Please use this identifier to cite or link to this item:
A Statistical Dynamic Modulus Model of Hot Mix Asphalt Using Joint Estimation and Mixed-Effects Accounting for Effects of Confinement, Moisture and Additives
|Title:||A Statistical Dynamic Modulus Model of Hot Mix Asphalt Using Joint Estimation and Mixed-Effects Accounting for Effects of Confinement, Moisture and Additives|
|Date Issued:||Dec 2016|
|Publisher:||[Honolulu] : [University of Hawaii at Manoa], [December 2016]|
|Abstract:||With the implementation of mechanistic-empirical pavement design methods, dynamic modulus (|E*|) has become the predominant characteristic of Hot Mix Asphalt (HMA) used as the elastic modulus in the computation of stresses and strains in pavement structures. The predictions of |E*| obtained with the Witczak models currently used in the Mechanistic Empirical Pavement Design Guide (MEPDG) do not account for some HMA characteristics such as polymer modified binders, fibers, confinement and aging effects related to climatic conditions. Therefore, development of models more representative of local materials and conditions are desirable.|
In this study, a predictive model for dynamic modulus of HMA was developed taking into consideration several HMA characteristics and testing conditions. 6821 observations of 257 mix specimens from three different laboratory datasets, two from Hawaii and one from Costa Rica, were used to estimate the model parameters. All three data sets contain information about some variables in common such as air voids, binder content, and gradation; however, some datasets contain mix characteristics and testing conditions not available in other datasets such as confinement level, available only in the Hawaiian datasets, and the number of freeze-thaw cycles, available only in the Costa Rican dataset. Important characteristics observed from these datasets include confinement, number of freeze-thaw cycles, binder modifiers (SBS polymers) and mixture additives (e.g. fibers and anti-stripping agents), all of which together with other commonly used variables were found to have statistically significant effects. The model was developed by using joint estimation and mixed-effects techniques. Joint estimation allowed the identification of model parameters available from only some of the databases and the determination of bias parameters. It also resulted in more efficient parameter estimates. The mixed-effects approach was used to account for unobserved heterogeneities between samples. Using these approaches, together with proper consideration of heteroscedasticity, allowed the estimation of a comprehensive statistical predictive model that satisfies closely all the regression assumptions and that provides accurate values of |E*| for Hawaiian and Costa Rican conditions for any combination of temperature and frequency. These can be used to generate the |E*| inputs that the MEPDG needs to compute |E*| master curves.
|Description:||Ph.D. University of Hawaii at Manoa 2016.|
Includes bibliographical references.
|Appears in Collections:||
Ph.D. - Civil and Environmental Engineering|
Please email firstname.lastname@example.org if you need this content in ADA-compliant format.
Items in ScholarSpace are protected by copyright, with all rights reserved, unless otherwise indicated.