TS-100
Modular Simulation of Marine Propulsion Systems
Using an Engineering Building Block Approach
Vassilios P. LAMBROPOULOS* and Nikolaos P. KYRTATOS**
ABSTRACT
This paper presents a method for modeling marine propulsion systems (including the hull dynamic behavior) using a scalable and extendable building block approach (were each building block models a physical subsystem of the installation) with a well defined interface between the modeled subsystems.
The modeling methodology presented breaks down the propulsion installation into small elementary subsystems which then are individually modeled isolated from the rest of the system. These elementary building blocks can then connected together through well defined interfaces in order to implement various installations.
The method proposed is suitable for the evaluation of the design's feasibility in the early stages, the fine tuning of the design and as an aid for taking operational decisions, by evaluating the performance of the propulsion system with economical and technical advantages. The method is equally suitable for higher level simulation languages and lower level programming tools as used in micro controllers.
The method has been successfully used for modeling a containership equipped with a large two stroke Diesel engine, thus demonstrating the advantages of the approach. The simulation results obtained gave good accuracy comparing with other simulation codes. Testing of the method accuracy also revealed the importance of subsystem isolation and definitive interfacing.
Key Words: Simulation, Modeling, Marine Propulsion, Modular, Method
1. INTRODUCTION
Today simulation pays a significant role in system design, because it provides insight and deeper understanding of the physical processes that are being modeled. As complex models are increasingly used in practice, modelers require ways of abstracting their models and having the ability to traverse levels of abstraction. The abstraction is an important design approach to break a system into hierarchical levels.
The tools used for simulation are generally divided in three categories. The programming languages (Fortran, C etc), the general simulation packages and the multidimensional analysis packages (like FEM). Each tool has specific advantages and disadvantages but the tool alone provides only the platform. It is the modeling method, which greatly determines the capabilities and limitations of the simulation. A careful investigation of the modeling approaches and techniques currently used, revealed that in general they do not provide adequate scalability and isolation characteristics required for the research projects undertaken in a modern marine engineering laboratory.
In this paper a method is presented which simplifies the modeling of marine propulsion systems (including the hull dynamic behavior) using a scalable and extendable building block approach (were each building block models a physical subsystem of the installation) with a well defined interface between the modeled subsystems.
For the modeling needs of this paper zero-dimensional models were used for the engine. Zero-dimensional models have several advantages and can be very useful even if adequate computational power is available. Generally a zero-dimensional model is considerably easier to build and it has fewer degrees of freedom to evaluate (less state variables and parameters), which makes it orders of magnitude lighter in required computing power. Also it requires less known data (or data needed to be manually entered) for the engine, which makes it easier to adapt for worn or failure operation (less data and fewer parameters, for example it is easier to simulate a blow-by valve than with fluid dynamic models).
2. CONCEPT
By breaking down the physical system (to be modeled) into smaller basic subsystems and then providing a model for each of the subsystems (called submodel), using a building block approach, we can greatly simplify the modeling process.
Since a physical system can be broken down in various ways (different boundaries for a subsystem can be chosen), a careful determination of the boundary of each subsystem is required in order to keep a minimum number of basic subsystems without loosing flexibility.
But the breakdown to smaller subsystems alone is not enough. A way for coupling these subsystems together is required in order to recreate the physical model (and/or others under investigation). The standardization of this enables fast model synthesis even for the most complex installations. Also the definitive boundary between the submodels allows the study to be focused to one or more of the subsystems by isolating them, thus enabling better investigation of their behavior.
