Design of Experiment Based on Full Factorial CFD Spray Modelling

Sammanfattning: Design of experiment is a powerful tool that can reduce the amount of physical testing needed to save resources, time and money. A CFD-setup was constructed in OpenFOAM-8, simulating injection of liquid propane with a standard multi-hole injector. The model was then validated against experimental results. Two design of experiment models were then created using Rstudio. The two models had four specified input parameters that were present in the CFD-model. The value of each parameter was then changed according to the respective DoE model and simulated. The two models were then evaluated and compared to each other. This was done by looking at two response variables and conducting ANOVA tests. Results showed some correlation between the two models. However significance factors and interactions differed to some degree between the two. A large effort was also put into comparing the models and evaluating tools for this purpose. This thesis lays a ground for future work developing a larger CFD-model, having a larger amount of changeable parameters. Furthermore it lays a ground for developing and optimising a DoE model as well as focusing on more tools for evaluating the acquired responses.