On 16/02/2013 18:56, Onetap wrote:
On Saturday, February 16, 2013 6:18:16 PM UTC, Colin Stamp wrote:
On 16/02/2013 15:28, Onetap wrote:



I still don't agree with that. Please note that I'm not saying it won't happen, I'm saying it wouldn't anticipate it happening.



The convection just generates a pressure difference due to the different masses of the H&C water columns.



With BBOE and no convection (air, rad, water at the same

temperature) water will still flow up the first channels and

down the last, the flow rate being such that the frictional pressure

loss equals the pressure difference across the inlet & outlet

provided by the pump.



I wouldn't expect that. The vast majority of the water will go whistling

straight through the bottom tank and out of the other end. To get to the

top of the rad, it would have to fight the restriction of one of the

channels, both up and then back down. in BBOE, it needs convection to

have a reason to do that.

It definitely does do that. It's the electrical resistance analogy with
multiple flow paths. You could work it out from the same principles, i.e.,
flow in = flow out, differential across all the flow paths is the same,
flow along any one path is determined by the resistance and the
differential.

That's why I said the vast majority would flow through the bottom tank.
Not *all* of it. The top and bottom tanks generally have a much wider
bore than the vertical channels. Any path in BBOE that includes a bit of
top tank, also has two vertical channels in it, along with the same
amount of tank as the straight through path and four right-angle turns
for good measure. The overall resistance to flow of a top-tank path will
be a lot higher than the straight-through path, so the flow will be much
lower.

Let's be charitable and assume that, say, 30% of the water does manage
to go through the top tank. The 70% that completely bypasses the core of
the rad would be absolutely disastrous in terms of system balancing.
Luckily, with the boiler turned on, convection sorts BBOE out.


With BBOE plus convection, it works as above, but pump and

convection forces work together.



Actually, the pump would work against convection with BBOE, but the

effect is very small because the velocity of water in the rad due to the

pump is so low. Convection easily overwhelms it.

Convection assists the pump, driving more of the flow up the alternative
flow paths, along the top header, rather than straight through the bottom
header.

The pump-only flow pattern has most of the flow going straight through
the bottom tank. Convection virtually completely wipes out that flow
pattern and replaces it with one where virtually all of the water goes
via the top tank. Hence convection fights the pump - and wins
comprehensively.

Note that this only applies *inside* the radiator, not in the pipes.

Cheers,

Colin.