Improving an inverse model of sheet metal forming by neural network based regression

An inverse model for a sheet meta l forming process aims to determine the initial parameter levels required to form the final formed shape. This is a difficult problem that is usually approached by traditional methods such as finite element analysis. Formulating the problem as a classification problem makes it possible to use well established classification algorithms, such as decision trees. Classification is, however, generally based on a winner-takes-all approach when associating the output value with the corresponding class. On the other hand, when formulating the problem as a regression task, all the output values are combined to produce the corresponding class value. For a multi-class problem, this may result in very different associations compared with classification between the output of the model and the corresponding class. Such formulation makes it possible to use well known regression algorithms, such as neural networks. In this paper, we develop a neural network based inverse model of a sheet forming process, and compare its performance with that of a linear model. Both models are used in two modes, classification mode and a function estimation mode, to investigate the advantage of re-formulating the problem as a function estimation. This results in large improvements in the recognition rate of set-up parameters of a sheet metal forming process for both models, with a neural network model achieving much more accurate parameter recognition than a linear model.

Contact pressure and wear in sheet metal forming - an FEM analysis

Wear is the principal cause of tool failure in most sheet metal forming processes. It is well known that the contact pressure between the blank and the tool has a large influence on the wear of the tool, and hence the tool life. This investigation utilises the finite element method to analyse the contact pressure distribution over the die radius for a particular deep drawing process. Furthermore, the evolution of the predicted contact pressure distribution throughout the entire stroke of the punch is also examined. It was found that the majority of the process shows a steady state pressure distribution, with two characteristic peaks over the die radius, at the beginning and end of the sheet contact area. Interestingly, the initial transient contact pressure response showed extremely high localised peak pressures; more than twice that of the steady state peaks. Results are compared to wear reported in the literature, during similar experimental deep drawing processes. Finally, the significance and effect of the results on wear and wear-testing techniques are discussed.

Modeling of contact pressure in sheet metal forming

For a given sheet metal forming process, an accurate determination of the contact pressure distribution experienced is an essential step towards the estimation of tool life. This investigation utilizes finite element (FE) analysis to determine the evolution and distribution of contact pressure over the die radius, throughout the duration of a channel forming process. It was found that a typical two-peak steady-state contact pressure response exists for the majority of the process. However, this was preceded by a transient  response, which produced extremely large and localized contact pressures. Notably, it was found that the peak transient contact pressure was more than double the steady-state peak. These contact pressure results may have a significant influence on the tool wear response and therefore impact current wear testing and prediction techniques. Hence, an investigation into the validity of the predicted contact pressure was conducted.

The influence of temperature on the forming behavior of metal/polymer laminates in sheet metal forming

The influence of temperature on the forming behavior of an aluminum/polypropylene/aluminum (APA) sandwich sheet was studied. Shear and tensile tests were performed to determine the mechanical properties of the laminate and the component materials as a function of process temperature. The forming limit diagram (FLD) of the laminate was established for two different temperatures, and its springback behavior was examined in four-point bend and channel bend tests. Cup forming tests were performed at various test temperatures to determine the limiting drawing ratio (LDR) and the tendency for wrinkling at these temperatures. Although there was only a minor influence of temperature on the mechanical properties and the FLD values of the laminate, the bend test results reveal that springback can be reduced by forming at higher temperature. The decreasing strength of the core material with rising process temperature led to an increased tendency of the laminate to wrinkle in the heated cup drawing tests.

The effect of clamping sequence on dimensional variability in sheet metal assembly

Finite element (FE) modelling techniques have become a popular tool for exploring welding and clamping sequence dependence in sheet metal assemblies. In the current paper, the dimensional variability associated with different assembly clamping sequences is investigated with a FE contact modelling approach implemented in the commercial code Abaqus. A simplified channel section assembly consisting of a top hat and bottom plate is the case study investigated. Expected variation modes of bow and twist were used to simulate key variability sources in the main structural component under investigation; the top hat of the channel section. It was found that final assembly variability can change considerably depending on clamp sequence selection. It was also found that different clamp sequences can control particular modes of variation better than others, and that there is not one particular clamping sequence that is the best for containing all variation modes. An adaptable assembly process is therefore suggested, where given the shape of input components the best available clamping sequence is selected. Comparison of the performance of the proposed adaptable clamping sequence to traditional fixed clamping sequences shows improvements for the dimensional control of variability in non-rigid components. While introducing such a method in production would require inspection of each component being assembled and investigation of the alternative clamping sequences, given access to fast and detailed dimensional inspection technology such as optical coordinate measuring machines (OCMM's), the approach shows promise for future application.

Knowledge based optimisation techniques for sheet metal forming die design

A knowledge based optimism technique has been developed to predict solutions for quality issues found in an initial draw die design. Post processing of the initial design yields all the features applying forces, and major quality issues. Using the geometric relationship between the two, a knowledge-base is interrogated to determine the possible corrective actions. These actions are then passed through a fast semi-analytical model to determine the level of change required. Results from a 2D forming are presented to highlight the advantage of the new algorithms over current optimisation techniques.

Computer aided ultrasonic thickness strain analysis of automotive sheet metal stampings using FEA information

This paper discusses a computer-aided methodology for thickness strain analysis (TSA) of sheet metal stampings using geometry and strain information that is extracted from finite element analysis (FEA) results. The system utilises both FEA results and an ultrasonic gauge capability expert system to assist press shop personnel, providing them with capabilities such as optimum measurement point location and an estimate of gauge error. Key advantages of this enhanced TSA methodology are related to overall efficiency and accuracy gains.

Galling reduction through die surface treatment - testing three treatments for sheet metal stamping

This brief article describes how tool galling occurs in stamping and describes results of tests on how TiCN, Hard-Cr, and CrCN were used on sample parts to help reduce galling.

Multivariate modelling of variability in sheet metal forming

The inherent variability in incoming material and process conditions in sheet metal forming makes quality control and the maintenance of consistency extremely difficult. A single FEM simulation is successful at predicting the formability for a given system, however lacks the ability to capture the variability in an actual production process due to the numerical deterministic nature. This paper investigates a probabilistic analytical model where the variation of five input parameters and their relationship to the sensitivity of springback in a stamping process is examined. A range of sheet tensions are investigated, simulating different operating windows in an attempt to highlight robust regions where the distribution of springback is small. A series of FEM simulations were also performed, to compare with the findings from the analytical model using AutoForm Sigma v4.04 and to validate the analytical model assumptions.

Results show that an increase in sheet tension not only decreases springback, but more importantly reduces the sensitivity of the process to variation. A relative sensitivity analysis has been performed where the most influential parameters and the changes in sensitivity at various sheet tensions have been investigated. Variation in the material parameters, yield stress and n-value were the most influential causes of springback variation, when compared to process input parameters such as friction, which had a small effect. The probabilistic model presented allows manufacturers to develop a more comprehensive assessment of the success of their forming processes by capturing the effects of inherent variation.

Simulation of contact pressure in sheet metal trimming/blanking process

In the car body stamping process, trim/blank die cutting edges are subjected to very high tribological loads that result in loss of tool material from both the punch and die cutting edges. According to Archard’s wear model, normal contact force and sliding distance directly affects the wear. Therefore, knowledge of the acting forces on local contact areas has a pivotal role towards the prediction of tool wear. This paper presents a finite element modelling approach to determining the contact pressure distribution on the tool cutting edges during a trimming/blanking process. Characteristic areas on sheared edge profile, variation of punch force and high contact pressures affected areas have also been analysed.

Contact pressure evolution and its relation to wear in sheet metal forming

For a given sheet metal forming process, an accurate determination of the contact pressure distribution is an essential step towards the estimation of tool life. This investigation utilizes finite element (FE) analysis to model and explain the evolution and distribution of contact pressure over the die radius, throughout the duration of a channel forming process. It was found that a typical two-peak steady-state contact pressure response exists for the majority of the process. However, this was preceded by an initial transient response, characterized by extremely large and localized contact pressures, which were more than double the magnitude of the steady-state peak pressure. The validity of the predicted contact pressure behavior was assessed via detailed numerical analysis and by examining the wear response of an experimental stamping operation. The experimental results revealed that the high contact pressure zones of the transient response corresponded to a severe galling wear mechanism. Therefore, the transient response may be of primary significance to the tool wear response; thus questioning the applicability of traditional bending-under-tension wear tests for sheet metal stamping processes. Finally, a parametric study was conducted, examining the influence of the major process parameters on the steady-state and peak transient contact pressures, using the developed FE model. It was found that the bend ratio and the blank material ultimate tensile strength had the most influence on the peak contact pressures. The main process-related parameters, friction coefficient and blank holder force, were found to have only a minor influence.

Contact pressure prediction in sheet metal forming using finite element analysis

Tool wear has become a significant issue associated with the forming of high strength sheet steels in the automotive industry. In order to combat this problem, recent research has been devoted to utilizing the contact results obtained from current sheet metal forming software predictions, in order to develop/apply tool wear models or tool material selection criteria for use in the stamping plant. This investigation aims to determine whether a specialized sheet metal forming software package can correctly capture the complex contact conditions that occur during a typical sheet metal stamping process. The contact pressure at the die radius was compared to results obtained using a general-purpose finite element software package, for a simple channel-forming process. Although some qualitative similarities between the two predictions were observed, it was found that significant differences in the magnitude and distribution of the contact pressure exists. The reasons for the discrepancies in results are discussed with respect to the simplifications and assumptions adopted in the finite element model definitions, and with regards to other results available in the literature.

Evaluating lubricants in sheet metal forming

The sheet metal forming process basically involves the shaping of sheet metal of various thickness and material properties into the desired contours. This metal forming process has been extensively used by the automotive industry to manufacture both car panels and parts. Over the years numerous investigations have been conducted on various aspects of the manufacturing process with varied success. In recent years the requirements on the sheet metal forming industry have headed towards improved stability in the forming process while lowering environmental burdens. Therefore the overall aim of this research was to identify a technique for developing lubricant formulations that are insensitive to the sheet metal forming process. Due to the expense of running experiments on production presses and to improve time efficiency of the process the evaluation procedure was required to be performed in a laboratory. Preliminary investigations in the friction/lubricant system identified several laboratory tests capable of measuring lubricant performance and their interaction with process variables. However, little was found on the correlation between laboratory tests and production performance of lubricants. Therefore the focus of the research switched to identifying links between the performance of lubricants in a production environment and laboratory tests. To reduce the influence of external parameters all significant process variables were identified and included in the correlation study to ensure that lubricant formulations could be desensitised to all significant variables. The significant process variables were found to be sensitive to die position, for instance: contact pressure, blank coating of the strips and surface roughness of the dies were found significant for the flat areas of the die while no variables affected friction when polished drawbeads were used. The next phase was to identify the interaction between the significant variables and the main lubricant ingredient groups. Only the fatty material ingredient group (responsible for the formation of boundary lubricant regimes) was found to significantly influence friction with no interaction between the ingredient groups. The influence of varying this ingredient group was then investigated in a production part and compared to laboratory results. The correlation between production performance and laboratory tests was found to be test dependant. With both the Flat Face Friction test and the Drawbead Simulator unaffected by changes in the lubricant formulation, while the Flat Bottom Cup test showing similar results as the production trial. It is believed that the lack of correlation between the friction tests and the production performance of the lubricant is due to the absence of bulk plastic deformation of the strip. For this reason the Ohio State University (OSU) friction test was incorporated in the lubricant evaluation procedure along with a Flat Bottom Cup test. Finally, it is strongly believed that if the lubricant evaluation procedure highlighted in this research is followed then lubricant formulations can be developed confidently in the laboratory.