[FLASH-USERS] extreme "sod"

[John Wohlbier]

Thanks to those who suggested the time step fix. I should have seen that.

Anyway, it's interesting that if you set dtinit but not dtmin the time 
step jumps up again after one cycle resulting in another crash. I 
suppose the default value of "dtmin" is something larger than I need.

Thanks again.

jgw


On 04/25/2011 02:32 PM, Sean Couch wrote:
> Hi John,
>
> Try setting dtinit=1.e-15 and dtmin=1.e-15.  Also, since there is no initial density jump, make sure you are refining on pressure: refine_var_2 = "pres".
>
> Hope this helps,
> Sean
>
> On Apr 25, 2011, at 2:57 PM, John Wohlbier wrote:
>
>> I'm trying to solve a 1-d planar "sod" problem with a very large pressure jump (motivated by the desire to do Sedov where the initial pressure grows without bound with decreasing grid size), but I'm getting a non convergence of the Riemann solver. I was wondering if anyone wanted to try their hand at a "Sod" problem with:
>>
>> gamma = 1.66667
>> pL = 1e19
>> pR = 1e-8
>> rhoL = rhoR = 1.0
>>
>>
>> Below is some of my output. Something has obviously gone wonky in the initial step. I would like to know if this is user error. If anyone wants input files, etc. let me know. I based this off the Sod that comes with the code, but changed the initialization for 1d.
>>
>> jgw
>>
>>
>> Initial dt verified
>> *** Wrote checkpoint file to sod_hdf5_chk_0000 ****
>> *** Wrote plotfile to sod_hdf5_plt_cnt_0000 ****
>> Initial plotfile written
>> Driver init all done
>>        n          t         dt |  dt_Hydro
>>        1 2.1404E-13 1.0000E-10 |  1.04255E-12
>> *** Wrote checkpoint file to sod_hdf5_chk_0001 ****
>>        2 2.0021E-10 1.0000E-10 |  6.68035E-37
>> *** Wrote checkpoint file to sod_hdf5_chk_0002 ****
>>
>> Nonconvergence in subroutine rieman
>>
>> Zone index       =           53
>> Zone center      =   0.48499999999999999
>> Iterations tried =           12
>> Pressure error   =   2.50736728133274093E-005
>> rieman_tol       =   1.00000000000000008E-005
>>
>> pL       =   1.00000000000000004E-010  pR       =  1.69945124795326129E+0
>> uL       =  -8.55361906563791716E+020  uR       = -2.95619597114146315E+0
>> cL       =   1.29112354172635239E-010  cR       =   53225794783526600.
>> gamma_eL =    1.6670000000000000       gamma_eR =   1.6670000000000000
>> gamma_cL =    1.6670000000000000       gamma_cR =   1.6670000000000000
>>
>> Iteration history:
>>
>>    n                    p*
>>    1    0.381681833473E+23
>>    2    0.666931951757E+39
>>    3    0.849680290662E+47
>>    4    0.490680234160E+51
>>    5    0.383003596149E+53
>>    6    0.367516282470E+54
>>    7    0.125015589879E+55
>>    8    0.226538136461E+55
>>    9    0.279461332702E+55
>>   10    0.290630268330E+55
>>   11    0.291332257483E+55
>>   12    0.291339562436E+55
>>
>>
>>
>> <wohlbier.vcf>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: wohlbier.vcf
Type: text/x-vcard
Size: 360 bytes
Desc: not available
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20110425/694b35da/attachment.vcf>