<p>Mesh size plays a critical role in finite element analysis (FEA), often introducing artificial sensitivity in failure predictions when element dimensions vary. Traditional mesh regularization techniques eliminate mesh dependency through adaptive meshing, smoothing operations, or energy minimization, but these methods are computationally expensive. An alternative approach involves scaling material failure parameters to achieve mesh-independent behavior. However, conventional scaling methods typically determine scaling factors through a trial-and-error process, which is time-consuming and inefficient. This study presents a fundamentally different approach: a predictive material property scaling framework that mathematically determines scaling factors based on material failure criteria and mesh size, eliminating the need for iterative calibration. Four mathematical models were developed to predict scaling factors for the Johnson-Cook material model, specifically applied to Ti-6Al-4&#xa0;V (TC4) titanium alloy. The approach utilizes scaling factors derived from tensile test data, which are then applied to impact simulations to improve both accuracy and computational efficiency. Simulations were performed with 11 different mesh sizes using unscaled models, experimentally determined scaling factors, and mathematically predicted scaling factors. Compared to unscaled models, incorporating the predictive scaling factors reduced residual velocity deviation from 17.81% to 1.67% and energy absorption deviation from 48.40% to 8.10% across all mesh sizes. These results demonstrate the effectiveness of the proposed predictive scaling approach in achieving mesh-independent behavior while requiring significantly less computational effort than traditional regularization methods. Overall, the study introduces a systematic and generalizable framework for predicting material-model-based scaling factors, offering a practical solution to improve simulation accuracy and reduce computational cost across a range of FEA problems.</p>