Hilding Elmqvist
Dassault Systèmes, Lund, Sweden
Sven Erik Matsson
Dassault Systèmes, Lund, Sweden
Christophe Chapuis
Dassault Systèmes, Velizy Villacoublay, France
Ladda ner artikel
http://dx.doi.org/10.3384/ecp09430113Ingår i: Proceedings of the 7th International Modelica Conference; Como; Italy; 20-22 September 2009
Linköping Electronic Conference Proceedings 43:63, s. 551-560
Publicerad: 2009-12-29
ISBN: 978-91-7393-513-5
ISSN: 1650-3686 (tryckt), 1650-3740 (online)
Traditionally; multibody systems have been defined in Modelica by connecting bodies and joints in a model diagram. Additionally the user must enter values for parameters defining masses; inertias and three dimensional vectors of positions and orientations. More convenient definition of multibody systems can be made using a 3D editor available; for example; in CATIA from Dassault Systèmes with immediate 3D viewing.
A tool has been developed that translates a CATIA model to Modelica by traversing the internal CATIA structure to get information about parts and joints and how they are related. This information is then used to generate a corresponding Modelica model. The traversal provides information about the reference coordinate system; the center of mass in the local coordinate system; the mass; the inertia; the shape and color of the body exported in VRML format for animation purposes and the icon exported as a PNG file to be used in the Modelica diagrams.
The Modelica diagram layout is automatically generated and is based on the spanning tree structure of the mechanism. Models obtained in this way often contain redundant constraints. A new method has been developed for Dymola to facilitate simulation of such models; i.e. the model reduction is performed automatically.
An important property of the translated model is the possibility to use Modelica extends (inheritance) for adding controllers and other features of the model for dynamic simulation. For instance; the engine model can be extended by introducing models of the gas forces of the combustion acting on the cylindrical joints of the pistons. In that way; the translated model is separated and can be changed independently of the added models.
Bowles P.; Tiller M.; Elmqvist H.; Brûck D.; Mattsson S.E.; Möller A.; Olsson H.; Otter M (2000).: Feasibility of Detailed Vehicle Modeling. SAE World Congress 2000.
Dymola (2009). Dymola Version 7.3. Dassault Systèmes; Lund; Sweden (Dynasim). Homepage: www.dymola.com.
Engelson V. (2000): Tools for Design; Interactive Simulation; and Visualization of Object-Oriented Models in Scientific Computing. Department of Computer and Information Science; Linköping University; Sweden.
Mattsson S.E.; Elmqvist H.; and Olsson H. (2000): Dynamic Selection of States in Dymola. Proceedings of Modelica Workshop 2000; pp. 61-67. Download: http://www.modelica.org/events/workshop2000/proceedings/Mattsson.pdf
Otter M.; Elmqvist H.; Mattsson S.E. (2003): The New Modelica MultiBody Library. Proceedings of Modelica’2003; ed. P. Fritzson; pp. 311-330. Download: http://www.modelica.org/events/Conference2003/papers/h37_Otter_multibody.pdf
Otter M.; Elmqvist H.; Mattsson S.E. (2007): Multidomain Modeling with Modelica. Handbook of Dynamic System Modeling; Chapter 36; Ed. P. Fishwick; Taylor & Francis Group LCC .
