5

I was screwing around lately in GeoGebra and I realized something.

Draw a $\triangle ABC$, and let the bisectors for $\angle A$ and $\angle C$ meet sides $BC$ and $AB$ at points $E$ and $D$, respectively. If the angle bisectors meet at the incenter, $F$, and if $FD \cong FE$, then either $\triangle ABC$ must be isosceles or $\angle B$ must be $60^\circ$.

However I was unable to prove why that is. Any help would be appreciated.

A demonstration with angle B = 60

Blue
  • 75,673
user587054
  • 1,245

6 Answers6

3

Let $a$ and $c$ be the half angles $ a = \frac{A}{2}$ and $ c = \frac{C}{2}$ respectively. Apply the sine rule to the triangles ΔBDF and ΔBEF

$$ DF = BF \frac{\sin\frac B2}{\sin(2a+c)}\tag{1}$$ $$ EF = BF \frac{\sin\frac B2}{\sin(2c+a)}\tag{2}$$

where ∠BDC = $2a + c$ and ∠BEA = $2c + a$ are recognized. Take the ratio of (1) and (2)

$$ \frac{DF}{EF} = \frac{\sin(2c+a)}{\sin(2a+c)}$$

If DF = EF, the equation yields,

$$\sin(2c+a)=\sin(2a+c).$$

which gives two solutions,

$$2c+a = 2a+c\tag{3}$$

$$2c+a = 180^\circ - (2a+c)\tag{4}$$

  • The solution (3) leads to $a = c$, which is the isosceles triangle case.

  • The solution (4) leads to $a + c = 60^\circ$ and in turn ∠B = $180^\circ -2(a+c)=60^\circ $, which is the 60-degree case.

Hence, both cases are proved.

Quanto
  • 97,352
3

I am trying to give a simple, purely geometric proof. Here is first a picture, notations are slightly changed, so that it is simpler to type formulas blindly:

Parts of the angle bisectors from incenter to the opposite side are equal, then an angle is 60 degrees, math stackexchange problem 3332190

$\Delta ABC$ is a triangle, $AA'$, $BB'$, $CC'$ are its angle bisector cevians, $I$ is their intersection, the incenter. We assume $IA'=IC'$, and that the perpendiculars $IU$ and $IS$ from $I$ on $BA$ and respectively $BC$, $U\in BA$, $S\in BC$, are so that one is inside the angle $\widehat{A'IC'}$, one outside. In our case, without loss of generality, $S$ is on the segment $BA'$, and $U$ is not on the segment $BC'$.

We assume $$IA'=IC'\ .$$

Then $\color{red}{\hat B=60^\circ}$.

Proof: The triangles $\Delta ISA'$ and $\Delta IUC'$ are congruent, having right angles in $S$, respectively $U$, and $IS=IU$, and the given relation $IA'=IC'$. So their two angles in $I$ are equal. So the two angles in $I$ obtained by adding $\widehat {SIC'}$ are equal. This implies: $$ \pi-\frac 12 \hat A-\frac 12C = \color{blue}{ \widehat{AIC} = \widehat{A'IC'} = \widehat{SIU} } = \pi-\hat B\ . $$ The blue part is the idea, the "mule" to connect the given information with the needed angle information. We obtain immediately $\hat B =\frac 12(\hat A+\hat C)=\frac12(\pi-B)$, so $3\hat B=\pi$.

$\square$


The other case, when the perpendiculars from $I$ on the sides $BA$, $BC$ are one "the same part" is arguably simpler. After drawing the picture and using the same notations, after using again the congruence of the two triangles $\Delta IA'S$ and $\Delta IC'U$, we get this time $$ \frac 12(\hat B+\hat A) = \widehat{ABI} + \widehat{IAB} = \color{blue} { \widehat{BIA'} = \widehat{BIC'} } = \widehat{IBC} + \widehat{ICB} =\frac 12(\hat B+\hat C) \ . $$ So $\hat A=\hat C$.

$\square$

dan_fulea
  • 32,856
1

Let $AB = c$, $BC = a$, $AC = b$. We will prove that if $EF = FD$, then $ABC$ is either isosceles or $\angle ABC = 60$.

We will first show that $\frac{DF}{FC} = \frac{c}{a+b}$. First, notice that by the angle bisector theorem, we have that $\frac{DF}{FC} = \frac{BD}{BC} = \frac{BD}{a}$. Now, note that $\frac{BD}{c - BD} = \frac{a}{b}$ by the angle bisector theorem, and expanding gives us $BD = \dfrac{\frac{a}{b} \cdot c}{1 + \frac{a}{b}}$. Thus, $\frac{BD}{a} = \dfrac{\frac{c}{b}}{\frac{b+a}{b}} = \frac{c}{b+a}$. Therefore, $\frac{DF}{DC} = \frac{c}{a+b+c}$.

Similarly, we have that $\frac{EF}{FA} = \frac{a}{b+c} \implies \frac{EF}{EA} = \frac{a}{a+b+c}$. Therefore, we have that, since $DF = EF$, $\frac{EA}{DC} = \frac{c}{a} \implies \frac{EA}{AB} = \frac{DC}{BC}$. Now, note that by the Law of Sines, $\frac{EA}{AB} = \frac{\sin{\angle B}}{\sin \angle{AEB}}$, and $\frac{DC}{BC} = \frac{\sin{\angle{B}}}{\sin{\angle{BDC}}}$, which implies that $\angle AEB = \angle BDC$ or $\angle AEB + \angle BDC = 180$.

We now consider cases. In the first case, $\angle AEB = 180 - (\angle B + \angle BAE)$ and $\angle BDC = 180 - (\angle B + \angle DCB)$ \implies $\frac{\angle A}{2} = \frac{\angle C}{2}$ which implies $\angle BAC = \angle BCA$, and in the second case we have $360 - 2\angle B - (\angle DCB + \angle BAE) = 180$, and since $\angle DCB + \angle BAE = 90 - \frac{\angle{B}}{2}$ \implies $\frac{3\angle{B}}{2} = 90 \implies \angle{B} = 60$, so we are done.

ETS1331
  • 1,335
1

Given a triangle as described in the question, where $FD \cong FE.$

We first show that $\angle FDB$ and $\angle FEB$ are either congruent or supplementary angles. There are two cases:

Case 1: The circle is tangent to $AB$ at $D.$ Then $FD$ is a radius of the circle, and $FD \cong FE$ implies that $FE$ also is a radius of the circle, hence the circle is tangent to $BC$ at $E.$ Then $\angle FDB$ and $\angle FEB$ both are right angles, hence are both congruent and supplementary in Case 1.

Case 2: The circle is not tangent to $AB$ at $D.$ Let $H$ be the point of tangency with $AB$ and let $K$ be the point of tangency with $BC.$ Then $\lvert FD\rvert > \lvert FH\rvert.$ Further, $FD \cong FE$ and $FH \cong FK$ imply that $\lvert FE\rvert > \lvert FK\rvert.$

Then $\triangle FHD$ and $\triangle FEK$ are right triangles with a congruent leg and congruent hypotenuse, so $\triangle FHD \cong \triangle FKE$ with $\angle FDH \cong \angle FEK.$

But $\angle FDB$ is either congruent to or supplementary to $\angle FDH$ and $\angle FEB$ is either congruent to or supplementary to $\angle FEK.$ It follows that $\angle FDB$ and $\angle FEB$ are either congruent or supplementary in Case 2.

This concludes the proof that $\angle FDB$ and $\angle FEB$ are either congruent or supplementary angles. Now consider the cases "congruent" and "supplementary" individually.

Suppose $\angle FDB \cong \angle FEB.$ Since $F$ is on the bisector of $\angle DBE,$ it follows that $\triangle BDF \cong \triangle BEF$ with $BD \cong BE.$ But also $\angle CEF \cong \angle ADF$ (since each is supplementary to $\angle FDB$ or $\angle FEB$) and $\angle CFE \cong \angle AFD$ (by vertical angles), and together with $FD \cong FE$ we then have $\triangle ADF \cong \triangle CEF$ with $AD \cong CE.$ Since $D$ is between $A$ and $B$ and since $E$ is between $B$ and $C,$ it follows that $AB \cong BC,$ that is, $\triangle ABC$ is isosceles.

On the other hand, suppose $\angle FDB$ and $\angle FEB$ are supplementary angles. Let $\angle BAC = \alpha$ and $\angle ACB = \gamma.$ Then, since an exterior angle of triangle $\triangle ACE$ or $\triangle BCD$ is the sum of the other two interior angles, $\angle FDB = \alpha+\gamma/2$ and $\angle FEB = \alpha/2+\gamma.$ But then, since $\angle FDB$ and $\angle FEB$ are supplementary, $$ (\alpha+\gamma/2) + (\alpha/2+\gamma) = 180^\circ, $$ which implies that $\alpha+\gamma = 120^\circ.$ Hence $\angle ABC = 60^\circ.$

Therefore either $\triangle ABC$ is isosceles, or $\angle ABC = 60^\circ.$


As an aside, if the circle is tangent to $AB$ at $D,$ then it follows both that $\triangle ABC$ is isosceles and that $\angle ABC = 60^\circ,$ that is, $\triangle ABC$ is equilateral.

But if the circle is not tangent to $AB$ at $D,$ then $\triangle FHD$ and $\triangle FKE$ are not degenerate. Either $\triangle FHD$ and $\triangle FKE$ have the same orientation, in which case $\angle ABC = 60^\circ,$ or $\triangle FHD$ and $\triangle FKE$ have opposite orientations (are mirror images), in which case $\triangle ABC$ is isosceles. In either case it is not possible both for $\triangle ABC$ to be isosceles and for $\angle ABC$ to be $60^\circ.$

David K
  • 98,388
  • 1
    If the triangle is isosceles with $\angle A = \angle C$, then $FD=FE$ by symmetry. I believe that case is encapsulated when $\Delta FHD \cong \Delta FKE$ inversely, when your angle chase does not hold. – Carl Schildkraut Aug 23 '19 at 20:44
  • 1
    @CarlSchildkraut Good point. So it turns out the "angle-chasing" involves a little less arithmetic and more congruent triangles in the isosceles case (and the angles turn out to be congruent, not supplementary). My previous answer was incorrect because it failed to correctly account for the orientation of $\triangle FHD$ and $\triangle FEK$ in the isosceles case, or to account for the case where those triangles are degenerate. – David K Aug 24 '19 at 01:05
0

Partial solution: If $ABC$ is isosceles, then $FE$ is equal to radius of the circle which also means that $FD$ is a radius and $FD$ is perpendicular to $AB$ (radius to tangent). Thus $DBC$ is a right triangle and $\angle DBC=2 \angle DCB$. From here we conclude that $\angle DCB =30°$ and $\triangle ABC$ is equilateral.

Vasili
  • 10,690
-1

In the standard notation by the angle bisector theorem we obtain: $$\frac{FE}{AF}=\frac{CF}{AC}=\frac{\frac{ab}{b+c}}{b}=\frac{a}{b+c}.$$ Thus, $$\frac{FE}{AE-FE}=\frac{a}{b+c},$$ which gives $$FE=\frac{a}{a+b+c}\cdot AE=\frac{a}{a+b+c}\cdot\frac{2bc\cos\frac{\alpha}{2}}{b+c}=\frac{2abc\cos\frac{\alpha}{2}}{(a+b+c)(b+c)}.$$ Similarly, $$FD=\frac{2abc\cos\frac{\gamma}{2}}{(a+b+c)(a+b)}.$$ Id est, $$\frac{2abc\cos\frac{\alpha}{2}}{(a+b+c)(b+c)}=\frac{2abc\cos\frac{\gamma}{2}}{(a+b+c)(a+b)}$$ or $$a(b+c-a)(a+b)^2=c(a+b-c)(b+c)^2$$ or $$(a-c)(a+b+c)(a^2+c^2-ac-b^2)=0.$$ Can you end it now?