4 layers) of particles indicating how much gravel, sand, silt and clay is present in the soil.

After 10 minutes.

This is the result of the soil sample I picked up.

This picture was taken 10 minutes after shaking the jar. You can see a clear water layer at the top and soil suspended in water at the bottom. Clear water layer indicates there are no fine particles at all. But the soil looked pretty fine. I expected some fine particles to remain suspended in water. But this result does not show fine particles. I thought - may be I did a mistake and did not take right sample or did not shake the jar well when mixing soil sample with water. Therefore I repeated the test with another sample from the same pile of soil.

After 24 hours

Second sample also showed similar result. Clear water layer appeared after 10 minutes. Everything settled down within first 30 minutes leaving almost clear water at the top.

Even then I waited till 24 hours to see whether more particles settle. Nothing happened - it did not show any changes after 24 hours.

Conclusion: No pebbles or large particles in the bottom layer. This soil probably has only sand and silt because it settles very fast. There is no clay in this soil - because if there is clay it should have remained suspended in water for longer duration. Clay is the binder to hold the soil particles together when we ram the soil. Without clay, the wall may not be strong enough to withstand the load (both vertical and lateral) and wear & tear of normal usage. Wall built from soil without clay tends to weather easily. So what to do now? Is this soil unfit for rammed earth walls?

Soil test : Cigar test

Cigar test

Remove pebbles / stone pieces from the soil. Don't sieve it, just hand pick pebbles from some soil; couple of fist full is enough. Mix enough water to make plastic dough. Take a medium lemon size dough ball and roll it between your palms to make cigars. When you push it out of your palm, it bends, hangs and ultimately breaks. It should break between 2-6 inch length. It is considered too sandy if it breaks below 2 inches (too fragile/brittle). It is considered too clayey if it breaks above 6 inches (too sticky). 

My soil sample never went beyond 3-4 inch. Sometimes broke at 2 inches too. Soil is not clayey, it is more sandy/silty. This is largely in agreement with mason jar test.

Soil test : Biscuit test

Prepare soil sample similar to cigar test by removing pebbles / stone pieces. Mix enough water to make malleable dough. Make approximately 2 inch diameter and slightly thicker then quarter inch biscuits.

Biscuit test

First dry them in shade for couple of days and later couple of days in sun so that they dry completely. Keeping them in sun first may lead to cracks since the soil dries too quickly. After the biscuits dry completely, take a biscuit and snap it in 2 pieces with your fingers. Soil is too sandy if it shatters or breaks into too small pieces or crumbles into powder. Soil is too clayey if it is hard to break it. Soil is considered good if it snaps into 2-3 pieces easily. 

My soil sample turned out to be good in this test because it snapped into 2-3 pieces without too much effort.

Soil Composition

Particle ratio of 40% sand + gravel, 40% silt and 20% clay is well suited (all ratios approximate) for rammed earth. Soil is a complex material. Science (science of soil is called Soil Mechanics; you Civil Engineers - remember 2nd/3rd year?) has not been able to understand soil properly because soil does not follow precise scientific rules and formulas. Formulas and rules mentioned in soil mechanics are empirical. Therefore any rule (including the soil composition rule) is only indicative, not deterministic. Therefore soil without having one or more types of these particles or particle types in very different ratios can also make good/acceptable rammed earth walls; especially with some stabilizers like lime or OPC (ordinary portland cement). Lime is used with clayey soils. OPC does not play well with clay because clay is soft and expands when wet. Cement is hard and brittle after setting. Therefore OPC should not be used for stabilizing clayey soil. OPC is best suited for sandy and/or gravely soils. OPC with sandy and/or gravely soil results in soilcrete (like concrete - but with soil instead of gravel and sand [called coarse aggregate and fine aggregate in structural engineering] in concrete).

Proof of the pudding ... rather proof of the wall!

Soil suitability for rammed earth construction can be verified by making a sample rammed earth block and testing it. I made a small box of 12 inch x 5 inch x 5 inch for ramming a test block. Don't jump on me for the unorthodox size - I ended up with this size because I used some wooden boards lying around.

Test block

I was too lazy to cut the boards to regular brick size. Box sides are held with screws so that I can disassemble the box to take the block out after ramming. Well, that turned out to be a bad decision in hind sight because pressure of ramming caused the screws to pull out and the box disassembled itself! After that I made a tourniquet setup to hold the boards of the box tightly in place.

I rammed first block without any stabilizer. Block came out looking well - in the sense there was no shrinkage and no cracks. Probably I ended up using less water than needed while ramming. Therefore this block did not come out that strong. Rammed layers of soil separated and broke when I dropped it on the ground from shoulder height. Conclusion: 1st attempt is a failure. No worries, failures are stepping stones of success!

I stabilized second block with 5% cement. Block looked better than the previous block. Probably I used less water for this block too. I allowed it to cure for 7 days and then dropped it from shoulder height. It fared far better than the block without stabilizer. Rammed earth layers did not separate out when dropped from shoulder height. But the block broke into 2 pieces. Rammed earth layers separated when one of those pieces was dropped from 15 ft. height. Conclusion: 2nd attempt is a moderate success. I told you - first attempt was the stepping stone ;-)

I stabilized 3rd block with 5% cement . I used more water than earlier blocks. After 3 days of curing it is looking really good. I did a drop test after 7 days of curing. Block passed drop tests from shoulder height and 15 ft. height without major damages. What I mean by that is - block was intact, rammed layers did not separate, only the corners got damaged when dropped; which is expected.

Later I had put the third block in the farm under irrigation sprinkler for testing the effect of water on the block. This irrigation sprinkler sprays water every day for 1 hour. I did not find major erosion of the block after keeping it under the sprinkler for 15 days. Looks good and I feel confident of using 5% cement stabilization with little higher water content than usually recommended in rammed earth literature. Slightly more water content is better than less water content. Lower water content does not allow us to compact the soil to needed extent. Higher water content also affects compaction, but not as much as lower water content. Therefore it is ok to have slightly higher water content than lower water content.

Note: After having rammed the whole building, I know slightly higher water content is better for sure. There are instances where my helpers put in little more water. After ramming such soil mixture we found it "sweating". We saw sweating on the inside of the form work when we removed the form work and also on the surface of the rammed wall. It is extra water oozing out of the rammed soil. Later, after curing, I found such walls to be dense and nice. On the other hand, walls with lower water content did not turn up that well. I found them to be less dense and porous. Therefore slightly higher water content is better than lower water content.