siyanger 发表于 2006-3-29 15:31

[求助]多自由度振动微分方程的求解！

用Matlab的ode45进行多自由度（15个自由度，激励为简谐力叠加）振动微分方程的求解（包括线性和非线性都不行），结果真是一踏糊涂！那实际计算中都用什么方法求解呢？

xj2070 发表于 2006-3-29 15:44

re[求助]多自由度振动微分方程的求解！

用Matlab的ode45进行多自由度（15个自由度，激励为简谐力叠加）振动微分方程的求解（包括线性和非线性都不行），结果真是一踏糊涂！那实际计算中都用什么方法求解呢？


Take it easy.:-)
You should be aware there are many reasons which may lead to
terrible results but they maybe irrevelant to matlab. For example, the integeration step you selected is not reasonable and so on.matlab is powerful.Result associates with tool but determined by user

siyanger 发表于 2006-3-29 16:06

呵呵，谢谢提醒。ode45是自己适应步长，不用人工设步长的。对于复杂的系统，定步长反而更不好求解了！

xj2070 发表于 2006-3-29 16:53

呵呵，谢谢提醒。ode45是自己适应步长，不用人工设步长的。对于复杂的系统，定步长反而更不好求解了！No thanks

IN the case of the structural response computation, the integeration step dt should be less than
T/10, T is the period of the highest frequency mode we care about. In pratice, the newmark-B method are generally empolied in the structural vibration response prediction

gghhjj 发表于 2006-3-30 07:06

回复：（siyanger）[求助]多自由度振动微分方程的求解...

看具体的方程程形势，想newmark等稳定性比较好的方法用得比较多
页: [1]
查看完整版本: [求助]多自由度振动微分方程的求解！