Examine the singular points of the equation y'=(4y-2x)/(x+y)
Examine the singular points of the equation $$y'=\frac{4y-2x}{x+y}.$$
Explanation of Method
The singular points are located by solving P(x,y) = 0 and Q(x,y) = 0 simultaneously. The associated planar autonomous system is formed (linearising about the singular point where necessary), its coefficient/Jacobian matrix is written down, and the point is classified from the sign and nature of that matrix's eigenvalues, using the standard classification of planar linear systems into nodes, saddles, foci, and centres.
Step 1 · Symmetric form
Write the equation as $$P(x,y)\,dx - Q(x,y)\,dy = 0,$$ with $$P(x,y)=4y-2x,\qquad Q(x,y)=x+y.$$
Step 2 · Locating the singular point
Set $P=0$ and $Q=0$ simultaneously: $$\begin{cases} 4y-2x=0,\\ x+y=0. \end{cases}$$ From the second equation $x=-y$. Substituting into the first: $$4y-2(-y)=4y+2y=6y=0\implies y=0,\quad x=0.$$ The unique singular point is $$(x^*,y^*)=(0,0).$$
Step 3 · Associated linear autonomous system
Since $P$ and $Q$ are homogeneous of degree 1, write the equivalent planar system $$\frac{dx}{dt}=P(x,y)=4y-2x,\qquad \frac{dy}{dt}=Q(x,y)=x+y,$$ with coefficient matrix $$A=\begin{pmatrix}-2&4\\1&1\end{pmatrix}.$$
Step 4 · Eigenvalues of $A$
The characteristic equation is $$\det(A-\lambda I)=0 \implies (-2-\lambda)(1-\lambda)-4\cdot1=0.$$ Expanding: $$-2+2\lambda-\lambda+\lambda^2-4=0 \implies \lambda^2+\lambda-6=0 \implies (\lambda+3)(\lambda-2)=0.$$ Thus $$\lambda_1=-3,\qquad \lambda_2=2.$$
Step 5 · Classification of the singular point
The two eigenvalues are real, distinct, and of opposite sign: $$\lambda_1=-3<0<\lambda_2=2.$$ By the standard classification of linear autonomous systems, the origin is a saddle point.
Step 6 · Eigenvectors and separatrices
For $\lambda_1=-3$: $$(A+3I)\mathbf{v}=0 \implies \begin{pmatrix}1&4\\1&4\end{pmatrix}\mathbf{v}=0.$$ Both rows give $v_1+4v_2=0$, so $$\mathbf{v}^{(1)}=\begin{pmatrix}4\\-1\end{pmatrix}.$$ The stable separatrix is tangent at the origin to the line $\ell_1:\; y=-\tfrac{1}{4}x$. Since $\lambda_1<0$, trajectories on this manifold are attracted to the origin as $t\to+\infty$. For $\lambda_2=2$: $$(A-2I)\mathbf{v}=0 \implies \begin{pmatrix}-4&4\\1&-1\end{pmatrix}\mathbf{v}=0.$$ Both rows give $v_1=v_2$, so $$\mathbf{v}^{(2)}=\begin{pmatrix}1\\1\end{pmatrix}.$$ The unstable separatrix is tangent at the origin to the line $\ell_2:\; y=x$. Since $\lambda_2>0$, trajectories on this manifold depart from the origin as $t\to+\infty$.
Step 7 · General solution of the linear system
$$\begin{pmatrix}x(t)\\y(t)\end{pmatrix} =C_1 e^{-3t}\begin{pmatrix}4\\-1\end{pmatrix} +C_2 e^{2t}\begin{pmatrix}1\\1\end{pmatrix}.$$ In components: $$x(t)=4C_1 e^{-3t}+C_2 e^{2t},\qquad y(t)=-C_1 e^{-3t}+C_2 e^{2t}.$$
Step 8 · Verification via straight-line solutions
Along $y=-\tfrac{1}{4}x$: substituting into the right-hand side gives $-4$, which does not equal the slope $-\tfrac14$, so the line is only an asymptotic (eigen-)direction of the linearised system, not an exact straight-line solution of the scalar ODE. Along $y=x$: substituting gives $$\frac{4y-2x}{x+y}=\frac{4x-2x}{x+x}=\frac{2x}{2x}=1,$$ which equals the slope of $y=x$, so $y=x$ is indeed an exact straight-line solution of the original scalar ODE, confirming $\ell_2$ as the true unstable separatrix.
Step 9 · Behaviour of integral curves near the origin
The origin is a saddle with a stable manifold tangent to $y=-\tfrac{1}{4}x$ and an unstable manifold that coincides exactly with the line $y=x$. All other trajectories approach the origin along $\ell_1$ as $t\to-\infty$ and escape to infinity along $\ell_2$ as $t\to+\infty$, forming the characteristic hyperbolic pattern of a saddle point.
Step 10 · Conclusion
$$\boxed{ (x^*,y^*)=(0,0)\text{ is the unique singular point; it is a \textbf{saddle point}.} }$$ The eigenvalues $\lambda_1=-3$ and $\lambda_2=2$ are real with opposite signs. The stable separatrix is tangent to the line $y=-\tfrac{1}{4}x$ and the unstable separatrix coincides with the exact straight-line solution $y=x$ of the ODE. All other trajectories exhibit the characteristic hyperbolic divergence of a saddle.
Summary Table
| Quantity | Value |
|---|---|
| Eigenvalues | $\det(A-\lambda I)=0 \implies (-2-\lambda)(1-\lambda)-4\cdot1=0.$ |
| Classification | saddle point |
Additional Notes
- The classification follows from the standard theory of planar linear (or linearised) autonomous systems, based on the sign and nature of the eigenvalues of the coefficient/Jacobian matrix.
Result
The final classification is: