JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

Riccati, Clairaut, and Singular Solutions

Two famous nonlinear equations that bend the rules — Riccati, which one known solution unlocks, and Clairaut, whose tidy family of lines hides an extra curve no constant can ever produce.

Where the linearizing tricks finally stall

In the last guide the Bernoulli equation y' + p(x) y = q(x) y^n looked nonlinear but quietly surrendered: the substitution v = y^(1-n) flattened it into a first-order linear equation you already know how to solve. That is the dream — find one change of variable and the nonlinearity dissolves. This guide meets two equations where the dream is harder to reach, and one of them refuses outright. Along the way we run into a genuinely new phenomenon: a solution the general solution cannot see.

The honest backdrop, which the whole rung keeps whispering, is that most first-order ODEs have no closed-form answer at all. The exact equations, integrating factors, and substitutions of this rung are a curated gallery of the lucky cases. Riccati and Clairaut belong in that gallery not because they are easy, but because each teaches a lesson that reaches far beyond itself.

Riccati: nonlinear, but one solution unlocks the rest

The Riccati equation is y' = q0(x) + q1(x) y + q2(x) y^2 — linear in y plus a single quadratic term q2 y^2. That lone y^2 is enough to put it firmly outside the linear world; in general the Riccati equation has no formula in elementary functions. This is not a gap in cleverness. It is a theorem-grade fact: there is no universal recipe.

But there is a beautiful conditional rescue. Suppose you can guess, or are handed, one particular solution y1(x). Then the substitution y = y1 + 1/v turns the whole thing into a first-order linear equation for v(x) — the very kind you can always integrate. So a single known solution acts like a key: it converts an unsolvable-looking quadratic into the friendliest family in the rung.

Riccati:   y' = q0(x) + q1(x) y + q2(x) y^2
know one solution y1  ->  set  y = y1 + 1/v
result:    v' + ( q1 + 2 q2 y1 ) v = - q2     (linear in v!)
One known solution y1 collapses Riccati into a linear equation for v.

Clairaut: differentiate, and the equation splits in two

The Clairaut equation has the curious shape y = x y' + f(y'), where f is some function of the slope y' alone. Write p for y' to keep your eyes clear: y = x p + f(p). The standard move is unusual — you differentiate both sides with respect to x. Doing that and tidying up gives (x + f'(p)) p' = 0. A product equals zero, so one factor must vanish, and the Clairaut equation cleaves neatly into two completely different cases.

First case: p' = 0, meaning the slope is constant, p = c. Substitute back and you get y = c x + f(c) — a one-parameter family of straight lines. That is the general solution, and remarkably it costs no integration at all: you simply replace y' by the constant c everywhere in the original equation. Each value of c is a different line in the family.

Second case is the strange and wonderful one: x + f'(p) = 0. This does not give a constant slope; it ties x to p through f'. Combined with the original y = x p + f(p), it traces out a single curve described parametrically by p. This curve is the singular solution, and it is the heart of this guide.

The singular solution: a curve no constant can reach

Picture the family of lines y = c x + f(c) drawn for every value of c. They tilt and shift, but they all stay tangent to one smooth curve — like the bright caustic edge a coffee cup throws onto the table when each reflected ray grazes the same boundary. That boundary curve is the singular solution. Geometrically it is the envelope of the line family: every line touches it, yet the envelope is not one of the lines.

Here is the punchline that makes singular solutions worth a whole guide: no choice of the constant c gives you the envelope. You can search through every value of c forever and never land on it. So the general solution, complete as it looks, is genuinely missing a solution. A method that only ever produces y = c x + f(c) is blind to a perfectly valid curve satisfying the very same equation.

Why singular solutions are allowed: uniqueness breaks down

It feels illegal. Earlier in the ladder, existence-and-uniqueness theory promised that through each point there runs exactly one solution. How can the envelope and a tangent line both pass through their shared touch-point? The answer is that uniqueness comes with fine print: it holds where the equation satisfies a Lipschitz condition in y. At every point where the envelope kisses a line, that condition quietly fails, and the guarantee of a single path evaporates.

You met this same crack once before, in the classic warning example y' = y^(2/3). At y = 0 the right side has an infinite slope in y, Lipschitz fails, and two solutions leave the origin: the flat line y = 0 and the curve y = (x/3)^3. That is failure of uniqueness in miniature. The singular solution of a Clairaut equation is the same phenomenon wearing a more elegant costume — an entire curve where solutions are free to branch.

So the takeaway is not 'be afraid of nonlinear equations', but 'respect them'. The clean superposition and tidy uniqueness of the linear world are luxuries you are now leaving behind. A nonlinear first-order ODE can hide solutions your substitution never reveals, and the only way to be sure you have them all is to check the equation's own structure — its envelope, its bad points, its broken Lipschitz spots.

A recipe for Clairaut, and where this rung has carried you

  1. Recognise the form y = x y' + f(y'): the equation is y equals x times the slope, plus a function of the slope alone.
  2. Get the line family for free: replace every y' with a constant c to write down y = c x + f(c). This is the general solution — no integration needed.
  3. Hunt the singular solution: differentiate the original once, factor out to get x + f'(p) = 0, and use it together with y = x p + f(p) to eliminate p.
  4. Check honestly: confirm the singular curve really solves the equation, and note it is the envelope tangent to every line in the family — a solution the general formula misses.

Step back and look at the whole rung. You learned to spot an exact equation and recover its hidden potential, to multiply by an integrating factor when exactness is just out of reach, to tame homogeneous equations with v = y/x, to linearize Bernoulli, and now to read Riccati and Clairaut. Every one of these is a way of bending a nonlinear or awkward equation back toward something you can integrate.

But Riccati's missing formula and Clairaut's missing solution both point the same way forward. When no substitution helps and no formula exists, you stop asking 'what is the exact answer?' and start asking 'what does the solution do?' — does it grow, settle, oscillate, blow up? That qualitative and numerical turn is the next great theme of the ladder, and these two equations are exactly the nudge that sends you there.