Modelling and Simulation of Propulsion Dynamics Due to Clutching
Hallvard ENGJA* and Oyvind BUNES**
The use of numerical models and simulation during the design of physical systems is becoming more and more important tools as modern computer capability increases rapidly. Numerical simulation support the designer in understanding his system better, and enables him to make modifications and test the response within seconds.
Thus, simulation is the natural link between design and operation.
In modelling and simulation of propulsion systems containing clutches, some challenging problems exist in that we have to deal with a discontinuity which may cause numerical stiffness problems and a variable system structure. This paper discuss the bond graph method as a modelling framework in order to discover causal difficulties and a type of modelling is presented which allows a constant structure for the overall system and minimises the numerical stiffness problem.
Key Words: Modelling, Bond Graph, Clutching, and Discontinuities.
Propulsor drives that use prime movers in combination may require a clutch to disconnect and reconnect main engines from the propulsor. This engagement of clutches may generate large dynamic loading in the torsional mode. Remember when you released the clutch pedal on the car to quickly and the engine stopped? The dynamic loads occurring due to clutching can produce damages in the propulsion system. To assure at the design stage, a performance with minimum transient loading during clutch operation, computer studies and computer simulations is the only realistic method of experimentation. Simulation is one of the most important and useful tools available to those responsible for the design and operation of systems. The component designer can experiment with his component in a laboratory, but for the system designer computer simulation is the only feasible method of experimentation. Therefore, simulation is an experimental and applied methodology, which can be used for the purpose to understand the behavior of the system or of evaluating various strategies for the operation of the system.
Computer simulation as a problem solving methodology consists of three important phases:
Modelling, which is by definition, the description of system behavior by means of suitably chosen mathematical relations or equations.
Numerical analysis and algorithm selection to solve the equations comprising the mathematical model.
Conducting experiments with a model for the purpose to either understand the system behavior or evaluating various design or operational strategies. This phase is what we often call simulation.
Thus, for modelling and simulation to be an effective problem solving methodology, it is required that the analyst has an insight into all three of the above-mentioned phases.
Only in this way can there be a degree of assurance that the simulation outputs constitute meaningful and useful solutions.
The idea of a model is thus a central idea involved in the study of the dynamics of real system simulation. To proceed from a schematic diagram to a reasonable model, a great deal of judgement is required. For simple systems, equation formulation according to well-established rules may pose no great difficulties. However, in many practical cases the analyst is required to use considerable ingenuity to organize the component models into a system model. This often proves to be a major impediment to those using simulations for system design. This is particularly true when dealing with clutches since the constitutive laws for such elements are difficult to describe usefully in a single input-output causal form. Part of the time the torque should be zero for any velocity and part of the time the velocity should be, zero for a given torque.
* Norwegian University of Science and Technology,
Department of Marine Engineering,
7491 Trondheim, NORWAY.
Fax: +47 73 59 59 83
MARINTEK, 7450 Trondheim, NORWAY.