For the best answers, search on this site https://shorturl.im/ayj0u
'Alexander,' your basic error is that Bernoulli's Principle applies along STREAMLINES, and only throughout a medium under various assumptions about conditions far upstream, or other restrictive conditions such as potential flow, or yet others too tedious to give here. (Think about how it is in fact derived --- by FOLLOWING the motion along a streamline!) Your supposed application of it simply violates the conditions under which it applies. That is the key point. Proceeding further, however: 'Bekki B' is correct in saying that "The water column is supported by pressure from below which differs because of Bernoulli's principle," but no-one seems to have seen whether the numerical details given in the picture are reasonable, or what they together imply. So let's look at that. For most practical purposes, water may be considered incompressible. Call its density 'ρ.' Although there will in practice be velocity differences across the pipe's cross-sections, let's assume that the velocity profiles are similar. (This could be the poorest assumption, but we need something like this, to proceed.) Then characterizing the flows by the values of the velocities along the central axis of (assumed) pipes of similar cross-sections A1, A2, as the figure invites us to do, we will have A1*v1 = A2*v2, or d1^2*v1 = d2^2*v2, that is v2 = (d1/d2)^2*v1. But d1/d2 = (90mm/40mm) = (3/2)^2, thus v2 = (3/2)^4 * v1, and v2^2 = (3/2)^8 * v1^2. Along the central streamline, with height h constant, we'll have v1^2 / 2 + P1/ρ = v2^2 / 2 + P2/ρ. Rearranging and inserting the above v1, v2 relationships we obtain: P1 - P2 = ρ * (v1^2 - v2^2) / 2, or numerically P1 - P2 = ρv1^2 [(3/2)^8 - 1]/2 = 12.31... ρv1^2. In cgs units, with ρ = 1gm/cm^3, this gives simply P1 - P2 = 12.31... v1^2. ......(A) VERTICALLY, the usual pressure support equation will hold. It does NOT have any reference to the horizontal velocity in it. That implies P1 - P2 = ρg (h1 - h2), where h1 - h2 = 37mm = 3.7cm. Then in cgs units, we have numerically P1 - P2 = 1*981*3.7 = 3630... cgs units. ......(B) Equations (A) and (B) imply: 12.31... v1^2 = 3630..., so that v1^2 = 294.8, or v1 = 17.2 cm/s. Then v2 = (3/2)^4 * v1 = 86.9 cm/s. These speeds, together with the pipe diameters given, seem reasonable for a laboratory experiment. There's a good chance that the flow would be laminar, which is of course implicitly built into the analysis. Live long and prosper. [Later edit: The algebraic result, before insertion of numerical values, follows from equating the two expressions for P1 and P2. That gives, simply: v2^2 - v1^2 = 2 g (h1 - h2). That is generally true, and shows very clearly how the speed-up from v1 to v2 is associated with a decrease in h2. It also shows how SMALL velocity differences at HIGH speeds are as effective as larger velocity differences at low speeds. To go further and derive the separate values of v1 and v2 requires an appeal to the ratios of the actual cross-sectional areas within which the flows are confined, such as that given by assuming d1^2*v1 = d2^2*v2, as I did above.] [ FURTHER EDIT: 'Alexander,' your second "Additional details" item misses the point, once again. You are in effect attempting to apply Bernoulli's Principle ACROSS rather than ALONG streamlines! There is NO streamline connecting the tops of those water columns, and the "Po's" in each of your three supposed applications of it will NOT be the same. THAT is what you have effectively assumed. ]