FAQ - Background

Q1. Can you provide me some hints about compressibility data for granular soils or those on which I cannot perform an odometer tests?
Sands and gravels are usually preconsolidated. Actually, most of the time, the compressibility for the full range of added load lies in the re-compression region. Typically the modulus number of a sand ranges from about 100 to about 400. For a gravel it ranges from about 300 through about 900+. These are wide ranges and in a given case where the low-range values could spell difficulty, you need to perform a test. Preferably in situ, say a small footing loading test at the actual depth and ensure that the initial stress conditions are those of your application (not a test on a ground surface where your foundation lies at depth). Then, back-calculate the measurements to obtain the soil parameters of the site.

Q2. UniSettle provides the option of using overconsolidation ratio, OCR, or preconsolidation margin. Are there conditions that would make using OCR produce more realistic settlement calculations and vice versa?
UniSettle offers the input option of the "preconsolidation margin" which was proposed by Bengt H. Fellenius many years ago. Most people report only the "overconsolidation ratio" (OCR) of a soil, determined, say, from a oedometer test. A UniSettle User can input the OCR directly in the settlement calculations, of course. However, often the OCR-value is only directly representative for the depth from where the sample was taken. Assume, now, a site with an upper layer of sand followed by compressible clay. Assume further that the sand thickness was originally 12 m (40 ft), but a long time ago about 7 m of the sand was removed (by erosion or similar) to a current thickness of 5 m. Therefore, the clay is preconsolidated by the unloading of 7 x 2,000 kg/m3, i.e., 140 KPa, which is the "preconsolidation margin". An oedometer test on a sample from 15 m, where the current effective overburden stress is 120 KPa, will then show an OCR of (120 +140)/120 = 2.2, say 2. However, at a depth of 9 m, where the effective overburden stress is 78 KPa, the OCR would have "increased" by 50% -to (78 + 140)/78 = 2.8, say 3. Of course, there is no particular difficulty in determining from the OCR value at 15 m depth to what the OCR would be at the top and bottom of the clay layer and then to let UniSettle interpolate linearly between the values. However, in accepting the input of the preconsolidation margin, the input effort is reduced. Moreover, a User may want to allow for a variation of the preconsolidation margin within the clay layer, which UniSettle easily accepts to do. If the input is realistic, the output (settlement) will be realistic, too.

Q3. With regard to the Immediate Recompression Modulus Eir, is the Eir value typically = 3 x Ei?
The reloading modulus Eir, is larger than the initial load modulus. For clays, typically, the ratio between Ei and Eir is 8 to 12, average of about 10. For coarse-grained material, the literature is scare and there is little consistency in the ratios proposed. This is probably due to the fact that coarse-grained soils are often precompressed and the modulus found is the Eir, already. Then, that modulus found, or determined, is often affected by sampling disturbance, or other effects. These effect will reduce early loading response that the intact Eir is less affected by, resulting in an apparent reloading modulus, which, in reality, is the undisturbed re-loading modulus. This said, for sand, a ratio of 3 to 5 is often proposed.

UniSettle treats the unloading modulus as the same as the reloading modulus. Calculations using an Eir-value larger than the Ei-value will always result in smaller deformation (heave for unloading and settlement for loading) than that for the Ei-value.

Q4. When I use an Eir value higher than Ei does that mean the rebound movement will be less than if I used the same value for Eir and Ei?
UniSettle treats the unloading modulus as the same as the reloading modulus. Calculations using an Eir-value larger than the Ei-value will always result in smaller deformation (heave for unloading and settlement for loading) than that for the Ei-value.