The Direct Numerical Simulation (DNS) approach to solving the fundamental transport equations down to the smallest scales of motion is favourable should the requirement be a truly predictive solution of fluid dynamic problems, but the simulation run times are unacceptable for most practical industrial applications. Despite the steadily increasing computational capabilities, Reynolds Averaged Navier-Stokes (RANS) based frameworks remain the only commercially viable option. The sub models within RANS simplify the description of key physical phenomena and include several numerical constants. These so-called “tuning constants” introduce multivariable dependencies that are almost impossible to untangle with local sensitivity studies. This paper addresses the prevailing difficulties in setting up an adequate Diesel spray simulation which arise from the mentioned multi-variable interactions of these “tuning constants”, by applying a statistical approach named Design of Experiments (DoE). Often combined with an optimiser, DoE is commonly used to find an optimum set of engine parameters for set criteria at reduced experimental effort. In this case, the methodology is applied to find a set of “tuning constants” in the simulations which best match experimental data gathered by the Engine Combustion Network (ECN). Offering a particularly rigorous dataset, the ECN provides high quality measurements of important metrics like mass fraction, droplet velocity and temperature distributions, as well as liquid and vapour penetration. Multi-variable DoEs were run to achieve spray matches for various boundary conditions. Multiple combinations of the “tuning-constants” were found to provide a match for each condition. These matches were studied with the objective of understanding how the required “tuning-constants” for a match changes with the boundary conditions. Such boundary conditions include ambient temperatures and injection pressures and this paper shows how to predictively change the “tuning constants” for changes in these and other conditions. The approach can be extended to combustion and emission models.