I don't know much about this subject at all; I'm only just getting into it. As it turns out, a physicist friend of mine asked me a formulation of the following:
Suppose $M$ is a surface in $\mathbb{R}^3$ and $M_t$ is a variation of $M$ with $M_0=M$. Under Mean Curvature Flow, is it possible for each (or any) of the $M_t$ to be minimal, including $M$?
I looked at some articles and in one by Colding et al, it says that $\nabla Vol(M_t) = H \mathbf{n}$ where $H=k_1+k_2$ (principle curvatures) and $\mathbf{n}$ is a unit normal. Then we have
$$\frac{d}{dt}Vol(M_t) = - \langle\nabla Vol,\nabla Vol \rangle = -\int_{M_t} H^2$$
For minimal surfaces, $H=0$ so it seems that the change in the volume of $M_t$ has to be constant. What if the surface undergoes a different sort of flow? Can all the $M_t$ be minimal surfaces? Does this extend to higher dimensions?
Thanks in advance.