The Taguchi optimisation method proved to be highly accurate and effective in predicting the tensile strength of the cow bone and cassava cortex hybrid composite. The study employed the Taguchi method using an L18 orthogonal array to systematically optimize three key factors:

• Reinforcement composition
• NaOH chemical treatment
• Particle size

🔹 Prediction Accuracy:

• The Taguchi model predicted the optimal tensile strength configuration (R4T4PS100), which consisted of: 20% cow bone + 20% cassava cortex reinforcement 4% NaOH treatment 100 µm particle size
• This same configuration also produced the highest experimental tensile strength of 91.80 MPa, indicating strong agreement between prediction and actual performance.

🔹 Regression and Error Analysis:

• A regression model was developed to quantify the relationship between input factors and tensile strength. The equation was: Tensile Strength (TS) = 87.04 – 1.615R + 0.408T – 0.00102PS
• The coefficient of determination (R²) for the model was 91.5%, indicating that over 91% of the variability in tensile strength could be explained by the model—a high level of predictive accuracy.
• The residual error, which is the difference between predicted and actual tensile strength, ranged from –7.9% to +11.25%, with most errors within ±5%, well below the commonly accepted threshold of 10% in materials engineering.

🔹 Validation with Residual Plots:

• Residual plots, including the normal probability plot, showed that residuals were normally distributed and symmetrically aligned, confirming that the model errors were random and not due to systematic bias.
• This validated the assumption of normality and homoscedasticity, which are crucial for the reliability of the regression model.

✅ Conclusion:

The Taguchi optimisation technique provided a robust, statistically reliable, and efficient predictive tool for identifying the optimal processing parameters of the hybrid composite. Its predictions closely matched the experimental results, proving its suitability for material design, testing, and development. The method's strength lies in its ability to optimize multiple variables simultaneously while minimizing the number of required experiments—making it a valuable tool for future composite development in automotive and other engineering applications.