Soil Testing and Data
Our main source of information, other than testing the soil itself, was
the USDA´s Soil and Water Conservation data.
According to their maps
our lot overlaps two soiltype areas -- Avonburg and Rossmoyne -- which are pretty similar except for the Rossmoyne´s very slight slope... we focus on the Avonburg since its drainage is a little slower... otherwise these two soil types have distributions and classifications that look basically no different... Similarly conservatively, the Health dpt chose to list the Ava type on their application data...
We did the jar test for samples from our land when we were first considering the site. The layers that precipitated out seemed fairly consistent with the ranges of sand, silt and clay in the official published data on the site... Instructions for the jar test as well as an array of other tests are given in the appendix of David Easton´s book The Rammed Earth House.
Three tables that describe our land as far as suitability for foundations and rammed earth are the tables for composition, type of soil and water table behavior...
The Chem/Phys properties have the percentages (by depth) of clay, the finest soil particles... Other factors of interest include the permeability rates for septic system design, and the shrink-swell potential for the absence/presence of difficult clays for ramming...
The Engineering properties have the classification data used by civil engineers as well as the older agricultural classification I'd seen in soil cement texts... For ramming, there are the distributions of soil particles by size, namely how many pass through sieves of increasing fineness. Larger than 3 inches are the stones and cobbles. Gratefully there aren´t any til we reach almost 3 foot down and then very few thereafter as well. The various sizes of gravel will not pass the No.10 sieve leaving only sand, silt and clay in that number. Sand is finally excluded by sieve No. 200 leaving silt and clay, of which we already know the clay portion so we can arrive at the silt portion by simple subtraction and have the whole picture...
Although this distribution for the site is within the broad limits of soils suitable for ramming, there was an opportunity we noted in evaluating the amounts and uses we needed. Adding an equal amount of clean sand to the distribution given in the table for soil below the organic layer (which should not be used for ramming and will need to be set aside for landscaping later), should give us an almost ideal ramming mix according to the UN publication on soil-cement construction... The title was Soil Cement: Its Use in Building... their optimum was 75% sand with 10% clay and our 1:1 mix would be over 60% sand with 15% clay..
According to the little monographs on soil-cement these soil classifications -- A4, A6, A7 -- listed in the table needed between 8-14% cement for stabilization to handle weather effects, and they advocated doing tests at 2%age point intervals to determine the optimal/strongest mix... we would do the stabilization for the outside walls to make them more weatherproof but that might be omitted for the inside walls...
Testing of the mix will be the key in any event since,
although the ranges are given for the original soil´s classification, what happens when we mix sand in...? then what classification would it be?
In the monographs, some correlations were given between the bearing values (useful for adequate foundations) for soil types and values for engineering parameters used in calculations... still trying to "translate" what bearing value or modulus of soil reaction could tell me about the stiffness (E) of the rammed earth shell...
For example, the bearing values for the Ava soil´s unified classifications of CL, ML alone predict a modulus of soil reaction of between 130-240psi so the 30" diameter disc used in the standard testing would deform the soil 1/10" with 9,000-17,000 lbs applied... comparably, the 24" thick rammed earth walls, which at their tallest weigh less than 4,500 lbs per linear foot and carry a roof share of roughly 4,250 lbs, would deform the Ava soil beneath them by 1/4"... which would suggest that the designed walls really need no special footer for soil strengthening, only for moisture protection... the psi is only 26psi... can´t imagine there being any foundation problem, especially since it´s all under the insulation umbrella... all I was concerned about was getting the french drains under the footer to empty into the circum-house trench, just like the insulation umbrella... didn´t figure they should be very deep, maybe 1.5 to 2ft down or they´ll be emptying warm water into the trench...
The compressive strength of rammed earth walls and roof was supposed to be optimized by the water content, which according to Construction of and on Compacted Fills (Monahan) should be about 5-7% for "typical allowable load-bearing borrow" to achieve "modified proctor" which is the target compaction for ramming.. of course the field written books simply talked about squeezing a handful and dropping it from waist-high and adding water if it shattered too much and describing the sound of the tamper hitting as being neither a whoosh nor a galumph!!
The text by Bruce King, PE has some sample rammed earth strengths and official specs of the daunting variety that look more rigorous to an engineer.. his example was "A-2-4 with 4.0 pH, stabilized to a liquid limit of 30, plastic limit of 21, and pH of 10.5 which achieved a cured density of 129pcf and compression tested at around 3000psi by day 56" which was his recommended testing protocol term... with the sand added as planned, the Ava mix is closer to an A2 but the stabilization to pH of 10.5 is more cement than most mix designs... other texts had a variety of compression test values as their expectation of normal, most less than King´s 3000psi... King also talks about earthquakes, wind and guidelines for size of openings and lots more...
The Water Table behavior is the third table and has flooding data as well as water table depth and movement timetable info...
The implications of the rising water table, though no problem structurally for a design built at grade, led to some figuring because of its possible impact on our planned heat storing design... which led to consultations with the regional soil scientist and the conclusion is detailed in
the errant watertable section.
