Math 555: Differential Equations I




Lecture Companion

§2.2: Separable FODE




A first order differential equation (FODE) is called separable if it can be written in the form

$$ M(x)\,dx + N(y)\, dy = 0. \tag{1} $$
In a real sense, this is the simplest possible form for a FODE to take. The solution curves are obtained as implicit functions $y = y(x)$ by merely integrating equation $(1)$. We obtain

$$ \int M(x)\,dx + \int N(y)\, dy = C. \tag{2} $$
When possible, we shall usually want to solve for $y$ as an honest function of $x$. In case the differential equation comes with an initial condition, then it will sometimes require careful attention in determining the proper form of the solution function and its domain.

We look at some examples below.




Example 1. Consider the differential equation

$$ \frac{dy}{dx} = \frac{xy}{x^2 + 1}. \tag{3} $$
This equation separates to become

$$ -\frac{x}{x^2 + 1}\,dx + \frac{1}{y}\,dy = 0. \tag{4} $$
Integrating equation $(4)$ yields

$$ -\tfrac{1}{2}\ln(x^2 + 1) + \ln y = C. \tag{5} $$
Solving equation $(5)$ for $y$ gives

$$ y(x) = C\sqrt{x^2 + 1}. \tag{6} $$
We plot the slope field and some integral curves below.






Example 2. Consider the initial value problem (IVP)

$$ \begin{cases} \dfrac{dy}{dt} = \dfrac{6t^2 + 4t}{2y - 4}, \\[0.5 ex] y(0) = -3. \end{cases} \tag{7} $$
This can be separated to become

$$ (2y - 4)\,dy = (6t^2 + 4t)\, dt, \tag{8} $$
and the Fundamental Theorem of Calculus may be applied with the initial values to give

$$ \int_{-3}^y 2u - 4 \, du = \int_{0}^t 6s^2 + 4s\, ds. \tag{9} $$
This is easy to integrate, but let's use Sage to do it for us. The left-hand integral is



The right-hand integral is



Therefore, the implicit solution is



And solving for $y$ explicitly yields



Notice that this IVP has two "solutions": one with a positive square root and one with a negative square root. Here "solutions" is in quotes because in reality there can only be one solution. The subtle issue here is that we are looking for a solution curve by trying to find a solution function. But what if the curve is not given by the graph of a function? Then you arrive at a conundrum like our current one.

We will interpret this situation as follows. Of the two functions that we've found, only one of them is acceptable as the solution function: the one that obeys the initial condition. However, we will regard the concatenation of the two graphs as a single solution curve to the IVP.

As always in math, context and perspective matter. Be mindful of whether you want a solution function or a solution curve when completing these problems in class. In practical applications, both points-of-view can be fruitful.

We check the initial values.




We see that only the negative square root obeys the initial condition. Therefore the solution is



We plot the slope field and both functions. The solution function is blue, while the entire curve may be regarded as the solution curve. ${\tt Sage}$ has trouble plotting the portion of the curve that has close-to-vertical tangent lines. You should regard this graph as connected in your mind.



Another option is to use ${\tt Sage}$ to find the domain of the solution function(s), and not let the $t$-values extend outside the domain.



Notice that the imaginary part of the middle solution is very small. Indeed, this is again a numerical artifact. A cubic polynomial with integer coefficients cannot have three imaginary solutions.

We use the real part of this solution as the left-hand bound for the $t$-values and re-plot the solution curves.






Back to main page


Your use of Wichita State University content and this material is subject to our Creative Common License.