And this one is too complex... Like a maze and because of Conservation of Momentum, everything is zero:
[PLAIN]http://www.2live.ro/demo/ball.bmp[/QUOTE]
No - actually, conservation of momentum works perfectly well in this case. All you need to do is look at the flow entering the pipe, and the flow exiting the pipe. In this case, the flow entering the pipe is going upwards, but the exit flow is downwards. Thus, by conservation of momentum, there must be a net upwards force on the pipe to counter the downwards force on the fluid. However, this all changes when you bend the above system:
http://www.2live.ro/demo/arc_open.bmp
Now, the incoming flow (I'll assume the inlet is on the left) is traveling to the left. The outcoming flow is also traveling to the left. Assuming the pipe is constant diameter, and the fluid is incompressible, the flow speed is the same at the inlet and outlet. Therefore, since the flow speed and direction are identical at both the inlet and the outlet, there is no net momentum change in the fluid, so there is no force on the pipe.
I think you're confusing yourself with the spiral pipe. It's true that in the first image above, the net force can be considered to be entirely from the U-shaped section, and the forces in the spirals all cancel out. However, this is no longer the case when you bend the spirals. The force nearly cancels out in the spirals when they are bent (so long as the bend radius is large relative to the spiral radius), but it doesn't entirely cancel out, and if you felt inclined to do the vector sum along the entire path, you would discover that the slight imbalance in the spiral is exactly enough to counteract the force in the U-shaped section. That method is absolutely valid, it simply isn't used very frequently since conservation of momentum allows us to arrive at the same result with much less effort (and it allows us to bypass errors of intuition more easily).