It is easy to model in COMSOL. At first non-dimensionalise your governing equations. Then find the coefficient in place of rho, mu, k, c_p etc and use it in the respective place. You will get non-dimensional solutions.
Defining a set of normalizing constants and ignoring any unit warnings that Comsol generates is sufficient to do non-dimensional analysis. But a piece of advice, the non-dimensional approach, especially if you have coupled physics, may lead you to unreasonable results. There are numerous threads on the Comsol forum discussing this, just google : non dimensional comsol, and go through them before proceeding.
Defining a set of normalizing constants and ignoring any unit warnings that Comsol generates is sufficient to do non-dimensional analysis. But a piece of advice, the non-dimensional approach, especially if you have coupled physics, may lead you to unreasonable results. There are numerous threads on the Comsol forum discussing this, just google : non dimensional comsol, and go through them before proceeding.
This is to S C Saha. If N-S equation is non dimensionalized, we will put rho=1 and eta= 1/Re. right? Re= DVrho/eta (with actual value of properties)
This is to George. Why there will be a unreasonable result? Because we are solving alll physics in no dimensional form with the equations coupled through non dimensional variale.
I could not understand your advice! The good think for COMSOL is, it is an equation solver unlike ANSYS. In Ansys we can not see the equations we are solving. But in COMSOL you can see all equations you are solving. If you have the correct non-dimensional equations with proper boundary conditions I don't see any problem.
Send me the dimensionless NS equations you want to solve (as different people use different dimensionless form). I will advise you how to choose parameters.
Define all parameters with values first. If we divide two momentum equations by Re then you have to use rho = 1/(Re*Ca), eta = 1/Re, body force = -1/(Re*Ca). Rests just use 1.
Just as a follow-up, it is now possible to define dimensionless Geometry and Physics (including boundary and initial conditions) in newer versions of Comsol.
In the component, unit system, simply choose "None". I hope this helps other researcher...