Equilateral triangle?

Geometry Level 1

In triangle A B C , ABC, there exists a point K K inside of it such that

A B K = B C K = C A K = 3 0 . \angle{ABK}= \angle{BCK}= \angle{CAK}=30^\circ.

Is it necessarily true that triangle A B C ABC is equilateral?

Yes No

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

8 solutions

Chan Tin Ping
Oct 22, 2017

Relevant wiki: Jensen's Inequality

(Notice that all the angle in this solution is degree not radians.) Let B A K = c , C B K = a , A C K = b \angle{BAK}=c,\angle{CBK}=a,\angle{ACK}=b

By ceva's theorem in trigonometric form, (Ceva thoerem states that s i n A B K s i n B C K s i n C A K = s i n B A K s i n C B K s i n A C K sin\angle{ABK}sin\angle{BCK}sin\angle{CAK}=sin\angle{BAK}sin\angle{CBK}sin\angle{ACK} ), ( s i n 30 ) 3 sin a sin b sin c = 1 \frac{(sin30)^3}{\sin{a}\sin{b}\sin{c}}=1 sin a sin b sin c = 1 8 \sin{a}\sin{b}\sin{c}=\frac{1}{8}

By AM-GM inequality, sin a + sin b + sin c 3 sin a sin b sin c 3 3 2 \begin{aligned} \sin{a}+\sin{b}+\sin{c} &\geq 3\sqrt[3]{\sin{a}\sin{b}\sin{c}} \\ &\geq\frac{3}{2} \end{aligned}

By Jensen inequality,(a+b+c+3 x 30=180) sin a + sin b + sin c 3 sin 180 3 × 30 3 3 2 \begin{aligned} \sin{a}+\sin{b}+\sin{c}&\leq3\sin{\frac{180-3\times30}{3}}\\ &\leq\frac{3}{2} \end{aligned}

Hence, sin a + sin b + sin c = 3 2 \sin{a}+\sin{b}+\sin{c}=\frac{3}{2} As AM-GM hold equality when all variables are same, which mean sin a = sin b = sin c \sin{a}=\sin{b}=\sin{c} a = b = c a=b=c Hence, all the angle of triangle ABC are same, which means triangle ABC is equilateral triange.

Erm, have you made a mistake? Surely the equation near the beginning should be ( sin 30 m ) 3 sin a sin b sin c = 1 \frac {\left(\sin \frac{30}m\right)^3}{\sin a \sin b \sin c} = 1

Stewart Gordon - 3 years, 7 months ago

Log in to reply

OK, actually my initial problem don't have the letter 'm', but someone change it with 'm', and I don't know what does the 'm' means. I also feel strange. I think the 'm' doesn't represent a number, maybe it's something like directed angle.

Chan Tin Ping - 3 years, 7 months ago

Log in to reply

I've just googled, and the pages I've found state that it means "measure", apparently as opposed to the angle itself regarded as a geometrical object. Though one of the pages can't seem to make up its mind whether m m\angle whatever equals a measure of angle or a dimensionless number. And besides, I wouldn't have thought it correct to write the 'm' in italics, though none of the pages seem to comment on this. Regardless of this, since it's confusing to people unfamiliar with the notation it's best avoided.

Stewart Gordon - 3 years, 7 months ago

And I cannot edit my problem now.

Chan Tin Ping - 3 years, 7 months ago

Google search m angle, you will find that m angle ABK = 30 means angle ABK = 30 degree. Sry.

Chan Tin Ping - 3 years, 7 months ago

What is Jensen inequality?

Anas Kudsi - 3 years, 7 months ago

Log in to reply

It's statement is very long, so you can just search it in Google.(or just click the relevant wiki.)

It's useful when treating some weird function such as sin, cos, √. To apply Jensen Inequality, you need to determine the function is convex or concave first.

Chan Tin Ping - 3 years, 7 months ago

(using < for angle symbol) If <BAC = 100, <ABC = 40, <BCA =40, a = 40-30 = 10, b = 40-30 = 10, c = 100 - 30 = 70, <CKA = 180 - 30 - b = 140, <CKB = 180 - 30 - a = 140, <BKA = 180 - 30 - c = 80, <BKA = 360 - <CKA - <CKB = 80,

So an obtuse 100-40-40 triangle fits. And obtuse triangles are not equilateral

cal Kruse - 3 years, 7 months ago

Log in to reply

(I don't know how to post diagram, so pls copy and paste it to see, sorry.)

https://m.facebook.com/photo.php?fbid=1713826635343455&id=100001484067680&set=p.1713826635343455&source=47&ref=bookmarks

The sum of <BKA + <AKC + <CKA = 360 didn't mean the BK, AK, CK concur. It can also satisfied when they are the external angle of triangle.

Chan Tin Ping - 3 years, 7 months ago

There is a much simpler way, I have no idea if youll see this but it is worth mentioning. Drop a parallel to AK through C and observe by (I believe) opposite interior angles the new angle is also 30 degrees. Thus the last angle is 90-30-30 leaving 30 in the missing spot. Similar arguments can be made for each side forcing this to be equilateral. Thanks.

Neima Ghandian - 3 years, 6 months ago

Log in to reply

Why the line which parallel to A K AK is perpendicular to the line B C BC ?

(Sry, I didn't notice your comment for a week.)

Chan Tin Ping - 3 years, 6 months ago

What if a and b = 25 and c= 40

moe levy - 3 years, 7 months ago

Log in to reply

Good point. These angles satisfy all four triangles' 180 agree requirement for a triangle on a plane. Thus, it is not necessary that triangle ABC is equilateral.

Sergey Sabirov - 3 years, 7 months ago

Log in to reply

No. Satisfying the 180 degree angles of the triangles is not enough. Problems will arise at the side lengths of the triangles.

Takeda Shigenori - 3 years, 7 months ago

Log in to reply

@Takeda Shigenori I have not heard that the sum of angles in a triangle is not a concrete requirement. The side length can be whatever in order to satisfy those angles.

Sergey Sabirov - 3 years, 7 months ago

Log in to reply

@Sergey Sabirov Take note of Sine Rule, which states that for a given triangle A B C \triangle ABC with side lengths a , b , c a, b, c opposite to the respective angles, we have a sin A = b sin B = c sin C \frac{a}{\sin A} = \frac{b}{\sin B} = \frac{c}{\sin C}

Try to use this formula to the three small triangles A B K , B C K , C A K \triangle ABK, \triangle BCK, \triangle CAK if you assume Moe's angles to be true, and you will find some bizarre statements, which basically means that the triangle does not exist (well, at least in Euclidean space).

Takeda Shigenori - 3 years, 7 months ago

Log in to reply

@Takeda Shigenori Yes, we cannot draw a larger triangle where vertices would coincide with those angles.

Sergey Sabirov - 3 years, 7 months ago

If a=b=25, c=40, then sin a×sin b×sin c isn't equal to 1/8. By ceva theorem, BK, AK and CK won't concur. But it is impossible.

Chan Tin Ping - 3 years, 7 months ago

Log in to reply

Ceva theorem doesn't apply, since K is not necessarily the same point as the intersection of the perpendicular lines.

moe levy - 3 years, 7 months ago

Log in to reply

@Moe Levy Ceva's Theorem holds for any point inside a triangle.

Takeda Shigenori - 3 years, 7 months ago

@Moe Levy Actually I don't know what do you mean. Your 'perpendicular lines' are refer to which lines?

Chan Tin Ping - 3 years, 7 months ago

You can draw your own triangle with those angles and see that it is doable. It is a fact and you cannot ignore it.

Sergey Sabirov - 3 years, 7 months ago

Log in to reply

@Sergey Sabirov Cannot, you draw it and show to me.

Chan Tin Ping - 3 years, 7 months ago

Log in to reply

@Chan Tin Ping Yeap, I can draw one of those triangles only but not the one with vertices coinciding in the same larger triangle. I was wrong.

Sergey Sabirov - 3 years, 7 months ago
Steven Yuan
Oct 29, 2017

Drop perpendiculars from K K to the side lengths of the triangle, as shown above. Since A Y K , B Z K , C X K \triangle AYK, BZK, CXK are all right triangles, we have

K X = K C sin 3 0 = 1 2 K C K Y = K A sin 3 0 = 1 2 K A K Z = K B sin 3 0 = 1 2 K B . KX = KC \sin 30^{\circ} = \dfrac{1}{2} KC \\ KY = KA \sin 30^{\circ} = \dfrac{1}{2} KA \\ KZ = KB \sin 30^{\circ} = \dfrac{1}{2} KB. \\

Thus,

K X + K Y + K Z = 1 2 ( K A + K B + K C ) . KX + KY + KZ = \dfrac{1}{2}(KA + KB + KC).

However, the Erdös-Mordell Inequality states that K X + K Y + K Z 1 2 ( K A + K B + K C ) , KX + KY + KZ \leq \frac{1}{2}(KA + KB + KC), with equality holding if and only if A B C \triangle ABC is equilateral. In this scenario, equality holds, so we conclude that yes , A B C \triangle ABC must be equilateral.

Moderator note:

One subtlety of this solution to note is that dropping the perpendiculars from K K does not prove collinearity - we can't assume, for instance, that B , B, K , K, and Y Y are all on the same line, even thought it's hard to draw the picture without this appearing to be the case. The solution here appropriately avoids this issue.

If we knew collinearity, we could easily use the fact we have six 30-60-90 triangles that share sides to prove that all of them are congruent, then note then that B X , X C , C Y , Y A , A Z , BX, XC, CY, YA, AZ, and Z B ZB must be all be congruent, and hence the triangle A B C ABC is equilateral.

Nice solution using Erdös-Mordell Inequality! One small typo, when you quote the inequality on the second to last line, the left hand side should be KX+KY+KZ, right?

Wei Chen - 3 years, 7 months ago

Log in to reply

I’ve edited it to the correct expression. Thanks for spotting the mistake!

Steven Yuan - 3 years, 7 months ago

Great solution .....Excellent use of Erdos-Mordell..

Anubhav Mahapatra - 3 years, 7 months ago

How have you deduced that K X = K C sin 3 0 KX = KC \sin 30^\circ etc.? Surely this would only be true if m = 1 m = 1 or m = 5 m = 5 .

Stewart Gordon - 3 years, 7 months ago

Log in to reply

The letter 'm' state for measure in degree. It doesn't represent a number.

Chan Tin Ping - 3 years, 7 months ago

There is a much simpler way, I have no idea if youll see this but it is worth mentioning. Drop a parallel to AK through C and observe by (I believe) opposite interior angles the new angle is also 30 degrees. Thus the last angle is 90-30-30 leaving 30 in the missing spot. Similar arguments can be made for each side forcing this to be equilateral. Thanks.

Neima Ghandian - 3 years, 6 months ago

This is only true for euclidean and regularly curved spaces. This breaks for all other spaces. There still exists a point k for all spaces where the angles are equal, but the triangle cannot be equilateral. A cylindrical space for instance.

Daniel Langstaff - 3 years, 7 months ago

Log in to reply

I think it's reasonable to assume we're in a Euclidean space for questions like this one.

Richard Desper - 3 years, 7 months ago
Takeda Shigenori
Oct 30, 2017

Let B A K = α , C B K = β , A C K = γ \angle BAK=α, \angle CBK=β, \angle ACK=γ

We have α + β + γ = 9 0 α+β+γ=90^{\circ} and from Ceva's Theorem in trigonometric form, sin 3 0 sin α sin 3 0 sin β sin 3 0 sin γ = 1 \frac{\sin 30^{\circ}}{\sin α} · \frac{\sin 30^{\circ}}{\sin β} · \frac{\sin 30^{\circ}}{\sin γ} = 1

Rearranging, sin α sin β sin γ = 1 8 \sin α \sin β \sin γ = \frac{1}{8}

Adding the natural logarithm to both sides, ln sin α + ln sin β + ln sin γ = ln 1 8 \ln \sin α + \ln \sin β + \ln \sin γ = \ln \frac{1}{8}

Now denote the function f : ( 0 , π 2 ) R , f ( x ) = ln sin x f:(0,\frac{π}{2}) \mapsto \mathbb{R}^{-}, f(x) = \ln \sin x , since f ( x ) = 1 sin x cos x = cot x f'(x) = \frac{1}{\sin x} · \cos x = \cot x and f ( x ) = csc 2 x < 0 f''(x) = - \csc^2 x < 0 , f ( x ) f(x) is a concave function. Therefore by Jensen's inequality, f ( α ) + f ( β ) + f ( γ ) 3 f ( α + β + γ 3 ) \frac{f(α)+f(β)+f(γ)}{3} \leqslant f(\frac{α+β+γ}{3})

ln sin α + ln sin β + ln sin γ 3 ln sin α + β + γ 3 = ln 1 2 \frac{\ln \sin α + \ln \sin β + \ln \sin γ}{3} \leqslant \ln \sin \frac{α+β+γ}{3} = \ln \frac{1}{2}

ln sin α + ln sin β + ln sin γ ln 1 8 \therefore \ln \sin α + \ln \sin β + \ln \sin γ \leqslant \ln \frac{1}{8}

Equality holds true if and only if α = β = γ α=β=γ or f ( x ) f(x) is linear, the latter of which is not the case as f ( x ) < 0 f''(x)<0

Thus α = β = γ = 3 0 α=β=γ=30^{\circ} and A B C \triangle ABC is obviously an equilateral triangle.

Note: With some modifications, the same method can also be applied to solve P5 of IMO 1991.

Does this method hold in irregular geometry spaces?

Daniel Langstaff - 3 years, 7 months ago

Log in to reply

I don't really have any knowledge of geometry outside Euclidean 😥

Takeda Shigenori - 3 years, 7 months ago
Rob Brott
Nov 4, 2017

Another approach is to use straight-forward analytic geometry.

Without loss of generality, put B at the origin, C on the x-axis one unit away (this distance simply sets the length scale), and the point A at { a x , a y } \{a_x, a_y\} . Further, restrict A to the first quadrant.

Then we can represent rays from each of the vertices towards the point K. First, consider the ray from B towards K. The direction for this ray is the vector from B to A rotated 30 degrees clockwise, and β \beta is a parameter along this ray.

{ 1 2 β ( 3 a x + a y ) , 1 2 β ( a x 3 a y ) } \left\{\frac{1}{2} \beta \left(\sqrt{3} a_x+a_y\right),-\frac{1}{2} \beta \left(a_x-\sqrt{3} a_y\right)\right\}

From A towards K, α \alpha is a parameter along the ray.

{ 1 2 3 α ( a x 1 ) + a x α a y 2 , α a x 2 1 2 α ( 3 a y + 1 ) + a y } \left\{-\frac{1}{2} \sqrt{3} \alpha \left(a_x-1\right)+a_x-\frac{\alpha a_y}{2},\frac{\alpha a_x}{2}-\frac{1}{2} \alpha \left(\sqrt{3} a_y+1\right)+a_y\right\}

From C towards K, γ \gamma is a parameter along the ray.

{ 1 3 γ 2 , γ 2 } \left\{1-\frac{\sqrt{3} \gamma }{2},\frac{\gamma }{2}\right\}

The point K is the intersection of these three rays, and we demand this be a single point. Setting these three rays equal to each other, and eliminating the parameters α \alpha , β \beta , and γ \gamma , yields the following relationship:

a x 2 + a y 2 + 1 = a x + 3 a y a_x^2+a_y^2+1=a_x+\sqrt{3} a_y

which can be rewritten, by completing the square, into the suggestive form

( a x 1 2 ) 2 + ( a y 3 2 ) 2 = 0 \left(a_x-\frac{1}{2}\right){}^2+\left(a_y-\frac{\sqrt{3}}{2}\right){}^2 = 0

This is a circle centered at { 1 2 , 3 2 } \{\frac{1}{2}, \frac{\sqrt{3}}{2} \} , of radius 0. In other words, a single point. But this point makes the triangle equilateral.

Bhaskar Pandey
Oct 31, 2017

Draw circles AKC, AKB, BKC, then we will see that angle KBC=KAB=KCA.

Slendy Slender
Nov 8, 2017

It appears to be that, since all mentioned angles (ABK, BCK, CAK) are equal to 30 degrees and an equilateral triangle has equal angles at all corners, we can infer that point K must be within the plane of an equilateral triangle.

Inscribed in a circle we take an arbitrary triangle ABC and from vertexes A, B and C draw lines forming an arbitrary angle w.
In our case w=30 these three new lines will form another triangle A1B1C1 congruent with the original one ABC . Then since we look for a triangle A1B1C1 which area must be null, then points A1, B1, C1 and K must been all coincident, the only solution is that our original points A,B,C inscribed in the circle form an equilateral triangle. Of course K is also the Fermat point

If angles' sum of a triangle is 180° so it's equilateral.

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...