Artificial Neural Networks to the Analysis of AISI 304 Steel Sheets through Limiting Drawing Ratio Test

Abstract

Drawbeads are employed in the formation of intricately shaped draw pieces to modify the stress state in specific areas of the drawpiece or to equalize the material's flow resistance along its perimeter. A unique drawbead simulator for calculating the factor of friction on the drawbead is presented in this article. The objective of this manuscript is to employ artificial neural networks (ANNs) to comprehend the impact of the primary friction process parameters, namely sample orientation concerning the rolling direction of the steel sheets, counter-sample surface roughness, and lubricant conditions, on the rate of friction. The goal was to create a database that would be utilized to train ANNs. This paper investigates the influence of temperature gradients developed in the tool profiles on the forming characteristics of AISI 304 steel sheets, primarily employing the cup drawing test. To examine the formability characteristics of the investigated steel sheets, limiting drawing ratios test was conducted by isothermal and non-isothermal heating conditions on the prepared circular blanks. The experimental results also highlighted that maintaining a uniform tooling temperature does not lead to an enhancement in formability. Specifically during non-isothermal forming conditions, punch was purposefully kept at a temperature lower than the die / blank holder, LDR increased by 16 % in non-isothermal and 8 % in isothermal forming conditions.