Optimization of Mortar Compressive Strength of Orange Peel Ash - Eggshell Powder Cement Blends

Abstract

This study explores the effect of replacing eggshell powder (ESP) in a cement blend with 
orange peel ash (OPA) on mortar compressive strength. OPA was produced by calcining 
orange peels at 650°C for 2 hours, followed by sieving through a 90µm mesh. Portland 
limestone cement (CEM II 42.5R A-L) was blended with ESP and partially replaced with 
OPA at 0–6 wt.% in 2 wt.% increments. The mortar was prepared using a 1:3:5 ratio (water, 
binder, sand) and tested using 50 mm cubes. Chemical analysis revealed that OPA primarily 
contained lime, silica, and potassium oxide, with a combined SiO2+Al2O3+Fe2O3 content of 
19.77 wt.%, classifying it as non-pozzolanic per ASTM C618-19. ESP, on the other hand, 
acted as a filler with high lime content (54.09 wt.%) with a CaCO3 composition of 96.55 
wt.%. The effect of OPA and ESP on mortar strength was assessed based on blending ratio, 
replacement level, and curing age, with optimization performed using Response Surface 
Methodology (RSM). Central Composite Design (CCD) and Box-Behnken Design (BBD) were 
employed for predictive modeling and compared for accuracy. Results indicated that 
increasing curing age (7–60 days) and either blending ratio or replacement level (0.25–0.75, 
2–6 wt.%) led to improved mortar strength. Analysis of variance (ANOVA) confirmed that 
curing age had a significant effect, with high F-values in both CCD and BBD models. CCD 
provided the best predictive accuracy, with a predicted R² (R²pred) of 0.8464, making it 
the most suitable model for describing the mortar compressive strength of these ternary 
cement blends. Optimization revealed that the ideal conditions for CCD were 0.75 blending 
ratio, 2.0 wt.% replacement, and 60 days curing period, achieving a mortar compressive 
strength of 42.17 N/mm² with a desirability of 1.0. For BBD, the optimal conditions were a 
0.25 blending ratio, 2.0 wt.% replacement, and a 60-day curing period, yielding 43.86 
N/mm² with a desirability of 0.885. The interactive effects of blending ratio, replacement 
level, and curing age were also analyzed, confirming their influence on mortar strength.