Font Size: a A A

The Research On The Cutting Performance Of Free-cutting Steel

Posted on:2015-05-16Degree:MasterType:Thesis
Country:ChinaCandidate:H ZhangFull Text:PDF
GTID:2181330467483797Subject:Mechanical Manufacturing and Automation
Abstract/Summary:PDF Full Text Request
With the development of domestic machinery industry especially automobile industry,free-cutting steel has been widely used. So it’s of great significance to study the cuttingperformance of free-cutting steel, optimization of process parameters for improvingmachining efficiency and quality and controlling cost. Based on the environmentallyfriendly free-cutting steel cutting process as the research object, the tool life, cuttingforce and surface roughness were studied. And with the help of the neural network,prediction model of surface roughness was established. The research content mainlyincludes:(1)The tool life and wear morphology of YT15cemented carbide tool were studiedwhen cutting the two kinds of free-cutting steel. The tool life was studied under thecondition of different cutting speeds by single factor experiments, and the Taylorequations of tool life were established. The tool wear morphology was mainly flankwear and crater wear. But under the cutting speed of150m/min, boundary wear wasalso serious.(2)The influence of cutting speed on cutting force and the influence of cutting speed,cutting depth, feed rate and corner radius on surface roughness were researched bysingle factor experiments, respectively. The empirical model of cutting force and surfaceroughness with cutting speed, cutting depth and feed as inputs were established usingmultiple linear regression method by orthogonal experiment. And the empirical modelsworked very well in the validation test.(3)The prediction model of surface roughness with cutting speed, cutting depth, feed,material, corner radius as inputs was established based on neutral network, and theforecast system was build using the Matlab GUI module. The weights of BP neuralnetwork was optimized by the genetic algorithm in this article, in order to solve theproblem of impropriety weight distribution caused by falling into local extremum andthe long training time. Experiment showed that the BP neural network could predict thesurface roughness of two kinds of free-cutting steel accurately.
Keywords/Search Tags:free-cutting steel, tool life, cutting force, the surface roughness. BPneural network
PDF Full Text Request
Related items