Willis Heating System

[JohnW2]

Do you accept, that a heating radiator on one wall, will cause the air to rise above it, makes it way across the ceiling, then fall down the far wall - the traditional theory of how a radiator in a room behaves?

Indeed, I do accept that.

However, as things are, I'm far less able to accept that a radiator mounted high up on the wall of a tall room would result in appreciable heating of the room below the level of the radiator - and that particularly if the radiator had aTRV (hence my belief that the location of the stat in a Willis system is crucal).

 

[Harry Bloomfield]

However, as things are, I'm far less able to accept that a radiator mounted high up on the wall of a tall room would result in appreciable heating of the room below the level of the radiator

The air circulates - rises from the radiator, across the ceiling, down the far wall, along the floor, and back to the radiator. Exactly the way the Willis, circulates water. Where the stat is located, in the Willis system, will make minimal difference to it's operation, due to the rapid movement of the water.

 

[JohnW2]

The air circulates - rises from the radiator, across the ceiling, down the far wall, along the floor, and back to the radiator. Exactly the way the Willis, circulates water.

I'm afraid I just don't get it. .... Air will only 'rise from the radiator' so long as the air above the radiator is colder than that air heated by the radiator. Once the air above the radiator is up to that temp, there is no reason for any further 'rising'.

"Hot air rises" is an over-simplification. Air which is hotter than that above it will rise, but if the air above it is just as hot, there will be no 'rising', no matter how 'hot' the air.

Where the stat is located, in the Willis system, will make minimal difference to it's operation, due to the rapid movement of the water.

That's surely nonsense, whether in relation to the Willis system (above or level with the cylinder), an internal immersion in a cylinder or the radiator mounted high up a wall, isn't it? In all cases, if a thermostat local to the heating source switches off the heating as soon as the local air is 'up to temp', then there will surely be no appreciable heating of water or air below that location (of heating source and stat), will there?

 

[Harry Bloomfield]

I'm afraid I just don't get it. .... Air will only 'rise from the radiator' so long as the air above the radiator is colder than that air heated by the radiator. Once the air above the radiator is up to that temp, there is no reason for any further 'rising'.

Which ignores the fact, that the rising air around the radiator - will always be hotter. The air at the ceiling can never get to that temperature, that the convection currents can ever come to a stop. Likewise, the Willis.

 

[JohnW2]

Which ignores the fact, that the rising air around the radiator - will always be hotter. The air at the ceiling can never get to that temperature, that the convection currents can ever come to a stop.

Literally true, but the temp of the air above a high radiator would fairly quickly get mighty close to that of air which was newly-heated by the rad, such that density differences, and hence 'buoyancy', would presumably be minimal, hence only slow and slight 'rising' at most, wouldn't it?

Likewise, the Willis.

Even worse with a Willis, since the only thing 'above' the heater is then small column of water in a short length of relatively small-bore pipe. The temp of water in that pipe would very quickly rise to that at the top of the Willis heater, at which point any 'rising' would essentially grind to a halt, wouldn't it? Furthermore, as I wrote, if there were a stat in the Willis, it would quite rapidly put a stop to any rising, unless it was set to a frighteningly high temp, wouldn't it?

 

[Harry Bloomfield]

Literally true, but the temp of the air above a high radiator would fairly quickly get mighty close to that of air which was newly-heated by the rad, such that density differences, and hence 'buoyancy', would presumably be minimal, hence only slow and slight 'rising' at most, wouldn't it?

No!

Even worse with a Willis, since the only thing 'above' the heater is then small column of water in a short length of relatively small-bore pipe. The temp of water in that pipe would very quickly rise to that at the top of the Willis heater, at which point any 'rising' would essentially grind to a halt, wouldn't it? Furthermore, as I wrote, if there were a stat in the Willis, it would quite rapidly put a stop to any rising, unless it was set to a frighteningly high temp, wouldn't it?

And again - No! Otherwise the Willis would not work at all, and absolutely no one would buy one.