Effective Technique to Estimate Cross Model Parameters

Abstract

The Cross model has been successfully used for modeling the shear thinning behavior of asphalt binders. Usually minimization of least squares/absolute error is used for obtaining model parameters. The curve fitted using these techniques shows more deviation at those parts of the domain where there are fewer observations. To improve the curve fit and eliminate this bias, an area minimization technique is proposed in this study. To the best of the authors’ knowledge, this technique is proposed for the first time in model parameter estimation. This paper discusses and compares least-squares and area minimization for fitting the complex viscosity master curves of four different asphalt binders at six different reference temperatures. The results clearly demonstrate that area minimization offers a solution that has less deviation throughout the domain of the observed data when compared with least-squares error.