[bram94]

Hello All,

With the help of this useful forum, I was able to do porous media simulations.
I've been trying to do more simulations. I got experimental data for a component. When I did a second-order polynomial curve fit (dP vs V) on this I got a negative inertial coefficient. It didn't really make physical sense to me. What would be some of the ways to check my data and get a better correlation, can I just use the absolute value of the inertial coefficient for my simulation? Thank you.

[fluid23]

It is not unheard of to have this happen. It is a result of the shape of the test data you are fitting to and the fact that you have to force the y intercept to be zero. In a strict physical sense, you are correct.... it doesn't make sense for them to be negative.

However, from a numerical standpoint it shouldn't matter. The goal is to assign coefficients which deliver your experimental pressure drop. I have defined negative values before and got what I believed to be valid results back, so unless your simulation blows up you should be fine.

I wouldn't assign the absolute value of the coefficient, that will change the result of the function. Ask yourself this, is ax^2-bx = ax^2+bx? Only if b=0.

[bram94]

That was a clear insight, thanks for the sharing it!