Uniqueness conjecture

102 Views Asked by At

Define an H-Triple (a, b, c), where a > b > c > 1, as one where a, b and c = $\frac{ab+1}{a+b}$ are all integers. (There are 'lots' of these!)

Conjecture; if (a, b, c) and (d, e, f) are distinct H-Triples, then abc$\neq$def.

That is, the product of the terms of an H-Triple is unique.

Can anybody help with this? Computers suggest any counterexample will not be small.

The reason the problem arises? Consider the family of elliptic curves $xy\frac{xy+1}{x+y}=k$ and the integer points that lie on them. If (a,b,c) is an H-Triple and k = abc then (a,b) is an integer point on this elliptic curve. The uniqueness conjecture would narrow down the possibility that there might be any others.

1

There are 1 best solutions below

3
On

Here is an idea to search for a counterexample. We have the following universal $H$-triples, namely $$ (a, b ,c)=(2k+1,2k-1,k),\; (r^2+r-1,r+1,r) $$ for all $k,r\ge 1$. Equality of their products means $$ 4k^3-k=r^4+2r^3-r $$ This Diophantine equation has the "trivial" solution $(k,r)=(2,2)$. In this case the triples coincide and are $(a, b,c)=(5,3,2)$ with product $abc=30$. With a computer one could try to find a non-trivial solution in positive integers, if it exists. This would yield two different $H$-triples with equal product. Of course one may use also different universal triples, like $(a,b,c)=(2l^2+4l+1,2l+3,2l+1)$ and others.