Solving Bessel Equation: Indicial Equation & Frobenius Solution

In summary: I'm having trouble understanding where I went wrong. Can you please explain it in simpler terms?In summary, the conversation discusses a problem involving the Bessel equation with p=0. The goal is to find the Frobenius solution, but the indicial equation only has one root at \alpha=0. The conversation then goes through the process of finding the solution and making iterations, but the correct answer is not reached. The expert summarizer notes that the lower limits were not changed correctly when changing n in the sums and provides the correct form of the solution as \sum \frac{(-1)^n}{ 2^{2n}(n!)^2 }a_0.
  • #1
Telemachus
835
30
Hi there. Well, I'm stuck with this problem, which says:
When p=0 the Bessel equation is: [tex]x^2y''+xy'+x^2y=0[/tex]

Show that its indicial equation only has one root and find the Frobenius solution correspondingly. (Answer: [tex]y=\sum \frac{(-1)^n}{ 2^{2n}(n!)^2 }x^{2n} [/tex]

Well, this is what I did:

At first I normalized the equation:
[tex]y''+\frac{y'}{x}+y=0[/tex]
Then
[tex]P(x)=\frac{1}{ x} \rightarrow xP(x)=1[/tex]
[tex]Q(x)=1 \rightarrow x^2Q(x)=x^2[/tex]
So x=0 is regular singular point.Then the indicial equation is: [tex]\alpha(\alpha-1)+p_0\alpha+q_0=0 \rightarrow \alpha^2-alpha+\alpha=0 \rightarrow \alpha=0[/tex]

[tex]y=\sum_{n = 0}^\infty a_n x^n \rightarrow y'=\sum_{n = 1}^\infty a_n n x^{n-1} \rightarrow y''=\sum_{n = 2}^\infty a_n n(n-1) x^{n-2}[/tex]

Then
[tex]x^2y''+xy'+x^2y=\sum_{n = 2}^\infty a_n n(n-1) x^n+ \sum_{n = 1}^\infty a_n n x^n+\sum_{n = 0}^\infty a_n x^{n+2} = \sum_{n = 2}^\infty a_n n(n-1) x^n+ \sum_{n = 1}^\infty a_n n x^n+\sum_{n = 0}^\infty a_{n-2}x^n[/tex]

So from here I took
[tex]a_n n(n-1)+a_{n-2}=0[/tex]

[tex]a_n=\frac{-a_{n-2}}{n(n-1)},n=2k \rightarrow a_{2k}=\frac{-a_{2k-2}}{2k(2k-1)}[/tex]

Then I've made some iterations, but I can't find the form that the problem gives as the answer, some of the iterations:

[tex]a_2=\frac{-a_0}{2 },a_4=\frac{a_0}{4.3.2 },a_6=\frac{-a_0}{6.5.4.3.2 } [/tex]

So the answer I seem to get is [tex]a_{2k}=\frac{(-1)^k}{(2k)!}a_0 [/tex]

But I should get [tex]\sum \frac{(-1)^n}{ 2^{2n}(n!)^2 }a_0[/tex] or something like that, which is how the answer the problem gives looks like.

I'm probably doing something wrong, but I couldn't figure it out what it is.

Bye there, thanks for helping!
 
Physics news on Phys.org
  • #2
Telemachus said:
Hi there. Well, I'm stuck with this problem, which says:
When p=0 the Bessel equation is: [tex]x^2y''+xy'+x^2y=0[/tex]

Show that its indicial equation only has one root and find the Frobenius solution correspondingly. (Answer: [tex]y=\sum \frac{(-1)^n}{ 2^{2n}(n!)^2 }x^{2n} [/tex]

Well, this is what I did:

At first I normalized the equation:
[tex]y''+\frac{y'}{x}+y=0[/tex]
Then
[tex]P(x)=\frac{1}{ x} \rightarrow xP(x)=1[/tex]
[tex]Q(x)=1 \rightarrow x^2Q(x)=x^2[/tex]
So x=0 is regular singular point.


Then the indicial equation is: [tex]\alpha(\alpha-1)+p_0\alpha+q_0=0 \rightarrow \alpha^2-alpha+\alpha=0 \rightarrow \alpha=0[/tex]

[tex]y=\sum_{n = 0}^\infty a_n x^n \rightarrow y'=\sum_{n = 1}^\infty a_n n x^{n-1} \rightarrow y''=\sum_{n = 2}^\infty a_n n(n-1) x^{n-2}[/tex]

Then
[tex]x^2y''+xy'+x^2y=\sum_{n = 2}^\infty a_n n(n-1) x^n+ \sum_{n = 1}^\infty a_n n x^n+\sum_{n = 0}^\infty a_n x^{n+2} = \sum_{n = 2}^\infty a_n n(n-1) x^n+ \sum_{n = 1}^\infty a_n n x^n+\sum_{n = 0}^\infty a_{n-2}x^n[/tex]
You have not changed the lower limits when you changed n in each sum.
In the last sum you have [itex]a_nx^{n+ 2}[/itex] with n going from 0 to infinity. If your new "n" is the old n+ 2, then you have [itex]a_{n- 2}x^n[/itex] with n going from 2 to infinity. Also, the first sum should have [itex]x^{n-2}[/itex] but you have written [itex]x^n[/itex]. If you let the new n be the old n-2, then the term in the first sum is [itex]a_{n+2}(n+2)(n+1)x^n[/itex] with the sum going from 0 to infinity. Finally, in the middle sum you have, again, [itex]x^n[/itex] when it should be [itex]x^{n-1}[/itex]. If you let the new n be the old n-1 then you should have [itex]a_{n+1}(n+1)x^n[/itex]
That is, you should have
[tex] \sum_{n = 0}^\infty a_{n+2} (n+2)(n+1) x^n+ \sum_{n = 1}^\infty a_{n+1} (n+1) x^n+\sum_{n = 2}^\infty a_{n-2}x^n[/tex]

So from here I took
[tex]a_n n(n-1)+a_{n-2}=0[/tex]

[tex]a_n=\frac{-a_{n-2}}{n(n-1)},n=2k \rightarrow a_{2k}=\frac{-a_{2k-2}}{2k(2k-1)}[/tex]

Then I've made some iterations, but I can't find the form that the problem gives as the answer, some of the iterations:

[tex]a_2=\frac{-a_0}{2 },a_4=\frac{a_0}{4.3.2 },a_6=\frac{-a_0}{6.5.4.3.2 } [/tex]

So the answer I seem to get is [tex]a_{2k}=\frac{(-1)^k}{(2k)!}a_0 [/tex]

But I should get [tex]\sum \frac{(-1)^n}{ 2^{2n}(n!)^2 }a_0[/tex] or something like that, which is how the answer the problem gives looks like.

I'm probably doing something wrong, but I couldn't figure it out what it is.

Bye there, thanks for helping!
 
Last edited by a moderator:
  • #3
Thanks HallsOfIvy.
 

FAQ: Solving Bessel Equation: Indicial Equation & Frobenius Solution

What is the Bessel equation?

The Bessel equation is a second order differential equation that arises in many areas of mathematical physics and engineering, particularly in problems involving circular or cylindrical symmetry. It is named after the mathematician Friedrich Bessel.

What is the Indicial Equation in relation to the Bessel equation?

The Indicial Equation is a tool used to determine the form of solutions to the Bessel equation. It is a quadratic equation that can be solved to find the two roots, which correspond to the exponents of the two linearly independent solutions to the Bessel equation.

What is the Frobenius solution to the Bessel equation?

The Frobenius solution is a method for finding solutions to differential equations when one of the roots of the Indicial Equation is a repeated root. It involves using a power series to find a solution that is valid in a certain range of values.

What are the applications of the Bessel equation?

The Bessel equation has many applications in physics and engineering, particularly in problems involving circular or cylindrical symmetry. It is used in the study of heat transfer, fluid mechanics, electrical engineering, and more.

Are there any alternative methods for solving the Bessel equation?

Yes, in addition to the Indicial Equation and Frobenius solution, there are other methods for solving the Bessel equation such as the power series method, the integral transform method, and the asymptotic approximation method. The most appropriate method depends on the specific problem and the desired level of accuracy.

Similar threads

Back
Top