Response surface methodology for optimization of stabilizer dosage rates of marginal sand stabilized with Sludge Ash and fiber based on UCS performances

Gullu H., Fedakar H. İ.

KSCE JOURNAL OF CIVIL ENGINEERING, vol.21, no.5, pp.1717-1727, 2017 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 21 Issue: 5
  • Publication Date: 2017
  • Doi Number: 10.1007/s12205-016-0724-x
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.1717-1727
  • Keywords: response surface methodology, optimization, sludge ash, fiber, sand, unconfined compressive strength, stabilization, UNCONFINED COMPRESSIVE STRENGTH, REINFORCED CEMENTED SAND, FLY-ASH, POLYPROPYLENE FIBER, PARTICLE-SIZE, CLAYEY SOIL, BOTTOM ASH, SOFT SOIL, BEHAVIOR, LIME
  • Abdullah Gül University Affiliated: Yes


Optimization of the stabilization materials in terms of their dosage rates specifically used for stabilization applications have become a great interest by the experimenters due to the concerns of strength performance, time and economy for the construction projects. Using a relatively recent optimization technique, Response Surface Methodology (RSM), this paper is mainly focused on investigation of the optimum amounts of stabilizers (Sewage Sludge Ash (SSA) and Polypropylene Fiber (PF)) and Curing Time (CT) that yield to maximum unconfined compressive strength (UCS) for stabilization of a marginal sand (poorly-graded sand). For this purpose, an experimental study has been carried out conducting UCS tests, where the stabilizer proportions are 0-30% for SSA and 0-1% for PF, by total dry weight of sand+SSA. Also, the curing times considered prior to testing are 0, 7 and 14 days. All UCS tests have been performed following the experimental program by central composite design that used the ranges of stabilizer proportions and curing times. On the basis of experimental data, a full quadratic model with natural log transformation and backward analysis has been built through RSM considering the factors of SSA, PF, CT, and the response of UCS. The results indicate that the mathematical model built in this study is statistically significant (p ae 0.05) through the analysis of variance (ANOVA), thus it is applicable for optimization process. The findings from the optimization effort demonstrate that the most potential values for SSA, PF and CT are 19.95%, 0.57% and 12.15-day, respectively. The proposed values could be beneficial for the experimenters in practice specifically for preliminary evaluations prior testings of stabilizations.