Locus of Incentre Problem: Find Ellipse

  • MHB
  • Thread starter Saitama
  • Start date
In summary, the problem asks for the locus of the center of an incircle for a triangle given the base and sum of its sides. This locus is shown to be an ellipse by using a coordinate geometry approach, where the vertices of the triangle are parametrized and the angle bisectors are used to determine the slope and equation of the lines containing the incenter. The resulting equation for the locus is an ellipse with foci at the endpoints of the base.
  • #1
Saitama
4,243
93
Problem:
Given the base of a triangle and sum of its sides, then the locus of the centre of its incircle is:

A)straight line
B)circle
C)ellipse
D)hyperbola

Attempt:
Since the given answer is ellipse, I tried showing that the sum of the distances of incentre from the end points of base is constant.

The side lengths opposite to vertex A, B and C are a, b and c respectively. As per the question, a (base) and a+b+c is given. The distance IB is $r/\sin(B/2)$ and IC is $r/\sin(C/2)$, where $r$ is the radius of incircle. Hence,

$$IB+IC=r\left(\frac{1}{\sin(B/2)}+\frac{1}{\sin(C/2)}\right)\,\,\,\, (*)$$
I use the following formulas:
$$r=\frac{\Delta}{s}=\sqrt{\frac{(s-a)(s-b)(s-c)}{s}}$$
$$\sin\left(\frac{B}{2}\right)=\sqrt{\frac{(s-a)(s-c)}{ac}}$$
$$\sin\left(\frac{C}{2}\right)=\sqrt{\frac{(s-a)(s-b)}{ab}}$$
Substituting in (*) and simplifying,
$$IB+IC=\sqrt{\frac{a}{s}}\left(\sqrt{c(s-b)}+\sqrt{b(s-c)}\right)$$
The terms outside the parentheses is constant, I am unable to prove that the terms inside the parentheses are a constant too. :(

Any help is appreciated. Thanks!
 

Attachments

  • triangle.png
    triangle.png
    4.1 KB · Views: 73
Mathematics news on Phys.org
  • #2
This is an interesting problem. The points $B$ and $C$ are fixed, and the triangle $ABC$ has a constant perimeter. So the point $A$ must lie on an ellipse with foci at $B$ and $C$. The incentre $I$ also lies on an ellipse, but the foci of that ellipse are not at $B$ and $C$, so you should not expect $\overline{IB} + \overline{IC}$ to be constant.

The only way I can solve the problem is to use coordinate geometry. Let $A$ be the point $(a\cos\theta,b\sin\theta)$, so that \(\displaystyle \frac{x^2}{a^2} + \frac{y^2}{b^2} = 1\) is the ellipse on which $A$ lies. Then $B$ and $C$ are the foci of that ellipse, with coordinates $B = (-ae,0)$, $C = (ae,0)$, where $e$ is the eccentricity of the ellipse, given by $b^2 = a^2(1-e^2)$. The lengths of the sides of the triangle are $\overline{BC} = 2ae$, $\overline{CA} = a(1-e\cos\theta)$, $\overline{AB} = a(1+e\cos\theta)$.

I then used a very useful formula for the coordinates of the incentre that I found here. In words, it says that you take the coordinates of the three vertices of the triangle, multiply each of them by the length of the opposite side, add them, and finally divide by the perimeter of the triangle. I leave you to apply that formula to our triangle here. I found that it gives $I = \Bigl(ae\cos\theta, \dfrac{be\sin\theta}{1+e}\Bigr).$ Thus $I$ lies on the ellipse \(\displaystyle \frac{x^2}{a^2e^2} + \frac{(1+e)^2y^2}{b^2e^2} = 1\), which passes through the points $B$ and $C$ (and so certainly does not have its foci at those points).
 
  • #3
Like Opalg, I found this problem interesting and find a coordinate geometry approach to be the most straightforward.

I decided to orient the fixed base of the triangle along the $x$-axis whose center is at the origin, and let its length be $2f$. From the definition of an ellipse, we know the locus of the movable vertex is an ellipse, which I describe by:

\(\displaystyle \frac{x^2}{a^2}+\frac{y^2}{b^2}=1\)

Please refer to the following diagram:

View attachment 1857

Point $P$ is an arbitrary point on the ellipse, and point $I$ is the associated incenter. It is the intersection of the two angle bisectors $\ell_1$ and $\ell_2$.

Let's parametrize point $P$ as follows:

\(\displaystyle x(t)=t\)

\(\displaystyle y(t)=\frac{b}{a}\sqrt{a^2-t^2}\)

where \(\displaystyle -a\le t\le a\)

While we are only considering the top half of the ellipse here, a similar argument could be used for the bottom half.

For these two lines, we know they pass through the foci of the ellipse, so we need to determine their slopes.

For line $\ell_1$, let's first look at the slope of the line through the points:

\(\displaystyle \left(t,\frac{b}{a}\sqrt{a^2-t^2} \right)\) and \(\displaystyle (f,0)\)

It's angle of inclination is therefore:

\(\displaystyle \theta_1=\tan^{-1}\left(\frac{b\sqrt{a^2-t^2}}{a(t-f)} \right)\) where \(\displaystyle t\ne f\)

And so the angle of inclination of $\ell_1$ is:

\(\displaystyle \frac{\theta_1}{2}=\frac{1}{2}\tan^{-1}\left(\frac{b\sqrt{a^2-t^2}}{a(t-f)} \right)\)

Hence, the slope of $\ell_1$ is found using a half-angle identity for tangent:

\(\displaystyle m_1=\frac{\sin\left(\tan^{-1}\left(\dfrac{b\sqrt{a^2-t^2}}{a(t-f)} \right) \right)}{1+\cos\left(\tan^{-1}\left(\dfrac{b\sqrt{a^2-t^2}}{a(t-f)} \right) \right)}\)

Using the fact that \(\displaystyle f^2=a^2-b^2\) we find:

\(\displaystyle m_1=-\frac{b}{a+f}\sqrt{\frac{a+t}{a-t}}\)

Armed now with the slope and the point through which it passes, line $\ell_1$ is given by (using the point-slope formula):

\(\displaystyle y=\frac{b}{a+f}\sqrt{\frac{a+t}{a-t}}(f-x)\)

For line $\ell_2$, let's first look at the slope of the line through the points:

\(\displaystyle \left(t,\frac{b}{a}\sqrt{a^2-t^2} \right)\) and \(\displaystyle (-f,0)\)

It's angle of inclination is therefore:

\(\displaystyle \theta_2=\tan^{-1}\left(\frac{b\sqrt{a^2-t^2}}{a(t+f)} \right)\) where \(\displaystyle t\ne -f\)

And so the angle of inclination of $\ell_2$ is:

\(\displaystyle \frac{\theta_2}{2}=\frac{1}{2}\tan^{-1}\left(\frac{b\sqrt{a^2-t^2}}{a(t+f)} \right)\)

Hence, the slope of $\ell_2$ is found using a half-angle identity for tangent:

\(\displaystyle m_2=\frac{\sin\left(\tan^{-1}\left(\dfrac{b\sqrt{a^2-t^2}}{a(t+f)} \right) \right)}{1+\cos\left(\tan^{-1}\left(\dfrac{b\sqrt{a^2-t^2}}{a(t+f)} \right) \right)}\)

Using the fact that \(\displaystyle f^2=a^2-b^2\) we find:

\(\displaystyle m_2=\frac{b}{a+f}\sqrt{\frac{a-t}{a+t}}\)

Armed now with the slope and the point through which it passes, line $\ell_2$ is given by (using the point-slope formula):

\(\displaystyle y=\frac{b}{a+f}\sqrt{\frac{a-t}{a+t}}(x+f)\)

Equating the two lines, and solving for $x$, we find:

\(\displaystyle =\frac{b}{a+f}\sqrt{\frac{a+t}{a-t}}(f-x)=\frac{b}{a+f}\sqrt{\frac{a-t}{a+t}}(x+f)\)

\(\displaystyle (a+t)(f-x)=(a-t)(x+f)\)

\(\displaystyle af-ax+ft-tx=ax+af-tx-ft\)

\(\displaystyle ft=ax\implies x=\frac{f}{a}t\implies y=\frac{b}{a+f}\sqrt{f^2-x^2}\)

Squaring this value for $y$, (which gives us the bottom half) we obtain the locus for $I$:

\(\displaystyle \frac{x^2}{f^2}+\frac{y^2}{\left(\dfrac{bf}{a+f} \right)^2}=1\)

which is clearly an ellipse.
 

Attachments

  • pranav.jpg
    pranav.jpg
    11.9 KB · Views: 55
  • #4
Hi Opalg and MarkFL! :)

Thank you both for your helpful solutions. I prefer to stay with parametric coordinates used by Opalg. :P

It was very silly of me to consider B and C as the foci. :eek:

About that incentre formula, it is also in my course and I need to memorise that (Doh).

Thank you once again.
 
  • #5


Hello,

Thank you for your attempt at solving this problem. It seems like you are on the right track, but there is a simpler way to prove that the locus of the incentre is an ellipse.

First, let's define the coordinates of the vertices of the triangle as A(0,0), B(a,0), and C(c,d). We can assume that the base of the triangle is on the x-axis, and the coordinates of the vertex C are (c,d) since the sum of the sides is given.

Next, let's find the coordinates of the incentre, I. We know that the incentre is the intersection of the angle bisectors of the triangle, which can be written as the equations:

$$y=\frac{dx}{c}$$
$$y=-\frac{d}{a-c}(x-a)$$

Solving these equations, we get the coordinates of the incentre as:

$$I\left(\frac{ac}{a+c},\frac{ad}{a+c}\right)$$

Now, let's find the distance of the incentre from the end points of the base, A(0,0) and B(a,0):

$$AI=\sqrt{\left(\frac{ac}{a+c}\right)^2+\left(\frac{ad}{a+c}\right)^2}=\frac{a\sqrt{c^2+d^2}}{a+c}$$

$$BI=\sqrt{\left(a-\frac{ac}{a+c}\right)^2+\left(\frac{ad}{a+c}\right)^2}=\frac{a\sqrt{c^2+d^2}}{a+c}$$

We can see that the sum of these distances is a constant, $a\sqrt{c^2+d^2}/(a+c)$, which proves that the locus of the incentre is an ellipse with foci at A and B and major axis of length a+c.

I hope this helps. Let me know if you have any further questions.
 

Related to Locus of Incentre Problem: Find Ellipse

1. What is the locus of the incentre in an ellipse?

The locus of the incentre in an ellipse is the set of all points that are equidistant from the three sides of the ellipse. This point is also known as the centre of the inscribed circle in the ellipse.

2. How can the locus of the incentre be found in an ellipse?

The locus of the incentre in an ellipse can be found by using the properties of an ellipse and the coordinates of its vertices. It can also be found by solving the equations of the three sides of the ellipse and finding the point of intersection.

3. What is the significance of the locus of the incentre in an ellipse?

The locus of the incentre in an ellipse is significant because it helps in determining the properties of the inscribed circle in the ellipse. It also helps in finding the centre of the ellipse and its eccentricity.

4. How does the locus of the incentre change with different orientations of the ellipse?

The locus of the incentre in an ellipse remains the same regardless of its orientation. This is because the property of the incentre being equidistant from the three sides of the ellipse holds true for all orientations.

5. Can the locus of the incentre be used to find other properties of an ellipse?

Yes, the locus of the incentre can be used to find the eccentricity, centre, and other properties of an ellipse. It can also be used in various geometric constructions and proofs involving ellipses.

Similar threads

Replies
2
Views
1K
Replies
19
Views
2K
Replies
6
Views
1K
Replies
2
Views
1K
Replies
1
Views
770
Replies
2
Views
10K
  • Introductory Physics Homework Help
Replies
10
Views
319
Replies
13
Views
1K
Replies
0
Views
9K
  • General Math
Replies
31
Views
1K
Back
Top