Given x≥y≥z, then
3x3=x3+y3+z3=nx2y2z2.∴3x=ny2z3.∵y,z∈N,∴1+y3z3≥y3+z3=nx2y2z2−x3=x2(ny2z2−x)⩾x3.
From (1) and (2), we get
9(1+y3z3)⩾9x2⩾n2y4z4.
If yz>1, then y4z4>1+y3z3.
From (3), we get 9>−n2. Therefore, n=1 or 2. That is, n⩽4.
Thus, it has been proven that (**) has no positive integer solutions. Now, we will discuss the cases for n=1,2,3,4 separately.
For n=1, the solution is
x=3;y=2,z=1.
For n=2, (**) becomes
x3+y3+z3=2x2y2z2.
From (1) and (3), we get 9(1+y3z3)⩾4y4z4.
∴yz⩽2.
i) If yz=1, then (4) becomes
x3+2=2x2.
This equation has no integer solutions, as 2x2−x3 is a multiple of 4, which cannot equal 2, leading to a contradiction. Therefore, equation (4) has no solutions.
ii) If yz=2, then (4) becomes
x3+9=8x2.
This simplifies to (x+1)(x2−9x+9)=0.
This equation has no positive integer solutions, so (4) has no solutions.
For n=3, (**) has the solution
x=y=z=1.
For n=4, (**) becomes
x3+y3+z3=4x2y2z2.
From (1), we know yz=1, so (5) becomes
x3+2=4x2.
If this equation has a solution, then x must be even. However, 4x2−x3 is a multiple of 4, leading to a contradiction. Therefore, (**) has no solutions in this case.
In summary, when n=1 or 3, (**) has positive integer solutions.