Characterization of asphalt concrete linear viscoelastic behavior utilizing Havriliak–Negami complex modulus model. (30th November 2015)

• Record Type: Journal Article
• Authors: Sun, Yiren; Chen, Jingyun; Huang, Baoshan
• Is Part Of: Construction & building materials. Volume 99(2015)
• Page Start: 226
• Page End: 234
• Publication Date: 2015-11-30
• Subjects: Asphalt concrete -- Master curve -- Continuous spectrum -- Discrete spectrum -- Prony series -- Havriliak–Negami model; Building materials -- Periodicals; 624.18
• DOI: 10.1016/j.conbuildmat.2015.09.016
• Languages: English
• ISSNs: 0950-0618

Abstract

Highlights: Both continuous relaxation and retardation spectra were analytically derived from a complex modulus model. A numerical interconversion method was proposed to determine one continuous spectrum from the other. High-quality discrete relaxation and retardation spectra were directly determined from the continuous ones. A two-step method for determining reduced master curves was suggested. Abstract: This study proposed a new procedure for characterizing linear viscoelastic (LVE) behavior of asphalt concrete using the Havriliak–Negami (HN) complex modulus model and corresponding continuous relaxation spectrum model. To extend its application to modeling retardation behavior of asphalt concrete, the continuous retardation spectrum was analytically derived from the HN model by performing an inverse Fourier–Laplace transform followed by a variable substitution. The two continuous spectra of the HN model allow the accurate construction of all the modulus and compliance function master curves in time and frequency domains as well as the efficient determination of the high-quality discrete relaxation and retardation spectra. The advantages of the proposed approach over existing ones were demonstrated through two different complex modulus test data sets. Also, a simple and practical numerical interconversion method was presented based on the LVE relations between the continuous relaxation and retardation spectra, which can effectively compute one continuous spectrum from the other for any characterization model. Further, for practical considerations, a two-step method for determining reduced master curves with fewer discrete spectrum lines was suggested.