Differential Equations › Section D · Laplace transforms › Topic D.3

Inverse Laplace transforms

Solving an equation with Laplace transforms ends with a function of s, and the answer you want is a function of t. Inverting a transform is a matching game: reshape the expression with algebra until every piece is a table entry.

Going back from s to t

The inverse transform runs the table in reverse: if \(\mathcal{L}\{f(t)\} = F(s)\), then \(\mathcal{L}^{-1}\{F(s)\} = f(t)\). For this to make sense, a transform must come from only one function. Lerch's theorem guarantees this for continuous functions: two continuous functions with the same transform are identical. Piecewise continuous functions with the same transform can differ only at isolated points, such as their jumps, which never matters in applications. There is a contour-integral formula for the inverse, but it needs complex analysis. In practice every inverse in this course is found by rearranging \(F(s)\) into table entries, using the transforms from the topic Computing Laplace transforms.

Linearity and fixing constants

The inverse is linear too: \(\mathcal{L}^{-1}\{aF + bG\} = af + bg\). So a sum is inverted term by term. The main trap is that table entries have their constants in particular places, and you must supply them. Take \(\frac{1}{s^2 + 9}\). The sine entry \(\frac{3}{s^2 + 9}\) needs a 3 on top, so write \(\frac{1}{3} \cdot \frac{3}{s^2 + 9}\), giving \(\frac{1}{3}\sin 3t\). Similarly \(\frac{1}{s^4}\) needs \(3! = 6\) on top to match \(t^3\), so its inverse is \(\frac{t^3}{6}\). Multiplying and dividing by the missing constant is always allowed.

Worked example: straight from the table

Find the inverse of

F(s) = \frac{3}{s - 2} + \frac{8}{s^3} + \frac{2s + 5}{s^2 + 9}

1. The first term matches \(e^{at}\) with \(a = 2\): it gives \(3e^{2t}\). 2. The second needs \(2! = 2\) on top: \(\frac{8}{s^3} = 4 \cdot \frac{2}{s^3}\), which gives \(4t^2\). 3. Split the third over its denominator: \(\frac{2s}{s^2 + 9}\) gives \(2\cos 3t\), and \(\frac{5}{s^2 + 9} = \frac{5}{3} \cdot \frac{3}{s^2 + 9}\) gives \(\frac{5}{3}\sin 3t\). 4. So

f(t) = 3e^{2t} + 4t^2 + 2\cos 3t + \frac{5}{3}\sin 3t

Check: transforming each term again returns the original three fractions.

Partial fractions: a review

Most transforms that arise from differential equations are rational functions, a polynomial over a polynomial, with the top of lower degree. Partial fractions split them into pieces the table recognises. Factor the denominator, then write one term per factor. A linear factor \(s - a\) gives \(\frac{A}{s - a}\). A repeated factor \((s - a)^k\) gives \(k\) terms, \(\frac{A_1}{s - a} + \dots + \frac{A_k}{(s - a)^k}\). An irreducible quadratic gives \(\frac{Bs + C}{s^2 + ps + q}\), with a linear numerator. Find the constants by clearing denominators and substituting convenient values of \(s\), or by matching coefficients.

Worked example: distinct linear factors

Find the inverse of \(F(s) = \frac{5s + 1}{(s - 1)(s + 2)}\). 1. Write

\frac{5s + 1}{(s - 1)(s + 2)} = \frac{A}{s - 1} + \frac{B}{s + 2}

2. Clear denominators: \(5s + 1 = A(s + 2) + B(s - 1)\). 3. At \(s = 1\): \(6 = 3A\), so \(A = 2\). At \(s = -2\): \(-9 = -3B\), so \(B = 3\). 4. Invert: \(f(t) = 2e^t + 3e^{-2t}\). Check: at \(s = 0\), the original is \(\frac{1}{-2} = -\frac{1}{2}\), and the split form gives \(\frac{2}{-1} + \frac{3}{2} = -\frac{1}{2}\). Each root of the denominator becomes an exponential in the answer.

Worked example: a repeated factor

Find the inverse of \(F(s) = \frac{s + 3}{s^2(s + 1)}\). 1. The factor \(s^2\) is repeated, so write \(\frac{A}{s} + \frac{B}{s^2} + \frac{C}{s + 1}\). 2. Clear denominators:

s + 3 = As(s + 1) + B(s + 1) + Cs^2

3. At \(s = 0\): \(B = 3\). At \(s = -1\): \(C = 2\). Comparing the \(s^2\) coefficients: \(0 = A + C\), so \(A = -2\). 4. Invert: \(f(t) = -2 + 3t + 2e^{-t}\). Check: over the common denominator the numerator is \(-2s(s + 1) + 3(s + 1) + 2s^2\), which simplifies to \(s + 3\). A repeated factor \((s - a)^2\) produces a term \(te^{at}\), just as \(\frac{1}{(s - 3)^2}\) inverts to \(te^{3t}\).

Completing the square

When a quadratic denominator has no real roots, complete the square so the shifting theorem applies. Then rewrite the numerator in terms of the same shifted variable. Sine and cosine entries both need \(s - a\) in the denominator, and the cosine entry needs exactly \(s - a\) on top, so the numerator must be expressed as a multiple of \(s - a\) plus a constant. The result is a damped oscillation \(e^{at}(\dots\cos bt + \dots\sin bt)\), where \(a\) is the real part of the complex roots of the denominator. Skipping this step, and trying to read sines straight off \(s^2 + 4s + 13\), is the most common error here.

Worked example: a damped oscillation

Find the inverse of \(F(s) = \frac{s + 7}{s^2 + 4s + 13}\). 1. Complete the square: \(s^2 + 4s + 13 = (s + 2)^2 + 9\). 2. Rewrite the numerator using \(s + 2\): \(s + 7 = (s + 2) + 5\). 3. Split:

F(s) = \frac{s + 2}{(s + 2)^2 + 9} + \frac{5}{3} \cdot \frac{3}{(s + 2)^2 + 9}

4. Each piece is a table entry shifted by \(a = -2\), so \(f(t) = e^{-2t}\big(\cos 3t + \frac{5}{3}\sin 3t\big)\). Check: recombining the two fractions puts \((s + 2) + 5 = s + 7\) back on top, and \(f(0) = 1\) because only the cosine is nonzero at \(t = 0\).

Mixing linear and quadratic factors

Find the inverse of \(F(s) = \frac{8}{s(s^2 + 4)}\). Write \(\frac{A}{s} + \frac{Bs + C}{s^2 + 4}\) and clear denominators: \(8 = A(s^2 + 4) + (Bs + C)s\). At \(s = 0\), \(A = 2\). Matching the \(s^2\) terms gives \(A + B = 0\), so \(B = -2\), and matching the \(s\) terms gives \(C = 0\). So \(F(s) = \frac{2}{s} - \frac{2s}{s^2 + 4}\) and \(f(t) = 2 - 2\cos 2t\). Check: \(f(0) = 0\), and \(f\) stays between 0 and 4, a pattern that returns when a constant force drives an undamped spring.

Common mistakes to avoid

Inverting \(\frac{1}{s^2 + 9}\) as \(\sin 3t\) instead of \(\frac{1}{3}\sin 3t\), or \(\frac{1}{s^4}\) as \(t^3\) instead of \(\frac{t^3}{6}\). Reading \(b\) from \(s^2 + b^2\) as 9 rather than 3. Giving a repeated factor only one term, or giving a quadratic factor a constant numerator instead of \(Bs + C\). Completing the square in the denominator but leaving the numerator in terms of \(s\). Getting a shift sign backwards: \(\frac{1}{s + 2}\) inverts to \(e^{-2t}\), not \(e^{2t}\).

Key takeaways

\(\mathcal{L}^{-1}\) is linear, and for continuous functions it is unique. Supply missing constants so each term matches a table entry exactly. Use partial fractions on rational transforms: one term per linear factor, \(k\) terms for a factor repeated \(k\) times, and a linear numerator over each irreducible quadratic. Complete the square on quadratics with complex roots and rewrite the numerator in the shifted variable. Real roots give exponentials, repeated roots give extra powers of \(t\), and complex roots give damped sines and cosines.

Flashcards

Flashcard

L⁻¹{F(s)}

The function f(t) whose transform is F(s)

Back from s to t.

Test your understanding

Question 1 of 5

What is the inverse Laplace transform of 6/(s² + 4)?

Question 2 of 5

What is the inverse Laplace transform of 2/s⁴?

Question 3 of 5

Use partial fractions to invert (3s + 1)/((s + 1)(s - 1)).

Question 4 of 5

What is the inverse Laplace transform of (s + 1)/(s² + 2s + 10)?

Question 5 of 5

Which set of partial fractions is the right starting point for 1/((s - 2)²(s² + 1))?

← Previous topicNext topic →