TY - JOUR

T1 - Mathematical prediction of the compressive strength of bacterial concrete using gene expression programming

AU - Algaifi, Hassan Amer

AU - Alqarni, Ali S.

AU - Alyousef, Rayed

AU - Bakar, Suhaimi Abu

AU - Ibrahim, M. H.Wan

AU - Shahidan, Shahiron

AU - Ibrahim, Mohammed

AU - Salami, Babatunde Abiodun

N1 - Publisher Copyright:
© 2021 THE AUTHORS

PY - 2021/11/20

Y1 - 2021/11/20

N2 - The impact of microbial calcium carbonate on concrete strength has been extensively evaluated in the literature. However, there is no predicted equation for the compressive strength of concrete incorporating ureolytic bacteria. Therefore, in the present study, 69 experimental tests were taken into account to introduce a new predicted mathematical formula for compressive strength of bacterial concrete with different concentrations of calcium nitrate tetrahydrate, urea, yeast extract, bacterial cells and time using Gene Expression Programming (GEP) modelling. Based on the results, statistical indicators (MAE, RAE, RMSE, RRSE, R and R2) proved the capability of the GEP 2 model to predict compressive strength in which minimum error and high correlation were achieved. Moreover, both predicted and actual results indicated that compressive strength decreased with the increase in nutrient concentration. In contrast, the compressive strength increased with increased bacterial cells concentration. It could be concluded that GEP2 were found to be reliable and accurate compared to that of the experimental results.

AB - The impact of microbial calcium carbonate on concrete strength has been extensively evaluated in the literature. However, there is no predicted equation for the compressive strength of concrete incorporating ureolytic bacteria. Therefore, in the present study, 69 experimental tests were taken into account to introduce a new predicted mathematical formula for compressive strength of bacterial concrete with different concentrations of calcium nitrate tetrahydrate, urea, yeast extract, bacterial cells and time using Gene Expression Programming (GEP) modelling. Based on the results, statistical indicators (MAE, RAE, RMSE, RRSE, R and R2) proved the capability of the GEP 2 model to predict compressive strength in which minimum error and high correlation were achieved. Moreover, both predicted and actual results indicated that compressive strength decreased with the increase in nutrient concentration. In contrast, the compressive strength increased with increased bacterial cells concentration. It could be concluded that GEP2 were found to be reliable and accurate compared to that of the experimental results.

KW - Bacterial concrete

KW - Bio-inspired self-healing

KW - Compressive strength prediction

KW - Gene expression programming modelling

KW - Microbial calcium carbonate

UR - http://www.scopus.com/inward/record.url?scp=85105319652&partnerID=8YFLogxK

U2 - 10.1016/j.asej.2021.04.008

DO - 10.1016/j.asej.2021.04.008

M3 - Article

AN - SCOPUS:85105319652

SN - 2090-4479

VL - 12

SP - 3629

EP - 3639

JO - Ain Shams Engineering Journal

JF - Ain Shams Engineering Journal

IS - 4

ER -