Project: Sheet metal roll forming process improvement by means of numerical simulation, process monitoring and artificial vision
The roll forming action is often defined like the action to form a sheet metal strip along straight, longitudinal, parallel bend lines with multiple pairs of contoured rolls without changing the thickness of the material at room temperature. Sheet metal forming processes are very high non-linear processes being very difficult to model. However, the acquired experience when forming classical materials, such as mild steels, has permit to guarantee the process stability and to form sound parts with very close tolerances._x000D_In the last few years and COly due to the constantly increasing market competitiveness there has been a trend towards getting more and more complex geometries, design free structures, using materials with higher ultimate tensile strengths and lower forming properties and trying to reduce the number of manufacturing steps due to economical reasons. The deformations that the materials suffer during the manufacturing processes have raised considerably being much closer to their forming limits. _x000D_New problems have appeared due to the novel stress states suffered by the material during the forming process, such as premature cracks, springback problems, distortion of the parts, bad final surface quality, scraps and defective parts Even with a good conventional design of the machines and setting up of the manufacturing process originates bad quality parts, profile warping and bow and machine safety problems._x000D__x000D_From the point of view of industrial roll forming process design, the critical steps are: the flower design and optimisation, manufacture and set-up of the roll forming line. From the point of view of industrial production, problems appear due to material properties changes, roll’s excessive wear, high set-up times and forming of bad quality parts and scrap._x000D_ _x000D_The finite element modelling is an essential tool to develop new forming processes and for a better understanding of their fundamentals, currently used to optimise the flower design, which enables time waste reduction and higher benefits by decreasing tooling costs and set-up times. _x000D_The existing numerical models do not include friction and do not take into account roll stands deformations. More accurate models are needed to totally improve the process, precisely calculate the needed motor’s power and to optimise the sizing of the industrial lines._x000D_On the other hand, artificial vision techniques, uses video cameras and computers to replace human vision in evaluation and inspection tasks that are precise, repetitive or high speed. One of the most important advantages of machine vision is improving the quality of the product while lowering its costs. _x000D_In visual inspection, a camera or a set of cameras are utilised to take a picture of a part and after that, an image treatment that could be pattern matching or edge detection finds if the quality of the part is within the predefined tolerances. Although not specific market solutions are known for the application of these techniques in the roll forming processes._x000D_Process control using force, torque and power signals has for many years proven to be reliable and relatively low cost. This has provided: machine and tool protection, increased productivity and improved product quality. The conventional sensors used in sheet metal forming processes monitoring systems are based COly on the measurement of two variables: forces/torque and acoustic emissions. Those kind of sensors have not been used in industrial roll forming lines, where process monitoring would provide stable processes and higher quality products._x000D_Due to the previous facts, the CO goal of the present project is to improve the current roll forming processes and overcome the new limits and problems that occur when forming new shapes and materials. For that, advanced numerical models will be developed to cover the current lack of knowledge. It is worth highlighting that data M GmbH, the coordinator of the project, is the world leader company developing their own software for roll forming simulation, called COPRA® Software._x000D__x000D_Force sensors will be constructed during the project and artificial vision techniques and hardware will be developed based on laser triangulation techniques for process monitoring and quality inspection of the parts. Those inputs will be used to build an expert system, using case-based reasoning techniques. Faults and causes will be recorded at two industrial companies and signals and images will be analysed to feed the expert system data base. An active user interface will permit to update the expert system by introducing operator’s information about online detected new faults and cases. All this information will be used to decrease the defective parts amount and the current high set-up times and breakdowns by means of early identification of the errors causes._x000D_As a whole, the project will develop the tools and the methodology for the global improvement of the roll forming processes._x000D__x000D_
|
Acronym
|
RFexpert
(Reference Number: 4745)
|
|
Duration
|
01/01/2010 - 01/01/2013
|
|
Project Topic
|
Improve quality of section and surface, finite element model, energy consumption in roll forming and time from design to the final profile, with means of Finite Element Analysis, Process Monitoring, Artificial Vision, Expert Systems and Implementation of outcomes in real production lines.
|
|
Network
|
Eurostars
|
|
Call
|
Eurostars Cut-Off 2
|
Project partner
