Talk:Pipe network analysis

Untitled
On the subject of Colebrook equation was said: "Given an initial guess value, usually two or three iterations are enough to give a good approximation with the following expression." Is there a source or a proof for this? 194.100.107.130 (talk) 06:40, 11 December 2008 (UTC)

I think the delta Q equation should have parenthesis in the denominator around the summations, i.e. 2(E+E).129.2.175.80 (talk) 19:19, 25 January 2010 (UTC)

An easier method
I found out this algorithm cumbersome, with having to find and track loops. It turns out that a simple algorithm with Navier-Stokes numerical calculation can solve arbitrary networks without assigning direction, structure or initial conditions, including pumps, free circulation (hot water has lower density, so gravity does the work), pipes with an open to atmosphere end, etc. Initialize all pressure and velocity to zero and it finds a highly precise solution in about 100 rounds. Initially conservation of mass or consistent head pressures are not present but the algorithm converges to such conditions. My solver is four lines of code in a loop, applying head pressures "by force" (between the ends of each component), doing a pressure correction step at nodes (pipe intersections), adjusting mass flows, and repeating the whole thing until done (Python Numpy). 93.108.178.201 (talk) 01:05, 2 December 2023 (UTC)