Page 61 - MODELING OF ASPHALT CONCRETE
P. 61

Modeling of Asphalt Binder Rheology and Its Application to Modified Binders      39


                         4.  Binders are homogeneous and isotropic (no sample geometry or particulate
                           additives effects).
                         5.  Similar time-temperature equivalency for all binders (one shift is used).
                         6.  Binders are not thixotropic (no effect of mechanical working).
                         7.  Stability of asphalts is affected mainly by oxidation.
                       The essence of the above assumptions is that asphalt binders are simple systems
                    that can be characterized using linear viscoelasticity and simple geometry within which
                    stress and strain fields are simple to calculate. Except for number 7, the assumptions in
                    the specification are related to rheological and thermorheological behavior. Assumption
                    number 7 may not be very important for neat asphalt, but it is critical for asphalts
                    modified with additives.
                       In a recent study conducted as part of the NCHRP 9-10 project, each one of the listed
                    assumption was challenged and data were collected to prove that many modified
                    binders used in practice do not satisfy these assumptions. There were, however,
                    important reasons why researchers in SHRP had considered these assumptions to be
                    reasonable among which the two main are: (1) Most asphalts tested during the SHRP
                    showed a relatively wide linear range. (2) Asphalt pavement structural design should
                    be selected such that materials will be exposed to small stresses and strains. Therefore,
                    the specification testing was designed to measure behavior within the linear range
                    expecting that asphalts in the field will be mainly performing within this region.
                       Some of the important behaviors that will need to be characterized were, however,
                    not addressed due to need for simplification. Of particular importance are the possible
                    thixotropic effect and the dissimilarity of the effect of repeated loading.
                       Figure 2-11 is an example of possible effects of repeated loading at two different
                                                                                            ∗
                    strain levels. It is clear that some modified asphalts can show significant reductions in G





























                    FIGURE 2-11  Thixotropic behavior of modifi ed binder at 10 percent strain.
   56   57   58   59   60   61   62   63   64   65   66