Professor Emeritus, University of Florida, Gainesville, Fla
University of Florida, Gainesville, Fla
State Materials and Research Engineering, Gainesville, Fla
Pages: 2 Published: Jan 1983
After 2000 years or more of using asphalts for engineering purposes that involve their rheological properties, we are finally beginning to understand their behavior. It is our opinion that the mystery of deformation of bitumen or its composites or both under stress has to be no different from any other viscoelastic engineering material. The principal differences one would expect should be only in the magnitude of the individual viscoelastic components contributions to affecting the response of a material to which type of stress (shear, compression, etc.) that is applied.
The viscosities of asphalts are not a single point measurement as seems to be the general understanding of technologists. This is true only for true Newtonian flow. There is no such material in the real world because all materials have some elasticity (even water). Accordingly, we must consider how the viscosity varies with stress and time to explain its real world behavior. The elasticity at a given stress is an instantaneous effect (Hooke spring, 1678) coincidental with a delayed decaying strain (Kelvin spring and dashpot, circa 1875) starting at time zero at which time also creep flow begins as a Burns-Schweyer dashpot. The latter is a variable dashpot resistance that responds to the stress in a manner necessary to account for whatever non-Newtonian flow response function is applicable for the material (pseudoplasticity, thixotropy, etc.). Therein the model differs from the Burger's model (1935) that uses a Maxwell dashpot that does not account for stress or strain susceptibility. Some concepts of a creep curve often include a St. Venant body (circa 1840) that is viewed as two parallel bodies held together by friction that is in series with the Burger model to allow for yield stress. The present authors believe yield flow for gelatin (or glass) at any stress (low or high) is only a matter of time, maybe eons. Therefore, there is no such property other than that a very strong spring in either the Hooke element or the Kelvin component that at low stresses is not observed until sufficient time has elapsed. Time is always a parameter in viscoelastic analysis.
Failure in the model is caused by breaking the spring or excessive flow that pulls the pistons out of the dashpots.
The objectives of this paper are to propose some new thoughts on using more recent developments in asphalt flow technology and attempt to demonstrate these concepts in a general way to asphalt pavement behavior over a range of ambient temperatures.
Accordingly the paper consists of three parts: (a) a short summary of the basic rheology concepts for viscoelastic flow of asphalts at an elementary level; (b) a demonstration of experimental data as to how Part A applies to bitumen and mixes; and (c) an illustration of the significance of the application in service.
In Part A the discussion provides a background for a new physical model of the viscous and elastic flow components that are shown to demonstrate good fit to experimental data in Part B. Some of this material is taken from polymer technology since asphalt bitumen is really a thermoplastic material.
Part C will discuss how the application of the rheological data and the physical model can aid in understanding design of better field performance for road paving applications by using rheological data on the bitumen and mixes.
pavements, flexibility, asphalts, rheology, flexible pavement
Paper ID: STP31798S