Strength verification methods for stabilised soil-cement columns: a laboratory investigation of PORT and PIRT

Dry soil mixing (DSM) is a form of ground improvement in which dry cementitious and/or pozzolanic binders are mixed in situ with a soft soil. The binders react with the soil's natural moisture to initiate hydration reactions, thus leading to improved strength and stiffness characteristics. The deep dry soil mixing method (DDSM) is one form of DSM in which stabilised soil columns are created in soft soil profiles. Although site-specific binder trials can help estimate achievable column strengths, variations in the soil profile, moisture content, organic content, curing temperature and mixing conditions may mean that in situ field strengths can differ from those obtained under laboratory conditions, thus requiring the need for in situ strength verification. Two field methods used to estimate the strength of a stabilised soil column are the Push-In Resistance Test (PIRT), where a winged penetrometer is pushed down through the stabilised column, and the Pull-Out Resistance Test (PORT), where a winged penetrometer is pulled up through the stabilised column from beneath its base. Using a semi-empirical equation, whose origins lie with the Iskymeter penetrometer, the probing force is related to undrained shear strength of the stabilised column using a bearing factor, N. While N values between 10 and 15 are quoted in the literature (empirical and Scandinavian experience), few field tests and no laboratory investigations have attempted to investigate the relationship and the factors upon which N depends, thus limiting international confidence in the method. In this thesis, the results from a unique series of one-quarter-scale laboratory tests are presented and discussed. Reduced-scale PORT and PIRT penetrometers were manufactured and stabilised column construction procedures, along with penetrometer testing methods, were developed after some preliminary trials. Stabilised columns were constructed by stabilising a soft organic silt with cement and allowed to cure for various durations before penetrometer testing. Once tested, the columns were exhumed and samples taken for unconfined compression strength (UCS) testing, with UCS values up to 800 kPa observed. Exhumed columns showed evidence of cracking caused during the penetrometer test, believed to be due to the increased brittleness of the stabilised material over the parent material, and this is an important feature when interpreting the test results. The following considerations were also pertinent when interpreting the results: (i) The effect of curing temperature was investigated through a comparison of separate mould samples cured at constant and ambient laboratory temperatures. In addition, thermistors within a specific column (and its test basin) were used to investigate temperature variations during curing. Temperature was found to have a significant effect and was accounted for using a framework in which the curing time was adjusted for temperature in the same manner as is applied for concrete. The framework also includes for the different binder contents and variations in the soil's moisture content. (ii) Based on best-fit relationships, corrections were applied to the column samples strength to account for any strength gain that occurred between the column penetrometer test and the UCS testing of the column samples. (iii) A supplementary series of column tests were carried out to establish the contribution of friction to both the PORT and PIRT column tests. Using the probing force profile, corrected for friction, and the corrected column strength, the actual bearing factor N, was calculated. The results show N to compare reasonably well with the guideline value of 10 for both PORT and PIRT however, the strength of the stabilised soil column has an influence on the N value. In the upper portion of the column, lower PORT and PIRT N values were observed due to lack of confinement around the top of the column, allowing it to crack open ahead of the penetrometer. Where a surcharge was applied to the top of the basin, increased N values were noted. In the middle and lower sections of the column, PIRT N values are seen to show a constant value with depth, while PORT N values tend to increase with depth due to frictional forces experienced by the PORT penetrometer and its wire. This is the first laboratory-scale study of its kind to be conducted and provides in-depth guidance on how the factors which influence the N value may be assessed in the laboratory. Furthermore, the work has international implications for increased confidence in the interpretation of the strength of dry soil mixed columns