On Virtual Commissioning of Manufacturing Systems
: Proposals for a systematic VC simulation study methodology and a new simulation model building approach

  • Peter Hoffmann

    Student thesis: Doctoral Thesis

    Abstract

    The development of manufacturing systems is faced with progressively tightening time frames, along with growing requirements on planning quality and engineering accuracy. These demands result from significant cost constraints, shortening of product life-cycles, increasing number of product variants and economic needs for rapid time-to-market. Thus, an efficient production ramp-up including the commissioning as the crucial part, becomes more and more important for engineering companies to stay profitable. Virtual Commissioning (VC) is widely considered as promising method to address the challenges associated with real commissioning, but the simulation model building necessary for VC is affiliated with considerable effort and required expertise. VC of manufacturing systems has been a research topic in academia and industry for far more than a decade. Positive results are reported from
    large companies e.g. from the automotive industry, which are mostly utilising the complex and costly suites of tools in the context of the Digital Factory, rarely from SMEs. However, in particular also SMEs are forced to improve their engineering and commissioning processes, but suites of tools and methodologies used in large companies are not reasonably transferable to SMEs. Rationale for the rare use of VC, besides its general complexity, are a high modelling effort to build the necessary virtual plant models and a lack of availability of methodologies for systematic implementation and reasonable
    execution of VC.

    Thus, the main goal of this research is the development of a new systematic simulation study methodology as general guideline for planning, implementation and execution of VC. It is intended to be notably beneficial for engineers from SMEs, as helpful guideline for planning, implementation and execution of VC and to facilitate the substantially high modelling effort required for VC of manufacturing systems. Besides clarifying the requirements and specifying an environment for VC, the criteria to select an appropriate simulation tool have been established. The proposed modular, component based simulation model building has been split into specified procedures for “Low-level Component Modelling”, to be conducted for the components of the decomposed real manufacturing system, and subsequent “High-level Plant Modelling” of the virtual manufacturing system. The applicability of these new approaches has been validated by planning, implementing and conducting a VC for a trackbound transportation system with self-driving transport cars on passive tracks, which is the major subsystem of the manufacturing system used as test-bed at the UASA Hannover.

    As one main result, a novel workflow for Low-level Component Modelling has been proposed that aims for the gradual relocation of this modelling task as far as possible to the origin of components, in the end the component manufacturers should provide together with the deliverable components their mechatronic component models. This is related to a novel proposal for exchangeable mechatronic component models and an outlined possible implementation with AutomationML.
    Date of Award30 May 2017
    Original languageEnglish
    SupervisorGiuliano Premier (Supervisor), Reimar Schumann (Supervisor) & Talal Maksoud (Supervisor)

    Keywords

    • Manufacturing processes
    • Automation
    • Production management
    • Software-in-the-Loop
    • SIL
    • Hardware-in-the-Loop
    • HIL
    • 3D CAD Data Preparation

    Cite this

    '