Variable separable equations

Some differential equations look tangled but are secretly simple: if every xx can be pushed to one side with dxdx and every yy to the other with dydy, the whole equation splits into two ordinary integrals. That is the variable separable method, the first and most fundamental technique in the ISC Class 12 Differential Equations chapter, and the building block that the homogeneous and linear methods eventually reduce to. This article rebuilds the method from the general solution down to a fully worked initial-value problem, then looks at the "reducible to separable" trick — a substitution that turns an equation that is not separable at first glance into one that is — and the traps that cost the most marks: forgetting the constant, dropping lost solutions, and mishandling a substitution's back-conversion.

The core insight

If dydx=f(x) g(y)\dfrac{dy}{dx} = f(x)\,g(y), rearranging to dyg(y)=f(x) dx\dfrac{dy}{g(y)} = f(x)\,dx and integrating both sides independently — with a single constant CC — gives the general solution, and a given initial condition then fixes CC to give the particular solution.

The core idea

A first-order differential equation is variable separable if the right side factors purely into a function of xx times a function of yy, with no term that mixes the two variables together:

dydx=f(x)⋅g(y)\frac{dy}{dx} = f(x)\cdot g(y)

where: f(x)f(x) depends only on xx and g(y)g(y) depends only on yy; if the right side cannot be split this way (for example x+yx+y, where the two variables are added rather than multiplied), the equation is not directly separable. Once recognised, rearrange so every yy-term (including dydy) sits on one side and every xx-term (including dxdx) sits on the other, then integrate each side on its own:

∫dyg(y)=∫f(x) dx+C\int \frac{dy}{g(y)} = \int f(x)\,dx + C

where: CC is a single arbitrary constant of integration — only one is needed, because the two separate constants produced by integrating each side individually can always be combined into one overall constant on either side of the equation. This relation between xx and yy is the general solution; substituting a given initial condition y(x0)=y0y(x_0) = y_0 pins down CC and yields the particular solution. Not every first-order equation is separable as written, but some become separable after a substitution. If the right side is a function of a linear combination of xx and yy, such as dydx=f(ax+by+c)\dfrac{dy}{dx} = f(ax+by+c), setting v=ax+by+cv = ax+by+c converts it into a separable equation in vv and xx, since dvdx=a+bdydx\dfrac{dv}{dx} = a + b\dfrac{dy}{dx} lets dydx\dfrac{dy}{dx} be replaced entirely in terms of dvdx\dfrac{dv}{dx}. Among the three first-order techniques in this syllabus — separable, homogeneous, linear — separable is the natural first check on any equation: the homogeneous method itself reduces to a separable equation after the substitution y=vxy = vx, so spotting separability early can save a great deal of algebra.

Figure
Diagram showing an equation split into an x side with dx and a y side with dy then each side integrated separately with one shared constant C
Separating variables: sort, then integrate independently dy divided by g of y all y-terms, one side f of x, dx all x-terms, other side = integrate integrate relation between x and y, plus one constant C general solution — initial condition fixes C
Derivation AThe general solution of a separable equation

The goal is to show, starting only from dydx=f(x)g(y)\dfrac{dy}{dx} = f(x)g(y) and the ordinary rules of differentiation and integration, why treating dydy and dxdx as separable pieces that can be integrated on opposite sides of the equation is legitimate and produces a genuine solution, not merely a notational trick — and why exactly one arbitrary constant survives even though two separate integrations are performed.

Step 1. Start from the defining form dydx=f(x) g(y)\dfrac{dy}{dx} = f(x)\,g(y), with g(y)≠0g(y) \neq 0. Divide both sides by g(y)g(y): 1g(y)dydx=f(x)\dfrac{1}{g(y)}\dfrac{dy}{dx} = f(x). Every term on the left now depends on yy (through the function 1/g(y)1/g(y)) and on xx only through the derivative dy/dxdy/dx; every term on the right depends only on xx.

Step 2. Integrate both sides with respect to xx: ∫1g(y)dydx dx=∫f(x) dx\displaystyle\int \frac{1}{g(y)}\frac{dy}{dx}\,dx = \int f(x)\,dx. On the left, this is exactly the pattern of integration by substitution with u=yu=y: since dydx dx=dy\dfrac{dy}{dx}\,dx = dy, the left integral becomes ∫dyg(y)\displaystyle\int \frac{dy}{g(y)}, purely in terms of yy.

Step 3. Write the result as ∫dyg(y)=∫f(x) dx+C\displaystyle\int \frac{dy}{g(y)} = \int f(x)\,dx + C. Two indefinite integrations were performed, each of which in principle produces its own constant, say C1C_1 on the left and C2C_2 on the right; but C1C_1 and C2C_2 can be combined into a single constant C=C2−C1C = C_2 - C_1 moved to one side, since both are arbitrary real numbers and only their difference matters for the relation between xx and yy. This is why only one constant is written, not two.

Step 4. Read off the meaning. The equation from Step 3 is an implicit (or, after simplification, explicit) relation H(x,y)=CH(x,y) = C that is satisfied by every solution curve of the original differential equation — the general solution, a whole family of curves indexed by CC. Given one point (x0,y0)(x_0, y_0) that the solution curve must pass through, substituting it into the relation solves for the single unknown CC, selecting exactly one curve from the family: the particular solution.

Worked example

Solve dydx=x2+1y\dfrac{dy}{dx} = \dfrac{x^2+1}{y}, given the initial condition y(0)=2y(0) = 2.

Step 1. Check the form and separate. The right side is 1y⋅(x2+1)\dfrac{1}{y}\cdot(x^2+1), a product of a function of yy alone and a function of xx alone, so the equation is separable. Cross-multiplying, y dy=(x2+1) dxy\,dy = (x^2+1)\,dx.

Step 2. Integrate both sides. ∫y dy=∫(x2+1) dx\displaystyle\int y\,dy = \int (x^2+1)\,dx gives y22=x33+x+C\dfrac{y^2}{2} = \dfrac{x^3}{3} + x + C, the general solution.

Step 3. Apply the initial condition. Substituting x=0, y=2x=0,\ y=2: 42=0+0+C\dfrac{4}{2} = 0+0+C, so C=2C=2.

Step 4. Write the particular solution and check it. y22=x33+x+2\dfrac{y^2}{2} = \dfrac{x^3}{3}+x+2, equivalently y2=2x33+2x+4y^2 = \dfrac{2x^3}{3}+2x+4. Differentiating the implicit form gives y dydx=x2+1y\,\dfrac{dy}{dx} = x^2+1, i.e. dydx=x2+1y\dfrac{dy}{dx} = \dfrac{x^2+1}{y} — matching the original equation, so the solution checks out. Final answer: y2=2x33+2x+4y^2 = \dfrac{2x^3}{3}+2x+4.

Competitive stretch

For dydx=ex−y\dfrac{dy}{dx} = e^{x-y} with y(0)=0y(0)=0: rewrite ex−y=exe−ye^{x-y} = e^x e^{-y}, separate as ey dy=ex dxe^y\,dy = e^x\,dx, integrate to ey=ex+Ce^y = e^x+C, apply x=0,y=0x=0,y=0 to get C=0C=0, giving ey=exe^y = e^x, so y=xy=x. The only new skill is splitting an exponential of a sum or difference before separating.

Common traps

  • Trying to separate an equation that isn't in product form. dydx=x+y\dfrac{dy}{dx} = x+y cannot be separated directly — the xx and yy are added, not multiplied — so it needs the linear-equation method instead.
  • Writing two constants instead of one. Integrating each side gives its own constant, but they must be combined into a single overall CC; keeping two separately is sloppy and confuses the step where the initial condition is applied.
  • Losing constant solutions when dividing by g(y)g(y). If g(y)=0g(y)=0 at some value y=ky=k, then y=ky=k is itself a solution that gets divided away in Step 1; ISC problems rarely probe this, but it is worth naming.
  • Forgetting to convert back after a reducing substitution. After solving in terms of v=ax+by+cv = ax+by+c, the final answer must be rewritten in terms of xx and yy, not left in vv.

Ask the tutor: Not sure whether your differential equation is separable, which substitution to try, or how to apply an initial condition correctly? Ask the AI tutor to walk through the structure and check each algebraic step with you.

How do I tell quickly whether an equation is separable before starting the algebra?

Check whether the right side of dydx=(something)\dfrac{dy}{dx} = \text{(something)} can be written as a product f(x)⋅g(y)f(x)\cdot g(y), with no addition or subtraction mixing the two variables inside a single factor. If xx and yy appear added together, as in x+yx+y or inside sin⁡(x+y)\sin(x+y), it is not directly separable — though a substitution like v=x+yv=x+y may reduce it to a separable equation in vv and xx. If xx and yy are multiplied, divided, or appear as separate factors, it usually is separable.

Why is only one constant of integration needed when both sides get integrated?

Each indefinite integration technically introduces its own constant, say C1C_1 on the yy-side and C2C_2 on the xx-side. But these are both arbitrary real numbers, and the equation only involves their difference, C2−C1C_2 - C_1, which is itself just another arbitrary constant. So writing a single CC (usually placed on the xx-side) loses no generality — it is a bookkeeping simplification, not an approximation.

What is the point of the reducing substitution if the equation could just be solved as it is?

Equations like dydx=f(ax+by+c)\dfrac{dy}{dx} = f(ax+by+c) genuinely cannot be separated as written, because xx and yy are locked together inside ff. The substitution v=ax+by+cv=ax+by+c is not a shortcut but a necessity: it converts dydx\dfrac{dy}{dx} into an expression involving only dvdx\dfrac{dv}{dx}, at which point the equation becomes separable in vv and xx and can be solved by the ordinary method — after which vv must be replaced back by ax+by+cax+by+c to give the answer in the original variables.

Before this

Where this leads