一维无迭代自适应网格技术在Burgers方程中的应用

ONE DIMENSIONAL ADAPTIVE GRID TECHNIQUE WITHOUT ITERATION AND ITS APPLICATION TO BURGERS EQUATION

  • 摘要: 基于变分原理的自适应网格技术需用迭代法求解,非常耗时.故提出了一种新的自适应网格技术,根据权函数曲线下面积相等的原则,重新分布格点位置,一步到位,无需迭代,大大提高了自适应网格技术的效率.利用具有解析解的一维Burgers方程对这一新技术进行验证,发现它能根据不同的权函数确定不同的格点分布;对同一种权函数,当其随时间变化时,对应的格点分布也相应变化.研究表明:该技术能根据问题的求解,在解的大梯度区自动加密网格,从而可计算出激波.

     

    Abstract: A new adaptive grid technique is proposed, which can redistribute grid points according to the equivalence of area under the curve of the weight function. So the iteration in the traditional technique through a variational approach is unnecessary here, and the efficiency is improved greatly. The new technique is applied to Burgers equation with analytic solutions. The results show that different weight functions have different distribution of grid points, and variational weight functions with time have variational distribution of grid points. Because the new technique can redistribute more grid points in the high gradient solution regions in response to the numerical solution, it can capture the shock efficiently.

     

/

返回文章
返回