No abstract

The shrinkage mechanisms of portland cement paste were investigated from shrinkage, weight loss, and pore structure measurements using nitrogen sorption and mercury intrusion porosimetry (MIP). Thin samples (2.3 mm) of well-hydrated (165 d) pastes of 0.4 and 0.6 water-to-cement (W/C) ratios were dried directly from saturated surface dry state to 75%, SO%, 11%, and 0% relative humidity (rh). From equilibrium shrinkage vs calculated increase in surface free energy curves two active stress mechanisms were identified. The GibbsBangham (surface free energy) effect is the major stress mechanism, which is active in the entire rh range investigated, whereas the capillary stress effect is active above 25% rh. From elastic modulus calculations it can be concluded that true Gibbs-Bangham shrinkage accounts for only 33% of total first drying shrinkage. Thus nearly 67% of first drying shrinkage may be due to a decrease in interlayer spacing caused by Gibbs-Bangham and capillary induced stresses. Further, nitrogen measures the true external surface area, and total external pore volume can be obtained from combined measurements using nitrogen sorption and MIP.

No abstract

No abstract

The

No abstract

No abstract

Temperature and moisture gradients can lead to significant tensile stresses at the slab top and bottom. Current techniques for assessing the internal stresses due to such gradients are based on the assumption that temperature and moisture distributions through the slab thickness are linear. However, the actual distributions of such gradients have been found to be highly nonlinear. A new closed form solution technique for calculating the stresses in a pavement slab due to nonlinear gradients is introduced. The analysis is separated into two parts. In the first, an expression is presented for calculating the self-equilibrated stresses within a cross section due to internal restraint (i.e., satisfying equilibrium conditions and continuity of the strain field within the cross section). These stresses are independent of slab dimensions and boundary conditions. In the second, the stresses due to external restraint (i.e., self-weight and subgrade reaction) are calculated using an equivalent linear temperature gradient obtained from the first part and existing closed form solutions by Westergaard or Bradbury. The solution to this step includes slab length and boundary conditions. Total internal stresses due to nonlinear gradients are obtained by using the superposition principle. The methodology has been applied to field data from two studies in which the temperature profiles were recorded throughout a 24-hr period. Linear gradient solution methods cannot accurately predict the curling stresses in concrete pavements. This is especially pronounced during nighttime and early morning hours, during which nonlinear analysis predicts tensile stress in both the slab bottom and top before the application of any traffic loading.