Modelling, simulation and experimental investigation of a rammer compactor machine

Abstract: This licentiate thesis considers modelling, simulation and experimental investigation of a rammer compactor machine. The purpose is to develop an efficient and verified method for simulation of rammer compactor machines that can be used in the product development process. The experience gained through this work is also intended to be useful for studying other types of dynamic compactor machines. Rammer compactor machines perform impact soil compaction, which is very efficient compared to static compaction. They are often used in places where a high degree of compaction is needed and the space for operation is limited. The complexity of this machine type makes design optimisation through traditional prototype testing impractical. This has pointed to the need for a theoretical model and simulation procedure for prediction of the dynamic behaviour of the machine. To be useful for optimisation as design parameters are changed during product development the theoretical model and simulation procedure must be verified. By concurrently working with theoretical modelling, simulations, experimental verifications, and optimisation an efficient analysis support for product development is achieved. This co-ordination works both ways in an iterative manner. Experimental investigations are used to verify theoretical models and simulations. Theoretical models and simulations are used to design good experiments. This Complete Approach concept makes better decisions possible earlier in the development process, resulting in decreased time-to-market and improved quality. In this thesis, the Complete Approach concept is applied on a rammer soil compactor machine. An introductory iteration is described. The good agreement between theoretical and experimental results indicates that the theoretical model and simulation procedure should be useful for introductory optimisation studies. Reasons for the discrepancy are discussed and suggestions for improvements of both the theoretical model and the experimental set-up in coming iterations are given.