Why Angular Bounds Need Care
A polar equation \(r=f(\theta)\) assigns a point in the plane to each angle in its domain. Selecting bounds for an integral over a polar curve is therefore not just a matter of choosing angles that appear in a sketch. The interval must cover the intended curve or region, and it must not trace the same part more than once.
Zeros of \(r\), changes in its sign, and symmetry provide useful clues. But none of these clues should be used without checking what the resulting points are. A negative radius points in the direction opposite to its angle, so two apparently different angular intervals can trace the same geometric curve.
Polar Coordinates and Repeated Points
For a polar equation \(r=f(\theta)\), the point traced at angle \(\theta\) has Cartesian coordinates
For a nonzero radius, the same point has the equivalent polar descriptions \((r,\theta)\) and \((-r,\theta+\pi)\). Angles can also differ by any integer multiple of \(2\pi\). Thus, comparing angles alone is not enough to decide whether two parameter values give different points.
For two nonzero polar descriptions \((r_1,\theta_1)\) and \((r_2,\theta_2)\) to give the same point, one of the following must hold: their radii are equal and their angles differ by a multiple of \(2\pi\), or their radii are opposites and their angles differ by an odd multiple of \(\pi\). In particular, if the radius keeps one strict sign on an interval whose length is less than \(2\pi\), distinct angles in that interval cannot describe the same nonzero point. This is a useful sufficient check, but it does not settle whether two different sign intervals overlap geometrically.
Use Zeros to Find Candidate Bounds
A zero of \(f(\theta)\) sends the curve to the pole. Consecutive zeros often mark a curve portion that begins and ends at the pole, so they are natural candidates for bounds that isolate one loop. They are candidates rather than an automatic answer: another interval might trace the same loop, or an interval might include more than the requested portion.
To locate these candidates, solve \(f(\theta)=0\) and order the resulting angles over a stated range. Then test the sign of \(f\) between consecutive zeros. A sign change means that the radius points in the opposite direction after the zero. That can correspond to another loop, but it can also produce a portion already traced elsewhere.
Solve \(f(\theta)=0\) over an angular range large enough to see the relevant repetitions.
Test \(f\) between consecutive zeros to identify intervals of positive and negative radius.
Use the zeros, any symmetry, and a sketch to isolate the requested curve portion.
Compare parameter values using the equivalence of polar coordinates and confirm the interval covers the intended portion once.
Worked Example: Isolating the Inner Loop of a Limacon
Worked Example: Isolating the Inner Loop of a Limacon
Choose bounds that trace the inner loop of \(r=1+2\cos\theta\) once.
First find the zeros:
In \([0,2\pi]\), the solutions are \(\theta=2\pi/3\) and \(\theta=4\pi/3\). Between them, test the midpoint \(\theta=\pi\):
The radius is negative between the zeros, so the candidate inner-loop interval is \([2\pi/3,4\pi/3]\). To confirm the sign throughout the open interval, note that \(\cos\theta<-1/2\) there, giving \(1+2\cos\theta<0\).
The interval length is \(4\pi/3-2\pi/3=2\pi/3\), less than \(2\pi\), and the radius is strictly negative in its interior. Distinct interior angles therefore cannot represent the same nonzero point: with radii of the same sign, equality would require their angles to differ by \(2\pi\), which this interval does not allow. The endpoints both give \(r=0\), so they meet at the pole and close the loop.
As a check on the orientation reversal caused by the negative radius, at the midpoint \(\theta=\pi\), the polar point \((-1,\pi)\) is Cartesian \((1,0)\):
Thus the selected interval runs from the pole around the inner loop to the pole, with its interior traced once.
Negative Radius Can Reveal a Duplicate Trace
A negative-radius interval is not automatically a new loop. The equivalence \((r,\theta)=(-r,\theta+\pi)\) means that the points from a negative-radius interval may coincide with points from a positive-radius interval. Checking the curve across the entire range of interest is therefore essential.
Worked Example: Detecting a Repeated Circle
Consider \(r=2\cos\theta\). Find the zeros in \([0,2\pi]\):
The radius is positive on \((-\pi/2,\pi/2)\) and negative on \((\pi/2,3\pi/2)\), with the intervals understood modulo \(2\pi\). To check whether they give different traces, compare \(\theta\) with \(\theta+\pi\):
The angle increases by \(\pi\) while the radius changes sign, so these two polar descriptions give the same Cartesian point. This identity shows that opposite-sign parameter values trace the same circle, rather than separate loops.
The circle can also be checked directly. From \(r=2\cos\theta\), use \(x=r\cos\theta\), \(y=r\sin\theta\), and \(r^2=x^2+y^2\):
This is the circle centered at \((1,0)\) with radius \(1\). The interval \([-\pi/2,\pi/2]\) traces it once: its Cartesian parametrization is \(x=1+\cos 2\theta\), \(y=\sin 2\theta\). As \(\theta\) runs from \(-\pi/2\) to \(\pi/2\), the angle \(2\theta\) runs from \(-\pi\) to \(\pi\), covering the circle once, with the two endpoints both at the pole. Using both sign intervals would trace that same circle twice.
Symmetry Helps, but Does Not Replace a Trace Check
Symmetry can reduce the work of identifying candidate bounds. If a curve has a reflection or rotational symmetry, one portion may generate other portions. However, symmetry can also make separate parameter intervals represent the same geometric piece. A symmetry statement about the graph does not by itself tell you whether an interval traces the graph once or repeatedly.
Worked Example: Choosing Bounds for One Petal of a Rose
For \(r=2\cos(3\theta)\), find bounds for the petal centered on the positive \(x\)-axis. The radius is largest at \(\theta=0\):
The neighboring zeros satisfy \(\cos(3\theta)=0\). The closest solutions on either side of \(0\) are \(3\theta=-\pi/2\) and \(3\theta=\pi/2\), so
Between these bounds, \(-\pi/2<3\theta<\pi/2\), so \(\cos(3\theta)>0\) and \(r>0\). The radius is zero at both endpoints. The interval length is \(\pi/3<2\pi\), so its interior cannot repeat a nonzero point: the radius stays positive, and equality at distinct angles would require a difference of \(2\pi\). Hence \([-\pi/6,\pi/6]\) traces this petal once.
There is also a symmetry check for repetitions. Since \(\cos(3(\theta+\pi))=-\cos(3\theta)\),
Thus the interval shifted by \(\pi\) can describe the same petal using negative radii. This confirms why choosing a convenient zero-to-zero interval alone is not enough: one must compare it with other intervals that may represent the same points.
A Practical Test for Bounds
When the requested object is a specific loop or region, state exactly which curve portion the bounds are meant to cover. Then use zeros and signs to identify the relevant parameter interval. Finally, check possible repetitions both within that interval and, when negative radii occur, against other angular intervals.
For a sign-constant interval of length less than \(2\pi\), the uniqueness check for its nonzero points is especially direct. If the interval includes both positive and negative radii, or if the candidate range spans several periods, the direct equivalence test matters: look for parameter values whose angles differ by \(\pi\) and whose radii are opposites, as well as values whose angles differ by \(2\pi\) and whose radii agree.
These checks are important whenever angular bounds will be used to measure a curve or region. If the bounds include a repeated trace, an integral over them counts that portion more than once; if they omit an interval, part of the intended object is missed. The bounds are therefore part of the mathematical setup, not a cosmetic choice.
Check Your Understanding
For each question, use zeros, signs, or polar-coordinate equivalence to justify your answer.
- For \(r=1+2\cos\theta\), find the two zeros in \([0,2\pi]\). On which interval between them is the radius negative?
- Why does a sign-constant interval of length less than \(2\pi\) not repeat a nonzero point of the curve?
- For \(r=2\cos\theta\), explain why shifting an angle by \(\pi\) can give the same Cartesian point.
- For \(r=2\cos(3\theta)\), verify that the zeros adjacent to the maximum at \(\theta=0\) are \(-\pi/6\) and \(\pi/6\).
- What extra check is needed when a candidate interval includes both positive and negative radii?