Change Propagation and Conceptual Package in the Early Concept Development Phases of Heavy-Duty Vehicles

Becker, Johannes K

July 2013

Abstract

This thesis addresses the issue of change propagation in heavy-duty vehicle architectures as well as solutions for geometric conflicts in conceptual vehicle packages. Both subjects are approached in support of the early concept development phase of a new truck architecture. The objective of this thesis is to support the early concept phase of this new architecture development in two ways. First of all, changes introduced to the system can propagate to unintended areas and cause additional unanticipated expenses. Analyzing the intended changes made to the product architecture can help to avoid decisions that would lead to change multiplication in critical parts of the system and can help to identify parts of the architecture that need to be made more flexible by providing them with sufficient contingency margins. Secondly, vehicle concepts generated the concept development tool for commercial vehicles (NuKET) were previously optimized in terms of weight distribution and functional requirements, but not examined for geometric feasibility. Another objective is to fill this gap and allow for fast package visualization and collision detection from within NuKET. On the subject of change propagation, this thesis develops a change prediction method which is based on change-related data provided by a truck manufacturer. By application of this method, this thesis aims to answer questions about the most likely change initiators as well as the change propagation paths to be expected. The subject of conceptual package modeling is approached by development of a component placement tool, which allows for a quick assessment of the geometric feasibility of a vehicle concept’s package and informs about potential conflict sources. Both topics treated in this thesis, although different in tools and general approach, aim for the same goal: to support the early concept phase of heavy-duty truck architecture development. As a re- sult of this research, change propagation paths were identified on two levels of detail of the product structure using two separate modeling approaches within the change prediction method. Important main component groups, which spread changes to large numbers of other main component groups were identified and further analyzed by zooming in on the component group level of detail. Analyzed component groups could be classified into behavioral classes regarding change propagation (e.g. change multiplier, change absorber) in order to optimize the architecture with regard to change prop- agation and changeability. The component placement tool was developed as an intended future part of NuKET and complements its feature set by providing visualization and collision detection even be- fore a conceptual digital mock-up needs to be generated. Insights gained from component placement can then lead to development of necessary and useful further architecture standards and placement guidelines, which might in turn impact the change dependencies found via change propagation anal- ysis in a positive way.

Bibtex

@mastersthesis{becker2013change,
  title={Change Propagation and conceptual package in the early concept development phases of heavy-duty vehicles},
  author={Becker, Johannes K},
  year={2013},
  school={Technische Universität München},
  abstract={This thesis addresses the issue of change propagation in heavy-duty vehicle architectures as well as solutions for geometric conflicts in conceptual vehicle packages. Both subjects are approached in support of the early concept development phase of a new truck architecture.
  The objective of this thesis is to support the early concept phase of this new architecture development in two ways. First of all, changes introduced to the system can propagate to unintended areas and cause additional unanticipated expenses. Analyzing the intended changes made to the product architecture can help to avoid decisions that would lead to change multiplication in critical parts of the system and can help to identify parts of the architecture that need to be made more flexible by providing them with sufficient contingency margins. Secondly, vehicle concepts generated the concept development tool for commercial vehicles (NuKET) were previously optimized in terms of weight distribution and functional requirements, but not examined for geometric feasibility. Another objective is to fill this gap and allow for fast package visualization and collision detection from within NuKET.
  On the subject of change propagation, this thesis develops a change prediction method which is based on change-related data provided by a truck manufacturer. By application of this method, this thesis aims to answer questions about the most likely change initiators as well as the change propagation paths to be expected. The subject of conceptual package modeling is approached by development of a component placement tool, which allows for a quick assessment of the geometric feasibility of a vehicle concept’s package and informs about potential conflict sources.
  Both topics treated in this thesis, although different in tools and general approach, aim for the same goal: to support the early concept phase of heavy-duty truck architecture development. As a re- sult of this research, change propagation paths were identified on two levels of detail of the product structure using two separate modeling approaches within the change prediction method. Important main component groups, which spread changes to large numbers of other main component groups were identified and further analyzed by zooming in on the component group level of detail. Analyzed component groups could be classified into behavioral classes regarding change propagation (e.g. change multiplier, change absorber) in order to optimize the architecture with regard to change prop- agation and changeability. The component placement tool was developed as an intended future part of NuKET and complements its feature set by providing visualization and collision detection even be- fore a conceptual digital mock-up needs to be generated. Insights gained from component placement can then lead to development of necessary and useful further architecture standards and placement guidelines, which might in turn impact the change dependencies found via change propagation anal- ysis in a positive way.}
}