[FLASH-USERS] Error of physical units in the eos_idealGamma implementation

Dean Townsley Dean.M.Townsley at ua.edu
Tue Feb 16 13:37:42 EST 2021


Hi Gabriel,

Interesting.  I'll admit to not a lot of familiarity with how the unit 
conversion stuff in flash is typically used, so I'm quite unsure how the 
"substance" dimension is treated.  (I'll admit to just doing everything 
in cgs myself.)

I still think the text in the user's guide is the thing that might need 
to change or be added to.  Making (Na k) have dimensions of 
energy/mass-temperature rather than energy/temperature seems like it is 
moving the wrong direction.

Dean

On 2/15/21 10:33 PM, RIGON Gabriel wrote:
> Hi,
>
> Actually there is a quick fix to this problem, but I'm not sure it is 
> 100% compatible with all of FLASH units.
> In the case of EosMain/Gamma, the parameter use for (Na k) is the 
> ideal gas constant define in PhysicalConstant_init.F90. Since this 
> parameter is always divided by A in this EoS implementation, we can 
> just change its units in PhysicialConstant_init going from : len 2, 
> time -2, mass 1, temp -1, charge 0, substance -1 to len 2, time -2, 
> mass 0, temp -1, charge 0, substance 0.
> But one should make sure this change won't negatively impact the other 
> units.
>
> Best,
> Gabriel
>
> On 16/02/2021 08:12, Dean Townsley wrote:
>>
>> I expect that rather than changing the implementation, many would 
>> rather a note be added to the user guide that if a non-cgs unit 
>> system is chosen A is the molar mass, not the atomic mass.
>>
>> Dean
>>
>> On 2/15/21 10:47 AM, Klaus Weide wrote:
>>> On Mon, 15 Feb 2021, RIGON Gabriel wrote:
>>>
>>>> The problem here is the actual definition of A in the code.
>>>> A is defined using the eos_singleSpeciesA variable which is 
>>>> initialized in
>>>> rp_initParameters to a dimensionless constant "1". The problem is 
>>>> the absence
>>>> of dimension.
>>> The "MKS" unit system in FLASH is not as much used (and not as well
>>> tested) as the default "cgs" unit system.  It would not be too 
>>> surprising
>>> to find problems in other physics units when using "MKS" mode.
>>>
>>> Klaus
>



More information about the flash-users mailing list