Thanks for your effort.

What  i want is to solve an unsteady state ode (du/dt=c, say c is a dependent variable in the subdomain next to the boundary) at the boundary of a geometry.

I have written a n weak form in the boundary setting using boundary weak form.

But it is giving an error failed to evaluate variable Jacobian.

Please read this and comment on this.

thanks

Vivek

If I'm getting it right... I think you should do a component coupling operator over that boundary an evaluate the desired variable, then activate the advance physics option and add an ODE, there you can add a name of the variable (u) and define the ODE as ut-c=0

Thanks all for your kind support.

Do you have any idea about  how to write species balance equation in weak form in comsol? with dirichlet boundary condition and indulation boundary condition.

Thanks

Vivek

I am solving Drift diffusion using PDE in comsol but getting negative value of co concentration?

Thanks

Hi Avinash Gupta , if this is occurring only at the beginning then use segregated approach and limit the minimum concentration to 0 (using segregated approach limit node). Make sure other things are correctly defined beforehand.

Thanks, Noman, I am trying what you recommended

No, Noman Iqbal its not working. I tried with what you suggested and even tried one of the recommendation given in this link ( https://www.researchgate.net/post/Why_does_my_COMSOL_model_not_converge ) by putting eps function instead of zero in places where concentration is near to zero. But there's no change reflecting still I am getting negative values in the initiation and while solving. Also, the max doping I have used is 4.5e27 but while solving the value crosses that limit which I guess is creating such negative value of charges. I am attaching the file for you to check.

Sorry Gupta I dont have the required modules. But checking ur model I think you better double check the units there are some concerns. Also first make only two nodes in the segregated approach. First add all variables. And second define like comp1.nox 0. Also for first try keep the default value for event tolerance.

Afterwards if it fails to solve try combining the variables to solve in different nodes. Or better check the theory if it provides u some hints how to solve the variables.

Hope it helps.