Assessing low shear viscosity as the new bitumen Softening Point test S.E. Zoorob a,⇑ , J.P. Castro-Gomes b , L.A. Pereira Oliveira b a Engineering & Computing Faculty, Sir John Laing Bldg., Coventry University, Coventry CV1 5FB, UK b C-MADE, Centre of Materials and Building Technologies, Department of Civil Engineering and Architecture, University of Beira Interior, 6201-001 Covilhã, Portugal article info Article history: Received 4 May 2011 Received in revised form 13 July 2011 Accepted 18 July 2011 Available online 25 August 2011 Keywords: Bitumen viscosity Bitumen stiffness Softening Point test Zero shear viscosity Low shear viscosity abstract A large amount of research has been published over the last decade in response to the inadequacy of the traditional Softening Point test and the more recent SHRP G / /sin d criteria to accurately characterise the rutting behaviour of bitumens for use in road construction. This is particularly evident when considering polymer modified bitumens with a high proportion of delayed elasticity. In this paper, the concept of Zero Shear Viscosity (ZSV) is explored and a newly proposed test variant on ZSV referred to as the Low Shear Viscosity (LSV) test has been examined in some detail. The LSV proposed test protocol, which essentially consists of measuring complex viscosity using dynamic oscillatory loading at a pre-determined equi-vis- cous temperature, was found to be suitable for characterising the creep performance of two penetration grade bitumens, whereas for an SBS modified bitumen used in this investigation, the protocol was shown to be entirely inadequate. An alternative method of data interpretation is proposed in this paper based on measurement of the phase angle values at low shear rates. The phase angle, or damping factor, has been shown to be a better candidate for characterising the three bitumens assessed in this investigation and is worthy of further investigation. Ó 2011 Elsevier Ltd. All rights reserved. 1. Zero-shear-viscosity Zero-Shear-Viscosity (ZSV), a theoretical concept, is the viscosity measured in shear deformation, when the shear rate is approaching zero. In other words, it is a measure of the viscosity of a material, when a shear stress is acting on it at a shear rate of almost zero. At such low shear rates, the bitumens undergo deformation so slowly, that it can adapt continuously to maintain equilibrium, de- spite the total amount of shear being large. At these conditions and at a given temperature, the viscosity of a viscoelastic liquid has a limiting value, referred to as the ZSV and is denoted by g 0 . The ZSV is said to be an indicator of two rutting related binder charac- teristics, namely the stiffness of the binder, and the binder’s resis- tance to permanent deformation under long term loading [1]. The most obvious method of determining ZSV consists of a creep test where a constant stress (r 0 ) is applied to a sample and the resulting strain (c) is then measured as a function of loading time. For sufficiently long times, the deformation rate is expected to reach a constant value, corresponding to steady state flow and ZSV can be determined as the inverse of the slope of the compli- ance curve at long creep time [1]. dc dt ! 0 t !1 ) dJðtÞ dt   1 ! g 0 ð1Þ The time required to attain steady state flow depends on the type of bitumen being analysed and goes from a few minutes for a conven- tional bitumens to in excess of 12 h for some elastomer-rich phase polymer modified bitumens exhibiting high delayed elasticity [2,3]. Testing time is thus a major barrier to the routine adoption of this form of testing for bitumen characterisation. Researchers have at- tempted to bypass this problem by seeking relationships between steady state shear flow and dynamic rheology. The section below describes one example selected from an immense body of literature from the field of polymer science tackling this subject matter. 2. Relation between steady shear flow and dynamic rheology Polymer fluids can be characterised in steady shear flow by; g ¼ f 1 ð _ cÞ and r 11  r 22 ¼ f 2 ð _ cÞ, and in dynamic measurements by; |g / |= f 3 (x), where: g = viscosity in steady shear flow at shear rate _ c,(r 11  r 22 ) = the first normal stress difference and |g / | = the absolute value of the complex dynamic viscosity at frequency x [1]. However it has long been known that similarities exist be- tween steady-state, non-linear shear flow material functions and linear viscoelastic properties as characterised by dynamic mea- surements of polymer fluids; as reflected by the following relation- ship; lim_ c!0 gð _ cÞ¼ lim x!0 g 0 ðxÞ where g 0 represents the real component of the dynamic viscosity [1,4]. These relations were predicted by molecular and phenomeno- logical theories and have been found to be followed experimentally for many polymer solutions and melts. Deviations from these 0950-0618/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2011.07.037 ⇑ Corresponding author. Tel.: +44 2476 887710; fax: +44 2476 888296. E-mail address: salah.zoorob@googlemail.com (S.E. Zoorob). Construction and Building Materials 27 (2012) 357–367 Contents lists available at SciVerse ScienceDirect Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat