How to integrate $int_{0}^{2pi}frac{mathrm dx}{(1-acos x)^2}$












4












$begingroup$


I'm having a hard time trying to resolve this integral :
$$int_{0}^{2pi}frac{mathrm dx}{(1-acos x)^2}$$
where $a$ is a positive real constant.



I tried using substitution, but I'm stuck by the fact that the integral must be computed between $0$ and $2pi$, which leads to integration between $X$ and $X$ (where $X$ can be $0$ or $1$ anything following the substitution in the boundaries).
I'm not looking for the full result if the computation is complicated, just some lead to start off...



Thanks



UPDATE : seeing the answers proposed, I checked that indeed $|a|<1$.










share|cite|improve this question











$endgroup$








  • 2




    $begingroup$
    Do you realize it essentially is the area enclosed by an ellipse?
    $endgroup$
    – Jack D'Aurizio
    Jan 16 at 17:24






  • 2




    $begingroup$
    Similar : math.stackexchange.com/q/1846774/321264.
    $endgroup$
    – StubbornAtom
    Jan 16 at 18:01






  • 2




    $begingroup$
    You can start with the standard result $$int_{0}^{2pi}frac{dx}{a-bcos x} =frac{2pi}{sqrt{a^2-b^2}}$$ and differentiate it with respect to $a$ and then put $a=1$ and replace $b$ by $a$.
    $endgroup$
    – Paramanand Singh
    Jan 17 at 5:24
















4












$begingroup$


I'm having a hard time trying to resolve this integral :
$$int_{0}^{2pi}frac{mathrm dx}{(1-acos x)^2}$$
where $a$ is a positive real constant.



I tried using substitution, but I'm stuck by the fact that the integral must be computed between $0$ and $2pi$, which leads to integration between $X$ and $X$ (where $X$ can be $0$ or $1$ anything following the substitution in the boundaries).
I'm not looking for the full result if the computation is complicated, just some lead to start off...



Thanks



UPDATE : seeing the answers proposed, I checked that indeed $|a|<1$.










share|cite|improve this question











$endgroup$








  • 2




    $begingroup$
    Do you realize it essentially is the area enclosed by an ellipse?
    $endgroup$
    – Jack D'Aurizio
    Jan 16 at 17:24






  • 2




    $begingroup$
    Similar : math.stackexchange.com/q/1846774/321264.
    $endgroup$
    – StubbornAtom
    Jan 16 at 18:01






  • 2




    $begingroup$
    You can start with the standard result $$int_{0}^{2pi}frac{dx}{a-bcos x} =frac{2pi}{sqrt{a^2-b^2}}$$ and differentiate it with respect to $a$ and then put $a=1$ and replace $b$ by $a$.
    $endgroup$
    – Paramanand Singh
    Jan 17 at 5:24














4












4








4





$begingroup$


I'm having a hard time trying to resolve this integral :
$$int_{0}^{2pi}frac{mathrm dx}{(1-acos x)^2}$$
where $a$ is a positive real constant.



I tried using substitution, but I'm stuck by the fact that the integral must be computed between $0$ and $2pi$, which leads to integration between $X$ and $X$ (where $X$ can be $0$ or $1$ anything following the substitution in the boundaries).
I'm not looking for the full result if the computation is complicated, just some lead to start off...



Thanks



UPDATE : seeing the answers proposed, I checked that indeed $|a|<1$.










share|cite|improve this question











$endgroup$




I'm having a hard time trying to resolve this integral :
$$int_{0}^{2pi}frac{mathrm dx}{(1-acos x)^2}$$
where $a$ is a positive real constant.



I tried using substitution, but I'm stuck by the fact that the integral must be computed between $0$ and $2pi$, which leads to integration between $X$ and $X$ (where $X$ can be $0$ or $1$ anything following the substitution in the boundaries).
I'm not looking for the full result if the computation is complicated, just some lead to start off...



Thanks



UPDATE : seeing the answers proposed, I checked that indeed $|a|<1$.







calculus integration analysis trigonometric-integrals






share|cite|improve this question















share|cite|improve this question













share|cite|improve this question




share|cite|improve this question








edited Jan 16 at 20:29







mocquin

















asked Jan 16 at 17:22









mocquinmocquin

213




213








  • 2




    $begingroup$
    Do you realize it essentially is the area enclosed by an ellipse?
    $endgroup$
    – Jack D'Aurizio
    Jan 16 at 17:24






  • 2




    $begingroup$
    Similar : math.stackexchange.com/q/1846774/321264.
    $endgroup$
    – StubbornAtom
    Jan 16 at 18:01






  • 2




    $begingroup$
    You can start with the standard result $$int_{0}^{2pi}frac{dx}{a-bcos x} =frac{2pi}{sqrt{a^2-b^2}}$$ and differentiate it with respect to $a$ and then put $a=1$ and replace $b$ by $a$.
    $endgroup$
    – Paramanand Singh
    Jan 17 at 5:24














  • 2




    $begingroup$
    Do you realize it essentially is the area enclosed by an ellipse?
    $endgroup$
    – Jack D'Aurizio
    Jan 16 at 17:24






  • 2




    $begingroup$
    Similar : math.stackexchange.com/q/1846774/321264.
    $endgroup$
    – StubbornAtom
    Jan 16 at 18:01






  • 2




    $begingroup$
    You can start with the standard result $$int_{0}^{2pi}frac{dx}{a-bcos x} =frac{2pi}{sqrt{a^2-b^2}}$$ and differentiate it with respect to $a$ and then put $a=1$ and replace $b$ by $a$.
    $endgroup$
    – Paramanand Singh
    Jan 17 at 5:24








2




2




$begingroup$
Do you realize it essentially is the area enclosed by an ellipse?
$endgroup$
– Jack D'Aurizio
Jan 16 at 17:24




$begingroup$
Do you realize it essentially is the area enclosed by an ellipse?
$endgroup$
– Jack D'Aurizio
Jan 16 at 17:24




2




2




$begingroup$
Similar : math.stackexchange.com/q/1846774/321264.
$endgroup$
– StubbornAtom
Jan 16 at 18:01




$begingroup$
Similar : math.stackexchange.com/q/1846774/321264.
$endgroup$
– StubbornAtom
Jan 16 at 18:01




2




2




$begingroup$
You can start with the standard result $$int_{0}^{2pi}frac{dx}{a-bcos x} =frac{2pi}{sqrt{a^2-b^2}}$$ and differentiate it with respect to $a$ and then put $a=1$ and replace $b$ by $a$.
$endgroup$
– Paramanand Singh
Jan 17 at 5:24




$begingroup$
You can start with the standard result $$int_{0}^{2pi}frac{dx}{a-bcos x} =frac{2pi}{sqrt{a^2-b^2}}$$ and differentiate it with respect to $a$ and then put $a=1$ and replace $b$ by $a$.
$endgroup$
– Paramanand Singh
Jan 17 at 5:24










3 Answers
3






active

oldest

votes


















8












$begingroup$

The polar equation of an ellipse (with respect to a focus, taking the aphelion as the point associated to $theta=0$) is
$$ rho(theta) = frac{ell}{1-ecostheta} $$
where, in terms of the standard parameters, $ell$ is the semi-latus rectum $frac{b^2}{a}$ and $e$ is the eccentricity $frac{c}{a}$.

The area enclosed by such ellipse is $pi a b $ or
$$ frac{1}{2}int_{0}^{2pi}frac{ell^2}{(1-ecostheta)^2},dtheta, $$
so we get that for any $ein[0,1)$ the following identity holds:
$$ int_{0}^{2pi}frac{dtheta}{(1-ecostheta)^2}= frac{2pi a b}{ell^2}=2pileft(frac{a}{b}right)^3=frac{2pi}{left(frac{b^2}{a^2}right)^{3/2}}={frac{2pi}{(1-e^2)^{3/2}}}. $$
The same holds if, in the LHS, we replace $e$ with $-e$, since $ rho(theta) = frac{ell}{1+ecostheta} $ is the polar equation of the same ellipse, with the perihelion being taken as the point associated to $theta=0$. So we get that for any $ain(-1,1)$
$$ int_{0}^{2pi}frac{dtheta}{(1-acostheta)^2}= color{blue}{frac{2pi}{(1-a^2)^{3/2}}}. $$






share|cite|improve this answer









$endgroup$













  • $begingroup$
    Thanks a lot for your answer, I did not realized this was the polar equation of an ellipse! This solution looks quite elegant. I also update my post, the coefficient $a$ is indeed always less than 1.
    $endgroup$
    – mocquin
    Jan 16 at 20:33






  • 1




    $begingroup$
    +1 for polar equation of ellipse. Although I believe computation of area of ellipse is easier if one uses rectangular coordinates. So in effect your answer uses the equivalence/transformation between these coordinate systems.
    $endgroup$
    – Paramanand Singh
    Jan 17 at 5:15



















3












$begingroup$

HINT:



Note that for $|a|<1$



$$begin{align}
int_0^{2pi}frac{1}{(1-acos(theta))^2},dtheta&=frac1{a^2}int_0^{2pi}frac{1}{(1/a-cos(theta))^2},dtheta\\
&=-frac1{a^2}frac{d}{d(1/a)}int_0^{2pi}frac{1}{1/a-cos(theta)},dthetatag1\\
&= -frac2{a^2}frac{d}{d(1/a)}int_0^{pi}frac{1}{1/a-cos(theta)},dthetatag2
end{align}$$



Now use the Weierstrass substitution to evaluate the integral on the right-hand side of $(2)$ and differentiate to obtain the coveted result.



Alternatively, one could use contour integration to evaluate the integral of interest or evaluate the integral on the right-hand side of $(1)$.






share|cite|improve this answer











$endgroup$













  • $begingroup$
    The Weierstrass substitution is going to fail if you apply it to this integral. You have to do something first (integrate on $[-pi,pi]$ and split, for instance).
    $endgroup$
    – Jean-Claude Arbaut
    Jan 16 at 17:51










  • $begingroup$
    @Jean-ClaudeArbaut This was meant to be a HINT only. But I've added another line to facilitate.
    $endgroup$
    – Mark Viola
    Jan 16 at 19:52



















1












$begingroup$

My goal with this answer is to give you a bunch of different really general integral formulas that you may find very useful.



$$F(a)=int_0^{pi}frac{mathrm dx}{(1+acos x)^2}$$
So your integral is $2F(-a)$. First off we will preform the tangent-half-angle substitution $t=tan(x/2)$. Hence we have that
$$F(a)=2int_0^infty frac1{(1+afrac{1-t^2}{1+t^2})^2}frac{mathrm dt}{1+t^2}$$
$$F(a)=2int_0^{infty}frac{1+t^2}{[(1-a)t^2+a+1]^2}mathrm dt$$
Which we can rewrite as
$$F(a)=frac2{1-a}int_0^inftyfrac{mathrm dt}{(1-a)t^2+a+1}+frac{4a}{a-1}int_0^{infty}frac{mathrm dt}{[(1-a)t^2+a+1]^2}$$





Next we consider the very general integral
$$I(m;a,b)=int_0^infty frac{mathrm dx}{(ax^2+b)^{m+1}}$$
First we integrate by parts with $mathrm dv=mathrm dx$ to produce
$$I(m;a,b)=frac{x}{(ax^2+b)^{m+1}}bigg|_0^{infty}+2(m+1)int_0^inftyfrac{ax^2}{(ax^2+b)^{m+2}}mathrm dx$$
$$I(m;a,b)=2(m+1)int_0^inftyfrac{ax^2+b-b}{(ax^2+b)^{m+2}}mathrm dx$$
$$I(m;a,b)=2(m+1)I(m;a,b)-2(m+1)bI(m+1;a,b)$$
Then solving for $I(m+1;a,b)$ and replacing $m+1$ with $m$,
$$I(m;a,b)=frac{2m-1}{2bm}I(m-1;a,b)$$
And for the base case:
$$I(0;a,b)=int_0^{infty}frac{mathrm dx}{ax^2+b}$$
The Trig sub $x=sqrt{frac{b}a}tan u$ gives
$$I(0;a,b)=fracpi{2sqrt{ab}}$$
Which is a special case of
$$int_{x_1}^{x_2}frac{mathrm dx}{ax^2+bx+c}=frac2{sqrt{4ac-b^2}}arctanfrac{2ax+b}{sqrt{4ac-b^2}}bigg|_{x_1}^{x_2}$$
Anyway, the recurrence has the solution
$$I(m;a,b)=fracpi{2^{2m+1}b^msqrt{ab}}{2mchoose m}$$
Which we will apply very shortly!





Recalling the definition of $I(m;a,b)$, we have that
$$F(a)=frac2{1-a}I(0;1-a,1+a)+frac{4a}{a-1}I(1;1-a,1+a)$$
$$F(a)=fracpi{(1-a^2)^{3/2}}$$
And since your integral is given by $2F(-a)$,
$$int_0^{2pi}frac{mathrm dx}{(1-acos x)^2}=frac{2pi}{(1-a^2)^{3/2}}$$





ADDENDUM



Consider the integral
$$C(n;a)=int_0^pi frac{mathrm dx}{(1+acos x)^n}$$
Starting with $t=tanfrac{x}2$,
$$C(n;a)=2int_0^infty frac1{left[1+afrac{1-t^2}{1+t^2}right]^n}frac{mathrm dt}{1+t^2}$$
$$C(n;a)=2int_0^infty frac{left(t^2+1right)^{n-1}}{left[(1-a)t^2+1+aright]^n}mathrm dt$$
Then using the binomial theorem,
$$C(n;a)=2sum_{k=0}^{n-1}{n-1choose k}int_0^inftyfrac{x^{2k}}{left[(1-a)x^2+1+aright]^n}mathrm dx$$
$$C(n;a)=frac2{(1-a)^n}sum_{k=0}^{n-1}{n-1choose k}int_0^inftyfrac{x^{2k}}{left[x^2+frac{1+a}{1-a}right]^n}mathrm dx$$
We then recall the integral due to my collaborator @DavidG,
$$int_0^inftyfrac{x^{q}}{left[x^w+bright]^p}mathrm dx=frac{b^{frac{1+q}{w}-p}}{w}frac{Gammaleft(p-frac{1+q}{w}right)Gammaleft(frac{1+q}{w}right)}{Gammaleft(pright)}$$
Here $Gamma(s)$ is the Gamma function. So with $w=2$, $b=frac{1+a}{1-a}$, $p=n$, and $q=2k$,
$$C(n;a)=frac1{(1-a)^n}sum_{k=0}^{n-1}{n-1choose k}left(frac{1+a}{1-a}right)^{frac{2k+1}{2}-n}frac{Gammaleft(n-frac{2k+1}{2}right)Gammaleft(frac{2k+1}{2}right)}{Gammaleft(nright)}$$
$$C(n;a)=frac1{(1-a)^n}sum_{k=0}^{n-1}left(frac{1+a}{1-a}right)^{frac{2k+1}{2}-n}frac{Gammaleft(n-frac{2k+1}{2}right)Gammaleft(frac{2k+1}{2}right)}{k!Gamma(n-k)}$$






share|cite|improve this answer











$endgroup$














    Your Answer





    StackExchange.ifUsing("editor", function () {
    return StackExchange.using("mathjaxEditing", function () {
    StackExchange.MarkdownEditor.creationCallbacks.add(function (editor, postfix) {
    StackExchange.mathjaxEditing.prepareWmdForMathJax(editor, postfix, [["$", "$"], ["\\(","\\)"]]);
    });
    });
    }, "mathjax-editing");

    StackExchange.ready(function() {
    var channelOptions = {
    tags: "".split(" "),
    id: "69"
    };
    initTagRenderer("".split(" "), "".split(" "), channelOptions);

    StackExchange.using("externalEditor", function() {
    // Have to fire editor after snippets, if snippets enabled
    if (StackExchange.settings.snippets.snippetsEnabled) {
    StackExchange.using("snippets", function() {
    createEditor();
    });
    }
    else {
    createEditor();
    }
    });

    function createEditor() {
    StackExchange.prepareEditor({
    heartbeatType: 'answer',
    autoActivateHeartbeat: false,
    convertImagesToLinks: true,
    noModals: true,
    showLowRepImageUploadWarning: true,
    reputationToPostImages: 10,
    bindNavPrevention: true,
    postfix: "",
    imageUploader: {
    brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
    contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
    allowUrls: true
    },
    noCode: true, onDemand: true,
    discardSelector: ".discard-answer"
    ,immediatelyShowMarkdownHelp:true
    });


    }
    });














    draft saved

    draft discarded


















    StackExchange.ready(
    function () {
    StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f3076005%2fhow-to-integrate-int-02-pi-frac-mathrm-dx1-a-cos-x2%23new-answer', 'question_page');
    }
    );

    Post as a guest















    Required, but never shown

























    3 Answers
    3






    active

    oldest

    votes








    3 Answers
    3






    active

    oldest

    votes









    active

    oldest

    votes






    active

    oldest

    votes









    8












    $begingroup$

    The polar equation of an ellipse (with respect to a focus, taking the aphelion as the point associated to $theta=0$) is
    $$ rho(theta) = frac{ell}{1-ecostheta} $$
    where, in terms of the standard parameters, $ell$ is the semi-latus rectum $frac{b^2}{a}$ and $e$ is the eccentricity $frac{c}{a}$.

    The area enclosed by such ellipse is $pi a b $ or
    $$ frac{1}{2}int_{0}^{2pi}frac{ell^2}{(1-ecostheta)^2},dtheta, $$
    so we get that for any $ein[0,1)$ the following identity holds:
    $$ int_{0}^{2pi}frac{dtheta}{(1-ecostheta)^2}= frac{2pi a b}{ell^2}=2pileft(frac{a}{b}right)^3=frac{2pi}{left(frac{b^2}{a^2}right)^{3/2}}={frac{2pi}{(1-e^2)^{3/2}}}. $$
    The same holds if, in the LHS, we replace $e$ with $-e$, since $ rho(theta) = frac{ell}{1+ecostheta} $ is the polar equation of the same ellipse, with the perihelion being taken as the point associated to $theta=0$. So we get that for any $ain(-1,1)$
    $$ int_{0}^{2pi}frac{dtheta}{(1-acostheta)^2}= color{blue}{frac{2pi}{(1-a^2)^{3/2}}}. $$






    share|cite|improve this answer









    $endgroup$













    • $begingroup$
      Thanks a lot for your answer, I did not realized this was the polar equation of an ellipse! This solution looks quite elegant. I also update my post, the coefficient $a$ is indeed always less than 1.
      $endgroup$
      – mocquin
      Jan 16 at 20:33






    • 1




      $begingroup$
      +1 for polar equation of ellipse. Although I believe computation of area of ellipse is easier if one uses rectangular coordinates. So in effect your answer uses the equivalence/transformation between these coordinate systems.
      $endgroup$
      – Paramanand Singh
      Jan 17 at 5:15
















    8












    $begingroup$

    The polar equation of an ellipse (with respect to a focus, taking the aphelion as the point associated to $theta=0$) is
    $$ rho(theta) = frac{ell}{1-ecostheta} $$
    where, in terms of the standard parameters, $ell$ is the semi-latus rectum $frac{b^2}{a}$ and $e$ is the eccentricity $frac{c}{a}$.

    The area enclosed by such ellipse is $pi a b $ or
    $$ frac{1}{2}int_{0}^{2pi}frac{ell^2}{(1-ecostheta)^2},dtheta, $$
    so we get that for any $ein[0,1)$ the following identity holds:
    $$ int_{0}^{2pi}frac{dtheta}{(1-ecostheta)^2}= frac{2pi a b}{ell^2}=2pileft(frac{a}{b}right)^3=frac{2pi}{left(frac{b^2}{a^2}right)^{3/2}}={frac{2pi}{(1-e^2)^{3/2}}}. $$
    The same holds if, in the LHS, we replace $e$ with $-e$, since $ rho(theta) = frac{ell}{1+ecostheta} $ is the polar equation of the same ellipse, with the perihelion being taken as the point associated to $theta=0$. So we get that for any $ain(-1,1)$
    $$ int_{0}^{2pi}frac{dtheta}{(1-acostheta)^2}= color{blue}{frac{2pi}{(1-a^2)^{3/2}}}. $$






    share|cite|improve this answer









    $endgroup$













    • $begingroup$
      Thanks a lot for your answer, I did not realized this was the polar equation of an ellipse! This solution looks quite elegant. I also update my post, the coefficient $a$ is indeed always less than 1.
      $endgroup$
      – mocquin
      Jan 16 at 20:33






    • 1




      $begingroup$
      +1 for polar equation of ellipse. Although I believe computation of area of ellipse is easier if one uses rectangular coordinates. So in effect your answer uses the equivalence/transformation between these coordinate systems.
      $endgroup$
      – Paramanand Singh
      Jan 17 at 5:15














    8












    8








    8





    $begingroup$

    The polar equation of an ellipse (with respect to a focus, taking the aphelion as the point associated to $theta=0$) is
    $$ rho(theta) = frac{ell}{1-ecostheta} $$
    where, in terms of the standard parameters, $ell$ is the semi-latus rectum $frac{b^2}{a}$ and $e$ is the eccentricity $frac{c}{a}$.

    The area enclosed by such ellipse is $pi a b $ or
    $$ frac{1}{2}int_{0}^{2pi}frac{ell^2}{(1-ecostheta)^2},dtheta, $$
    so we get that for any $ein[0,1)$ the following identity holds:
    $$ int_{0}^{2pi}frac{dtheta}{(1-ecostheta)^2}= frac{2pi a b}{ell^2}=2pileft(frac{a}{b}right)^3=frac{2pi}{left(frac{b^2}{a^2}right)^{3/2}}={frac{2pi}{(1-e^2)^{3/2}}}. $$
    The same holds if, in the LHS, we replace $e$ with $-e$, since $ rho(theta) = frac{ell}{1+ecostheta} $ is the polar equation of the same ellipse, with the perihelion being taken as the point associated to $theta=0$. So we get that for any $ain(-1,1)$
    $$ int_{0}^{2pi}frac{dtheta}{(1-acostheta)^2}= color{blue}{frac{2pi}{(1-a^2)^{3/2}}}. $$






    share|cite|improve this answer









    $endgroup$



    The polar equation of an ellipse (with respect to a focus, taking the aphelion as the point associated to $theta=0$) is
    $$ rho(theta) = frac{ell}{1-ecostheta} $$
    where, in terms of the standard parameters, $ell$ is the semi-latus rectum $frac{b^2}{a}$ and $e$ is the eccentricity $frac{c}{a}$.

    The area enclosed by such ellipse is $pi a b $ or
    $$ frac{1}{2}int_{0}^{2pi}frac{ell^2}{(1-ecostheta)^2},dtheta, $$
    so we get that for any $ein[0,1)$ the following identity holds:
    $$ int_{0}^{2pi}frac{dtheta}{(1-ecostheta)^2}= frac{2pi a b}{ell^2}=2pileft(frac{a}{b}right)^3=frac{2pi}{left(frac{b^2}{a^2}right)^{3/2}}={frac{2pi}{(1-e^2)^{3/2}}}. $$
    The same holds if, in the LHS, we replace $e$ with $-e$, since $ rho(theta) = frac{ell}{1+ecostheta} $ is the polar equation of the same ellipse, with the perihelion being taken as the point associated to $theta=0$. So we get that for any $ain(-1,1)$
    $$ int_{0}^{2pi}frac{dtheta}{(1-acostheta)^2}= color{blue}{frac{2pi}{(1-a^2)^{3/2}}}. $$







    share|cite|improve this answer












    share|cite|improve this answer



    share|cite|improve this answer










    answered Jan 16 at 17:31









    Jack D'AurizioJack D'Aurizio

    292k33284673




    292k33284673












    • $begingroup$
      Thanks a lot for your answer, I did not realized this was the polar equation of an ellipse! This solution looks quite elegant. I also update my post, the coefficient $a$ is indeed always less than 1.
      $endgroup$
      – mocquin
      Jan 16 at 20:33






    • 1




      $begingroup$
      +1 for polar equation of ellipse. Although I believe computation of area of ellipse is easier if one uses rectangular coordinates. So in effect your answer uses the equivalence/transformation between these coordinate systems.
      $endgroup$
      – Paramanand Singh
      Jan 17 at 5:15


















    • $begingroup$
      Thanks a lot for your answer, I did not realized this was the polar equation of an ellipse! This solution looks quite elegant. I also update my post, the coefficient $a$ is indeed always less than 1.
      $endgroup$
      – mocquin
      Jan 16 at 20:33






    • 1




      $begingroup$
      +1 for polar equation of ellipse. Although I believe computation of area of ellipse is easier if one uses rectangular coordinates. So in effect your answer uses the equivalence/transformation between these coordinate systems.
      $endgroup$
      – Paramanand Singh
      Jan 17 at 5:15
















    $begingroup$
    Thanks a lot for your answer, I did not realized this was the polar equation of an ellipse! This solution looks quite elegant. I also update my post, the coefficient $a$ is indeed always less than 1.
    $endgroup$
    – mocquin
    Jan 16 at 20:33




    $begingroup$
    Thanks a lot for your answer, I did not realized this was the polar equation of an ellipse! This solution looks quite elegant. I also update my post, the coefficient $a$ is indeed always less than 1.
    $endgroup$
    – mocquin
    Jan 16 at 20:33




    1




    1




    $begingroup$
    +1 for polar equation of ellipse. Although I believe computation of area of ellipse is easier if one uses rectangular coordinates. So in effect your answer uses the equivalence/transformation between these coordinate systems.
    $endgroup$
    – Paramanand Singh
    Jan 17 at 5:15




    $begingroup$
    +1 for polar equation of ellipse. Although I believe computation of area of ellipse is easier if one uses rectangular coordinates. So in effect your answer uses the equivalence/transformation between these coordinate systems.
    $endgroup$
    – Paramanand Singh
    Jan 17 at 5:15











    3












    $begingroup$

    HINT:



    Note that for $|a|<1$



    $$begin{align}
    int_0^{2pi}frac{1}{(1-acos(theta))^2},dtheta&=frac1{a^2}int_0^{2pi}frac{1}{(1/a-cos(theta))^2},dtheta\\
    &=-frac1{a^2}frac{d}{d(1/a)}int_0^{2pi}frac{1}{1/a-cos(theta)},dthetatag1\\
    &= -frac2{a^2}frac{d}{d(1/a)}int_0^{pi}frac{1}{1/a-cos(theta)},dthetatag2
    end{align}$$



    Now use the Weierstrass substitution to evaluate the integral on the right-hand side of $(2)$ and differentiate to obtain the coveted result.



    Alternatively, one could use contour integration to evaluate the integral of interest or evaluate the integral on the right-hand side of $(1)$.






    share|cite|improve this answer











    $endgroup$













    • $begingroup$
      The Weierstrass substitution is going to fail if you apply it to this integral. You have to do something first (integrate on $[-pi,pi]$ and split, for instance).
      $endgroup$
      – Jean-Claude Arbaut
      Jan 16 at 17:51










    • $begingroup$
      @Jean-ClaudeArbaut This was meant to be a HINT only. But I've added another line to facilitate.
      $endgroup$
      – Mark Viola
      Jan 16 at 19:52
















    3












    $begingroup$

    HINT:



    Note that for $|a|<1$



    $$begin{align}
    int_0^{2pi}frac{1}{(1-acos(theta))^2},dtheta&=frac1{a^2}int_0^{2pi}frac{1}{(1/a-cos(theta))^2},dtheta\\
    &=-frac1{a^2}frac{d}{d(1/a)}int_0^{2pi}frac{1}{1/a-cos(theta)},dthetatag1\\
    &= -frac2{a^2}frac{d}{d(1/a)}int_0^{pi}frac{1}{1/a-cos(theta)},dthetatag2
    end{align}$$



    Now use the Weierstrass substitution to evaluate the integral on the right-hand side of $(2)$ and differentiate to obtain the coveted result.



    Alternatively, one could use contour integration to evaluate the integral of interest or evaluate the integral on the right-hand side of $(1)$.






    share|cite|improve this answer











    $endgroup$













    • $begingroup$
      The Weierstrass substitution is going to fail if you apply it to this integral. You have to do something first (integrate on $[-pi,pi]$ and split, for instance).
      $endgroup$
      – Jean-Claude Arbaut
      Jan 16 at 17:51










    • $begingroup$
      @Jean-ClaudeArbaut This was meant to be a HINT only. But I've added another line to facilitate.
      $endgroup$
      – Mark Viola
      Jan 16 at 19:52














    3












    3








    3





    $begingroup$

    HINT:



    Note that for $|a|<1$



    $$begin{align}
    int_0^{2pi}frac{1}{(1-acos(theta))^2},dtheta&=frac1{a^2}int_0^{2pi}frac{1}{(1/a-cos(theta))^2},dtheta\\
    &=-frac1{a^2}frac{d}{d(1/a)}int_0^{2pi}frac{1}{1/a-cos(theta)},dthetatag1\\
    &= -frac2{a^2}frac{d}{d(1/a)}int_0^{pi}frac{1}{1/a-cos(theta)},dthetatag2
    end{align}$$



    Now use the Weierstrass substitution to evaluate the integral on the right-hand side of $(2)$ and differentiate to obtain the coveted result.



    Alternatively, one could use contour integration to evaluate the integral of interest or evaluate the integral on the right-hand side of $(1)$.






    share|cite|improve this answer











    $endgroup$



    HINT:



    Note that for $|a|<1$



    $$begin{align}
    int_0^{2pi}frac{1}{(1-acos(theta))^2},dtheta&=frac1{a^2}int_0^{2pi}frac{1}{(1/a-cos(theta))^2},dtheta\\
    &=-frac1{a^2}frac{d}{d(1/a)}int_0^{2pi}frac{1}{1/a-cos(theta)},dthetatag1\\
    &= -frac2{a^2}frac{d}{d(1/a)}int_0^{pi}frac{1}{1/a-cos(theta)},dthetatag2
    end{align}$$



    Now use the Weierstrass substitution to evaluate the integral on the right-hand side of $(2)$ and differentiate to obtain the coveted result.



    Alternatively, one could use contour integration to evaluate the integral of interest or evaluate the integral on the right-hand side of $(1)$.







    share|cite|improve this answer














    share|cite|improve this answer



    share|cite|improve this answer








    edited Jan 16 at 19:51

























    answered Jan 16 at 17:33









    Mark ViolaMark Viola

    134k1278177




    134k1278177












    • $begingroup$
      The Weierstrass substitution is going to fail if you apply it to this integral. You have to do something first (integrate on $[-pi,pi]$ and split, for instance).
      $endgroup$
      – Jean-Claude Arbaut
      Jan 16 at 17:51










    • $begingroup$
      @Jean-ClaudeArbaut This was meant to be a HINT only. But I've added another line to facilitate.
      $endgroup$
      – Mark Viola
      Jan 16 at 19:52


















    • $begingroup$
      The Weierstrass substitution is going to fail if you apply it to this integral. You have to do something first (integrate on $[-pi,pi]$ and split, for instance).
      $endgroup$
      – Jean-Claude Arbaut
      Jan 16 at 17:51










    • $begingroup$
      @Jean-ClaudeArbaut This was meant to be a HINT only. But I've added another line to facilitate.
      $endgroup$
      – Mark Viola
      Jan 16 at 19:52
















    $begingroup$
    The Weierstrass substitution is going to fail if you apply it to this integral. You have to do something first (integrate on $[-pi,pi]$ and split, for instance).
    $endgroup$
    – Jean-Claude Arbaut
    Jan 16 at 17:51




    $begingroup$
    The Weierstrass substitution is going to fail if you apply it to this integral. You have to do something first (integrate on $[-pi,pi]$ and split, for instance).
    $endgroup$
    – Jean-Claude Arbaut
    Jan 16 at 17:51












    $begingroup$
    @Jean-ClaudeArbaut This was meant to be a HINT only. But I've added another line to facilitate.
    $endgroup$
    – Mark Viola
    Jan 16 at 19:52




    $begingroup$
    @Jean-ClaudeArbaut This was meant to be a HINT only. But I've added another line to facilitate.
    $endgroup$
    – Mark Viola
    Jan 16 at 19:52











    1












    $begingroup$

    My goal with this answer is to give you a bunch of different really general integral formulas that you may find very useful.



    $$F(a)=int_0^{pi}frac{mathrm dx}{(1+acos x)^2}$$
    So your integral is $2F(-a)$. First off we will preform the tangent-half-angle substitution $t=tan(x/2)$. Hence we have that
    $$F(a)=2int_0^infty frac1{(1+afrac{1-t^2}{1+t^2})^2}frac{mathrm dt}{1+t^2}$$
    $$F(a)=2int_0^{infty}frac{1+t^2}{[(1-a)t^2+a+1]^2}mathrm dt$$
    Which we can rewrite as
    $$F(a)=frac2{1-a}int_0^inftyfrac{mathrm dt}{(1-a)t^2+a+1}+frac{4a}{a-1}int_0^{infty}frac{mathrm dt}{[(1-a)t^2+a+1]^2}$$





    Next we consider the very general integral
    $$I(m;a,b)=int_0^infty frac{mathrm dx}{(ax^2+b)^{m+1}}$$
    First we integrate by parts with $mathrm dv=mathrm dx$ to produce
    $$I(m;a,b)=frac{x}{(ax^2+b)^{m+1}}bigg|_0^{infty}+2(m+1)int_0^inftyfrac{ax^2}{(ax^2+b)^{m+2}}mathrm dx$$
    $$I(m;a,b)=2(m+1)int_0^inftyfrac{ax^2+b-b}{(ax^2+b)^{m+2}}mathrm dx$$
    $$I(m;a,b)=2(m+1)I(m;a,b)-2(m+1)bI(m+1;a,b)$$
    Then solving for $I(m+1;a,b)$ and replacing $m+1$ with $m$,
    $$I(m;a,b)=frac{2m-1}{2bm}I(m-1;a,b)$$
    And for the base case:
    $$I(0;a,b)=int_0^{infty}frac{mathrm dx}{ax^2+b}$$
    The Trig sub $x=sqrt{frac{b}a}tan u$ gives
    $$I(0;a,b)=fracpi{2sqrt{ab}}$$
    Which is a special case of
    $$int_{x_1}^{x_2}frac{mathrm dx}{ax^2+bx+c}=frac2{sqrt{4ac-b^2}}arctanfrac{2ax+b}{sqrt{4ac-b^2}}bigg|_{x_1}^{x_2}$$
    Anyway, the recurrence has the solution
    $$I(m;a,b)=fracpi{2^{2m+1}b^msqrt{ab}}{2mchoose m}$$
    Which we will apply very shortly!





    Recalling the definition of $I(m;a,b)$, we have that
    $$F(a)=frac2{1-a}I(0;1-a,1+a)+frac{4a}{a-1}I(1;1-a,1+a)$$
    $$F(a)=fracpi{(1-a^2)^{3/2}}$$
    And since your integral is given by $2F(-a)$,
    $$int_0^{2pi}frac{mathrm dx}{(1-acos x)^2}=frac{2pi}{(1-a^2)^{3/2}}$$





    ADDENDUM



    Consider the integral
    $$C(n;a)=int_0^pi frac{mathrm dx}{(1+acos x)^n}$$
    Starting with $t=tanfrac{x}2$,
    $$C(n;a)=2int_0^infty frac1{left[1+afrac{1-t^2}{1+t^2}right]^n}frac{mathrm dt}{1+t^2}$$
    $$C(n;a)=2int_0^infty frac{left(t^2+1right)^{n-1}}{left[(1-a)t^2+1+aright]^n}mathrm dt$$
    Then using the binomial theorem,
    $$C(n;a)=2sum_{k=0}^{n-1}{n-1choose k}int_0^inftyfrac{x^{2k}}{left[(1-a)x^2+1+aright]^n}mathrm dx$$
    $$C(n;a)=frac2{(1-a)^n}sum_{k=0}^{n-1}{n-1choose k}int_0^inftyfrac{x^{2k}}{left[x^2+frac{1+a}{1-a}right]^n}mathrm dx$$
    We then recall the integral due to my collaborator @DavidG,
    $$int_0^inftyfrac{x^{q}}{left[x^w+bright]^p}mathrm dx=frac{b^{frac{1+q}{w}-p}}{w}frac{Gammaleft(p-frac{1+q}{w}right)Gammaleft(frac{1+q}{w}right)}{Gammaleft(pright)}$$
    Here $Gamma(s)$ is the Gamma function. So with $w=2$, $b=frac{1+a}{1-a}$, $p=n$, and $q=2k$,
    $$C(n;a)=frac1{(1-a)^n}sum_{k=0}^{n-1}{n-1choose k}left(frac{1+a}{1-a}right)^{frac{2k+1}{2}-n}frac{Gammaleft(n-frac{2k+1}{2}right)Gammaleft(frac{2k+1}{2}right)}{Gammaleft(nright)}$$
    $$C(n;a)=frac1{(1-a)^n}sum_{k=0}^{n-1}left(frac{1+a}{1-a}right)^{frac{2k+1}{2}-n}frac{Gammaleft(n-frac{2k+1}{2}right)Gammaleft(frac{2k+1}{2}right)}{k!Gamma(n-k)}$$






    share|cite|improve this answer











    $endgroup$


















      1












      $begingroup$

      My goal with this answer is to give you a bunch of different really general integral formulas that you may find very useful.



      $$F(a)=int_0^{pi}frac{mathrm dx}{(1+acos x)^2}$$
      So your integral is $2F(-a)$. First off we will preform the tangent-half-angle substitution $t=tan(x/2)$. Hence we have that
      $$F(a)=2int_0^infty frac1{(1+afrac{1-t^2}{1+t^2})^2}frac{mathrm dt}{1+t^2}$$
      $$F(a)=2int_0^{infty}frac{1+t^2}{[(1-a)t^2+a+1]^2}mathrm dt$$
      Which we can rewrite as
      $$F(a)=frac2{1-a}int_0^inftyfrac{mathrm dt}{(1-a)t^2+a+1}+frac{4a}{a-1}int_0^{infty}frac{mathrm dt}{[(1-a)t^2+a+1]^2}$$





      Next we consider the very general integral
      $$I(m;a,b)=int_0^infty frac{mathrm dx}{(ax^2+b)^{m+1}}$$
      First we integrate by parts with $mathrm dv=mathrm dx$ to produce
      $$I(m;a,b)=frac{x}{(ax^2+b)^{m+1}}bigg|_0^{infty}+2(m+1)int_0^inftyfrac{ax^2}{(ax^2+b)^{m+2}}mathrm dx$$
      $$I(m;a,b)=2(m+1)int_0^inftyfrac{ax^2+b-b}{(ax^2+b)^{m+2}}mathrm dx$$
      $$I(m;a,b)=2(m+1)I(m;a,b)-2(m+1)bI(m+1;a,b)$$
      Then solving for $I(m+1;a,b)$ and replacing $m+1$ with $m$,
      $$I(m;a,b)=frac{2m-1}{2bm}I(m-1;a,b)$$
      And for the base case:
      $$I(0;a,b)=int_0^{infty}frac{mathrm dx}{ax^2+b}$$
      The Trig sub $x=sqrt{frac{b}a}tan u$ gives
      $$I(0;a,b)=fracpi{2sqrt{ab}}$$
      Which is a special case of
      $$int_{x_1}^{x_2}frac{mathrm dx}{ax^2+bx+c}=frac2{sqrt{4ac-b^2}}arctanfrac{2ax+b}{sqrt{4ac-b^2}}bigg|_{x_1}^{x_2}$$
      Anyway, the recurrence has the solution
      $$I(m;a,b)=fracpi{2^{2m+1}b^msqrt{ab}}{2mchoose m}$$
      Which we will apply very shortly!





      Recalling the definition of $I(m;a,b)$, we have that
      $$F(a)=frac2{1-a}I(0;1-a,1+a)+frac{4a}{a-1}I(1;1-a,1+a)$$
      $$F(a)=fracpi{(1-a^2)^{3/2}}$$
      And since your integral is given by $2F(-a)$,
      $$int_0^{2pi}frac{mathrm dx}{(1-acos x)^2}=frac{2pi}{(1-a^2)^{3/2}}$$





      ADDENDUM



      Consider the integral
      $$C(n;a)=int_0^pi frac{mathrm dx}{(1+acos x)^n}$$
      Starting with $t=tanfrac{x}2$,
      $$C(n;a)=2int_0^infty frac1{left[1+afrac{1-t^2}{1+t^2}right]^n}frac{mathrm dt}{1+t^2}$$
      $$C(n;a)=2int_0^infty frac{left(t^2+1right)^{n-1}}{left[(1-a)t^2+1+aright]^n}mathrm dt$$
      Then using the binomial theorem,
      $$C(n;a)=2sum_{k=0}^{n-1}{n-1choose k}int_0^inftyfrac{x^{2k}}{left[(1-a)x^2+1+aright]^n}mathrm dx$$
      $$C(n;a)=frac2{(1-a)^n}sum_{k=0}^{n-1}{n-1choose k}int_0^inftyfrac{x^{2k}}{left[x^2+frac{1+a}{1-a}right]^n}mathrm dx$$
      We then recall the integral due to my collaborator @DavidG,
      $$int_0^inftyfrac{x^{q}}{left[x^w+bright]^p}mathrm dx=frac{b^{frac{1+q}{w}-p}}{w}frac{Gammaleft(p-frac{1+q}{w}right)Gammaleft(frac{1+q}{w}right)}{Gammaleft(pright)}$$
      Here $Gamma(s)$ is the Gamma function. So with $w=2$, $b=frac{1+a}{1-a}$, $p=n$, and $q=2k$,
      $$C(n;a)=frac1{(1-a)^n}sum_{k=0}^{n-1}{n-1choose k}left(frac{1+a}{1-a}right)^{frac{2k+1}{2}-n}frac{Gammaleft(n-frac{2k+1}{2}right)Gammaleft(frac{2k+1}{2}right)}{Gammaleft(nright)}$$
      $$C(n;a)=frac1{(1-a)^n}sum_{k=0}^{n-1}left(frac{1+a}{1-a}right)^{frac{2k+1}{2}-n}frac{Gammaleft(n-frac{2k+1}{2}right)Gammaleft(frac{2k+1}{2}right)}{k!Gamma(n-k)}$$






      share|cite|improve this answer











      $endgroup$
















        1












        1








        1





        $begingroup$

        My goal with this answer is to give you a bunch of different really general integral formulas that you may find very useful.



        $$F(a)=int_0^{pi}frac{mathrm dx}{(1+acos x)^2}$$
        So your integral is $2F(-a)$. First off we will preform the tangent-half-angle substitution $t=tan(x/2)$. Hence we have that
        $$F(a)=2int_0^infty frac1{(1+afrac{1-t^2}{1+t^2})^2}frac{mathrm dt}{1+t^2}$$
        $$F(a)=2int_0^{infty}frac{1+t^2}{[(1-a)t^2+a+1]^2}mathrm dt$$
        Which we can rewrite as
        $$F(a)=frac2{1-a}int_0^inftyfrac{mathrm dt}{(1-a)t^2+a+1}+frac{4a}{a-1}int_0^{infty}frac{mathrm dt}{[(1-a)t^2+a+1]^2}$$





        Next we consider the very general integral
        $$I(m;a,b)=int_0^infty frac{mathrm dx}{(ax^2+b)^{m+1}}$$
        First we integrate by parts with $mathrm dv=mathrm dx$ to produce
        $$I(m;a,b)=frac{x}{(ax^2+b)^{m+1}}bigg|_0^{infty}+2(m+1)int_0^inftyfrac{ax^2}{(ax^2+b)^{m+2}}mathrm dx$$
        $$I(m;a,b)=2(m+1)int_0^inftyfrac{ax^2+b-b}{(ax^2+b)^{m+2}}mathrm dx$$
        $$I(m;a,b)=2(m+1)I(m;a,b)-2(m+1)bI(m+1;a,b)$$
        Then solving for $I(m+1;a,b)$ and replacing $m+1$ with $m$,
        $$I(m;a,b)=frac{2m-1}{2bm}I(m-1;a,b)$$
        And for the base case:
        $$I(0;a,b)=int_0^{infty}frac{mathrm dx}{ax^2+b}$$
        The Trig sub $x=sqrt{frac{b}a}tan u$ gives
        $$I(0;a,b)=fracpi{2sqrt{ab}}$$
        Which is a special case of
        $$int_{x_1}^{x_2}frac{mathrm dx}{ax^2+bx+c}=frac2{sqrt{4ac-b^2}}arctanfrac{2ax+b}{sqrt{4ac-b^2}}bigg|_{x_1}^{x_2}$$
        Anyway, the recurrence has the solution
        $$I(m;a,b)=fracpi{2^{2m+1}b^msqrt{ab}}{2mchoose m}$$
        Which we will apply very shortly!





        Recalling the definition of $I(m;a,b)$, we have that
        $$F(a)=frac2{1-a}I(0;1-a,1+a)+frac{4a}{a-1}I(1;1-a,1+a)$$
        $$F(a)=fracpi{(1-a^2)^{3/2}}$$
        And since your integral is given by $2F(-a)$,
        $$int_0^{2pi}frac{mathrm dx}{(1-acos x)^2}=frac{2pi}{(1-a^2)^{3/2}}$$





        ADDENDUM



        Consider the integral
        $$C(n;a)=int_0^pi frac{mathrm dx}{(1+acos x)^n}$$
        Starting with $t=tanfrac{x}2$,
        $$C(n;a)=2int_0^infty frac1{left[1+afrac{1-t^2}{1+t^2}right]^n}frac{mathrm dt}{1+t^2}$$
        $$C(n;a)=2int_0^infty frac{left(t^2+1right)^{n-1}}{left[(1-a)t^2+1+aright]^n}mathrm dt$$
        Then using the binomial theorem,
        $$C(n;a)=2sum_{k=0}^{n-1}{n-1choose k}int_0^inftyfrac{x^{2k}}{left[(1-a)x^2+1+aright]^n}mathrm dx$$
        $$C(n;a)=frac2{(1-a)^n}sum_{k=0}^{n-1}{n-1choose k}int_0^inftyfrac{x^{2k}}{left[x^2+frac{1+a}{1-a}right]^n}mathrm dx$$
        We then recall the integral due to my collaborator @DavidG,
        $$int_0^inftyfrac{x^{q}}{left[x^w+bright]^p}mathrm dx=frac{b^{frac{1+q}{w}-p}}{w}frac{Gammaleft(p-frac{1+q}{w}right)Gammaleft(frac{1+q}{w}right)}{Gammaleft(pright)}$$
        Here $Gamma(s)$ is the Gamma function. So with $w=2$, $b=frac{1+a}{1-a}$, $p=n$, and $q=2k$,
        $$C(n;a)=frac1{(1-a)^n}sum_{k=0}^{n-1}{n-1choose k}left(frac{1+a}{1-a}right)^{frac{2k+1}{2}-n}frac{Gammaleft(n-frac{2k+1}{2}right)Gammaleft(frac{2k+1}{2}right)}{Gammaleft(nright)}$$
        $$C(n;a)=frac1{(1-a)^n}sum_{k=0}^{n-1}left(frac{1+a}{1-a}right)^{frac{2k+1}{2}-n}frac{Gammaleft(n-frac{2k+1}{2}right)Gammaleft(frac{2k+1}{2}right)}{k!Gamma(n-k)}$$






        share|cite|improve this answer











        $endgroup$



        My goal with this answer is to give you a bunch of different really general integral formulas that you may find very useful.



        $$F(a)=int_0^{pi}frac{mathrm dx}{(1+acos x)^2}$$
        So your integral is $2F(-a)$. First off we will preform the tangent-half-angle substitution $t=tan(x/2)$. Hence we have that
        $$F(a)=2int_0^infty frac1{(1+afrac{1-t^2}{1+t^2})^2}frac{mathrm dt}{1+t^2}$$
        $$F(a)=2int_0^{infty}frac{1+t^2}{[(1-a)t^2+a+1]^2}mathrm dt$$
        Which we can rewrite as
        $$F(a)=frac2{1-a}int_0^inftyfrac{mathrm dt}{(1-a)t^2+a+1}+frac{4a}{a-1}int_0^{infty}frac{mathrm dt}{[(1-a)t^2+a+1]^2}$$





        Next we consider the very general integral
        $$I(m;a,b)=int_0^infty frac{mathrm dx}{(ax^2+b)^{m+1}}$$
        First we integrate by parts with $mathrm dv=mathrm dx$ to produce
        $$I(m;a,b)=frac{x}{(ax^2+b)^{m+1}}bigg|_0^{infty}+2(m+1)int_0^inftyfrac{ax^2}{(ax^2+b)^{m+2}}mathrm dx$$
        $$I(m;a,b)=2(m+1)int_0^inftyfrac{ax^2+b-b}{(ax^2+b)^{m+2}}mathrm dx$$
        $$I(m;a,b)=2(m+1)I(m;a,b)-2(m+1)bI(m+1;a,b)$$
        Then solving for $I(m+1;a,b)$ and replacing $m+1$ with $m$,
        $$I(m;a,b)=frac{2m-1}{2bm}I(m-1;a,b)$$
        And for the base case:
        $$I(0;a,b)=int_0^{infty}frac{mathrm dx}{ax^2+b}$$
        The Trig sub $x=sqrt{frac{b}a}tan u$ gives
        $$I(0;a,b)=fracpi{2sqrt{ab}}$$
        Which is a special case of
        $$int_{x_1}^{x_2}frac{mathrm dx}{ax^2+bx+c}=frac2{sqrt{4ac-b^2}}arctanfrac{2ax+b}{sqrt{4ac-b^2}}bigg|_{x_1}^{x_2}$$
        Anyway, the recurrence has the solution
        $$I(m;a,b)=fracpi{2^{2m+1}b^msqrt{ab}}{2mchoose m}$$
        Which we will apply very shortly!





        Recalling the definition of $I(m;a,b)$, we have that
        $$F(a)=frac2{1-a}I(0;1-a,1+a)+frac{4a}{a-1}I(1;1-a,1+a)$$
        $$F(a)=fracpi{(1-a^2)^{3/2}}$$
        And since your integral is given by $2F(-a)$,
        $$int_0^{2pi}frac{mathrm dx}{(1-acos x)^2}=frac{2pi}{(1-a^2)^{3/2}}$$





        ADDENDUM



        Consider the integral
        $$C(n;a)=int_0^pi frac{mathrm dx}{(1+acos x)^n}$$
        Starting with $t=tanfrac{x}2$,
        $$C(n;a)=2int_0^infty frac1{left[1+afrac{1-t^2}{1+t^2}right]^n}frac{mathrm dt}{1+t^2}$$
        $$C(n;a)=2int_0^infty frac{left(t^2+1right)^{n-1}}{left[(1-a)t^2+1+aright]^n}mathrm dt$$
        Then using the binomial theorem,
        $$C(n;a)=2sum_{k=0}^{n-1}{n-1choose k}int_0^inftyfrac{x^{2k}}{left[(1-a)x^2+1+aright]^n}mathrm dx$$
        $$C(n;a)=frac2{(1-a)^n}sum_{k=0}^{n-1}{n-1choose k}int_0^inftyfrac{x^{2k}}{left[x^2+frac{1+a}{1-a}right]^n}mathrm dx$$
        We then recall the integral due to my collaborator @DavidG,
        $$int_0^inftyfrac{x^{q}}{left[x^w+bright]^p}mathrm dx=frac{b^{frac{1+q}{w}-p}}{w}frac{Gammaleft(p-frac{1+q}{w}right)Gammaleft(frac{1+q}{w}right)}{Gammaleft(pright)}$$
        Here $Gamma(s)$ is the Gamma function. So with $w=2$, $b=frac{1+a}{1-a}$, $p=n$, and $q=2k$,
        $$C(n;a)=frac1{(1-a)^n}sum_{k=0}^{n-1}{n-1choose k}left(frac{1+a}{1-a}right)^{frac{2k+1}{2}-n}frac{Gammaleft(n-frac{2k+1}{2}right)Gammaleft(frac{2k+1}{2}right)}{Gammaleft(nright)}$$
        $$C(n;a)=frac1{(1-a)^n}sum_{k=0}^{n-1}left(frac{1+a}{1-a}right)^{frac{2k+1}{2}-n}frac{Gammaleft(n-frac{2k+1}{2}right)Gammaleft(frac{2k+1}{2}right)}{k!Gamma(n-k)}$$







        share|cite|improve this answer














        share|cite|improve this answer



        share|cite|improve this answer








        edited Jan 18 at 1:24

























        answered Jan 17 at 2:49









        clathratusclathratus

        5,1141439




        5,1141439






























            draft saved

            draft discarded




















































            Thanks for contributing an answer to Mathematics Stack Exchange!


            • Please be sure to answer the question. Provide details and share your research!

            But avoid



            • Asking for help, clarification, or responding to other answers.

            • Making statements based on opinion; back them up with references or personal experience.


            Use MathJax to format equations. MathJax reference.


            To learn more, see our tips on writing great answers.




            draft saved


            draft discarded














            StackExchange.ready(
            function () {
            StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f3076005%2fhow-to-integrate-int-02-pi-frac-mathrm-dx1-a-cos-x2%23new-answer', 'question_page');
            }
            );

            Post as a guest















            Required, but never shown





















































            Required, but never shown














            Required, but never shown












            Required, but never shown







            Required, but never shown

































            Required, but never shown














            Required, but never shown












            Required, but never shown







            Required, but never shown







            Popular posts from this blog

            Human spaceflight

            Can not write log (Is /dev/pts mounted?) - openpty in Ubuntu-on-Windows?

            張江高科駅