ODE: Series Solutions About Singular Points

Many of the most important differential equations in engineering—governing heat transfer, vibrations, and wave propagation—feature coefficients that become infinite at a point. The standard power series method fails at these singular points, leaving you seemingly stuck. The Frobenius method is the powerful, systematic extension that unlocks solutions near these critical locations, allowing engineers to model physical systems with inherent singularities, such as those with cylindrical symmetry or at boundaries.

1. Classifying Singular Points

For a second-order linear ODE in standard form:

$$y^{\prime \prime }+P(x)y^{\prime }+Q(x)y=0$$

a point $x=x_0$ is classified as an ordinary point if both $P(x)$ and $Q(x)$ are analytic (i.e., expressible as a convergent power series) at $x_0$. If either $P(x)$ or $Q(x)$ is not analytic at $x_0$, then $x_0$ is a singular point.

This classification is the crucial first step. You must first identify all singular points before attempting a series solution. For example, in the equation $x^2{y}^{\prime \prime }+x{y}^{\prime }+(x^2-{\nu }^2)y=0$ (Bessel's equation), the coefficients are $P(x)=1/x$ and $Q(x)=(x^2-{\nu }^2)/x^2$. Both functions blow up at $x=0$, making it a singular point.

2. Regular versus Irregular Singular Points

Not all singularities are created equal. A singular point $x=x_0$ is a regular singular point if the functions $(x-x_0)P(x)$ and $(x-x_0{)}^{2}Q(x)$ are both analytic at $x_0$. In simpler terms, the singularity is "mild" or "removable" when multiplied by these factors. If a singular point does not satisfy this condition, it is an irregular singular point.

The Frobenius method, our primary tool, applies only to equations with regular singular points. For Bessel's equation at $x=0$, we check: $xP(x)=x\ast (1/x)=1$ and $x^{2}Q(x)=x^2\ast ((x^2-{\nu }^2)/x^2)=x^2-{\nu }^2$. Both results are polynomials, and therefore analytic, confirming $x=0$ is a regular singular point. Equations with irregular singular points require more advanced asymptotic methods.

3. The Frobenius Method Procedure

The Frobenius method proposes a solution of the form:

$$y(x)=(x-x_0{)}^r\sum _{n=0}^{\infty }{a}_n(x-x_0{)}^n=\sum _{n=0}^{\infty }{a}_n(x-x_0{)}^{n+r}$$

where $x_0$ is the regular singular point, $r$ is a number (real or complex) to be determined, and $a_0\ne 0$. This differs from a standard power series by the factor $(x-x_0{)}^r$, which accommodates the singularity.

The procedure is systematic:

1. Assume the Solution Form: Substitute $y$, $y^{\prime }$, and $y^{\prime \prime }$ from the Frobenius series into the ODE.
2. Combine Series: Align all series to a common power of $(x-x_0)$, typically $x^{n+r}$.
3. Derive the Indicial Equation: The coefficient of the lowest power of $x$ (usually $x^r$) yields an equation involving only $r$ and $a_0$. Since $a_0\ne 0$, this simplifies to the indicial equation, a quadratic in $r$.
4. Solve the Recurrence Relation: The coefficients of higher powers ($x^{n+r}$) give a recurrence relation that defines $a_n$ in terms of $a_{n-1},a_{n-2},...$ and $r$.
5. Construct the Solution(s): Use the roots $r_1$ and $r_2$ from the indicial equation with the recurrence relation to build one or two linearly independent series solutions.

4. The Indicial Equation and Series Solution Forms

The indicial equation is the cornerstone of the method. Its roots, $r_1$ and $r_2$, dictate the fundamental form of the solutions. There are three distinct cases, which you must check in this order:

1. Case I: Roots Differing by a Non-Integer. If $r_1-r_2$ is not an integer, you obtain two linearly independent solutions directly from the two values of $r$:

$$y_1(x)=|x{|}^{r_1}\sum _{n=0}^{\infty }{a}_n(r_1)x^n,y_2(x)=|x{|}^{r_2}\sum _{n=0}^{\infty }{a}_n(r_2)x^n.$$

1. Case II: Double Root ($r_1=r_2$). When the indicial equation has a repeated root $r$, the first solution is $y_1(x)=|x{|}^r\sum a_n{x}^n$. The second, linearly independent solution always contains a logarithmic term:

$$y_2(x)=y_1(x)\ln |x|+|x{|}^r\sum _{n=1}^{\infty }{b}_n{x}^n.$$

1. Case III: Roots Differing by an Integer. If $r_1-r_2=N$, a positive integer, the larger root $r_1$ typically yields the first solution $y_1$. The solution for the smaller root $r_2$ may either generate a second series or it may fail because the recurrence relation becomes undefined. When it fails, the second solution again involves a logarithmic term:

$$y_2(x)=c{y}_1(x)\ln |x|+|x{|}^{r_2}\sum _{n=0}^{\infty }{d}_n{x}^n,$$ where $c$ is a constant (which may be zero).

5. Bessel's Equation: A Prime Engineering Example

Bessel's equation, $x^2{y}^{\prime \prime }+x{y}^{\prime }+(x^2-{\nu }^2)y=0$, is the canonical engineering application of the Frobenius method. Applying the method at the regular singular point $x=0$ yields an indicial equation of $r^2-{\nu }^2=0$, so $r_{1,2}=\pm \nu$.

• If $\nu$ is not an integer (Case I), the two independent solutions are Bessel functions of the first kind, $J_{\nu }(x)$ and $J_{-\nu }(x)$.
• If $\nu =m$ is an integer (Case III, with roots differing by $2m$), the functions $J_m(x)$ and $J_{-m}(x)$ are linearly dependent. The standard second solution is then the Bessel function of the second kind $Y_m(x)$, which contains a logarithmic term. These functions are foundational for problems with cylindrical symmetry, such as heat flow in a cylinder or the vibrations of a circular drumhead.

6. Handling Logarithmic Terms in Second Solutions

Constructing the second solution when a logarithmic term is required (Cases II and III) is the most technically demanding part of the Frobenius method. The systematic approach is to:

1. Start with the general Frobenius series including a logarithmic term: $y_2(x)=y_1(x)\ln |x|+x^{r_2}{\sum }_{n=0}^{\infty }{b}_n{x}^n$.
2. Substitute $y_2$, $y_2^{\prime }$, and $y_2^{\prime \prime }$ into the original ODE.
3. Crucially, use the fact that $y_1$ is already a solution. This substitution will simplify the resulting equation.
4. Solve for the new coefficients $b_n$, often finding that the constant $c$ (the coefficient of the log term) is determined by the recurrence relation breakdown for the smaller root $r_2$. This process is methodical but algebraically intensive.

Common Pitfalls

1. Misclassifying a Singular Point: Attempting the Frobenius method at an irregular singular point will lead to failure. Always verify that $(x-x_0)P(x)$ and $(x-x_0{)}^{2}Q(x)$ are analytic.

• Correction: Perform the analyticity check formally. If the limit of these functions as $x\to x_0$ does not exist or is infinite, the point is irregular.

1. Incorrectly Applying the Indicial Equation Cases: Jumping to a logarithmic solution when it isn't needed, or missing one when it is, is a common error.

• Correction: Always calculate $r_1-r_2$. Proceed strictly in order: Non-integer -> Standard series for both roots. Equal roots -> Logarithmic solution required. Integer difference -> Attempt the series for the smaller root; if the recurrence relation breaks down (e.g., demands division by zero), a logarithmic term is required.

1. Mishandling the Recurrence Relation for the Second Solution: When solving for coefficients in a logarithmic case (like finding $b_n$), errors often arise from improper differentiation of the $\ln |x|$ term or from not correctly substituting the known $y_1$ solution.

• Correction: Write out the derivatives carefully: $y_2^{\prime }=y_1^{\prime }/y_1+...$ and $y_2^{\prime \prime }=...$. Substitute into the ODE and group terms methodically, leveraging the fact that $L{y}_1=0$ to cancel large sections of the equation.

1. Assuming $a_0$ is Arbitrary in All Cases: While $a_0$ is always arbitrary for the first solution, when constructing the second solution with a logarithm, the first coefficient $b_0$ (or sometimes $b_N$) is often not arbitrary; it is determined by the need to satisfy the recurrence relation.

• Correction: Follow the algebra of the recurrence relations precisely. Let the equations dictate which coefficients are free parameters.

Summary

• The Frobenius method is the essential technique for solving linear ODEs about regular singular points, extending the standard power series approach with a factor of $(x-x_0{)}^r$.
• Correctly classifying a point as an ordinary point, regular singular point, or irregular singular point is the mandatory first step.
• The indicial equation, derived from the lowest power of $x$ after substitution, determines the possible values of $r$ and dictates which of three cases applies for the solution's form.
• Bessel's equation serves as a paradigm, producing Bessel functions $J_{\nu }(x)$ and $Y_m(x)$, which are indispensable in engineering applications involving cylindrical geometries.
• When the roots of the indicial equation are equal or differ by an integer, the second linearly independent solution often requires the introduction of a logarithmic term, constructed via a careful, systematic extension of the Frobenius method.