Cauchy-Euler Equation

April 13, 2026

Problem

Solve x^2 y'' - 2 x y' + 2 y = 0. Substitute y = x^r to reduce to an algebraic equation in r and show the resulting power-law solutions.

Explanation

What is a Cauchy-Euler equation?

A Cauchy-Euler equation (also equidimensional or just Euler equation) is a linear ODE where each term has matching "dimensions" of xx and derivative, like anxny(n)+an1xn1y(n1)++a1xy+a0y=0.a_n \, x^{n} y^{(n)} + a_{n-1} \, x^{n-1} y^{(n-1)} + \ldots + a_1 \, x \, y' + a_0 \, y = 0.

The second-order case: ax2y+bxy+cy=0.a \, x^{2} y'' + b \, x \, y' + c \, y = 0.

It is not constant-coefficient, but it has a similar clean structure because of the scaling symmetry: rescaling xλxx \to \lambda x leaves the equation form-invariant. This symmetry is why the characteristic ansatz y=xry = x^{r} works so cleanly here, and why "erte^{r t} for constant-coefficient ODEs" generalises to "xrx^{r} for Cauchy-Euler ODEs".

Either ansatz is a special case of the observation: the solution space of a linear ODE with a continuous symmetry is spanned by functions that diagonalize that symmetry.

The given equation

x2y2xy+2y=0.x^{2} y'' - 2 x y' + 2 y = 0.

Coefficients a=1a = 1, b=2b = -2, c=2c = 2. Work on x>0x > 0.

Step-by-step solution

Step 1 — Ansatz y=xry = x^{r}.

Compute derivatives: y=xr,y=rxr1,y=r(r1)xr2.y = x^{r}, \qquad y' = r \, x^{r-1}, \qquad y'' = r(r-1) \, x^{r-2}.

Plug into the ODE: x2r(r1)xr22xrxr1+2xrx^{2} \cdot r(r-1) x^{r-2} - 2 x \cdot r \, x^{r-1} + 2 \, x^{r} =r(r1)xr2rxr+2xr=xr[r(r1)2r+2]= r(r-1) x^{r} - 2 r \, x^{r} + 2 \, x^{r} = x^{r}\bigl[r(r-1) - 2 r + 2\bigr] =xr[r23r+2].= x^{r}\bigl[r^{2} - 3 r + 2\bigr].

Step 2 — Characteristic (indicial) equation.

Since xr0x^{r} \ne 0 on x>0x > 0: r23r+2=0    (r1)(r2)=0    r1=1,r2=2.r^{2} - 3 r + 2 = 0 \implies (r - 1)(r - 2) = 0 \implies r_1 = 1, \, r_2 = 2.

Step 3 — General solution.

Two distinct real roots → two power-law solutions: y(x)=C1x+C2x2(x>0)\boxed{\, y(x) = C_1 \, x + C_2 \, x^{2} \,} \qquad (x > 0)

Verification

y=C1x+C2x2y = C_1 x + C_2 x^{2} y=C1+2C2xy' = C_1 + 2 C_2 x y=2C2y'' = 2 C_2

x2y=2C2x2x^{2} y'' = 2 C_2 x^{2}, 2xy=2C1x4C2x2-2 x y' = -2 C_1 x - 4 C_2 x^{2}, 2y=2C1x+2C2x22 y = 2 C_1 x + 2 C_2 x^{2}.

Sum: 2C2x22C1x4C2x2+2C1x+2C2x2=(24+2)C2x2+(2+2)C1x=02 C_2 x^{2} - 2 C_1 x - 4 C_2 x^{2} + 2 C_1 x + 2 C_2 x^{2} = (2 - 4 + 2) C_2 x^{2} + (-2 + 2) C_1 x = 0. ✓

The three cases of the indicial equation

Exactly like constant-coefficient ODEs, three flavors depending on the discriminant Δ=(ba)24ac\Delta = (b - a)^{2} - 4 a c (after rearranging the indicial equation ar(r1)+br+c=0a r(r-1) + b r + c = 0):

  • Two real distinct roots r1r2r_1 \ne r_2: y=C1xr1+C2xr2y = C_1 x^{r_1} + C_2 x^{r_2}. (This problem.)
  • Repeated real root rr: y=xr(C1+C2lnx)y = x^{r}(C_1 + C_2 \ln x). The lnx\ln x replaces the xx factor we saw in constant-coefficient repeated-root cases (#185).
  • Complex conjugate roots r=α±iβr = \alpha \pm i \beta: y=xα(C1cos(βlnx)+C2sin(βlnx))y = x^{\alpha}\bigl(C_1 \cos(\beta \ln x) + C_2 \sin(\beta \ln x)\bigr). Oscillations in the logarithm of xx — the spacing between zeros gets tighter as x0+x \to 0^+ or xx \to \infty.

Mnemonic. Replace ertxre^{r t} \leftrightarrow x^{r} and tlnxt \leftrightarrow \ln x. The substitution t=lnxt = \ln x (so x=etx = e^{t}) literally transforms a Cauchy-Euler ODE into a constant-coefficient one — see next section.

The substitution t=lnxt = \ln x (optional derivation)

Set t=lnxt = \ln x, so x=etx = e^{t} and dx=etdt=xdtdx = e^{t} dt = x \, dt, i.e. d/dx=(1/x)d/dtd/dx = (1/x) \, d/dt.

xdydx=dydt,x2d2ydx2=d2ydt2dydt.x \frac{dy}{dx} = \frac{dy}{dt}, \qquad x^{2} \frac{d^{2} y}{dx^{2}} = \frac{d^{2} y}{dt^{2}} - \frac{dy}{dt}.

The Cauchy-Euler ODE ax2y+bxy+cy=0a x^{2} y'' + b x y' + c y = 0 transforms to a(yttyt)+byt+cy=0    aytt+(ba)yt+cy=0,a (y_{tt} - y_{t}) + b y_{t} + c y = 0 \implies a y_{tt} + (b - a) y_{t} + c y = 0, a constant-coefficient ODE in tt. Solve it by the usual characteristic equation, then substitute t=lnxt = \ln x back.

For our problem: ytt+(21)yt+2y=0    ytt3yt+2y=0y_{tt} + (-2 - 1) y_{t} + 2 y = 0 \implies y_{tt} - 3 y_{t} + 2 y = 0, characteristic roots r=1,2r = 1, 2, solutions et,e2te^{t}, e^{2 t}x,x2x, x^{2}. Matches step 3.

For x<0x < 0

The ansatz y=xry = x^{r} is problematic for non-integer rr when x<0x < 0. Use y=xry = |x|^{r} instead, or consider the two half-lines separately. Solutions on x<0x < 0 are C1x+C2x2C_1 |x| + C_2 |x|^{2} in this example.

Initial value problem

y(1)=3y(1) = 3, y(1)=5y'(1) = 5. From y=C1x+C2x2y = C_1 x + C_2 x^{2}, y=C1+2C2xy' = C_1 + 2 C_2 x: 3=C1+C2,5=C1+2C2    C2=2,  C1=1.3 = C_1 + C_2, \quad 5 = C_1 + 2 C_2 \implies C_2 = 2, \; C_1 = 1. y(x)=x+2x2.y(x) = x + 2 x^{2}.

Where Cauchy-Euler shows up

  • Spherical or cylindrical coordinates with separation of variables. The radial part of Laplace's equation is often Cauchy-Euler.
  • Scaling-invariant systems in physics and economics — anytime the law of the system is invariant under xλxx \to \lambda x, expect power-law solutions.
  • Boundary layer problems in fluid dynamics (inner scaling) often reduce to Cauchy-Euler form.

Common mistakes

  • Using erxe^{r x} ansatz for a Cauchy-Euler ODE. The right ansatz is xrx^{r} — it's the scaling eigenfunction, not the translation eigenfunction.
  • Forgetting the - in y(y1)y(y - 1) when expanding y=r(r1)xr2y'' = r(r-1) x^{r-2}. It's the second derivative, not the first squared.
  • Missing lnx\ln x for repeated roots. Analogous to the xx factor for repeated roots in constant-coefficient ODEs, but with lnx\ln x because of the log-substitution bridge.
  • Applying at x=0x = 0. The Cauchy-Euler equation has a singularity at the origin; solutions often behave badly there. Stick to x>0x > 0 or x<0x < 0.

Try it in the visualization

Slide the coefficients a,b,ca, b, c and watch the two indicial roots move. See the solution shape morph between power laws, logarithms (repeated root), and log-spaced oscillations (complex roots) — three qualitatively different regimes, all glued by the xrx^{r} / lnx\ln x machinery.

Interactive Visualization

Parameters

1.00
-2.00
2.00
1.00
0.50
5.00
30.00
Your turn

Got your own math or physics problem?

Turn any problem into an interactive visualization like this one — powered by AI, generated in seconds. Free to try, no credit card required.

Sign Up Free to Try It30 free visualizations every day
Cauchy-Euler Equation | MathSpin