Hot strip roughing mill: camber measurement

To understand and to overcome the camber problem in the Hot Strip Mill, no good instrument was available on the market to measure this phenomenon. It was decided between EES, IRSID, SOLLAC Dunkerque (now ARCELOR Dunkerque) and with the ECSC financial help, to develop and to test a first system on line in the Hot Strip Mill. Sollac Dunkerque was the pilot plant and coordinator. IRSID helped us to specify the gauge architecture and to select the best possible algorithm for this problem and did this as follows: IRSID has built a simulator, designed different types of well-known camber defects (sleeve, arch), and possible disturbance (deviation rotation), made a lot of simulations to compare the advantages and disadvantages of each type of algorithm, and to find the optimal configuration and characteristics for an industrial gauge. Using the two prototype camera systems in the Hot Strip Mill in Dunkerque, Irsid analysed the real measurement to improve defect and disturbance knowledge. Only after that, and confirmed by manual measurementslchecks, the best combination was chosen: A two-camera gauge using "Derive" algorithm EES was chosen as a well-known width gauge supplier, and supplied a suitable hardware and software solution, enabling camber measurement with a resolution σ (shape's deviation) to 5 mm. To succeed in this task, Trials were made with PC-104+ format industrial processing cards, A new camera data acquisition card in PC-104+ format was developed and tested QNX was tried and selected as a new operating system The real time processing of each gauge has been implemented in a processing module EES have designed and manufactured the electronics cabinet Integrated system tests were performed. This required setting up the cabinet with two stereoscopic cameras spaced 750 mm above an illuminated and moving strip to simulate the site conditions. The gauge was installed at Dunkerque site EES had to come back on several occasions to Dunkerque to solve small / bigger problems, especially to improve the software reliability of this complete new design. ARCELOR Dunkerque plant (coordinator) had to implement a initial simple prototype, to be ready to host properly the industrial prototype, and to carry out the validation tests. The following actions were done: A second camera optical bed was installed at the exit of the last Roughing Mill A simple prototype was implemented to start the study and to supply raw data for the IRSID simulator, and to test the two points algorithm on site. The measuring room was prepared to host the new industrial prototype. Providing air-cooling, power supply and a fibre optics link with HSM computer. Different tests and calibrations were carried out to validate the industrial prototype A special alignment tool was designed and built to guarantee an accurate and user-friendly alignment of the two optical beds. Tests and manual measurements were carried out on real products to validate the algorithm selection.