Design improvement
 

As promised some piccies and explanation.
What I was aiming at was increase of area and actually some increase in pressure (velocity) due to maintained (roughly) cross sectional area after stream leaves nozzle.
Apologies for crappy quality of renderings due to lack of time... :(

It's only schematic for starters. Red stuff represents blocks base, blue are nozzles and greyish bits represent my 'innovation'.


http://www.hpphotos.com/servlet/Link...71833be3&size=

TOP VIEW
http://www.hpphotos.com/servlet/Link...24f2458d&size=

LEVEL VIEW
http://www.hpphotos.com/servlet/Link...5b6040d2&size=

Last One
http://www.hpphotos.com/servlet/Link...50066f13&size=

The only problem is in manufactruing the damn thing. One could machine it form one piece of copper but costs probably would be greatly prohibitive. Then I thought about drilling hole for each pin. If nozzle is 1mm dia and is spaced by 2mm from each other then pins could be 2.5mm in dia if conoidal or 2.5 square base tapering to a point, which is much easier to make - just take 2.5mm dia hard copper or brass wire , grind it to shape leaving around 3mm for thread at one end, cut thread and screw it into ready threaded hole in baseplate.

What do you think?

Les, I always thought that we are dealing here with conductive energy transfer, not convectional (vector oriented, usually in static flow, or totally static conditions, in liquids that is) thermal energy transfer.
I may be totally wrong here since I only atrted doing this stuff recently :)

Think again and again.

Nice design indeed.

The pins act as the required fins, and there's still an inpingement effect on the baseplate and, to some extend, up the sides of the pins.

But the proportion of those pins leave a much smaller "fin equivalent". The "Cascade" being the reverse design, the fins are more massive, in higher proportion, and most important: connected.

I'm sure it would be an excellent performer, but I wouldn't bet that it would beat a Cascade just yet.

The main issue I see is that by reversing the design, you have the incoming flow interfere with the second inpingement effect, and adding an almost useless laminar flow parallel to the pins. The secondary inpingement may also end up forming on a plane perpendicular to the baseplate, between the pins, which would be another loss of pressure, without any thermal benefit.

On a scale so small, I'd consider press fitting the copper pins into the baseplate. Got an Enermax around the shop?

Made a cut-out that is about 1.5mm larger on each edge over the Barton AthlonXP core size (i.e. an 17x10mm rectangle). It effectively blocks off about half of the tubes. The blocked off tubes do receive a small amount of flow as the water flows through the chamfers between the jet intakes.

After some testing, I'm seeing basically at least a full 0.5C improvement with the Iwaki MD30-RZ pump, which I guess is about as much as I would expect.

Re: the XXX. The program is written up and I fixed/tweaked a few things. The machinists won't get time to run until about 10 days away, and even then it's a "maybe". They are being pretty nice to me with the prices they're charging so I can't complain too loudly if higher paying customer jobs hold things up a bit.

Blocking off 50% of the jets should make even a bigger difference at lower flow rates. You should get double the velocity at the same flow rate so the cascade may give near the same performance at .5 gpm that it does at 1 gpm without the blockage. This would amount to 1C according to pHaestus' results

No, not quite. Blocking off 50% of the jets results in close to double the restriction. Given the way that mag-drive pumps work this would equate to maybe 25% less total flow rates, or about a 50% increase in jet velocity.

Mind you, the MD30-RZ was already pushing about 10LPM (~2.65GPM) through the complete system without the restriction plate, so there's not much correlation that could be drawn with Phaestus's results.

I wasn't refering to how it would perform with a specific pump, but how it would plot on a graph which shows performance at specific flow rates.