PROJECTS: Simulation of Energy Demand of Additive ManufacturingLi Yi Development of a coupled simulation model to solve the equations of motion of an air bearing spindle’s rotorAndreas Lange Molecular Simulation of Nanoscopic Contact ProcessesSimon Stephan Phase field simulations for fatigue failure prediction in processingChristoph Schreiber Physical Modeling of grinding forcesPraveen Sridhar Applying Human Centered Design to Complex Augmented Reality Environments
Carol Ximena Naranjo Valero Physical attributes in material flow modelingAnna-Pia Lohfink Dynamics model, analysis, and experiment design of inertia impact router robotic end-effectorThomas Kuo Extension of a white light interferometer by an optical coherence interferometer and applications for layer thickness determination and porous structure measurementAndrej Keksel Physical Modeling of Material Flows in Cyber-Physical Production SystemsMoritz Friedrich Glatt NSF/DFG Collaboration to Understand the Prime Factors Driving Distortion in Milled Aluminum Workpieces
Daniel WeberVisualization and Analytics for Supercomputing Performance DataAlfredo Giménez, PhD A particle finite element method for cutting of solidsXialong Ye Molecular Dynamics Modeling of Machining Processes of Coposite MaterialsVardanyan Vardan Optimization of the cooling channel outlet conditions of internally cooled drillsDaniel Müller Ultrasonic air bearing spindle for micro cuttingSebastian Greco Sustainable Grinding Machine Tool DesignIan Garretson Adaptive Surface Reconstruction for 3D CT-Data based on Geometric ModellingDennis Mosbach Algorithmic Curvature Theories in Image ProcessingMarkus Kronenberger Similarity in Visualization und Visual AnalyticsPatrick Rüdiger Feature Detection and Tracking: A Topological Visual Analytics ApproachJonas Lukasczyk Interactive Realistic Rendering on Massively Parallel Computer SystemsValentin Fuetterling Experimental investigations and simulation of micro grinding process
Dr. Dinesh Setti User-centered Approaches for Featur Extraction in CT Scanned DataDr. Christina Gillmann On the Design of Particle Functionalized FluidsZeyad Zaky, PhD Data-driven model for energy prediction in precision manufacturingRaunak Bhinge, PhD A Data-Driven Approach to Analyzing Machining Energies of Various MaterialsMaxwell Micali, PhD Numerical Modeling of Laser Additive Manufacturing ProcessesMarc Russell STAR: A Simulation Tool for Automation and RoboticsKristopher Wehage, PhD Surface Integrity & Process Optimization during Finishing OperationsJayanti Das, PhD Analysis and Visualization Flow Networks Using a Flow Topology GraphGarrett Aldrich Machining Distortion due to Residual Stresses in Quenched AluminumDestiny R. Garcia Physical Modeling of Time and Motion Estimation in Manual AssemblyChao Wang, PhD Moment Invariants in Flow VisualizationDr. Roxana Bujack Physical attributes in material flow modelingDr. Georg Kasakow Virtual test field for sustainability assessment of cybertronic production systemsDr. Rebecca Ilsen Planning and controlling of multiple, parallel engineering changes in manufacturing systemsDr. Daniel Cichos Influence of life cycle oriented services on the energy efficiency of machine toolsDr. Gülsüm Mert Interactive Specification and Visualization of Features to Support Factory PlanningDr. Diana Fernandez-Prieto Virtual measuring instrumentsDr. François Torner Ensemble Visualization and Analysis of Machine ToolsMathias Hummel A particle finite element method for cutting of solidsDr. Matthias Sabel Dynamic phase field simulations for failure prediction in processingDr. Alexander Schlüter Turning of Al-MMCs – 3D FE SimulationJosé Leonardo York Duran Molecular Dynamics Modeling of Machining Processes of Composite MaterialsDr. Zhibo Zhang Simulation of machining with cutting liquids based on atomistic force fieldsMartin Lautenschläger Computational Steering Technologies based on Parameter Space ExplorationDr. Tobias Post Visualization and interaction techniques scaling on process and environment levelDr. Franca Alexandra Rupprecht Description of the core research idea and the IRTG’s resulting main focus
The use of computer-based virtual models and simulations in manufacturing planning has become common engineering practice. However, these models are often purely geometrical such as in a material flow simulation or in a virtual reality model of a factory during the planning phase. Simulations on the manufacturing process level already include some real physical phenomena. Among such simulations are finite element (FE) methods, multibody (MB) simulations and molecular dynamics (MD) methods, which are used to model metal cutting processes.
The central research idea of the IRTG is to include physical properties and simulations based on physical models in a multi-level and multi-scale virtual manufacturing system. Developing the techniques to simulate, visualize, and analyze manufacturing processes on different levels and scales will form the core of the research program. In order to achieve the research goal of this IRTG, physical models of manufacturing processes, machine tools, and measurement machines, as well as advanced simulation methods and new principles to visualize the results are needed. Furthermore, new methods and algorithms will be developed to realize an interaction with the physical models in real-time scenarios. The research will be complemented by methods to evaluate and assess the simulation tools from the virtual environment.
The outcome of the research activities in this IRTG has the potential to change the way in which manufacturing systems are planned, operated, and optimized. Since the approach will be oriented on very basic principals and methods, the results can be applied to many different application fields such as mechanical engineering, automotive engineering, aerospace engineering, or medical technology.
Research program structure
Manufacturing systems are complex structures that produce a single component, module, or a complete product as a result of highly correlated processes. The individual processes – for example milling, grinding, or electro-discharge machining (EDM) – with their interdependencies and process parameters are combined to create so-called manufacturing process chains. A complete factory – or manufacturing system – typically consists of several manufacturing process chains. Virtual models and simulations are an adequate means to better understand the processes and correlations within the manufacturing system.
Manufacturing systems are commonly structured into different levels of details in order to allow simulations on different scales. The research program will address manufacturing modeling and simulation on the factory level, the machine level, and the process level. It will include the important issue of interaction between the levels through several level bridging projects (see figure 1).
On the factory level, physical features and material properties will be included in the material and information flow to describe the operating sequence of the manufacturing process chain accurately. On the machine level, machine tools and their components as well as measuring instruments will be specified as virtual physical models. Within the process level, various manufacturing technologies will be investigated. Their detailed characteristics will be described by physically-based simulations. To attain a comprehensive view of the manufacturing system, the correlations and interdependencies between the elements on each level as well as between different levels will be part of the research.
Figure 1: Structure of the research program
On the factory level, physical properties and features of the factory and of the products which are manufactured in the factory will be incorporated in the models. The main research questions on this level include the development of novel physics-based factory models (e.g. describing material flow, sustainability indicators, or process planning), the integration of physical attributes into VR models, and the implementation of links between the various physics-based simulations and the VR systems. The integration of physical attributes into the virtual models will help to carry out a realistic virtual analysis of the manufacturing environment.
Virtual sensors will monitor the process cycles and parameters at critical points of the manufacturing process chain. They provide the possibility to trace the virtual flow of material and information, monitor the execution of simulations, and integrate human expertise into the process. In our approach, the users will always have the possibility to decide how to adapt the process parameters in order to maintain proper operation of the manufacturing system.
Major algorithmic challenges in the simulation and visualization of manufacturing systems arise from the distribution, execution, and integration of large amounts of heterogeneous data, which are typically aggregated on the factory level. Therefore, new descriptions of processes at different scales will be created in order to establish an overview of the entire system.
The transition from factory level to machine level to process level and back will be integrated on the factory level with input from the level bridging projects. The simulation and visualization of multiple processes on the factory level is essential since they are influenced by incoming parameters from the lower levels and in turn influence them. Information such as temperature, pressure, material, object mass, power usage, etc. have to be transferred between the levels.
On the machine level, machine tools and their components, tooling systems, and measuring instruments will be specified as virtual physical models. Several research tasks will be tackled. First, it is planned to create physical models of machine tools and measurement instruments. Second, machine tool components and tooling systems will be designed, simulated, and tested using physics-based models. Third, the integration of the virtual models will be realized. They will be visualized in VR, and improved interaction techniques with the virtual models are part of the research.
During the development of machine tools and measurement instruments, virtual prototypes help to reduce the number of physical prototypes and to detect failure in the early phases of the product creation process. In order to realize model-based machine tool-, measurement instrument-, and component-simulations within a virtual environment, a physical model is required. Physics-based models of a desktop machine tool, of machine tool components and of tooling systems will be implemented. Physical characteristics to be modeled include - but are not limited to - deformation, stress, and temperature. The machine tools and components will be optimized and virtually tested against their original design requirements.
The machine tools and components will be validated and observed by virtual measurement technologies to achieve a good and stable result of the manufacturing process - just like in real manufacturing systems. Research on the machine level will therefore include virtual measurement systems. The main challenge here is to describe the sensor behavior by modeling and to numerically estimate the transfer functions of the sensor. The applicability and resolution of the virtual sensors will be enhanced and calibrated by numerical simulations of the workpiece and the tools.
Finally, the integration of the virtual models as well as new visualization and interaction techniques will be investigated in order to provide a suitable user interface for machine analysis in VR systems.
On the process level, various manufacturing technologies will be investigated at all participating universities of this IRTG. The integrated research program will include complementary as well as partly overlapping projects. Cutting processes will for instance be investigated in Kaiserslautern, Davis, and Berkeley whereas research on EDM will exclusively be conducted in Davis. Thus, a broad range of manufacturing technologies will be covered and it will be possible to generate manufacturing process chains which could produce real parts in an industrial environment. It is, however, neither necessary nor intended to cover all currently available manufacturing technologies.
The detailed characteristics of the manufacturing processes will be described through physics-based simulations. The physical models will be used to simulate the machining process itself as well as the material behavior and the structural response of the workpiece. The simulation of manufacturing processes invokes many length and time scales on one system level. The interaction of these length and time scales in homogenization and coarse graining methods is the subject of current research in applied physics and engineering. The core methods which will be used for the numerical realization are finite element, particle finite element, molecular dynamics, and phase field simulations. New variants as well as new applications of these methods will form the core of the research on the process level. By conducting such research on physical modeling of manufacturing processes and integrating the simulations into one overall system, the findings of this IRTG will create a new level of understanding of various manufacturing technologies.
The manufacturing technologies which will be in the focus of modeling and simulation on the process level are metal cutting (with a scale ranging from nano over micro to conventional machining) and grinding. For both cutting and grinding, the physics-based simulation of the coolant will be part of the research program. The research program will also address non-conventional processes such as EDM, spraying, and plasma-sintering for which comparatively little research on physical modeling has been reported up to now.
In order to obtain a comprehensive view of the manufacturing system, the correlations and interdependencies between the levels need to be assessed. This is a highly challenging research task, since very different simulation tools and algorithms are applied on each level. The need to integrate different simulation strategies (e.g. discrete vs. continuous modeling) contributes further to the research challenge of the level-bridging projects. In order to link different time and space scales, homogenization methods will be used, while data extraction of relevant interactions can be obtained by model reduction techniques.
To obtain a comprehensive view of the entire manufacturing system, the interfaces between the three levels must be defined allowing for the transfer of - typically aggregated - information from one level to another. This definition, the implementation of the links, the implementation of scalable manufacturing process simulations, and the visualization of the results are the main research tasks of the level bridging projects. The influencing parameters and their interactions will be detected and visualized. With this approach, the impact of one model onto another will become apparent. In this context, high importance will be placed on user-friendly and target driven visualization of machining processes, machines, virtual sensors, and entire manufacturing process chains. As a long term goal of this IRTG, the comprehensive visualization will allow to assess characteristics of the manufacturing process itself such as productivity, quality, or sustainability indicators.
Parallel to the scaling of the manufacturing process simulation, there is also a demand for scaling the visualization and interaction. For a complete simulation of a manufacturing process at factory level, a large virtual environment, for example a CAVE, is needed. For many applications, the simulation results as well as the current parameters and measurements of the real process should be available on-site. It is obvious that environments like a CAVE are neither portable nor can they be integrated in the real manufacturing system. Engineers will rather need mobile devices (e.g. tablet computers or data glasses) to control and steer the manufacturing processes online. Therefore, we will develop visualization and interaction techniques that scale both the level of detail of the manufacturing system and the devices used.
The level bridging projects are of high importance for the scientific program of this IRTG. However, level bridging research on manufacturing systems requires experience and knowledge in many modeling and simulation techniques as well as an excellent overview of a manufacturing system from factory level to process level. Therefore, part of the level bridging research must be carried out by experienced postdoctoral researchers from Mechanical Engineering and Computer Science. The postdoc from Mechanical Engineering will work on integrating the physical models of the three levels from the engineering point of view. He/she will closely collaborate with a postdoc from Computer Science who is focusing on the computational integration. At UC Berkeley, a postdoctoral researcher from Mechanical Engineering will be part of this program, at UC Davis a postdoctoral researcher from Computer Science. Thus, the complete postdoc researcher group will consist of four researchers, two from Mechanical Engineering and two from Computer Science.