| Abstract: |
The flexible pavements primarily employ bituminous concrete (BC) as the surface course which field performance critically depends on the quantity of bitumen added and its proper blending with a well-graded mineral aggregate skeleton [1], [2]. This paper provides a systematic review and meta-analysis of previous experimental work on the parametric effect of binder content on the Marshall properties of dense-graded bituminous concrete. Using the PRISMA protocol [3], qualitative synthesis returned 41 studies and quantitative meta-analysis, 27 studies (yielding 74 data sets of Marshall). The synthesized findings validate that Marshall stability improves with bitumen content until an optimum is attained and decreases beyond that; the pooled random-effects optimum bitumen content (OBC) was estimated at 5.21% by mass of mix (95% confidence interval: 5.02–5.40%) with moderate heterogeneity (I² = 64%), in close agreement to the national specification region of 5.0–5.5% [1]. Flow values are monotonically increasing with bitumen content, unit weight is a maximum at or very near the optimum, and air voids fall to the optimum (or as close as possible). Optimum content and peak-stability are significantly modulated with the influencing factors of gradation fineness, filler type and compaction energy as shown by meta-regression. Sensitivity assessments in a critical appraisal highlighted source of between-study scatter as inconsistent compaction energy, non-uniform reporting of volumetric properties and variable conditioning regimes. Furthermore, the pooled data reveal that deviations of ±1% from the optimum decrease stability within 10–25%, which quantifies a practical penalty for poor bitumen control during production [2]. The review concludes with a recommendation for standardized Marshall testing and reporting, and the utilization of pooled optimum-content ranges during preliminary mix design. |
