东北大学学报（自然科学版）

东北大学学报(自然科学版) ›› 2010, Vol. 31 ›› Issue (8): 1113-1116+1125.DOI: -

• 论著 • 上一篇    下一篇

1.5维刚塑性有限元快速求解板带轧制过程

梅瑞斌;李长生;刘相华;包立;   

  1. 东北大学秦皇岛分校;东北大学轧制技术及连轧自动化国家重点实验室;
  • 收稿日期:2013-06-20 修回日期:2013-06-20 出版日期:2010-08-15 发布日期:2013-06-20
  • 通讯作者: -
  • 作者简介:-
  • 基金资助:
    国家自然科学基金重点资助项目(50534020)

Fast 1.5-dimension rigid-plastic FEM solution to strip rolling process

Mei, Rui-Bin (1); Li, Chang-Sheng (2); Liu, Xiang-Hua (2); Bao, Li (1)   

  1. (1) Northeastern University at Qinhuangdao, Qinhuangdao 066004, China; (2) State Key Laboratory of Rolling and Automation, Northeastern University, Shenyang 110004, China
  • Received:2013-06-20 Revised:2013-06-20 Online:2010-08-15 Published:2013-06-20
  • Contact: Mei, R.-B.
  • About author:-
  • Supported by:
    -

RichHTML

0

PDF (PC)

617

摘要： 为提高有限元计算效率,实现其在线应用,从减少未知数和自由度个数出发,推导了1.5维有限元的形函数、B矩阵和Hessian矩阵,建立了板带轧制过程1.5维刚塑性有限元快速求解模型,并利用FORTRAN语言开发求解程序RF-1.5D.针对某钢厂典型精轧过程,对轧制力、计算迭代次数和计算时间进行了求解分析.结果显示:轧制力计算值与实测值吻合良好,计算误差小于10%,计算精度令人满意;1.5维有限元求解单道次轧制过程迭代步数少于35次,计算时间少于100 ms,相比2维单元,每迭代步耗费的计算时间明显减少,提高了计算效率.可见,1.5维有限元的计算精度和计算时间满足在线应用的初步要求.

关键词: 1.5维刚塑性有限元, 板带轧制, 计算时间, 迭代步数, 轧制力

Abstract: To improve the FEM computational efficiency and implement its online application, an 1.5-dimension rigid-plastic FEM (RPFEM) model was developed as the fast solution to the strip rolling process, where the shape function and the B and Hessian matrixes of 1.5-dimension RPFEM were derived to decrease the number of unknown values and degrees of freedom. The different strip rolling processes were solved successfully by the software RF-1.5D we developed, and the rolling force, iteration steps and computing time were investigated with the data acquired from a certain plant. The calculated value of rolling force is basically in agreement with the measured value, i.e., an error less than 10%. The iteration steps in the solution for a single pass rolling are thus less than 35 and take 100 ms only, which means that the computing time for every iteration step is shortened by 1.5-dimension RPFEM in comparison with that by 2D. The precision and high efficiency of the 1.5-dimension RPFEM are therefore proved available to meet the basic requirements for online prediction.

中图分类号: