Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

1. I start with simple differential inequality: find $u\in C^1[0,1]$ such that $u(0) = 0$ and $$ u'(t)\leq -u(t) $$ for all $t\in [0,1]$. Using Gronwall's lemma one can see that $u\leq 0$. On the other hand it seems to be the only solution, since this inequality keep $u$ non-decreasing for $u<0$ and non-increasing for $u>0$. Is it right that only $u=0$ satisfies this inequality?

2. With Gronwall's lemma one can see that any solution of $$ u'(t)\leq\beta(t)u(t) $$ is bounded from above by the solution of $$ u'(t) = \beta(t)u(t). $$

So there are two main results:

(i) there is a solution of $u'\leq \beta u$ which dominates any other solution.

(ii) this solution is attained on the correspondent equation.

Are there any similar results on PD inequalities of the type $$ u_t(t,x)\leq L_x u(t,x) $$ where $L_x$ is a differential operator in $x$ variable (of first or second order).

The main question for me if (i) is valid for such inequalities.

share|cite|improve this question
up vote 5 down vote accepted

Related to your second question: I assume your functions are of the form $u: (a,b) \to L^2(\Omega)$. If you can get a bound on $(Lw, w)_2$ in terms of $\|w\|_2^2$ (as you can typically do by linearity, Holder's Inequality, Green's Identity, etc.) then assuming positivity you can do the following:

$$u_t(t, x) \le Lu(t, x)$$ $$1/2 u(t,x) u_t(t,x) \le 1/2 Lu(t,x) u(t,x)$$

$$\int_{\Omega} 1/2 u(t,x) u_t(t,x) dx \le 1/2(Lu(t), u(t))_2$$

$$d/dt \|u(t)\|_2^2 \le 1/2(Lu, u)_2 \le C\|u(t)\|_2^2$$

and now use Gronwall and the initial condition to get bounds on the $L^2$ norm of the solution. I don't know if that helps.

share|cite|improve this answer

Try $u(t)=\rm{e}^{-t}-1$.........................

share|cite|improve this answer
Actually, $u(t)=-t$ will do. @Gortaur: I think $|u'(t)|\le u(t)$ makes for a more interesting problem -- is this perhaps what you had in mind? – joriki Jun 10 '11 at 10:11
No, I was just confused and forget that sign of $u$ does not influence a sign of inequality. What about the second question? – Ilya Jun 10 '11 at 12:18
@Gortaur: I don't know anything about that, sorry. – joriki Jun 11 '11 at 6:40

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.