What Are the Possible Values of k Given Specific Conditions?

  • Context: MHB 
  • Thread starter Thread starter Albert1
  • Start date Start date
Click For Summary

Discussion Overview

The discussion revolves around determining the possible values of \( k \) defined by the equation \( k=\dfrac {a^2+ab+b^2}{ab-1} \) under the conditions that \( a, b, k \in \mathbb{N} \) and \( a > b \). Participants explore various approaches to find the maximum and all possible values of \( k \), including specific cases and mathematical reasoning.

Discussion Character

  • Exploratory, Mathematical reasoning, Debate/contested

Main Points Raised

  • Some participants report finding specific values of \( k \), such as \( k=4 \) with \( (a,b) = (10,4) \) and \( k=7 \) with \( (a,b) = (2,1) \).
  • One participant suggests that the minimum value of \( k \) is 4 and the maximum is 7, proposing that these are the only possible values for \( k \) under 20.
  • Another participant presents a mathematical approach by rewriting the equation as a quadratic in \( a \) and identifies a necessary condition for integer solutions involving a Diophantine equation, finding solutions for \( k=4 \) and \( k=7 \) but not for other values up to 20.
  • There is a discussion about the parity of \( a \) and \( b \), noting that if both are odd, \( k \) does not exist, while if both are even, \( k \) is also even.
  • Participants express uncertainty about proving certain conditions that would confirm the proposed values of \( k \), particularly in relation to the derived equations and inequalities.

Areas of Agreement / Disagreement

Participants generally agree on the values \( k=4 \) and \( k=7 \) being possible, but there is no consensus on whether these are the only values or how to conclusively prove the conditions leading to these results.

Contextual Notes

Participants note limitations in their approaches, including unresolved mathematical steps and the dependence on specific assumptions about the parity of \( a \) and \( b \). The discussion reflects ongoing exploration rather than definitive conclusions.

Albert1
Messages
1,221
Reaction score
0
(1)$a,b,k\in\mathbb{N}$

(2)$a>b$

(3)$k=\dfrac {a^2+ab+b^2}{ab-1}$

please find :

(i) $\max(k)$

(ii) all possible values of $k$
 
Last edited:
Mathematics news on Phys.org
I am getting nowhere with this, but [sp]there are solutions for $k=4$ (for example $(a,b) = (10,4)$) and $k=7$ ($(a,b) = (2,1)$). I think that those are the only values of $k$ for $k\leqslant 20$, but I do not see how to exclude higher values of $k$. (Headbang)[/sp]
 
Albert said:
(1)$a,b,k\in\mathbb{N}$

(2)$a>b$

(3)$k=\dfrac {a^2+ab+b^2}{ab-1}$

please find :

(i) $\max(k)$

(ii) all possible values of $k$
$k>\dfrac{\sqrt[3]{3a^3b^3}}{ab}=3---(1)$(AP >GP)
if a,b both are odd numbers then k does not exist
if a,b both are even numbers then k is also a even number (ex:a=4,b=2 then k=4)
so min(k)=4
let a=2m, b=2n,here $m,n \in N$
then :
$k=\dfrac {4(m^2+mn+n^2)}{(4mn-1)}=4x (x\in N)$
for all a,b being even numbers ,we want to prove k=4 then we must prove x=1,that is to prove :
$x=\dfrac{(m^2+mn+n^2)}{(4mn-1)}=1$------(2)(if $x\in N)$
if a odd and b even then k must be odd (ex:a=11 ,b=2 then k=7)
also if a even and b odd then k must be odd
let a=b+d (here b:even and d odd ,so a must be odd)
$k=\dfrac {3b^2+3bd-3+d^2+3}{b^2+bd-1}=3+\dfrac{d^2+3}{b^2+bd-1}$
here $\dfrac{d^2+3}{b^2+bd-1}$ must be even
now we must prove :
$y=\dfrac{d^2+3}{b^2+bd-1}=4$-----(3)(if $y\in N)$

if (2) and (3) can be proved then all is done ,that is min(k)=4 ,and max(k)=7
all the possible valus of k=4 and 7
the proof (2) and (3):I am still thinking ----
By using a program (10001>a>b>0) the possible valus of k=4 and 7
I know (2) and (3) must be true , but how to prove it ?:confused:
 
Last edited:
Albert said:
(1)$a,b,k\in\mathbb{N}$

(2)$a>b$

(3)$k=\dfrac {a^2+ab+b^2}{ab-1}$
My approach was to write this as $a^2+ab+b^2 = k(ab-1)$ and to solve it as a quadratic in $a$, namely $a^2 - b(k-1)a + (b^2+k)$, with solutions $$a = \tfrac12\Bigl(b(k-1) \pm\sqrt{b^2(k-1)^2 - 4(b^2+k)}\Bigr).$$ For integer solutions a necessary condition is $b^2(k-1)^2 - 4(b^2+k) = c^2$ for some $c\in\mathbb{N}.$ So we need the Diophantine equation $(k-3)(k-1)b^2 - 4k = c^2$ to have solutions. I could find solutions when $k=4$ or $7$ but not for other values of $k\leqslant20$. That was as far as I could get.
 

Similar threads

  • · Replies 3 ·
Replies
3
Views
1K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 8 ·
Replies
8
Views
3K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 21 ·
Replies
21
Views
1K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 12 ·
Replies
12
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K