Re: Under-tank heating coils.

> From: Kevin Conlin <kcconlin at zola_cae.ca>
> If I were to uniformly heat a container of water from below, I would
> expect to see the formation of roughly hexagonal convection cells,
> each with warm water rising in a narrow column at the center and cool
> water subsiding around it.  Buried wires presumably give rise to
> linear cells.  For a given amount of heat applied to the system I
> would expect the same rate of water transport through the substrate.
> Why would either system have any advantage over the other?  How can a
> plant distinguish between linear and hexagonal convection cells?  -

I was reading through a thermodynamics book a few months ago trying to
find some way to calculate how much flow would result from how much
heat (this is a very difficult problem - way beyond my capability) but
I did find that below a certain heat density, heat will travel by
CONDUCTION in a liquid and will not produce movement.  Above that
threshold and below another, CONVECTION will occur and produce a
laminar flow.  Above the second threshold, turbulent flow will begin,
something like boiling.

IMHO, coils can produce enough localized heating to generate a
CONVECTION flow around them.  To heat the entire bottom surface enough
to create the same amount of heat density would take a lot of heat. 

So it's not a matter of linear vs. whatever, it's how much heat
density can be generated.

Assuming a Dupla 100w coil is 1/4" dia. by 22 ft long, it has about
200 sq in of surface area, or 0.5 watts per square in.  A 75 gallon 
tank (48"x18") has about 860 sq in of bottom area which would require
430 watts to produce the same heat density as the coil.  This may tend
to overheat the tank.