Optimizing Sample Size for Concrete Strength, Uncertainty and Cost Reduction

For the experimental assessment of concrete material properties, estimating compressive strength is crucial for the calculation of structural safety and optimising testing campaigns. However, the optimal number of tests required to comply with a target reliability while balancing effort has rarely been analysed in the past. This work introduces an approach combining statistical modelling and uncertainty analysis to evaluate how sample size influences prediction accuracy and structural reliability, risks and testing costs. Based on core samples of concrete from an existing structure, prediction intervals are propagated into reliability calculations against a target level. A parametric study shows uncertainty decreases rapidly with an increasing number of tests, up to a threshold beyond which further gains are not cost-efficient. The procedure identifies the sample size that balances effort and uncertainty reduction, providing a first step towards a tool for planning and optimising testing campaigns.