Skip to content
Topos Uranos

Summary

This lesson solves the equations and inequalities in which the unknown appears inside trigonometric functions. It first proves the general solutions of sin⁡x=c\sin x = c, cos⁡x=c\cos x = c and tan⁡x=c\tan x = c on the whole real line, which extend by periodicity the solutions in one turn obtained with the inverse functions, and counts how many of them fall in a given interval. It then teaches how to reduce to those elementary equations the ones that are solved with identities, by factoring or with multiple angles, through the equality of two sines or of two cosines; it solves the equations linear in sine and cosine with the harmonic form, together with their solvability condition, and the homogeneous equations, and shows how dividing loses solutions and squaring adds them. Finally, it solves the inequalities in one turn with the unit circle and with sign charts, and extends them periodically.

Learning objectives

  1. Prove and apply the general solutions of the elementary equations of the sine, the cosine and the tangent, and count their solutions in a given interval.
  2. Reduce a trigonometric equation to elementary equations by means of identities, factoring, the equality of two sines or of two cosines and the harmonic form, and decide when an equation linear in sine and cosine has a solution.
  3. Recognize the solutions that are lost when dividing and those that are added when squaring or multiplying, and check each solution obtained in the original equation.
  4. Solve trigonometric inequalities in one turn with the unit circle and with sign charts, and write their general solution.

The general solutions of the elementary equations

A trigonometric equation is an equation in which the unknown appears in the argument of trigonometric functions, such as 2sin⁡2x−3sin⁡x+1=02\sin^2 x - 3\sin x + 1 = 0 or sin⁡3x=sin⁡x\sin 3x = \sin x; solving it means finding the set of all real numbers that satisfy it. In the lesson on the inverse trigonometric functions the theorem on elementary equations was proved, which gives the solutions of sin⁡x=c\sin x = c and of cos⁡x=c\cos x = c in [0,2π)\ico{0}{2\pi}. Now, the sine and the cosine have period 2π2\pi, and the tangent has period π\pi, according to the lesson on the trigonometric functions and their graphs; it follows that each solution repeats infinitely many times, and that to describe them all it is enough to know them in an interval whose length is one period. The fact that allows us to pass from such an interval to the whole line is a variant of the theorem of reduction to the first turn, proved in the lesson on angles and their measure.

LemmaReduction to one period

Let p>0p > 0 and aa be real numbers. For every real xx there is a unique integer kk such that x−kp∈[a,a+p)x - kp \in \ico{a}{a + p}.

A change of scale carries the interval [a,a+p)\ico{a}{a + p} onto the first turn, [0,2π)\ico{0}{2\pi}, and the theorem of reduction to the first turn gives the required integer. The figure uses a=1a = 1, p=52p = \frac{5}{2} and x=8x = 8.

Proof

  1. θ=2πp(x−a)=θ0+2kπ,0≤θ0<2π\theta = \frac{2\pi}{p} (x - a) = \dato{t}{\theta_0 + 2k\pi, \quad 0 \leq \theta_0 < 2\pi}

    Let θ\theta be the number obtained from xx by subtracting aa and multiplying by 2πp\frac{2\pi}{p}. By the theorem of reduction to the first turn, there exist a unique integer kk and a unique θ0\theta_0 in [0,2π)\ico{0}{2\pi} such that θ=θ0+2kπ\theta = \theta_0 + 2k\pi.

  2. x−kp=a+p2πθ0→a≤x−kp<a+px - kp = a + \frac{p}{2\pi} \theta_0 \rightarrow \resaltar{a \leq x - kp < a + p}

    We multiply by p2π\frac{p}{2\pi}, which is positive and preserves the inequalities, and add aa: the number x−kpx - kp lies in the required interval. In the figure, k=2k = 2 and x−2p=3x - 2p = 3.

  3. a≤x−jp<a+p→0≤θ−2jπ<2π→j=ka \leq x - jp < a + p \rightarrow 0 \leq \resaltar{\theta - 2j\pi} < 2\pi \rightarrow \resaltar{j = k}

    Uniqueness: if an integer jj also satisfies a≤x−jp<a+pa \leq x - jp < a + p, subtracting aa and multiplying by 2πp\frac{2\pi}{p} shows that θ−2jπ\theta - 2j\pi lies in [0,2π)\ico{0}{2\pi}, so that θ=(θ−2jπ)+2jπ\theta = \left( \theta - 2j\pi \right) + 2j\pi is another decomposition like the one in the first step; by the uniqueness given by the reduction theorem, j=kj = k.

TheoremGeneral solutions

Let cc be a real number with ∣c∣≤1\abs{c} \leq 1. The solutions of sin⁡x=c\sin x = c are the numbers arcsin⁡c+2kπ\arcsin c + 2k\pi and π−arcsin⁡c+2kπ\pi - \arcsin c + 2k\pi, and those of cos⁡x=c\cos x = c are the numbers arccos⁡c+2kπ\arccos c + 2k\pi and −arccos⁡c+2kπ-\arccos c + 2k\pi, where kk ranges over the integers. For every real cc, the solutions of tan⁡x=c\tan x = c are the numbers arctan⁡c+kπ\arctan c + k\pi, with kk an integer. If ∣c∣>1\abs{c} > 1, the sine and cosine equations have no solution.

In the notation of sets and logical connectives, the theorem is written as follows:

sin⁡x=c↔x=arcsin⁡c+2kπ∨x=π−arcsin⁡c+2kπ,k∈Z\sin x = c \leftrightarrow x = \arcsin c + 2k\pi \vee x = \pi - \arcsin c + 2k\pi, \quad k \in \mathbb{Z}
cos⁡x=c↔x=±arccos⁡c+2kπ,tan⁡x=c↔x=arctan⁡c+kπ,k∈Z\cos x = c \leftrightarrow x = \pm \arccos c + 2k\pi, \qquad \tan x = c \leftrightarrow x = \arctan c + k\pi, \quad k \in \mathbb{Z}

We work first on [−π2,3π2)\ico{-\frac{\pi}{2}}{\frac{3\pi}{2}}, which is one turn, split in two: on [−π2,π2]\icc{-\frac{\pi}{2}}{\frac{\pi}{2}} the sine is strictly increasing, by the theorem of monotonicity on the principal intervals, which the lesson on the inverse functions derives from the axiom of traversal by arc length, and on (π2,3π2)\ioo{\frac{\pi}{2}}{\frac{3\pi}{2}} the symmetry of supplementary angles acts, sin⁡(π−x)=sin⁡x\sin(\pi - x) = \sin x, from the lesson on trigonometric ratios. The lemma extends the result to the whole line, and the cosine equation is reduced to the sine equation. The figures use c=12c = \frac{1}{2}, with arcsin⁡c=π6\arcsin c = \frac{\pi}{6}.

Proof

  1. −π2≤x≤π2→(sin⁡x=c↔x=α)\dato{a}{-\frac{\pi}{2} \leq x \leq \frac{\pi}{2} \rightarrow \left( \sin x = c \leftrightarrow x = \alpha \right)}

    Let α=arcsin⁡c\alpha = \arcsin c, which lies in [−π2,π2]\icc{-\frac{\pi}{2}}{\frac{\pi}{2}} and satisfies sin⁡α=c\sin \alpha = c. Since the sine is strictly increasing on that interval, it does not take the same value twice there: α\alpha is the only solution in the first half of the turn. On the unit circle, P(α)P(\alpha) is the point of the right half at height cc.

  2. sin⁡x=sin⁡(π−x)=c↔π−x=α↔x=π−α\sin x = \resaltar{\sin(\pi - x)} = c \leftrightarrow \resaltar{\pi - x = \alpha} \leftrightarrow \dato{b}{x = \pi - \alpha}

    If xx lies in (π2,3π2)\ioo{\frac{\pi}{2}}{\frac{3\pi}{2}}, the angle π−x\pi - x lies in (−π2,π2)\ioo{-\frac{\pi}{2}}{\frac{\pi}{2}} and has the same sine as xx. By the preceding step, xx is a solution exactly when π−x=α\pi - x = \alpha: the horizontal line at height cc cuts the left half at the symmetric point P(π−α)P(\pi - \alpha).

  3. sin⁡x=sin⁡x0=c↔x=α+2kπ∨x=π−α+2kπ\sin x = \resaltar{\sin x_0} = c \leftrightarrow \dato{g}{\resaltar{x = \alpha + 2k\pi \vee x = \pi - \alpha + 2k\pi}}

    Now let xx be any real number. By the lemma, with p=2πp = 2\pi and a=−π2a = -\frac{\pi}{2}, we write x=x0+2kπx = x_0 + 2k\pi with x0x_0 in the turn studied, and by periodicity sin⁡x=sin⁡x0\sin x = \sin x_0. Therefore, xx is a solution if and only if x0x_0 is α\alpha or π−α\pi - \alpha.

  4. α=±π2→π−α=α∨π−α=α+2π\alpha = \pm \frac{\pi}{2} \rightarrow \resaltar{\pi - \alpha = \alpha} \vee \resaltar{\pi - \alpha = \alpha + 2\pi}

    Conversely, every number of those two families is a solution, again by periodicity. On the graph of the sine, the line y=cy = c cuts the curve at two points in each period. If c=1c = 1 or c=−1c = -1, the two families coincide, because π−α\pi - \alpha differs from α\alpha by a multiple of 2π2\pi, and there is a single intersection point per period.

  5. π2−x=π2−arccos⁡c+2kπ∨π2−x=π2+arccos⁡c+2kπ\frac{\pi}{2} - x = \resaltar{\frac{\pi}{2} - \arccos c} + 2k\pi \vee \frac{\pi}{2} - x = \resaltar{\frac{\pi}{2} + \arccos c} + 2k\pi

    For the cosine we use the symmetry of complementary angles, cos⁡x=sin⁡(π2−x)\cos x = \sin\left( \frac{\pi}{2} - x \right), and the theorem on complementary angles from the lesson on the inverse functions, arcsin⁡c=π2−arccos⁡c\arcsin c = \frac{\pi}{2} - \arccos c. Applying the preceding step to the angle π2−x\frac{\pi}{2} - x yields two families.

  6. cos⁡x=c↔x=±arccos⁡c+2kπ\cos x = c \leftrightarrow \resaltar{x = \pm \arccos c + 2k\pi}

    Solving for xx, the families are arccos⁡c−2kπ\arccos c - 2k\pi and −arccos⁡c−2kπ-\arccos c - 2k\pi; since −k-k ranges over the integers when kk does, they are those of the statement. On the circle, the points P(β)P(\beta) and P(−β)P(-\beta), with β=arccos⁡c\beta = \arccos c, are the two intersections of the vertical line with abscissa cc.

  7. tan⁡x=tan⁡x0=c↔x=arctan⁡c+kπ\tan x = \resaltar{\tan x_0} = c \leftrightarrow \resaltar{x = \arctan c + k\pi}

    For the tangent, with period π\pi, the lemma is applied with p=πp = \pi and a=−π2a = -\frac{\pi}{2}: if x=x0+kπx = x_0 + k\pi with x0x_0 in (−π2,π2)\ioo{-\frac{\pi}{2}}{\frac{\pi}{2}}, where the tangent is a bijection onto R\mathbb{R}, the only solution is x0=arctan⁡cx_0 = \arctan c; the point −π2-\frac{\pi}{2} does not count, because the tangent is not defined there. If ∣c∣>1\abs{c} > 1, neither the sine nor the cosine reaches the value cc, because their values lie in [−1,1]\icc{-1}{1}.

For example, since arcsin⁡(−12)=−π6\arcsin\left( -\frac{1}{2} \right) = -\frac{\pi}{6}, the solutions of sin⁡x=−12\sin x = -\frac{1}{2} are those of the two families below; with k=1k = 1 in the first and k=0k = 0 in the second we obtain the two that lie in [0,2π)\ico{0}{2\pi}, namely 11π6\frac{11\pi}{6} and 7π6\frac{7\pi}{6}, as in the lesson on the inverse functions.

sin⁡x=−12↔x=−π6+2kπ∨x=7π6+2kπ,k∈Z\sin x = -\frac{1}{2} \leftrightarrow x = -\frac{\pi}{6} + 2k\pi \vee x = \frac{7\pi}{6} + 2k\pi, \quad k \in \mathbb{Z}

The solutions of sin⁡x=0\sin x = 0 form two families, 2kπ2k\pi and π+2kπ\pi + 2k\pi, which together are the integer multiples of π\pi; those of cos⁡x=0\cos x = 0 are ±π2+2kπ\pm\frac{\pi}{2} + 2k\pi, which together are the numbers π2+kπ\frac{\pi}{2} + k\pi. These two compact descriptions are often used:

sin⁡x=0↔x=kπ,cos⁡x=0↔x=π2+kπ,k∈Z\sin x = 0 \leftrightarrow x = k\pi, \qquad \cos x = 0 \leftrightarrow x = \frac{\pi}{2} + k\pi, \quad k \in \mathbb{Z}

The number of solutions in an interval

Questions about trigonometric equations are usually restricted to an interval, almost always [0,2π)\ico{0}{2\pi}. The lemma says that each family of the form x1+2kπx_1 + 2k\pi has exactly one member in [a,a+2π)\ico{a}{a + 2\pi}, whatever aa may be; from this follows the number of solutions of the elementary equations in any turn.

TheoremSolutions in one turn

Let aa and cc be real numbers. If ∣c∣<1\abs{c} < 1, each of the equations sin⁡x=c\sin x = c and cos⁡x=c\cos x = c has exactly two solutions in [a,a+2π)\ico{a}{a + 2\pi}, and if ∣c∣=1\abs{c} = 1, exactly one. The equation tan⁡x=c\tan x = c has exactly one solution in [a,a+π)\ico{a}{a + \pi}.

Each family of solutions contributes a single member to the turn, by the lemma; it suffices to know when the two sine families, or the two cosine families, are one and the same family. The figures use a=1a = 1, with c=12c = \frac{1}{2} and c=1c = 1.

Proof

  1. x1+2jπ∈[a,a+2π)↔j=−k\dato{u}{x_1 + 2j\pi \in \ico{a}{a + 2\pi} \leftrightarrow j = -k}

    Let x1x_1 be a real number. By the lemma, with p=2πp = 2\pi, there is a unique integer kk such that x1−2kπx_1 - 2k\pi lies in [a,a+2π)\ico{a}{a + 2\pi}: the family of the numbers x1+2jπx_1 + 2j\pi, with jj an integer, has exactly one member in the turn.

  2. α+2kπ=π−α+2jπ↔α=π2+(j−k)π↔α=±π2↔c=±1\alpha + 2k\pi = \pi - \alpha + 2j\pi \leftrightarrow \resaltar{\alpha = \frac{\pi}{2} + (j - k)\pi} \leftrightarrow \alpha = \pm \frac{\pi}{2} \leftrightarrow \resaltar{c = \pm 1}

    Let α=arcsin⁡c\alpha = \arcsin c, in [−π2,π2]\icc{-\frac{\pi}{2}}{\frac{\pi}{2}}. If the two sine families have a common member, they are the same family, and this happens exactly when α\alpha is π2\frac{\pi}{2} or −π2-\frac{\pi}{2}, that is, when c=±1c = \pm 1. Therefore, if ∣c∣<1\abs{c} < 1 the two families contribute two distinct solutions to the turn, and if ∣c∣=1\abs{c} = 1, only one.

  3. β+2kπ=−β+2jπ↔β=(j−k)π↔β∈{0,π}↔c=±1\beta + 2k\pi = -\beta + 2j\pi \leftrightarrow \resaltar{\beta = (j - k)\pi} \leftrightarrow \beta \in \left\{ 0, \pi \right\} \leftrightarrow \resaltar{c = \pm 1}

    Let β=arccos⁡c\beta = \arccos c, in [0,π]\icc{0}{\pi}. The two cosine families have a common member exactly when β\beta is an integer multiple of π\pi, that is, when β\beta is 00 or π\pi, which is equivalent to c=±1c = \pm 1; the count is the same as for the sine.

  4. −π2<arctan⁡c<π2→arctan⁡c+kπ≠π2+jπ-\frac{\pi}{2} < \arctan c < \frac{\pi}{2} \rightarrow \resaltar{\arctan c + k\pi \neq \frac{\pi}{2} + j\pi}

    For the tangent there is a single family, with spacing π\pi, and the lemma, with p=πp = \pi, gives it exactly one member in [a,a+π)\ico{a}{a + \pi}. That member is a solution, because arctan⁡c+kπ\arctan c + k\pi is never of the form π2+jπ\frac{\pi}{2} + j\pi, where the tangent is not defined: arctan⁡c\arctan c lies in (−π2,π2)\ioo{-\frac{\pi}{2}}{\frac{\pi}{2}}.