### 【思雨凝风】求助：热变形与流体的双向流固耦合

http://forum.simwe.com/thread-984839-1-1.html

http://forum.simwe.com/thread-934761-1-1.html

http://forum.simwe.com/thread-911031-1-1.html

2011年6月3日

If let me solve this problem. The time integral method definitely is explicit, The PDE for thermal fluid is Advection-Diffusion equation. The shell element has to be used to solve such thin cylinder, note that such thin shell element should contain the thermal expansion term, \alpha*deltaT as the thermal strain. The shell part gorvening equation could be time-dependent or not, but apparently dropping off the time rate term my reduce the computation a lot and improve the stability of whole algorithm. just my $0.02 回复楼上的三位前辈 北鹰南飞前辈 您的思路给了我启示，我有点明白我的问题出在哪里了。 我一直想去搞明白，到底谁在前面，其实这个东西是同步的，没必要去纠结谁先谁后。 wild_field前辈 adina我也研究过，但是放弃了，因为我用cfx和awb已经很久了，再换软件不是很划算。 我们实验室有个用adina的高人，也是做流体的，没有跟进也考虑了避免和她冲突 caoer前辈 您回答问题大部分都是英文，能用到这种地步让人敬佩。 不过您的英文对我来说还是有点难度，需要慢慢读，所以不能马上回复您，见谅。 感谢三位前辈对我的帮助，我已经有点开窍了，呵呵 谢谢，祝三位工作愉快 引用Caoer版主英文原文 I dont know the commercial software for sure. If let me solve this problem. The time integral method definitely is explicit, The PDE for thermal fluid is Advection-Diffusion equation. The shell element has to be used to solve such thin cylinder, note that such thin shell element should contain the thermal expansion term, \alpha*delta T as the thermal strain. The shell part governing equation could be time-dependent or not, but apparently dropping off the time rate term my reduce the computation a lot and improve the stability of whole algorithm. just my$0.02

just my \$0.02

2）建立CFX分析，做流固耦合，确定网格变形和传热耦合；
3）求解，查看结果~ 三场耦合难度较大，计算时间也长很多，况且耦合的精度不是很高

http://www.comsol.com/products/tutorials/introduction/page6/
http://www.cax.cn/comsol/version4/basic-module/CM_Fluid_Valve.html

I dont know the commercial software for sure. If let me solve this problem. Th ...

"Just my two cent" is a slang which means "It is just my opinion" . 楼上是高手，呵呵 不知道实际模型是不是很复杂，如果是这个模型，完全可以单独计算热流耦合与热固耦合，单项迭代几次或者就分开计算也可，精度不会差太多，因为就这个模型固体的热变形对流道影响很小。单独计算精度反而会高。