Development of a simulation-based method for the assessment of decomposition strategies of building structures into individual modules from a production engineering point of view in the early stage of module or building set development

Applicants

Prof. Dr.-Ing. habil. Thorsten Schmidt Technische Universität Dresden, Professur für Technische Logistik

Dr.-Ing. Mathias Kühn Technische Universität Dresden, Professur für Technische Logistik

Scientific staff

Dipl.-Wi.-Ing. Vincent Betker Technische Universität Dresden, Professur für Technische Logistik

Project description

For the structural design of high performance concrete structures and disassembling into individual modules (Module = single component with physical connections ), there are a large number of equivalent variants (Options with regard to structure: e.g. shape, reinforcement proportion, layer thickness etc.). From a manufacturing point of view, the modules can be produced with different manufacturing processes with a similar level of quality. In order to manufacture the modules with minimized costs, there are numerous possibilities for production organization. From a constructional and production-technical point of view, there are therefore thus numerous variants to realize a product, e.g. a building. (see Figure 1).

Figure 1: Complexity of the modularization of high-performance concrete structures

A preselection of disassembly strategies or derivation of constraints must be made in order to estimate whether a module is suitable for the module kits. For this purpose, it is necessary to valuate individual module variants from the manufacturing point of view. Currently, there are no options for the efficient assessment of segmentation variants and their forecast production units with regard to the target of a high output at low unit costs. Therefore, the research objective is a method for holistic evaluation and pre-selection of decomposition strategies and technologies for high-performance concrete structures in the early stages of the development of modular systems.
Due to the high number of variants and complex interdependencies between module, manufacturing process and workflow organization, as well as the influence of uncertain parameters, simulation-based optimization is a suitable method to choose the best parameters. First, technology and process relevant production system components must be defined. Subsequently, any production system variants are configured as a simulation model by means of parameter-based model generation. In order to reduce the simulation effort, simulation data are used by machine learning models, so that suggestions for the basic module design and unit cost prognosis are possible for similar building structures (see Figure 2). Furthermore, the aim is to determine restrictions on the module dismantling process based on the manufacturing process and workflow organization, in order to achieve a holistic optimization and to derive generally valid design guidelines supported by machine learning methods.

Poster on the project contents

Figure 2: Linking simulation-based optimization with machine learning models

Publications

[1] Müller, M.; Völker, M.; Schmidt, T.:
Entwicklung einer simulationsbasierten Methode zur Bewertung von Zerlegungsstrategien von Baustrukturen in einzelne Module aus produktionstechnischer Sicht in der frühen Phase der Modul- bzw. Baukastenentwicklung.
In: BetonWerk International Nr. 6, 2020, S. 26-27
Link zum Artikel

Supervised theses

2021

[3] Tang, M.
Multikriterielle Optimierungsverfahren für die automatische Modellgenerierung
Diplomarbeit, Technische Universität Dresden, Professur für Technische Logistik, Betreuer: Marie Klitzke, M. Sc. , Madlin Weise, M. Sc.

[2] Mu, Y.
Entwicklung eines Kostenmodells zur simulationsbasierten Bewertung von Produktionssystemvarianten zur Herstellung von Betonmodulen
Diplomarbeit, Technische Universität Dresden, Professur für Technische Logistik, Betreuer: Marie Klitzke, M. Sc. , Madlin Weise, M. Sc.

[1] Long, Z.
Auswahl und Erprobung von Verfahren zur multikriteriellen Optimierung von Produktionssystem- sowie Beton-Modulvarianten
Diplomarbeit, Technische Universität Dresden, Professur für Technische Logistik, Betreuer: Marie Klitzke, M. Sc. , Madlin Weise, M. Sc.