@李东岳 在 请教各位如何用tecplot360(EX2022R1)处理OpenFOAMv2312中颗粒分布 中说：

还真是。我刚才看了一下。应该是我的虚拟机上面没有装tecplot，所以没找到tec的头文件。

dyfluid@dyfluid-virtual-machine:~/OpenFOAM/OpenFOAM-8/applications/utilities/postProcessing/dataConversion/foamToTecplot360$ wmake wmakeLnIncludeAll: running wmakeLnInclude on dependent libraries: wmakeLnInclude error: base directory /home/dyfluid/OpenFOAM/ThirdParty-8/tecio/tecsrc/ does not exist Making dependency list for source file foamToTecplot360.C could not open file TECIO.h for source file foamToTecplot360.C due to No such file or directory could not open file MASTER.h for source file foamToTecplot360.C due to No such file or directory could not open file GLOBAL.h for source file foamToTecplot360.C due to No such file or directory Making dependency list for source file vtkMesh.C Making dependency list for source file tecplotWriter.C could not open file TECIO.h for source file tecplotWriter.C due to No such file or directory could not open file MASTER.h for source file tecplotWriter.C due to No such file or directory could not open file GLOBAL.h for source file tecplotWriter.C due to No such file or directory g++ -std=c++11 -m64 -Dlinux64 -DWM_ARCH_OPTION=64 -DWM_DP -DWM_LABEL_SIZE=32 -Wall -Wextra -Wold-style-cast -Wnon-virtual-dtor -Wno-unused-parameter -Wno-invalid-offsetof -Wno-attributes -O3 -DNoRepository -ftemplate-depth-100 -I/home/dyfluid/OpenFOAM/ThirdParty-8/tecio/tecsrc/lnInclude -I/home/dyfluid/OpenFOAM/OpenFOAM-8/src/lagrangian/basic/lnInclude -I/home/dyfluid/OpenFOAM/OpenFOAM-8/src/finiteVolume/lnInclude -I/home/dyfluid/OpenFOAM/OpenFOAM-8/src/dynamicMesh/lnInclude -I/home/dyfluid/OpenFOAM/OpenFOAM-8/src/meshTools/lnInclude -IlnInclude -I. -I/home/dyfluid/OpenFOAM/OpenFOAM-8/src/OpenFOAM/lnInclude -I/home/dyfluid/OpenFOAM/OpenFOAM-8/src/OSspecific/POSIX/lnInclude -fPIC -c tecplotWriter.C -o /home/dyfluid/OpenFOAM/OpenFOAM-8/platforms/linux64GccDPInt32Opt/applications/utilities/postProcessing/dataConversion/foamToTecplot360/tecplotWriter.o In file included from tecplotWriter.C:26: tecplotWriter.H:39:10: fatal error: TECIO.h: 没有那个文件或目录 39 | #include "TECIO.h" | ^~~~~~~~~ compilation terminated. make: *** [/home/dyfluid/OpenFOAM/OpenFOAM-8/wmake/rules/General/transform:26：/home/dyfluid/OpenFOAM/OpenFOAM-8/platforms/linux64GccDPInt32Opt/applications/utilities/postProcessing/dataConversion/foamToTecplot360/tecplotWriter.o] 错误 1 dyfluid@dyfluid-virtual-machine:~/OpenFOAM/OpenFOAM-8/applications/utilities/posttProcessing/tProcessing/dataCotProttProcessing/dataConversion/foamToTecplot360$

应该确实是，我看了一下时间，我openfoam-8装的时候电脑上已经有装tecplot2023了。但openfoam-7装的时候没有，也许就是编译记录的Log文件里就没有foamToTec的原因。

@dyb 我的理解哈，这个视频我看过的。这里其实就是常规的合成湍流入口，只不过作者给它套了个风资源里面常见的名词（阵风），很有可能它根本就不是我们经常理解的阵风的概念。

并没有把阵风加到湍流里面这种说法，阵风本来就是风流大幅变化的一种叫法而已。既然做LES，时间尺度肯定已经远小于阵风所在的秒级尺度了。

@李东岳 好的好的，谢谢李老师

torch.set_flush_denormal(True)

https://discuss.pytorch.org/t/training-time-gets-slower-and-slower-on-cpu/145483/4

我在计算固体冲压发动机内两相流颗粒二次燃烧时，C颗粒燃烧采用openfoam自带的模型，B颗粒的燃烧自己编写的表面燃烧模型，但计算时总是在空气和颗粒接触不久之后发散。MaxCo为1时，自适应时间步长大概在2e-6s左右，多次尝试后存在以下几种情况：
（1）两种颗粒的Cloud都启用表面燃烧时，不管如何调整species的Cp等物性参数，均会在0.0068s时出现Energy -> temperature conversion failed to converge，输出的报错信息显示，温度在1000±1K之间来回振荡；
（2）两种颗粒的Cloud都启用表面燃烧，但B颗粒所有反应的速率都设为0时，计算在0.0116s发散，从报错信息来看，应该是ReactingMultiphaseParcel中的calc函数调用后，MomentumParcel的move()函数计算出错；
（3）C颗粒启用表面燃烧模型，B颗粒的表面燃烧模型选择none，此时计算不会发散。

对于情况（1），在将thermo.C文件中的tol_从1e-4改为2e-4后能继续往后计算，到0.0073s时还是发散了。但情况（2）反应速率均为0的情况下，在我理解中应该跟情况（3）是等价的，不知道为什么一个报错，一个却能顺利计算。

@李东岳 明白了，谢谢老师

@李东岳
谢谢老师，一开始我认为是入口的速度梯度计算的不准确，所以用差分方法直接计算。后面对照下看来是接近的。
我认为可能是不可压缩的算法里不考虑严格的能量守恒，所以fb4e0ad7-cdcb-4682-ab71-5ada18146f9d-image.png 直接计算并不为0.
但是上面公式推导中使用高斯方法转换体面积分的时候默认了不可压缩时这项为零。应该写出完整的形式计算0ceb3ad3-73d3-4c9e-81b8-f3f473dfcbbb-image.png

Total viscous power in the domain (volume integral): -1.50451907 Total integrated tau : gradU: 1.48509576 Total integrated viscous power (surface): 0.00384099619

这样计算后数值是接近的，等稳定后误差应该还会减小。

@李东岳 testSimplefoam.zip 谢谢老师，我这边上传了，分别是internalField为边界条件和为0的两个0文件，然后我用的是singleGraph3监测x=2m位置处的速度和湍动能信息。网格数量不大。

@xuebao1989 大概率是网格问题，你就用最常用的数值格式（比如参考商业软件的默认设置），先把网格优化下。

谢谢老师，太牛逼了。

挺好，这玩意之前我也遇到，一直没法解决

收藏一下，感觉有用

@李东岳 好的李老师，感谢感谢

@五好青年 老哥，我也会有这个问题，单核正常，并行发散，这是不是与分配核数有关呀？

@李东岳 是可以不选择的，但怕设置错了，这玩意用的人确实不多啊，老师还有啥推荐的网格软件吗

VIV.zip 这是我配置的openFoam动网格算VIV的算例，从目前的结果来看，程序能正常运行，但程序运行结束以后，圆柱不动，想请各位大佬帮我看看，哪儿出问题了，谢谢大家。

@疏影横斜水清浅 重力以及体积力

@sgggggg 硕士论文我也想拜读一下，可以加个联系方式吗2132306535

@李东岳 请问这可能是是网格太小导致吗？

各位大佬好，我是CFDEM新手，最近尝试参照twoSpheresGlowinskiMPI构建自己的算例，采用dynamicRefine，在计算过程中断构建的算例基本与参照算例一致，但试跑算例出现中断，不知道是否是哪里没有注意到，求大佬指导
报错描述
当不使用自适应网格时，算例正常计算；当采用自适应网格功能interDyMFoam，网格未加密前（未达到设置加密条件），正常计算;网格开始加密计算中断。
报错代码如下：

Time = 0.0001 Selected 0 cells for refinement out of 1613716. Selected 0 split points out of a possible 0. Courant Number mean: 0.000199582 max: 0.0804916 - evolve() timeStepFraction() = 2 Starting up LIGGGHTS Executing command: 'run 10 ' Setting up run at Fri Jul 26 12:59:28 2024 Memory usage per processor = 6.75634 Mbytes Step Atoms KinEng rke Volume dragtota dragtota dragtota 1 1 2.5261073e-08 0 0.0260508 0 0 0 CFD Coupling established at step 10 11 1 3.5918126e-08 0 0.0260508 0 0 0 Loop time of 0.000962037 on 16 procs for 10 steps with 1 atoms, finish time Fri Jul 26 12:59:28 2024 Pair time (%) = 4.70437e-06 (0.489001) Neigh time (%) = 0 (0) Comm time (%) = 4.65044e-06 (0.483395) Outpt time (%) = 6.52643e-05 (6.78397) Other time (%) = 0.000887418 (92.2436) Nlocal: 0.0625 ave 1 max 0 min Histogram: 15 0 0 0 0 0 0 0 0 1 Nghost: 0 ave 0 max 0 min Histogram: 16 0 0 0 0 0 0 0 0 0 Neighs: 0 ave 0 max 0 min Histogram: 16 0 0 0 0 0 0 0 0 0 Total # of neighbors = 0 Ave neighs/atom = 0 Neighbor list builds = 0 Dangerous builds = 0 LIGGGHTS finished Foam::cfdemCloudIB::reAllocArrays() nr particles = 1 evolve done. DILUPBiCG: Solving for Ux, Initial residual = 1, Final residual = 1.06984e-07, No Iterations 1 DILUPBiCG: Solving for Uy, Initial residual = 1, Final residual = 4.99282e-07, No Iterations 1 DILUPBiCG: Solving for Uz, Initial residual = 1, Final residual = 6.95219e-10, No Iterations 2 DICPCG: Solving for p, Initial residual = 1, Final residual = 9.26695e-07, No Iterations 573 time step continuity errors : sum local = 3.69904e-10, global = -1.28591e-13, cumulative = -1.28591e-13 DICPCG: Solving for p, Initial residual = 0.0476412, Final residual = 9.26405e-07, No Iterations 513 time step continuity errors : sum local = 3.16922e-08, global = -3.49546e-12, cumulative = -3.62406e-12 DICPCG: Solving for p, Initial residual = 0.0055667, Final residual = 9.51869e-07, No Iterations 500 time step continuity errors : sum local = 3.3706e-08, global = -6.10801e-12, cumulative = -9.73206e-12 DICPCG: Solving for p, Initial residual = 0.00178549, Final residual = 9.94251e-07, No Iterations 473 time step continuity errors : sum local = 3.53268e-08, global = -5.71629e-12, cumulative = -1.54484e-11 No finite volume options present particleCloud.calcVelocityCorrection() DICPCG: Solving for phiIB, Initial residual = 1, Final residual = 9.63924e-07, No Iterations 501 ExecutionTime = 30.7 s ClockTime = 32 s Time = 0.0002 Selected 784 cells for refinement out of 1613716. Refined from 1613716 to 1619204 cells. Selected 0 split points out of a possible 784. Courant Number mean: 0.126921 max: 9.46381 - evolve() timeStepFraction() = 2 Starting up LIGGGHTS Executing command: 'run 10 ' Setting up run at Fri Jul 26 12:59:56 2024 Memory usage per processor = 6.75634 Mbytes Step Atoms KinEng rke Volume dragtota dragtota dragtota 11 1 3.5918126e-08 0 0.0260508 0 0 0.017347056 CFD Coupling established at step 20 21 1 3.6972707e-08 0 0.0260508 0 0 0.017347056 Loop time of 0.00149142 on 16 procs for 10 steps with 1 atoms, finish time Fri Jul 26 12:59:56 2024 Pair time (%) = 3.75399e-06 (0.251706) Neigh time (%) = 0 (0) Comm time (%) = 3.65471e-06 (0.24505) Outpt time (%) = 8.7527e-05 (5.86871) Other time (%) = 0.00139648 (93.6345) Nlocal: 0.0625 ave 1 max 0 min Histogram: 15 0 0 0 0 0 0 0 0 1 Nghost: 0 ave 0 max 0 min Histogram: 16 0 0 0 0 0 0 0 0 0 Neighs: 0 ave 0 max 0 min Histogram: 16 0 0 0 0 0 0 0 0 0 Total # of neighbors = 0 Ave neighs/atom = 0 Neighbor list builds = 0 Dangerous builds = 0 LIGGGHTS finished nr particles = 1 -------------------------------------------------------------------------- MPI_ABORT was invoked on rank 15 in communicator MPI_COMM_WORLD with errorcode 1. NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes. You may or may not see output from other processes, depending on exactly when Open MPI kills them. --------------------------------------------------------------------------