Font Size: a A A

Research On Warping Phenomenon Control In Plate And Strip Rolling

Posted on:2004-05-13Degree:MasterType:Thesis
Country:ChinaCandidate:B ChenFull Text:PDF
GTID:2121360095955418Subject:Materials Processing Engineering
Abstract/Summary:PDF Full Text Request
Warping phenomenon in plate and strip rolling is one of important technique problems in rolling field which haven't be solved . Warping in rolling causes the decrease in productivity, the deterioration in product quality, and damage to the equipment. Technical personnel home and abroad have been investigating in the field for long time. But this problem still exists in many steel rolling mills and theory of which haven't be explained correctly.In order to solve the problem , we analyzed deformation field in symmetrical rolling and asymmetrical rolling, auto-adjust steady process in asynchronous rolling and temperature' influence on warping.Because warping is caused by a number of factors. It is difficult to quantify the effect of the each factor on warping. In order to verify the proportion of each factor's influence, we carry on a simulative experiment. The experiment simulates the actual statement in the rolling mill and arranges experiment with orthogonal test . According to experimental result, we find out the each factor 's influence on warping. Besides, we put forward that the form parameter of deformed field and rollers speed ratio is good means to control warping.In order to predict warping, we apply Artificial Neural Network to it. The Artificial Neural Network is a rising interscience. It's consist of neural joints and links between joints as cerebra. The training process is divided into two parts: Forward and backward. In forward process, The data inputed to the input layer propagate to the output layer through the hidden layer. The error propagate reversely and modify weights of each layer to minimize error. Process will be stop until error become small enough after calculating repeatedly. As a result, one weights matrix is got. Forecasting process is just a forward process. The outputs of output layer is forecast value. The chosen neural network architecture consisted of one input layer, one hidden layer and the output layer. The input layer had 6 inputs, such as diameter proportion of rollers, speed proportion of rollers, surface roughness of rollers, form parameter of deformed field, thickness of rolled piece as well as biting angel. The output layer gave the warping forecast value..In order to apply the technology to practice. We program a medium and heavy plate rolling schedule design soft with control of warping phenomenon in rolling process. The soft can be used to instruct practice production. This soft is easy to use. No special course is needed. The soft can be used to instruct practice production if installed in operation room. The plate rolling schedule result from the soft had been used in the plate rolling mill of ShaoGuan steel and iron corporation. It is work well. And plate warping is under the control.This paper aim to find the cause of the warping and it's resolvent. It's attained. By this way, producer increase the productivity, improve product quality, and protect the equipment. So it has great significance on theory and practice. In view of the great plate and strip yield in our country, economic benefits of the research shouldn't be underestimated.
Keywords/Search Tags:Phenomenon
PDF Full Text Request
Related items