11
$\begingroup$

$$\lim_{x\rightarrow 0 } \dfrac{1}{x} \int_{x}^{2x} e^{-t^2}\,\mathrm dt$$

I don't have any idea to solve this integral.

$\endgroup$
  • 6
    $\begingroup$ Welcome to Maths.SE :). This is not an improper integral. Hint: Fundamental Theorem of Calculus. $\endgroup$ – Taladris Dec 14 '15 at 5:29
15
$\begingroup$

Perhaps, an easy way is using l'Hopital's rule:

$$\lim_{x\to 0}\frac{\int_x^{2x}e^{-t^2}\,dt}{x}=\lim_{x\to 0} (2e^{-4x^2}-e^{-x^2})=1$$

How do I differentiate de numerator?

You use the rule

$$g(x)=\int_{a(x)}^{b(x)}f(t)\,dt\implies\\\implies g'(x)=b'(x)f(b(x))-a'(x)f(a(x))$$

which is valid whenever $f$ is continuous on a suitably big interval (say, $\mathbb R$) and $a,b$ are differentiable.

Indeed, $\int_{a(x)}^{b(x)}f(t)\,dt=F(b(x))-F(a(x))$ for a differentiable function such that $F'(x)=f(x)$. And you get the formula above using the chain rule.

$\endgroup$
  • $\begingroup$ How do you differentiate the numerator ? $\endgroup$ – Rising Star Dec 14 '15 at 5:31
  • $\begingroup$ I added it to the answer, @RisingStar. I also found and corrected a small mistake I made. $\endgroup$ – Gae. S. Dec 14 '15 at 5:38
10
$\begingroup$

Hint: Write down the series $1-t^2+\cdots$ for $e^{-t^2}$, integrate term by term, and divide by $x$.

Remark: If your question was about finding an antiderivative of $e^{-t^2}$ as a first step in solving the problem, that will not work. For $e^{-t^2}$ does not have an antiderivative that can be expressed in terms of elementary functions.

$\endgroup$
10
$\begingroup$

This is not an improper integral ($x\to 0$)

Now denote $f(x)=\int_0^xe^{-t^2}dt$

Keeping in mind $f(0)=0$ and $f'(x)=e^{-x^2}$, we're looking for:

$$\begin{align}\lim_{x\to 0}{f(2x)-f(x)\over x}&=2f'(0)-f'(0)\\&=1\end{align}$$

$\endgroup$
4
$\begingroup$

If you don't want to use L'Hôpital's rule, you can rewrite the limit as $$ \lim_{h\rightarrow 0 } \frac{\int_{h}^{2h} e^{-t^2} dt-\int_{0}^{2\cdot0} e^{-t^2} dt}{h} $$ and recognize the derivative of $$ f(x):=\int_{x}^{2x} e^{-t^2} dt $$ at $x=0$, which is (by the chain rule and the Fundamental Theorem of Calculus) $$ f'(0)=2e^{0}-e^0=1 $$

$\endgroup$
3
$\begingroup$

In a neighbourhood of the origin: $$ e^{-t^2} = 1-t^2 + o(t^3) \tag{1}$$ hence: $$ \int_{x}^{2x}e^{-t^2}\,dt = x-\frac{7}{3}x^3+o(x^4) \tag{2}$$ and: $$\frac{1}{x}\int_{x}^{2x}e^{-t^2}\,dt = \color{red}{1}-\frac{7}{3}x^2+o(x^3).\tag{3}$$

$\endgroup$
1
$\begingroup$

Solution based in squeezing: for $x>0$, $$e^{-4x^2} = \frac1x (2x-x)e^{-(2x)^2}\le\dfrac{1}{x}\int_{x}^{2x}e^{-t^2}\, dt\le\frac1x(2x-x) e^{-x^2} = e^{-x^2}.$$ The case $x<0$ is similar.

$\endgroup$
0
$\begingroup$

More generally, for small $x$, if $f$ is differentiable at zero, $f(x) \approx f(0)+xf'(0) $ so

$\begin{array}\\ \int_{ax}^{bx} f(t) dt &\approx \int_{ax}^{bx} (f(0)+tf'(0)) dt\\ &= (tf(0) + t^2f'(0)/2)\mid_{ax}^{bx}\\ &= (x(b-a)f(0) + x^2(b^2-a^2)f'(0)/2)\\ \text{or}\\ \frac1{x}\int_{ax}^{bx} f(t) dt &\approx (b-a)f(0) + x(b^2-a^2)f'(0)/2)\\ &\to (b-a)f(0) \qquad\text{as } x \to 0\\ \end{array} $

If $f(x) = e^{-x^2}$, $a=1, b=2$, then $f(0) = 1$ so the limit is $(2-1)1 = 1 $.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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