大型铸锻件
大型鑄鍛件
대형주단건
HEAVY CASTING AND FORGING
2015年
5期
4-8
,共5页
2.25Cr-1Mo-0.25V钢%晶粒尺寸%模型%热变形%BP神经网络
2.25Cr-1Mo-0.25V鋼%晶粒呎吋%模型%熱變形%BP神經網絡
2.25Cr-1Mo-0.25V강%정립척촌%모형%열변형%BP신경망락
2.25Cr-1Mo-0.25V steel%grain size%model%thermal deformation%BP neural network
采用不同初始晶粒尺寸的锻态2．25Cr-1Mo-0．25V钢试样，在Gleeble-1500D热模拟试验机上进行热压缩实验，通过定量金相法测得各变形条件下的平均晶粒尺寸。以实验数据为基础，采用BP神经网络对实验数据进行训练，获得了2．25Cr-1Mo-0．25V钢的热变形平均晶粒尺寸模型。结果表明预测值与实验值吻合良好，验证了所建晶粒尺寸模型的准确性，实现对2．25Cr-1Mo-0．25V钢热锻后平均晶粒度的预测。
採用不同初始晶粒呎吋的鍛態2．25Cr-1Mo-0．25V鋼試樣，在Gleeble-1500D熱模擬試驗機上進行熱壓縮實驗，通過定量金相法測得各變形條件下的平均晶粒呎吋。以實驗數據為基礎，採用BP神經網絡對實驗數據進行訓練，穫得瞭2．25Cr-1Mo-0．25V鋼的熱變形平均晶粒呎吋模型。結果錶明預測值與實驗值吻閤良好，驗證瞭所建晶粒呎吋模型的準確性，實現對2．25Cr-1Mo-0．25V鋼熱鍛後平均晶粒度的預測。
채용불동초시정립척촌적단태2．25Cr-1Mo-0．25V강시양，재Gleeble-1500D열모의시험궤상진행열압축실험，통과정량금상법측득각변형조건하적평균정립척촌。이실험수거위기출，채용BP신경망락대실험수거진행훈련，획득료2．25Cr-1Mo-0．25V강적열변형평균정립척촌모형。결과표명예측치여실험치문합량호，험증료소건정립척촌모형적준학성，실현대2．25Cr-1Mo-0．25V강열단후평균정립도적예측。
By adopting the forged 2.25Cr-1Mo-0.25V steel samples with different initial grain size, the hot com-pression test has been performed on the Gleeble-1500D thermal simulation testing machine.Through the quantitative metallographic method, the average grain size under various deformation conditions has been measured.Based on the experimental data and the BP neural network training with these data, the model of average grain size of thermal deformation of 2.25Cr-1Mo-0.25V steel has been obtained.It turned out that the predicted value fit the tested value well, and the accuracy of built grain size model was verified.Therefore, the prediction of average grain size after hot forging of 2.25Cr-1Mo-0.25V steel has been realized.
