## Electric potential scaling in case of two temperature electric arc simulation

Questions and remarks about Code_Saturne usage
Forum rules
rodion
Posts: 89
Joined: Wed Jan 11, 2017 4:13 pm

### Electric potential scaling in case of two temperature electric arc simulation

Hello evereone,

I am developing a two temperature model of electric arc in Code_Saturne. In this model I have two thermal scalars: electron enthalpy and heavy species enthalpy. These two scalars have diffferent diffusivity and different source terms. The source term for electrons is Joule power and radiation loss while for heavy species it is the heat transfer from electrons. Hence I have to reach the balance between them, The transport properties (thermal and electrical conductivities, viscosity) are defined by the electrons temperature Te and the ratio of temperatures Te/Th.

When I switched my electric arc model from the LTE simulation with just one thermal scalar to two thermal scalars I found that the arc fades away every time. The total Joule power goes down as well as the voltage. I started looking for a reason and found that in the arc voltage scaling algorithm we have the following formulas for the correction coefficient:

Kcorrection = coepoa = Iimposed/Icomputed = Iimposed*Vimposed/Qtotal_computed

but if coepoa >1.05 then

Kcorrection = coepot = SQRT(20*MIN{(Enthalpy per kg in cell i)/(Joule power per kg in cell i)})

So I corrected it in order to take into account both enthalpies, but the arc still fades away.

Could you please tell me what am I doing wrong with this scaling algorithm and what is the origin of the second formula for the correction coefficient? Probably it must be changed to implement the 2T model? I am a little concerned especially by the number 20 in the formula.

Rodion
Attachments

Yvan Fournier
Posts: 2611
Joined: Mon Feb 20, 2012 3:25 pm

### Re: Electric potential scaling in case of two temperature electric arc simulation

Hello Rodion,

I am not sure i am of much help here, as you are probably more knowledgeable in the electric arcs model than I am. So I can only suggest general troubleshooting recommendations.

I may see a colleague who worked on this module tomorrow, so I can ask for recommendations, but I would be tempted to disable the correction factor altogether (forcing it to 1), or limit it to values near 1 (as I assume it is here to compensate for numerical/rounding issues which should remain small). A first test would be to check (for example print) its value.

I'll keep you informed if I have any news on the "20" factor.

Best regards,

Yvan