On universal sums $x(ax+b)/2+y(cy+d)/2+z(ez+f)/2$

Zhi-Wei Sun
2015-01-01
Abstract:Let $a,b,c,d,e,f$ be integers with $age cge eu003e0$, $bu003e-a$ and $bequiv apmod2$, $du003e-c$ and $dequiv cpmod 2$, $fu003e-e$ and $fequiv epmod2$. Suppose that $bge d$ if $a=c$, and $dge f$ if $c=e$. When $b(a-b)$, $d(c-d)$ and $f(e-f)$ are not all zero, we show that if each $ninmathbb N={0,1,2,ldots}$ can be written $x(ax+b)/2+y(cy+d)/2+z(ez+f)/2$ with $x,y,zinmathbb N$ then the tuple $(a,b,c,d,e,f)$ must be on our list of $473$ candidates, and show that 56 of them meet our purpose. When $bin[0,a)$, $din[0,c)$ and $fin[0,e)$, we investigate the universal tuples $(a,b,c,d,e,f)$ over $mathbb Z$ for which any $ninmathbb N$ can be written $x(ax+b)/2+y(cy+d)/2+z(ez+f)/2$ with $x,y,zinmathbb Z$, and show that there are totally 12082 such candidates some of which are proved to be universal tuples over $mathbb Z$. For example, we show that any $ninmathbb N$ can be written as $x(x+1)/2+y(3y+1)/2+z(5z+1)/2$ with $x,y,zinmathbb Z$, and conjecture that each $ninmathbb N$ can be written as $x(x+1)/2+y(3y+1)/2+z(5z+1)/2$ with $x,y,zinmathbb N$.
What problem does this paper attempt to address?