¿10^n-1 is a divisor of 11^n-1?

by EmersonSoriano, Apr 5, 2025, 6:32 PM

Determine if there exists a positive integer $n$ such that $10^n - 1$ is a divisor of $11^n - 1$.
This post has been edited 1 time. Last edited by EmersonSoriano, 31 minutes ago
Reason: change

Diagonal of a pentagon that divides it into a triangle and a cyclic quadrilatera

by EmersonSoriano, Apr 5, 2025, 6:30 PM

We say that a diagonal of a convex pentagon is good if it divides the pentagon into a triangle and a circumscribable quadrilateral. What is the maximum number of good diagonals that a convex pentagon can have?

Clarification: A polygon is circumscribable if there is a circle tangent to each of its sides.

Two midpoints and the circumcenter are collinear.

by ricarlos, Apr 5, 2025, 5:52 PM

Let $ABC$ be a triangle with circumcenter $O$. Let $P$ be a point on the perpendicular bisector of $AB$ (see figure) and $Q$, $R$ be the intersections of the perpendicular bisectors of $AC$ and $BC$, respectively, with $PA$ and $PB$. Prove that the midpoints of $PC$ and $QR$ and the point $O$ are collinear.
Attachments:

geometry party

by pnf, Apr 5, 2025, 5:36 PM

All heads to tails?

by smartvong, Apr 5, 2025, 5:14 PM

An equilateral triangle is formed using $n$ rows of coins. There is 1 coin in the first row, 2 coins in the second row, 3 coins in the third row, and so on, up to $n$ coins in the $n$th row. Initially, all of the coins show heads (H). Carley plays a game in which, on each turn, she chooses three mutually adjacent coins and flips these three coins over. To win the game, all of the coins must be showing tails (T) after a sequence of turns. An example game with 4 rows of coins after a sequence of two turns is shown.

[asy]
import graph;

size(100);

// Define constants
real radius = 0.3;
real v_dist = sqrt(3)/2;
pen coinPen = black+0.8;
pen linePen = gray+dotted+0.6;
pen labelPen = black;
pen fillPen = lightgray+opacity(0.5); // Semi-transparent fill
real labelScale = 0.8;

// Function to get coin position
pair P(int i, int j) {
    return ( (j - (i+1.0)/2.0), -(i-1)*v_dist );
}

int n = 4;
pair[][] coinPos = new pair[n+1][];
for(int i=1; i<=n; ++i) {
    coinPos[i] = new pair[i+1];
    for(int j=1; j<=i; ++j) {
        coinPos[i][j] = P(i,j);
    }
}

// Highlight the two flipped triangles T(2,2) and T(3,5)
// T(2,2) involves (2,1), (2,2), (3,2) - Downward
pair pA1 = P(2,1); pair pA2 = P(2,2); pair pA3 = P(3,2);
fill(pA1--pA2--pA3--cycle, fillPen);

// T(3,5) involves (3,3), (4,3), (4,4) - Upward (using problem's naming convention)
// Or more standard: (i=3, j=3) up -> (3,3), (4,3), (4,4)
pair pB1 = P(3,3); pair pB2 = P(4,3); pair pB3 = P(4,4);
fill(pB1--pB2--pB3--cycle, fillPen);

// Draw connecting lines
for(int i=1; i<n; ++i) {
    for(int j=1; j<=i; ++j) {
        draw(coinPos[i][j] -- coinPos[i+1][j], linePen);
        draw(coinPos[i][j] -- coinPos[i+1][j+1], linePen);
    }
}
for(int i=1; i<=n; ++i) {
    for(int j=1; j<i; ++j) {
        draw(coinPos[i][j] -- coinPos[i][j+1], linePen);
    }
}

// Draw coins (after fills)
for(int i=1; i<=n; ++i) {
    for(int j=1; j<=i; ++j) {
        draw(Circle(coinPos[i][j], radius), coinPen);
    }
}

// Add H/T labels based on the example image
label(scale(labelScale)*"H", coinPos[1][1], labelPen);
label(scale(labelScale)*"T", coinPos[2][1], labelPen);
label(scale(labelScale)*"T", coinPos[2][2], labelPen);
label(scale(labelScale)*"H", coinPos[3][1], labelPen);
label(scale(labelScale)*"T", coinPos[3][2], labelPen);
label(scale(labelScale)*"T", coinPos[3][3], labelPen);
label(scale(labelScale)*"H", coinPos[4][1], labelPen);
label(scale(labelScale)*"H", coinPos[4][2], labelPen);
label(scale(labelScale)*"T", coinPos[4][3], labelPen);
label(scale(labelScale)*"T", coinPos[4][4], labelPen);
[/asy]

Below (a), (b) and (c), you will find instructions about how to refer to these turns in your solutions.

(a) If there are 3 rows of coins, give a sequence of 4 turns that results in a win.

(b) Suppose that there are 4 rows of coins. Determine whether or not there is a sequence of turns that results in a win.

(c) Determine all values of $n$ for which it is possible to win the game starting with $n$ rows of coins.

Note: For a triangle with 4 rows of coins, there are 9 possibilities for the set of three coins that Carley can flip on a given turn. These 9 possibilities are shown as shaded triangles below:

[asy]
import graph;

size(150); // Increased size for clarity

// Define constants
real radius = 0.3;
real v_dist = sqrt(3)/2;
pen coinPen = black+0.8;
pen linePen = gray+dotted+0.6;
pen labelPen = black;
pen fillPen = lightgray;
real labelScale = 0.6; // Smaller label scale

// Function to get coin position
pair P(int i, int j) {
    return ( (j - (i+1.0)/2.0), -(i-1)*v_dist );
}

// Function to draw the grid and empty coins
void drawGrid(int n) {
    pair[][] coinPos = new pair[n+1][];
    for(int i=1; i<=n; ++i) {
        coinPos[i] = new pair[i+1];
        for(int j=1; j<=i; ++j) {
            coinPos[i][j] = P(i,j);
        }
    }

    // Draw connecting lines
    for(int i=1; i<n; ++i) {
        for(int j=1; j<=i; ++j) {
            draw(coinPos[i][j] -- coinPos[i+1][j], linePen);
            draw(coinPos[i][j] -- coinPos[i+1][j+1], linePen);
        }
    }
    for(int i=1; i<=n; ++i) {
        for(int j=1; j<i; ++j) {
            draw(coinPos[i][j] -- coinPos[i][j+1], linePen);
        }
    }
    // Draw coins (empty)
    for(int i=1; i<=n; ++i) {
        for(int j=1; j<=i; ++j) {
            draw(Circle(coinPos[i][j], radius), coinPen);
        }
    }
}

// Function to draw a shaded UP triangle move and its label
void drawMoveUp(int i, int j, string lbl) {
    pair p1 = P(i,j);
    pair p2 = P(i+1,j);
    pair p3 = P(i+1,j+1);
    fill(p1--p2--p3--cycle, fillPen);
    // Redraw circles on top
    draw(Circle(p1, radius), coinPen);
    draw(Circle(p2, radius), coinPen);
    draw(Circle(p3, radius), coinPen);
    // Label
    label(scale(labelScale)*lbl, (p1+p2+p3)/3, labelPen);
}

int n=4;
drawGrid(n);

// Draw the 6 moves shown in the first diagram
drawMoveUp(1, 1, "$T(1,1)$");
drawMoveUp(2, 1, "$T(2,1)$");
drawMoveUp(2, 2, "$T(2,3)$"); // Note index from problem
drawMoveUp(3, 1, "$T(3,1)$");
drawMoveUp(3, 2, "$T(3,3)$"); // Note index from problem
drawMoveUp(3, 3, "$T(3,5)$"); // Note index from problem
[/asy]

[asy]

import graph;

size(150); // Increased size for clarity

// Define constants
real radius = 0.3;
real v_dist = sqrt(3)/2;
pen coinPen = black+0.8;
pen linePen = gray+dotted+0.6;
pen labelPen = black;
pen fillPen = lightgray;
real labelScale = 0.6; // Smaller label scale

// Function to get coin position
pair P(int i, int j) {
    return ( (j - (i+1.0)/2.0), -(i-1)*v_dist );
}

// Function to draw the grid and empty coins (same as before)
void drawGrid(int n) {
    pair[][] coinPos = new pair[n+1][];
    for(int i=1; i<=n; ++i) {
        coinPos[i] = new pair[i+1];
        for(int j=1; j<=i; ++j) {
            coinPos[i][j] = P(i,j);
        }
    }
    // Draw connecting lines
    for(int i=1; i<n; ++i) {
        for(int j=1; j<=i; ++j) {
            draw(coinPos[i][j] -- coinPos[i+1][j], linePen);
            draw(coinPos[i][j] -- coinPos[i+1][j+1], linePen);
        }
    }
    for(int i=1; i<=n; ++i) {
        for(int j=1; j<i; ++j) {
            draw(coinPos[i][j] -- coinPos[i][j+1], linePen);
        }
    }
    // Draw coins (empty)
    for(int i=1; i<=n; ++i) {
        for(int j=1; j<=i; ++j) {
            draw(Circle(coinPos[i][j], radius), coinPen);
        }
    }
}

// Function to draw a shaded DOWN triangle move and its label
void drawMoveDown(int i, int j, string lbl) {
    // Coins involved are (i,j), (i,j+1), (i+1, j+1)
    pair p1 = P(i,j);
    pair p2 = P(i,j+1);
    pair p3 = P(i+1,j+1);
    fill(p1--p2--p3--cycle, fillPen);
    // Redraw circles on top
    draw(Circle(p1, radius), coinPen);
    draw(Circle(p2, radius), coinPen);
    draw(Circle(p3, radius), coinPen);
    // Label
    label(scale(labelScale)*lbl, (p1+p2+p3)/3, labelPen);
}

int n=4;
drawGrid(n);

// Draw the 3 moves shown in the second diagram
drawMoveDown(2, 1, "$T(2,2)$"); // Corresponds to coins (2,1), (2,2), (3,2)
drawMoveDown(3, 1, "$T(3,2)$"); // Corresponds to coins (3,1), (3,2), (4,2)
drawMoveDown(3, 2, "$T(3,4)$"); // Corresponds to coins (3,2), (3,3), (4,3)
[/asy]

[You should use the names for these moves shown inside the 9 shaded triangles when answering (b). You should adapt this naming convention in a suitable way when answering parts (a) and (c).]

Valuable subsets of segments in [1;n]

by NO_SQUARES, Apr 3, 2025, 8:34 PM

The integer $n \geqslant 2$ is given. Let $A$ be set of all $n(n-1)/2$ segments of real line of type $[i, j]$, where $i$ and $j$ are integers, $1\leqslant i<j\leqslant n$. A subset $B \subset A$ is said to be valuable if the intersection of any two segments from $B$ is either empty, or is a segment of nonzero length belonging to $B$. Find the number of valuable subsets of set $A$.

2025 Caucasus MO Juniors P4

by BR1F1SZ, Mar 26, 2025, 12:57 AM

In a convex quadrilateral $ABCD$, diagonals $AC$ and $BD$ are equal, and they intersect at $E$. Perpendicular bisectors of $AB$ and $CD$ intersect at point $P$ lying inside triangle $AED$, and perpendicular bisectors of $BC$ and $DA$ intersect at point $Q$ lying inside triangle $CED$. Prove that $\angle PEQ = 90^\circ$.

functional equations over positive rationals make me big sad

by bryanguo, Apr 25, 2023, 1:50 AM

Let $\mathbb{Q}^{+}$ denote the set of positive rational numbers. Find, with proof, all functions $f:\mathbb{Q}^+ \to \mathbb{Q}^+$ such that, for all positive rational numbers $x$ and $y,$ we have \[f(x)=f(x+y)+f(x+x^2f(y)).\]
This post has been edited 1 time. Last edited by bryanguo, Apr 25, 2023, 1:53 AM

Collinearity of Kiepert perspectors

by TelvCohl, Sep 1, 2018, 7:38 AM

Preliminaries

Lemma 1 : Given a $ \triangle ABC $ with isogonal conjugate $ (P,P^*), (Q,Q^*). $ Then $ S $ $ \equiv $ $ PQ $ $ \cap $ $ P^*Q^* $ and $ T $ $ \equiv $ $ PQ^* $ $ \cap $ $ P^*Q $ are isogonal conjugate WRT $ \triangle ABC. $

Proof : Let $ X, Y $ be the intersection of $ AQ $ with $ PQ^*, P^*Q^*, $ respectively. Since $$ A(Q,Q^*;T,P) = (X,Q^*;T,P) \stackrel{Q}{=} (Y,Q^*;P^*,S) = A(Q,Q^*;P^*,S) = A(Q^*,Q;S,P^*), $$so $ (AS, AT) $ are isogonal conjugate WRT $ \angle A. $ Similarly, $ (BS, BT), (CS, CT) $ are isogonal conjugate WRT $ \angle B, \angle C, $ respectively, so we conclude that $ S, $ $ T $ are isogonal conjugate WRT $ \triangle ABC. $ $  \qquad \blacksquare  $

Lemma 2 : Given a $ \triangle ABC $ and a point $ U $ lying on the perpendicular bisector of $ BC. $ Let $ V $ be the point such that $ \measuredangle ABU = \measuredangle VCB, \measuredangle ACU = \measuredangle VBC. $ Then $ A, U, V $ are collinear.

Proof : Let $ W $ be the reflection of $ A $ in the perpendicular bisector of $ BC, $ then $ \measuredangle UCW = \measuredangle VCB, \measuredangle UBW = \measuredangle VBC $ $ \Longrightarrow $ $ V,W $ are isogonal conjugate WRT $ \triangle BUC, $ so we conclude that $ A, U, V $ are collinear. $  \qquad \blacksquare  $

Lemma 3 : Given a $ \triangle ABC $ with circumcircle $ \odot (O, R) $ and 9-point center $ N. $ Let $ P,Q $ be the isogonal conjugate WRT $ \triangle ABC $ and $ M $ be the midpoint of $ PQ. $ Then $ MN = \frac{OP \cdot OQ}{2R}. $

Proof : Let $ \triangle A_1B_1C_1 $ be the circumcevian triangle of $ P $ WRT $ \triangle ABC $ and $ A_2, B_2, C_2 $ be the reflection of $ A_1, B_1, C_1 $ in $ BC, CA, AB, $ respectively, then from Properties of Hagge circle (Property 1, Corollary 1.1, Property 3) we get $ H \in \odot (A_2B_2C_2), $ the reflection $ T $ of $ Q $ in $ N $ is the circumcenter of $ \triangle A_2B_2C_2 $ and $ \triangle A_1B_1C_1 \cup P \stackrel{-}{\sim} \triangle A_2B_2C_2 \cup P $ where $ H $ is the orthocenter of $ \triangle ABC, $ so $$ \frac{OP}{2MN} = \frac{OP}{TP} = \frac{\text{circumradius of }\triangle A_1B_1C_1}{\text{circumradius of }\triangle A_2B_2C_2} = \frac{R}{HT} = \frac{R}{OQ}. \qquad \blacksquare $$
Main result

Notation : Given a $ \triangle ABC $ with centroid $ G, $ circumcenter $ O, $ orthocenter $ H, $ 9-point center $ N, $ symmedian point $ K $ and $ \theta \in \left ( -\frac{\pi}{2}, \frac{\pi}{2} \right ]. $ Let $ \triangle A_{\theta}B_{\theta}C_{\theta} $ be the Kiepert triangle of $ \triangle ABC $ with angle $ \theta, $ $ K_{\theta} $ be the perspector of $ \triangle ABC, \triangle A_{\theta}B_{\theta}C_{\theta}, $ and let $ A^*_{\theta}, B^*_{\theta}, C^*_{\theta}, K^*_{\theta} $ be the isogonal conjugate of $ A_{\theta}, B_{\theta}, C_{\theta}, K_{\theta} $ WRT $ \triangle ABC, $ respectively.

Property 1 : $ \color{blue} K \in K_{\theta}K_{-\theta}. $

Proof : Let $ \triangle T_aT_bT_c $ be the tangential triangle of $ \triangle ABC. $ Note that $ \measuredangle CBC_{-\theta} = \measuredangle T_bCB_{\theta}, \measuredangle BCB_{-\theta} = \measuredangle T_cBC_{\theta}, $ so we conclude that $$ B(C,T_b; B_{\theta},B_{-\theta}) = C(B,T_b; B_{\theta},B_{-\theta}) = B(C,T_c;C_{\theta},C_{-\theta}) = C(B,T_c;C_{\theta},C_{-\theta}). $$i.e. $ K \equiv BT_b \cap CT_c, K_{\theta} \equiv BB_{\theta} \cap CC_{\theta}, K_{-\theta} \equiv BB_{-\theta} \cap CC_{-\theta} $ are collinear. $ \qquad \blacksquare $

Corollary 1.1 : $ \color{blue} G \in K_{\theta}K^*_{-\theta}. $

Proof : From Property 1 we get $ K $ is the intersection of $ K_{\theta}K_{-\theta}, K^*_{\theta}K^*_{-\theta}, $ so by Lemma 1 we conclude that $ G $ lies on $ K_{\theta}K^*_{-\theta}, K_{-\theta}K^*_{\theta}. $ $ \qquad \blacksquare $

Property 2 : $ \color{blue} H \in A_{\theta}A^*_{\frac{\pi}{2}-\theta}, B_{\theta}B^*_{\frac{\pi}{2}-\theta}, C_{\theta}C^*_{\frac{\pi}{2}-\theta}, K_{\theta}K^*_{\frac{\pi}{2}-\theta}. $

Proof : Simple angle chasing yields $ \measuredangle HBA_{\theta} = \measuredangle A^*_{\frac{\pi}{2} - \theta}CB, \measuredangle HCA_{\theta} = \measuredangle A^*_{\frac{\pi}{2} - \theta}BC, $ so by Lemma 2 we get $ H $ lies on $ A_{\theta}A^*_{\frac{\pi}{2}-\theta}. $ Similarly, we can prove $ H \in B_{\theta}B^*_{\frac{\pi}{2}-\theta}, C_{\theta}C^*_{\frac{\pi}{2}-\theta}. $ By Desargues's theorem for $ \triangle BC_{\theta}C^*_{\frac{\pi}{2}-\theta} $ and $ \triangle CB_{\theta}B^*_{\frac{\pi}{2}-\theta} $ $ \Longrightarrow $ $ BC, B_{\theta}C_{\theta}, B^*_{\frac{\pi}{2}-\theta}C^*_{\frac{\pi}{2}-\theta} $ are concurrent, so by Desargues's theorem for $ \triangle BB_{\theta}B^*_{\frac{\pi}{2}-\theta} $ and $ \triangle CC_{\theta}C^*_{\frac{\pi}{2}-\theta} $ we conclude that $ H $ lies on $ K_{\theta}K^*_{\frac{\pi}{2}-\theta}. $ $ \qquad \blacksquare $

Corollary 2.1 : $ \color{blue} O \in A^*_{\theta}A^*_{\frac{\pi}{2}-\theta}, B^*_{\theta}B^*_{\frac{\pi}{2}-\theta}, C^*_{\theta}C^*_{\frac{\pi}{2}-\theta}, K_{\theta}K_{\frac{\pi}{2}-\theta}. $

Proof : From Property 2 we get $ H $ is the intersection of $ K_{\theta}K^*_{\frac{\pi}{2}-\theta}, K_{\frac{\pi}{2}-\theta}K^*_{\theta}, $ so by Lemma 1 $ \Longrightarrow $ $ O $ lies on $ K_{\theta}K_{\frac{\pi}{2}-\theta}. $ Analogously, we can prove $ O \in A^*_{\theta}A^*_{\frac{\pi}{2}-\theta}, B^*_{\theta}B^*_{\frac{\pi}{2}-\theta}, C^*_{\theta}C^*_{\frac{\pi}{2}-\theta}. $ $ \qquad \blacksquare $

Property 3 : Given $ \color{blue} \phi, \sigma \in \left ( -\frac{\pi}{2}, \frac{\pi}{2} \right ]. $ Then $ \color{blue} K^*_{-(\phi + \sigma)} \in K_{\phi} K_{\sigma}. $

Proof : Note that for a fixed $ \tau \in \left ( -\frac{\pi}{2}, \frac{\pi}{2} \right ] $ the mapping $ \mathbb{I}: K_{\theta} \mapsto K_{\tau - \theta} $ is an involution on the Kiepert hyperbola of $ \triangle ABC, $ so $ K_{\theta} K_{\tau - \theta} $ passes through the pole $ X $ of $ \mathbb{I} $ for all $ \theta \in \left ( -\frac{\pi}{2}, \frac{\pi}{2} \right ]. $ Consider the case when $ \theta = 0 $ and $ \theta = \frac{\pi}{2} $ we conclude that $ X = K^*_{-\tau} $ by Corollary 1.1 and Property 2 $ \Longrightarrow $ $ K^*_{-\tau} \in K_{\theta} K_{\tau - \theta}. $ $ \qquad \blacksquare $

Property 4 : Let $  \color{blue} T $ be the intersection of $  \color{blue} OH $ with $  \color{blue} A_{\theta}A^*_{\theta}. $ Then $  \color{blue} \frac{OH}{OT} = 4\cos^2\theta-1. $

Proof : Let $ D $ be the intersection of $ AA^*_{\theta} $ with the perpendicular bisector of $ BC $ and let $ BD = CD = \eta, $ then from $ \measuredangle ADO = \measuredangle DAH = \measuredangle OAA_{\theta} $ we get $ OA^2 = OD \cdot OA_{\theta} $ $ \Longrightarrow $ $ D $ is the image of $ A_{\theta} $ under the inversion WRT $ \odot(O), $ so $ \eta = \frac{BC \cdot R}{2 \cdot OA_{\theta} \cdot \cos \theta}. $ On the other hand, by Lemma 2 we get $ \measuredangle A^*_{\theta}CD = \measuredangle ABC,  \measuredangle A^*_{\theta}BD = \measuredangle ACB, $ so $$ \frac{HA^*_{\theta}}{A_{\frac{\pi}{2}-\theta}A^*_{\theta}} = \frac{AA^*_{\theta}}{DA^*_{\theta}} = \frac{2R \cdot \sin\theta}{\eta} = \frac{4 \cdot OA_{\theta} \cdot \sin \theta \cdot \cos \theta}{BC},$$hence by Menelaus' theorem for $ \triangle OHA_{\frac{\pi}{2}-\theta} $ and $ \overline{A^*_{\theta}A_{\theta}T} $ we conclude that $$ \frac{HT}{OT} = \frac{HA^*_{\theta}}{A_{\frac{\pi}{2}-\theta}A^*_{\theta}} \cdot \frac{A_{\frac{\pi}{2}-\theta}A_{\theta}}{OA_{\theta}} = -2\cos 2\theta = -2\left(2\cos^2\theta -1 \right ) \Longrightarrow \frac{OH}{OT} = 4\cos^2\theta-1. \qquad \blacksquare $$
Corollary 4.1 : $ \color{blue} A_{\theta}A^*_{\theta}, A_{-\theta}A^*_{-\theta}, B_{\theta}B^*_{\theta}, B_{-\theta}B^*_{-\theta}, C_{\theta}C^*_{\theta}, C_{-\theta}C^*_{-\theta}, K_{\theta}K^*_{\theta}, K_{-\theta}K^*_{-\theta} $ are concurrent on $  \color{blue} OH. $

Proof : From Property 4, $ A_{\theta}A^*_{\theta}, A_{-\theta}A^*_{-\theta}, B_{\theta}B^*_{\theta}, B_{-\theta}B^*_{-\theta}, C_{\theta}C^*_{\theta}, C_{-\theta}C^*_{-\theta} $ are concurrent at $ V. $ By Desargues's theorem for $ \triangle BC_{\theta}C^*_{\theta} $ and $ \triangle CB_{\theta}B^*_{\theta} $ we get $ BC, B_{\theta}C_{\theta}, B^*_{\theta}C^*_{\theta} $ are concurrent, so by Desargues's theorem for $ \triangle BB_{\theta}B^*_{\theta} $ and $ \triangle CC_{\theta}C^*_{\theta} $ we conclude that $ V $ lies on $ K_{\theta}K^*_{\theta}. $ Analogously, we can prove $ V \in K_{-\theta}K^*_{-\theta}. $ $ \qquad \blacksquare $

Property 5 : $ \color{blue} N \in K_{\theta}K_{\frac{\pi}{2}+\theta}. $

Proof : Let $ D_{-\theta}, E_{-\theta}, F_{-\theta} $ be the midpoint of $ AA_{-\theta}, BB_{-\theta}, CC_{-\theta}, $ respectively. First, note that $ AB_{\theta}A_{-\theta}C_{\theta}, BC_{\theta}B_{-\theta}A_{\theta}, CA_{\theta}C_{-\theta}B_{\theta} $ are parallelogram, so $ \triangle D_{-\theta}E_{-\theta}F_{-\theta} $ is the medial triangle of $ \triangle A_{\theta}B_{\theta}C_{\theta}.$ On the other hand, simple angle chasing yields the second intersection of $ AA_{\frac{\pi}{2}-\theta} $ with $ \odot (A_{\frac{\pi}{2}-\theta}BC) $ lies on $ \odot (B_{\theta}), \odot (C_{\theta}) $ where $ \odot (B_{\theta}), \odot (C_{\theta}) $ is the circle with center $ B_{\theta}, C_{\theta} $ and containing $ A, $ so $ AK_{\frac{\pi}{2}-\theta} \perp E_{\theta} F_{\theta}. $ Analogously, we can prove $ BK_{\frac{\pi}{2}-\theta} \perp F_{\theta} D_{\theta}, CK_{\frac{\pi}{2}-\theta} \perp D_{\theta} E_{\theta}, $ so by Sondat's theorem for $ \triangle ABC $ and $ \triangle D_{-\theta}E_{-\theta}F_{-\theta} $ we conclude that $ N \in K_{-\theta} K_{\frac{\pi}{2}-\theta}. $ $ \qquad \blacksquare $

Application

Notation : Let $ G, O, N, K, F_1, F_2, S_1, S_2 $ be the centroid, circumcenter, 9-point center, symmedian point, 1st Fermat point, 2nd Fermat point, 1st Isodynamic point, 2nd Isodynamic point of $ \triangle ABC, $ respectively.

Theorem (Basic properties of Fermat points and Isodynamic points) :

(1) $ \qquad $ $ G \in F_1S_2, F_2S_1. $ $ \qquad $ $ K \in F_1F_2, S_1S_2. $ $ \qquad $ $ \overline{GON} \parallel F_1S_1 \parallel F_2S_2. $ $ \qquad $ (2) $ \qquad $ $GK $ is the G-symmedian of $ \triangle GF_1F_2. $ $ \qquad $ (3) (Lester circle) $ O, N, F_1, F_2 $ are concyclic.

Proof : (1) is a consequence of the results in the previous section. Let $ T $ be the intersection of $ GO, F_1F_2 $ and $ \triangle^{1}_{\mathbf{N}}, \triangle^{2}_{\mathbf{N}} $ be the 1st Napoleon triangle, 2nd Napoleon triangle of $ \triangle ABC, $ respectively. First, note that $ O $ lies on $ S_1S_2, $ so $ T $ is the reflection of $ O $ in $ G. $ Since $ G $ is the center of $ \triangle^{i}_{\mathbf{N}}, $ $ F_{j} $ lies on the circumcircle of $ \triangle^{i}_{\mathbf{N}} $ and $ O,F_i $ are isogonal conjugate of $ \triangle^{i}_{\mathbf{N}} $ for $ \{ i, j \} = \{ 1, 2 \}, $ so by Lemma 3 we get $$ \frac{KF_1}{KF_2} = \frac{TF_1}{TF_2} = \frac{GO \cdot GF_1}{GO \cdot GF_2} \cdot \frac{\text{circumradius of }\triangle^{2}_{\mathbf{N}}}{\text{circumradius of }\triangle^{1}_{\mathbf{N}}} = \left ( \frac{GF_1}{GF_2} \right )^2, $$hence $ GK $ is the G-symmedian of $ \triangle GF_1F_2 $ and (2) is proved. Finally, from the previous conclusion and $ (T,K; F_1, F_2) = -1 $ we get $ \overline{GON} $ is tangent to $ \odot (GF_1F_2) $ at $ G, $ so $ TO \cdot TN = {TG}^2 = TF_1 \cdot TF_2 $ and hence we conclude that $ O, N, F_1, F_2 $ are concyclic. i.e. (3) is proved. $ \qquad \blacksquare $

Remark : It is possible to prove (2) or (3) directly.

1. Proof of (2) without using (1) :

Since $ F_1, F_2 $ are antigonal conjugate WRT $ \triangle ABC, $ so the center of the Kiepert hyperbola $ \mathcal{K} $ of $ \triangle ABC $ is the midpoint of $ F_1F_2 $ $ \Longrightarrow $ the isogonal conjugate of $ \mathcal{K} $ WRT $ \triangle GF_1F_2 $ is the perpendicular bisector of $ F_1F_2, $ hence the tangent of $ \mathcal{K} $ at $ G $ is the G-symmedian of $ \triangle GF_1F_2. $ On the other hand, $ \mathcal{K} $ is the isotomic conjugate of $ GK $ WRT $ \triangle ABC, $ so $ GK $ is tangent to $ \mathcal{K} $ at $ G $ and hence $ GK $ is the G-symmedian of $ \triangle GF_1F_2. $ $ \qquad \blacksquare $

2. Proof of (3) without using (1) and (2) :

Lemma : Given a $ \triangle ABC $ and a circle $ \Omega $ passing through $ B $ and $ C. $ Let $ E, F $ be the intersection of $ \Omega $ with $ CA, AB, $ respectively and let $ Y \in BE, Z \in CF $ be the points such that $ \frac{BE}{EY} = \frac{CF}{FZ} = \kappa. $ Suppose that $ U, V $ is the intersection of the tangent of $ \odot (AYZ) $ at $ A $ with $ BC, YZ, $ respectively, then $ \frac{UA}{AV} = \kappa. $

Proof : Let $ B^* \in CA, C^* \in AB $ be the points such that $ \frac{BA}{AC^*} = \frac{CA}{AB^*} = \kappa $ and let $ A^* $ be the intersection of $ B^*Z, C^*Y. $ Clearly, $ \triangle ABC, \triangle A^*B^*C^* $ are homothetic and $ \triangle AEY \stackrel{-}{\sim} \triangle AFZ, $ so $$ \frac{VY}{VZ} = \left ( \frac{AY}{AZ} \right )^2 = \left ( \frac{AE}{AF} \right )^2 =  \frac{AE}{AF} \cdot \frac{AB}{AC} = \frac{C^*Y}{B^*Z} \cdot \frac{A^*B^*}{A^*C^*}, $$hence by Menelaus' theorem we conclude that $ V \in B^*C^* $ $ \Longrightarrow $ $ \frac{UA}{AV} = \frac{\text{dist}(A,BC)}{\text{dist}(A,B^*C^*)} = \kappa. $ $ \qquad \blacksquare $

Back to the main problem :

Let $ \triangle N^{1}_aN^{1}_bN^{1}_c, \triangle N^{2}_aN^{2}_bN^{2}_c $ be the 1st Napoleon triangle, 2nd Napoleon triangle of $ \triangle ABC, $ respectively and let $ T^* $ be the intersection of $ F_1F_2 $ with the tangent of $ \odot (GF_1F_2) $ at $ G. $ Since $ F_1, F_2 $ is the reflection of $ N^{2}_a, N^{1}_a $ in $ GN^{1}_a, GN^{2}_a, $ respectively, so by Lemma for $ \triangle GN^{1}_aN^{2}_a $ we get the reflection $ \overline{T} $ of $ T^* $ in $ G $ lies on the perpendicular bisector of $ BC. $ Similarly, we can prove $ \overline{T} $ lies on the perpendicular bisector of $ CA, AB, $ so $ \overline{T} \equiv O $ and hence we conclude that $ T^*N \cdot T^*O = T^*G^2 = T^*F_1 \cdot T^*F_2 $ $ \Longrightarrow $ $ O, N, F_1, F_2 $ are concyclic. $ \qquad \blacksquare $

Inequality with three variables

by crazyfehmy, May 31, 2013, 7:29 PM

Factorial: n!|a^n+1

by Nima Ahmadi Pour, Apr 24, 2006, 10:55 AM

Find all positive integers $ n$ such that there exists a unique integer $ a$ such that $ 0\leq a < n!$ with the following property:
\[ n!\mid a^n + 1
\]

Proposed by Carlos Caicedo, Colombia
Tags
About Owner
  • Posts: 2312
  • Joined: Oct 8, 2014
Blog Stats
  • Blog created: Jun 15, 2016
  • Total entries: 26
  • Total visits: 54408
  • Total comments: 5
Search Blog
a