Page 1 of 1

Graham condenser

Posted: Thu Jan 22, 2009 6:40 pm
by pHneutral
Just wondering.. is there any reason, other than difficulty in assembly, that more people don't use a graham style condenser? This is one with the tube through the jacket coiled around like a spring, as is seen in reflux condensers, only the vapor flows through the coil.

http://www.uicoglass.com/condensers/ui3525.jpg" onclick="window.open(this.href);return false;" rel="nofollow

I would guess the first reason would be vapor flow obstruction, which is why some prefer a shotgun style instead, but if you had enough cooling you should be able to knock it down pretty well given the increase in surface area and vapor path length.

Re: Graham condenser

Posted: Thu Jan 22, 2009 7:05 pm
by rad14701
I have one half built to use in place of a liebig as a take off condenser for my reflux column... The spiral should slow the product speed and keep it in contact for a longer time due to the added internal length...

Re: Graham condenser

Posted: Thu Jan 22, 2009 7:36 pm
by snuffy
It's a way of getting a long skinny tube in a short fat space. A Graham condenser is a big step over a Leibig condenser of the same outside length. Unlike a Leibig, the Graham has to be installed vertically to prevent liquid pooling in the tube and acting like a trap.

All condensers are about the surface area of the cooled surface. More surface area, more heat transfer.

An aspect that is ignored too often is the distance the vapor must travel to touch the cooled surface. There is no action-at-a-distance, it has to touch. So the smaller (thinner) the vapor space, the more efficient the heat exchanger.

Alternatives to skinny tubes: fat tubes nested closely inside each other, turbulators to prevent the flow channeling into the center of a large tube, corrugated surfaces that both turbulate and increase the surface area over a cylindrical shape, packing the space with copper mesh, etc, etc...

Assuming a reasonably good design (mostly no big voids in the vapor space) and using copper plumbing tube and pipe, it really comes down to the mass of the copper. More copper, more cooling capacity.

Long skinny tubes can get overloaded and plug with liquid, causing burping and spitting (and sometimes also causing surge boiling from the fluctuating pressure.) Many skinny parallel tubes are one solution, like in shell and tube heat exchangers. 1/4" ID is pretty much the lower limit before plugging with liquid becomes an issue.

Two 3/8" tubes in parallel spirals (a double coil condenser turned inside out) would have adequate capacity for up to about 1500-1800 Watts. The rule of thumb for tube bending is the spiral diameter needs to be at least 6 tube diameters to avoid flattening or crimping, so you'd end up with (3/8 * 6) + (3/8 * 2) = (18 + 2) / 8 = 2.5" OD on the spiral, which would fit nicely in a 3" pipe.

You'd want about 4' in each run. Each turn would be about 6.5" in length, so you'd need about 7 or 8 turns, with double helices wound on the same radius, you'd get about 8" length for the shell. Might be worth a try. A glass shell would be cool, it would look like a short fat strand of DNA...

*mumbles... wanders off*

Re: Graham condenser

Posted: Thu Jan 22, 2009 10:55 pm
by HookLine
Love your work, snuffy.
snuffy wrote:The rule of thumb for tube bending is the spiral diameter needs to be at least 6 tube diameters to avoid flattening or crimping, so you'd end up with (3/8 * 6) + (3/8 * 2) = (18 + 2) / 8 = 2.5" OD on the spiral, which would fit nicely in a 3" pipe.
Some flattening ain't necessarily a bad thing. The reason is to do with the cross-sectional shape of the condenser tube, specifically a circle versus an ellipse (or indeed any 2 dimensional shape other than a circle). For an equal cross-sectional-area (CSA), an ellipse has a higher ratio of perimeter (ie heat exchange surface area) to internal area (ie coolant volume), and less distance between any point in the coolant and the heat exchange surface. This means the condenser is more efficient at exchanging heat, which in turn means that it takes less condenser to exchange the same amount of heat.

The trade-off is that any non-circular CSA gives higher resistance to coolant flow.

Re: Graham condenser

Posted: Fri Jan 23, 2009 8:32 am
by snuffy
HookLine wrote:Some flattening ain't necessarily a bad thing. ...
The trade-off is that any non-circular CSA gives higher resistance to coolant flow.
Totally right. The heat transfer is by surface area and flattening doesn't change that a bit.

My 3/8" in 1/2" coiled heat exchangers flatten the 1/2" a little, but not enough to impede the liquid flow.

Re: Graham condenser

Posted: Fri Jan 23, 2009 9:32 am
by pHneutral
Fantastic!!! :)

I see.. hence the "cold finger core double helix" condenser I have seen around. I didn't immediately think of that as a graham variant, but it is, although it is a reflux style one.

Thanks guys, and especially you snuffy, for the info!!! It's got me thinking about a lot of things.