Josef Meinhardt

Stamping Plant 4.0 – Basics for the Application of Data Mining Methods in Manufacturing Car Body Parts

Data-driven quality evaluation in the stamping process of car body parts is quite promising because dependencies in the process have not yet been sufficiently researched. However, the application of data mining methods for the process in stamping plants would require a large number of sample data sets. Today, acquiring these data represents a major challenge, because the necessary data are inadequately measured, recorded or stored. Thus, the preconditions for the sample data acquisition must first be created before being able to investigate any correlations. In addition, the process conditions change over time due to wear mechanisms. Therefore, the results do not remain valid and a constant data acquisition is required. In this publication, the current situation in stamping plants regarding the process robustness will be first discussed and the need for data-driven methods will be shown. Subsequently, the state of technology regarding the possibility of collecting the sample data sets for quality analysis in producing car body parts will be researched. At the end of this work, an overview will be provided concerning how this data collection was implemented at BMW as well as what kind of potential can be expected.

A modular modeling approach for describing the in-plane forming behavior of unidirectional non-crimp-fabrics

Various commercially available unidirectional (UD) non-crimp-fabrics (NCFs) are currently used for manufacturing carbon fiber reinforced plastic (CFRP) parts. These UD NCFs can differ significantly in their forming behavior. For optimizing and ensuring the manufacturability of the forming process of CFRP parts manufactured from UD NCFs these differences have to be taken into account. This motivates developing an efficient and universally applicable modular modeling approach for describing the in-plane forming behavior of various UD NCFs. The first component of this modular approach is a hyperelastic material model that accurately predicts the fiber orientation of UD NCFs during forming. This material model is implemented via a user-defined material subroutine in the commercial finite element package LS-DYNA. The second component is a simple truss structure that allows modeling the various stitch patterns of the different UD NCFs. This modular model can be calibrated via simple tensile tests. To demonstrate the versatility of this approach, the in-plane forming behavior of three different UD NCFs is validated by comparing experimental data and simulation results of the common picture frame test.

Experimental and numerical investigation of blankholder’s vibration in a forming tool: a coupled MBS-FEM approach

In order to achieve the energy and efficiency goals in modern automotive press shops, press systems with increasingly high stroke rates are being implemented (Meinhardt in proceedings of ACI forming in car body engineering. Bad Nauheim, Germany 2012). As a side effect, the structural dynamic loads on the press and especially on the forming tool increase. Hence, to design reliable and withstanding forming tools, a detailed knowledge of the vibrations and resulting critical loads is essential. In this paper, the main focus is put on the vibration of the blankholder—the heaviest moving component in the forming tool. To predict those vibrations, a coupled multibody-finite element simulation (MBS-FEM) is conducted, which combines rigid and elastic modeling approaches. Also, an experimental validation of the blankholder vibration under operational load is carried out. To compare the numerical and experimental results—both in time and frequency domain—an 1/3-octave analysis of a blankholder’s vibrational speed is performed. The test measurements agree well with the MBS-FEM simulation.

Data-driven inline optimization of the manufacturing process of car body parts

The manufacturing process of car body parts needs to be adaptable during production because of fluctuating variables; finding the most suitable settings is often expensive. The cause-effect relation between variables and process results is currently unknown; thus, any measure taken to adjust the process is necessarily subjective and dependent on operator experience. To investigate the correlations involved, a data mining system that can detect influences and determine the quality of resulting parts is integrated into the series process. The collected data is used to analyze causes, predict defects, and optimize the overall process. In this paper, a data-driven method is proposed for the inline optimization of the manufacturing process of car body parts. The calculation of suitable settings to produce good parts is based on measurements of influencing variables, such as the characteristics of blanks. First, the available data are presented, and in the event of quality issues, current procedures are investigated. Thereafter, data mining techniques are applied to identify models that link occurring fluctuations and appropriate measures to adapt the process so that it addresses such fluctuations. Consequently, a method is derived for providing objective information on appropriate process parameters.

A review on process-induced distortions of carbon fiber reinforced thermosets for large-scale production

During the manufacturing process of carbon fiber reinforced plastics, residual stresses and shape distortions occur. Analytical or numerical models can be employed to accurately predict these stresses and deformations. For a virtual compensation of a part’s geometry, an understanding of the main driving mechanisms behind process-induced distortions is essential. The present study shows main sources as well as influencing factors on stresses and distortions that arise during large-scale production of composites based on thermoset resin systems. Besides basic information on the thermal, chemical and mechanical behavior of thermoset resins and carbon fibers, state-of-the-art numerical approaches are reviewed with a focus on finite element discretization and constitutive modeling. Current virtual compensation strategies as well as possible influences of the forming operation are discussed. Finally, a conclusion is drawn that includes recommendations for future developments in the field of process-induced distortions.

Towards Data Driven Process Control in Manufacturing Car Body Parts

The manufacturing process of car body parts is a complex industrial process where many machine parameters and material measurements are involved in establishing the quality of the final product. Data driven models have shown great advantages in helping decision makers to optimize this kind of complex processes where good physical models are hard to build. In this paper a framework for on-line process monitoring and predictive modeling is proposed to optimize a car body part production process. Anomaly detection plays an important role in this framework as it can provide an early alert for operators on the production line using a complex set of machine parameters and material properties. In this paper an anomaly detection algorithm, Gloss, that is successfully implemented as the first module in the process, is introduced. Gloss finds local outliers in high dimensional mixed data-sets using a relative density measure that takes the global neighborhood into account while searching for outliers in subspaces of the data. An overview of the application and implementation of the algorithm in the car body part press shop is presented.

Evaluation and Validation of Methods to Determine Limit Strain States with Focus on Modeling Ductile Fracture

In the manufacturing process of body in white components made from sheet metal it is state of the art to accompany the process by means of finite element analysis. A main criterion for determining a feasible tool design and production process parameters is the prediction of material failure, which can be categorized in instability and ductile fracture. The ductile fracture failure mode is more likely to occur, as more advanced high strength steels and aluminum alloys are used for body in white components. Therefore different approaches have been presented to model ductile fracture over the past years. This task is more challenging when the material is exposed to arbitrary loading paths that can occur in deep drawing processes. However there is no guideline for sheet metal forming applications to determine which models for predicting ductile fracture are suitable, which experiments are necessary and how calibration of model parameters and validation of model prediction can be performed. Additionally there is no standard established that prescribes the evaluation of limit strain states from experiments. Suitable limit strain states are a basic requirement for prediction of ductile fracture as they are used for calibration of fracture models. In this paper, two methods for evaluation of limit strains are discussed and applied to tensile specimens with circular hole and circular cut outs made from aluminum alloy AlSi0.6Mg0.5. One validation experiment is used to investigate failure prediction that is based on limit strain states from different evaluation methods.