Problem 256
Hard Further Methods

Given the differential equation (x^3 - 1)y' = 2xy^2 - x^2y…

Given the differential equation \[(x^3 - 1)y' = 2xy^2 - x^2y - 1.\] \(1^\circ\) Determine a particular solution of the form \(y = ax + b\) and then find the general solution. \(2^\circ\) Find the particular solution \(y = \varphi(x)\) that satisfies \(y(0) = -\frac{1}{4}\).

Explanation of Method

This is a Riccati equation. First a linear particular solution \(y_1 = ax+b\) is found by substituting into the equation and matching powers of \(x\). Then the standard Riccati substitution \(y = y_1 + 1/v\) reduces the equation to a first-order linear ODE in \(v(x)\), which is solved with an integrating factor.

Find the particular solution $y=ax+b$

Substituting \(y=ax+b\), \(y'=a\) into \((x^3-1)y' = 2xy^2-x^2y-1\) and expanding the right side gives \[(2a^2-a)x^3+(4ab-b)x^2+2b^2x-1,\] while the left side is \(ax^3-a\). Matching coefficients of \(x^3,x^2,x^1,x^0\) gives \(a(a-1)=0\), \(b(4a-1)=0\), \(2b^2=0\), and \(a=1\). The only consistent choice is \(a=1,\ b=0\), so \[y_1 = x.\] Direct substitution confirms: with \(y=x\), \(y'=1\), both sides of the ODE equal \(x^3-1\).

Riccati substitution $y=x+1/v$

Let \(y = x + \dfrac{1}{v}\), so \(y' = 1 - \dfrac{1}{v^2}\dfrac{dv}{dx}\). Substituting into the ODE and simplifying (the \(x^3\) and constant terms cancel using the fact that \(y_1=x\) already solves the equation) leaves \[(x^3-1)\frac{dv}{dx} = -3x^2 v - 2x.\]

Solve the linear ODE for $v$

Writing this as \[\frac{dv}{dx}+\frac{3x^2}{x^3-1}v = \frac{-2x}{x^3-1},\] the integrating factor is \[\mu = e^{\int \frac{3x^2}{x^3-1}dx} = x^3-1.\] Then \[\frac{d}{dx}\big[v(x^3-1)\big] = -2x \quad\Longrightarrow\quad v(x^3-1) = C - x^2 \quad\Longrightarrow\quad v = \frac{C-x^2}{x^3-1}.\]

Back-substitute for the general solution

\[y = x + \frac{1}{v} = x + \frac{x^3-1}{C-x^2} = \frac{x(C-x^2)+x^3-1}{C-x^2} = \frac{Cx-1}{C-x^2}.\]

Apply the initial condition $y(0)=-1/4$

At \(x=0\): \[-\frac{1}{4} = \frac{C\cdot 0 - 1}{C-0} = -\frac{1}{C} \quad\Longrightarrow\quad C=4.\] So \[\varphi(x) = \frac{4x-1}{4-x^2}.\]

Independent symbolic verification

Both claimed solutions were re-checked from scratch with a computer algebra system: substituting \(y=\dfrac{Cx-1}{C-x^2}\) (with \(C\) a free symbolic parameter) and its derivative into \((x^3-1)y'-2xy^2+x^2y+1\) simplifies identically to \(0\) for all \(C\), confirming the general solution satisfies the original Riccati equation exactly (not just the linearized version). The particular solution \(y=\dfrac{4x-1}{4-x^2}\) was checked the same way and also gives residual \(0\), with \(y(0)=-\tfrac14\) confirmed directly. No errors were found anywhere in this problem's derivation.


Result

Final answer:

Particular linear solution: \(\boxed{y_1 = x}\)

General solution: \[\boxed{y = \frac{Cx-1}{C-x^2}}\]

Particular solution with \(y(0)=-\frac14\): \[\boxed{y = \varphi(x) = \frac{4x-1}{4-x^2}}\]

All results were confirmed correct as stated by independent symbolic substitution back into the original (non-linearized) Riccati equation.