We are all taught the derivation of the mass conservation using a fixed Eulerian control volume in a typical fluid dynamics course.
That is, first we think about the total rate of change of mass in that control volume. $$ \frac{d}{dt}\int_V\rho dV $$ Next, we say that in the absence of sinks or sources, the fluid that enters this volume contributes to the increase. The expression is : $$ -\int_{\partial V}\rho\vec{u}\cdot \hat{n}dA $$
Now these expressions are equal, $$ \frac{d}{dt}\int_V\rho dV = -\int_{\partial V}\rho\vec{u}\cdot \hat{n}dA $$
to give us the integral form of the mass equation.
Now, to get the differential form, we use the gauss-divergence theorem and do some strange manipulation to the total derivative on the LHS :
$$ \int_{V}\left[\frac{\partial\rho}{\partial t} + \nabla\cdot(\rho \vec{u})\right]dV = 0 $$
So, my question is how did the substantial derivative for density turn into a partial derivative?
I understand that if you consider the total mass of the volume, it's essentially changing only with time. i.e. $$ \frac{dM}{dT} = \frac{\partial M}{\partial T} $$
where M is the mass of the control volume.
But how do you explain the use material derivative with density above?
Shouldn't this be true :
$$ \frac{d}{dt}\int_V\rho dV = \int_V\frac{d\rho}{dt}dV = \int_V\left[\frac{\partial\rho}{\partial t} + \vec{u}\cdot\nabla\rho\right]dV $$
Can this only be explained by $$\vec{u}\cdot\nabla\rho=0$$?
I figured it out. Understand that the substantial derivative exists for only field variables i.e. functions that vary with position vector x and time t.
Thus, $$ \frac{DM}{DT} $$
doesn't really make sense.
For a fixed control volume the mass can only vary with time.