Skip to main content
MathOverflow

Return to Question

added 124 characters in body
Source Link
Pace
Pace
  • 51
  • 3

Let $u$ be a weak solution (i.e. $u \in C([0,T];L^1(\Omega))$ of some degenerate or nondegenerate parabolic equation $u' - Au = f$ on a bounded domain. For(For my purpose it is enough to have this for the porous medium equation, in which also $u^m \in L^2(0,T;H^1(\Omega))$.)

I am seeking a estimate of the form $$\lVert u(t+h)-u(t) \rVert_{L^1(\Omega)} \leq \frac{Ch}{t}$$ for $h >0$ small (want to send $h$ to zero later). Here $C$ is some constant not depending on $h$, $t$ or $u$. Note that the power of $h$ must be one, not a half.

This type of estimate is useful in order to show time regularity.

My question is, what techniques are available and can one use to in order to derive such an estimate?My question is, what techniques are available and can one use to in order to derive such an estimate, apart from the one I describe below?

I know of one way, which is detailed in the book on PME by Vazquez. This way comes out of rescaling the solution so that it satisfies the PME again with a rescaled initial data and then using the comparison principle after picking the scaling parameter appropriately.

Is there some other method? I'd prefer variational methods and not semigroup.

Why I want to do this? Because I work on a similar such equation where the rescaling I described above doesn't work so I am looking for different methods.

Let $u$ be a weak solution of some degenerate or nondegenerate parabolic equation $u' - Au = f$ on a bounded domain. For my purpose it is enough to have this for the porous medium equation.

I am seeking a estimate of the form $$\lVert u(t+h)-u(t) \rVert_{L^1(\Omega)} \leq \frac{Ch}{t}$$ for $h >0$ small (want to send $h$ to zero later). Here $C$ is some constant not depending on $h$, $t$ or $u$. Note that the power of $h$ must be one, not a half.

This type of estimate is useful in order to show time regularity.

My question is, what techniques are available and can one use to in order to derive such an estimate?

I know of one way, which is detailed in the book on PME by Vazquez. This way comes out of rescaling the solution so that it satisfies the PME again with a rescaled initial data and then using the comparison principle after picking the scaling parameter appropriately.

Is there some other method? I'd prefer variational methods and not semigroup.

Why I want to do this? Because I work on a similar such equation where the rescaling I described above doesn't work so I am looking for different methods.

Let $u$ be a weak solution (i.e. $u \in C([0,T];L^1(\Omega))$ of some degenerate or nondegenerate parabolic equation $u' - Au = f$ on a bounded domain. (For my purpose it is enough to have this for the porous medium equation, in which also $u^m \in L^2(0,T;H^1(\Omega))$.)

I am seeking a estimate of the form $$\lVert u(t+h)-u(t) \rVert_{L^1(\Omega)} \leq \frac{Ch}{t}$$ for $h >0$ small (want to send $h$ to zero later). Here $C$ is some constant not depending on $h$, $t$ or $u$. Note that the power of $h$ must be one, not a half.

This type of estimate is useful in order to show time regularity.

My question is, what techniques are available and can one use to in order to derive such an estimate, apart from the one I describe below?

I know of one way, which is detailed in the book on PME by Vazquez. This way comes out of rescaling the solution so that it satisfies the PME again with a rescaled initial data and then using the comparison principle after picking the scaling parameter appropriately.

Is there some other method? I'd prefer variational methods and not semigroup.

Why I want to do this? Because I work on a similar such equation where the rescaling I described above doesn't work so I am looking for different methods.

Source Link
Pace
Pace
  • 51
  • 3

Getting an estimate of the form $\lVert u(t+h)-u(t) \rVert_{L^1(\Omega)} \leq \frac{Ch}{t}$ on solution of PDE

Let $u$ be a weak solution of some degenerate or nondegenerate parabolic equation $u' - Au = f$ on a bounded domain. For my purpose it is enough to have this for the porous medium equation.

I am seeking a estimate of the form $$\lVert u(t+h)-u(t) \rVert_{L^1(\Omega)} \leq \frac{Ch}{t}$$ for $h >0$ small (want to send $h$ to zero later). Here $C$ is some constant not depending on $h$, $t$ or $u$. Note that the power of $h$ must be one, not a half.

This type of estimate is useful in order to show time regularity.

My question is, what techniques are available and can one use to in order to derive such an estimate?

I know of one way, which is detailed in the book on PME by Vazquez. This way comes out of rescaling the solution so that it satisfies the PME again with a rescaled initial data and then using the comparison principle after picking the scaling parameter appropriately.

Is there some other method? I'd prefer variational methods and not semigroup.

Why I want to do this? Because I work on a similar such equation where the rescaling I described above doesn't work so I am looking for different methods.