TU Darmstadt / ULB / TUbiblio

Feature Recognition and Parameterization Methods for Algorithm-based Product Development Process

Weber Martins, Thiago and Anderl, Reiner (2017):
Feature Recognition and Parameterization Methods for Algorithm-based Product Development Process.
In: 37th Computers and Information in Engineering Conference, In: International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, Cleveland, OH, 6-9 August 2017, [Conference or Workshop Item]

Abstract

The algorithm-based product development process applies mathematical optimization tools in the conceptual steps of the product development process. It relies on formalized data such as initial loads and boundary conditions to find the best product solution for optimized bifurcated sheet metal parts. Previous research efforts focused on the automation of CAD modeling steps. Current algorithms are able to generate the CAD models of optimized bifurcated sheet metal products automatically, however, they are rough with low-level of detail and abstraction. Consequently, CAD models are embodied and detailed manually in a partly iterative and time-consuming process to include parameters, constraints and design features. Hence, this paper introduces feature recognition and parametrization methods for the algorithm-based product development of bifurcated sheet metal products. It proposes the inclusion of a pre-processor to analyze the solution graph resulted from topology optimization before the generation of CAD models. Algorithms that derive the geometric shape from the solution graph by recognizing features as well as assigning parameters are introduced. Then, feature-based CAD models of bifurcated sheet metal products are automatically generated. The proposed methods and algorithms are implemented with Python and validated with a use-case. Benefits and limitations of the proposed methods are discussed.

Item Type: Conference or Workshop Item
Erschienen: 2017
Creators: Weber Martins, Thiago and Anderl, Reiner
Title: Feature Recognition and Parameterization Methods for Algorithm-based Product Development Process
Language: English
Abstract:

The algorithm-based product development process applies mathematical optimization tools in the conceptual steps of the product development process. It relies on formalized data such as initial loads and boundary conditions to find the best product solution for optimized bifurcated sheet metal parts. Previous research efforts focused on the automation of CAD modeling steps. Current algorithms are able to generate the CAD models of optimized bifurcated sheet metal products automatically, however, they are rough with low-level of detail and abstraction. Consequently, CAD models are embodied and detailed manually in a partly iterative and time-consuming process to include parameters, constraints and design features. Hence, this paper introduces feature recognition and parametrization methods for the algorithm-based product development of bifurcated sheet metal products. It proposes the inclusion of a pre-processor to analyze the solution graph resulted from topology optimization before the generation of CAD models. Algorithms that derive the geometric shape from the solution graph by recognizing features as well as assigning parameters are introduced. Then, feature-based CAD models of bifurcated sheet metal products are automatically generated. The proposed methods and algorithms are implemented with Python and validated with a use-case. Benefits and limitations of the proposed methods are discussed.

Journal or Publication Title: International Design Engineering Technical Conferences and Computers and Information in Engineering Conference
Title of Book: 37th Computers and Information in Engineering Conference
Volume: Volume 1B
Divisions: 16 Department of Mechanical Engineering > Department of Computer Integrated Design (DiK)
DFG-Collaborative Research Centres (incl. Transregio) > Collaborative Research Centres > CRC 666: Integral Sheet Metal Design with Higher Order Bifurcations
DFG-Collaborative Research Centres (incl. Transregio) > Collaborative Research Centres
16 Department of Mechanical Engineering
DFG-Collaborative Research Centres (incl. Transregio)
Event Title: International Design Engineering Technical Conferences and Computers and Information in Engineering Conference
Event Location: Cleveland, OH
Event Dates: 6-9 August 2017
Date Deposited: 09 Oct 2017 14:05
Alternative Abstract:
Alternative abstract Language
Der algorithmen-basierte Produktentwicklungsprozess wendet mathematische Optimierungswerkzeuge in den konzeptionellen Schritten des Produktentwicklungsprozesses an. Es basiert auf formalisierten Daten wie Anfangslasten und Randbedingungen, um die beste Produktlösung für optimierte verzwigte Blechteile zu finden. Frühere Forschungsergebnisse konzentrierten sich auf die Automatisierung von CAD-Modellierungsschritten. Derzeitige Algorithmen sind in der Lage, die CAD-Modelle von optimierten verzweigte Blechprodukten automatisch zu generieren, sind jedoch grob und mit geringem Detailgrad und Abstraktion. Folglich werden CAD-Modelle in einem teilweise iterativen und zeitaufwendigen Prozess ergänzt und detailliert, um Parameter, Fertigungsrestriktionen und Konstruktionsmerkmale einzuschließen. Daher werden in diesem Beitrag Funktionen zur Erkennung und Parametrisierung von Merkmalen für die algorithmen-basierte Produktentwicklung von verzweigten Blechprodukten vorgestellt. Es schlägt die Einbeziehung eines Vorprozessors vor, um den Lösungsgraphen zu analysieren, der sich aus der Topologieoptimierung vor der Generierung von CAD-Modellen ergibt. Es werden Algorithmen eingeführt, die die geometrische Form aus dem Lösungsgraphen durch Erkennen von Features sowie die Zuweisung von Parametern ableiten. Dann werden feature-basierte CAD-Modelle von verzweigten Blechprodukten automatisch generiert. Die vorgeschlagenen Methoden und Algorithmen werden mit Python implementiert und mit einem Use-Case validiert. Vorteile und Grenzen der vorgeschlagenen Methoden werden diskutiert.German
Export:

Optionen (nur für Redakteure)

View Item View Item