在OpenFoam上使用cantera计算甲烷燃烧出现异常错误

NewKid

不好意思，开始把这个帖子发在“燃烧论坛”了，现在才发现有专门的OpenFoam论坛，故重新发过来，请各位帮忙解决一下，谢谢。
大家好，我在OF1.5上使用Cantera计算GRI30甲烷燃烧机理的时候，出现了如下的错误：正常使用并行计算，算到第3步的时候，提示浮点错误，然后我使用unset FOAM_SIGFEP，再计算，还是在第3步出现其他的异常，似乎是调用什么过程的异常。后来我跟踪代码，发现是在thermo->correct()这里出错，为了能够计算下去，我暂时屏蔽了这条语句，能够计算出第4步，第5步又会报错，其部分stack-trace文件内容如下：
Time = 5

Solving chemistry

[CVODES ERROR]  CVode
  At t = 4.5122e-06 and h = 1.71955e-13, the error test failed repeatedly or with |h| = hmin.



************************************************
                   Cantera Error!
************************************************


Procedure: CVodesIntegrator
Error:      CVodes error encountered.


[1]
[1]
[1]  With the state: T:298 p:9.52039e+19  and the mixture
53
(
3.12581e-12
1e-15
1e-15
1e-15
1e-15
9.80655e-13
1e-15
1e-15
1e-15
1e-15
1e-15
1e-15
1e-15
9.40244e-14
1e-15
1e-15
1e-15
7.32012e-13
1e-15
8.59131e-13
2.17475e-13
3.4921e-13
5.2162e-14
1.0809e-13
1.68415e-11
1e-15
1.32435e-11
1e-15
1.6594e-15
1e-15
1e-15
1e-15
1e-15
1e-15
1e-15
1e-15
3.76105e-14
1e-15
1.88184e-13
1e-15
3.42296e-13
1.4961e-13
4.10128e-12
1e-15
1e-15
5.86822e-13
2.85311e-13
1
1e-15
1e-15
1.30254e-14
1e-15
1e-15
)
Cantera complained in cell 0

#0  Foam::error::printStack(Foam::Ostream&) in "/OpenFOAM/OpenFOAM-1.5-dev/lib/linuxGccDPOpt/libOpenFOAM.so"
#1  Foam::error::abort() in "/OpenFOAM/OpenFOAM-1.5-dev/lib/linuxGccDPOpt/libOpenFOAM.so"
#2  Foam::canteraLocalTimeChemistryModel::solve(double, double) in "/home/slax/OpenFOAM/slax-1.5-dev/lib/linuxGccDPOpt/libcanteraThermosChemistry.so"
#3  main in "/home/slax/OpenFOAM/slax-1.5-dev/applications/bin/linuxGccDPOpt/alternateSteadyReactingFoam"
#4  __libc_start_main in "/lib/libc.so.6"
#5  _start in "/home/slax/OpenFOAM/slax-1.5-dev/applications/bin/linuxGccDPOpt/alternateSteadyReactingFoam"
[1]
[1]
[1]     From function canteraLocalTimeChemistryModel::solve
[1]     in file canteraLocalTimeChemistryModel.C at line 127.
[1]
FOAM aborting (FOAM_ABORT set)
[1]
[slax:10054] *** Process received signal ***
[slax:10054] Signal: Aborted (6)
[slax:10054] Signal code:  (-6)
[slax:10054] [ 0] [0xffffe410]
[slax:10054] [ 1] /lib/libc.so.6(abort+0x101) [0xb693b571]
[slax:10054] [ 2] /OpenFOAM/OpenFOAM-1.5-dev/lib/linuxGccDPOpt/libOpenFOAM.so(_ZN4Foam5error5abortEv+0x199) [0xb6cb3e89]
[slax:10054] [ 3] /home/slax/OpenFOAM/slax-1.5-dev/lib/linuxGccDPOpt/libcanteraThermosChemistry.so(_ZN4Foam30canteraLocalTimeChemistryModel5solveEdd+0xd13) [0xacd3c8f3]
[slax:10054] [ 4] alternateSteadyReactingFoam [0x805d29e]
[slax:10054] [ 5] /lib/libc.so.6(__libc_start_main+0xe0) [0xb6925390]
[slax:10054] [ 6] alternateSteadyReactingFoam [0x805a9d1]
[slax:10054] *** End of error message ***
--------------------------------------------------------------------------
mpirun noticed that process rank 1 with PID 10054 on node slax exited on signal 6 (Aborted).
--------------------------------------------------------------------------

这可能是什么引起的？有帖子在说稳态计算容易出这样的问题，尤其是初始条件设定。另外，我使用cantera已经成功计算出氢氧燃烧的算例。