SHRINKAGE STRESSES IN SOIL-CEMENT STABILIZED BASE COURSE

Abstract

Soil-cement is a useful low cost highway construction material having a great promise of being adopted for stage construction in India. It imbibes the advantage of better load spreading capability, comparatively high modulus of elasticity, effectiveness against mud pumping in concrete pavements and speedy construction employing construction machinery. Uncertainty in the availability of bituminous binders due to the recent oil crisis has enhanced the importance of soil-cement as an alternative construction material, at a time when a large road kilometrage is immediately needed for catering to the needs of medium to he^vy traffic. Numerous- performance studies although have established the suitability of the material, in many instances failures have been reported and are attributed to shrinkage cracking. Cracking does not immediately impair the serviceability of the pavement, however, the cracks being progressive in nature, open out,penetrating the upper layers causing material fai lure. These cracks look unsightly and pose severe maintenance problem. Interior of the pavement is exposed to weathering effects causing progressive deterioration. Thus there is a need to examine the occurrence of shrinkage in soil-cement stabilized base courses. Mechanics of shrinkage cracking in soil-cement base 11 courses have not been sufficiently investigated. A few studies conducted were based on visible crack intensity followed by efforts based on the identification of shrinkage as the major affecting parameter. Shrinkage characteristics of soil-cement have been studied and employed for crack control. Simultaneously use of admixtures in crack minimiza-* tion has been investigated resulting in limited achievement. With a view to examining the problem of shrinkage crack ing rationally, computation of shrinkage stresses is under taken. Based on principles of mechanics, crack width and spacing are calculated. However, there is no agreement between the computed and observed values. The excessive computed crack width is attributed to stress relaxation due to creep exhibited by the soil-cement material. Thus viscoelastic nature of soil-cement simulated by rheological model is employed to compute shrinkage stresses. Consequent to shrinkage stresses defying reasonable results of computation and with a view to designing soilcement base courses on the principle of acceptable cracking frequency concept, present study is planned. In order to analyse the problem of detrimental shrinkage cracking, the mechanics of soil-cement rerction and the structure of soilcement have been studied in CHAPTER I. It is inferred that in granular soil-cement mixes shrinkage, occurs due to loss of moisture from the cement gel whereas the cement-clay mineral interaction is the additional contributing factor Ill in the case of clayey soil-cement mixes. Cracking in soil-cement which is the result of shrinkage, gets reduced due to the compensating influence of creep exhibited by the material which has to be adequately accounted for. CHAPTER II deals with the preliminary computation of shrinkage stresses based on the existing elastic theories. Thin slab theory and subsequently thick slab theory are employed to compute shrinkage stresses in soil-cement base courses. An assessment of the nature and magnitude of shrinkage stresses are achieved by employing general -equations of theory of elasticity with assumed displace ment functions. CHAPTER III is devoted to developing an analytical model for use in the computation of viscoelastic shrinkage stresses. Model simulation is performed for soil-cement material responses. However, actual creep response is employed in the formulation of the numerical computation of relaxation modulus. To achieve efficient computation technique, the concept of modified strains obtained by the combination of shrinkage strains and corresponding relaxa tion moduli is adopted. CHAP'iER IV deals with developing finite element formulation for initial strain problem. Modified strains obtained earlier are employed as initial strains by time incremental procedure to yield stresses at time nodes in the iv continuum of soil-cement base courses. The computer programme consisting of a main programme drawn to run on IBM 370/145 with seventeen subroutines is developed for obtaining shrinkage stresses in the continuum of soilcement base courses. CHAPTER V of the study presents experimental investi gation including creep of soil-cement mix materials. Labora tory experimentation is explained and observations recorded to identify important factors for use in the design of soil-cement mixtures for semi-full scale base courses. CHAPTER VI presents the investigation on semi-full scale base courses, carried out with the primary aim of simulating field conditions closely and deriving data for the computation of shrinkage stresses in base courses. Add itional observations of crack initiation time and dimensions are also recorded. CHAPTER VII describes the computation of shrinkage stresses using input data derived from semi-full scale study. Inputs of creep response and free shrinkage are obtained from the creep response curve of the material and from moisture variation along the depth of the drying base course respectively. Creep response numerically provides relaxation modulus. The free shrinkage at time nodes is computed and used as intermediate input which subsequently yields modi fied strains in combination with relaxation modulus. Initial strains after time increments are in turn converted to V shrinkage stresses across the depth of the base course at time nodes. Finally, CHAPTER VIII lists various conclusions drawn from the present study. The soil-cement base courses develop tensile stresses at the top which are of higher order for full base restraint condition and extend to maximum of one third the depth. The maximum tensile stresses attain an optimum value and then decrease leading to a logical conclusion of importance of curing during initial period and that of extended curing. Predicted crack dimensions provide an excellent agreement with the observed values. Thus, it is asserted that soil-cement base courses can be designed on the concept of acceptable frequency of shrinkage cracking.