This paper presents a novel and efficient method for generating three-dimensional (3D) mesoscale structures of concrete using the discrete element method (DEM). The proposed approach enables the flexible and precise simulation of various aggregate shapes and volume fractions, which is crucial for sensitivity analysis and computational studies. Unlike traditional digitalization methods, this technique focuses on creating adjustable synthetic models to investigate the impact of different parameters on the mechanical properties of concrete. This technique demonstrates significant advantages in generating complex concave aggregates and high aggregate volume fraction models while allowing for flexible control of particle spacing, thus enhancing computational efficiency and model accuracy. Numerical simulations using the proposed method show excellent agreement with laboratory experimental results, validating its reliability. This method not only facilitates deeper sensitivity analysis but also aids in optimizing concrete designs and applications by providing insights into the effects of various parameters on concrete performance.