| 查看: 1087 | 回复: 1 | |||
[求助]
Fluent UDF 改单机多核
|
|
最近在学习udf的并行，也看了很多资料，仍然不能掌握，所以给出一个例子请求帮助。 以下是一个简单的定义物体x方向上的速度每一个时间步后速度增加0.001，如何改成单机多核，或者说不用改就能在单机多核上运行？ 谢谢！ #include "udf.h" #include "mem.h" #include "math.h" #include "dynamesh_tools.h" static real init_vel=0; DEFINE_CG_MOTION(sway,dt,vel,omega,time,dtime) { FILE *fd; init_vel+=0.001;/*迭代出新的速度*/ vel[0]=init_vel; fd=fopen("vel.dat","a"  ;fprintf(fd,"%f,%f\n", time, vel[0]); fclose(fd); } |
回复此楼» 猜你喜欢
40+中年人失业后，还能找到好工作吗？
已经有11人回复
-
用NBS溴化时溶剂除了CCl4外还有什么
已经有6人回复
-
申请科研项目纯靠实力能成功吗？
已经有11人回复
-
极限拉扯，如何抉择
已经有15人回复
-
二审4个多月了还是Awaiting Reviewer Scores
已经有4人回复
-
OpenAI又攻克数学难题了，离AI做有机合成还有多远？
已经有13人回复
-
课题组会帮你通过考核么？
已经有7人回复
-
售一区SCI文章T0P，我:8O.551.O54,科目全,可十急
已经有9人回复
-
售SCI一区文章，我:8O5.5.1.O5.4,科目全,可伽急
已经有10人回复
-
售SCI一区T0P文章，我:8.O55.1.O.54,科目全,可十急
已经有8人回复
|
补充一点： 我的电脑是单机4核，用的15.0，采用的串行，编译上述udf得到的速度更新结果如下：它并不是每一个时间步速度增加，而是每个时间步内速度已经叠加16次，并且我也尝试过和每一步的迭代次数无关。 不知道问题出在哪里了，请有经验的朋友指点！感谢！ timestep 速度 0.001000,0.001 0.001000,0.002 0.001000,0.003 0.001000,0.004 0.001000,0.005 0.001000,0.006 0.001000,0.007 0.001000,0.008 0.001000,0.009 0.001000,0.010 0.001000,0.011 0.001000,0.012 0.001000,0.013 0.001000,0.014 0.001000,0.015 0.002000,0.016 0.002000,0.017 0.002000,0.018 0.002000,0.019 0.002000,0.020 0.002000,0.021 0.002000,0.022 0.002000,0.023 0.002000,0.024 0.002000,0.025 0.002000,0.026 0.002000,0.027 0.002000,0.028 0.002000,0.029 0.002000,0.030 0.002000,0.031 0.002000,0.032 |
2楼2015-11-14 17:23:16
10