FWIW - Copper Cascade BP thickness is 0.70mm.

That should throw a rather large cat amongst the calculations, and is certainly something for the theorists to chew on. ;)

Oh, and something else to consider. Muse on the possibility that degree of heat-spread can be very heavily affected by the cooling that's occuring on the wetted side of the copper.

Bottom of the jet holes I presume? What about between the jet holes?

...giving an "effective baseplate thickness" of: for a 5mm thick block ~2.85mm, for a 6mm thick block ~3.35mm. Not such a big cat, assuming a hole to space area ratio of 1:1. ;)

The Rotor block is a bit odd though.

Incoherent

"effective baseplate thickness" - okay - that sort of over-simplification really is going WAY too far.

What I'm saying is that whatever track you're attempting to run with on some of these simplistic assumptions you're making, you're WAY off base. Bill already said that, but now it's my turn to reinforce that sentiment. I don't mind so much the estimation of "average h", mostly because many convection research papers deal with that sort of simplistic terminology in their concluding statements (a bit like the generic 3DMark2001 score of the cooling world as a generic assessment of cooling performance), but attempting to reduce block designs down to an "effective bp thickness" is taking it too far.

Well it seems to fit.

K.I.S.S. :)

But point taken. Planets have been discovered based on assumptions which were later proved totally wrong.
I've dug a deep enough hole, I think I'll stop here.

Cheers

Incoherent

I will point the cat in your direction though to wrestle with.

One of the older prototype blocks I built had, by your "effective bp thickness" terminology, an effective thickness of 1.35mm, 55% of the "effective surface area" of the Cascade, and yet performed mid-way between a Cascade and a White Water.

The copper Cascade itself, by the same "effective bp thickness" definition, is 1.90mm thick.

Am interested to see how you account for those sorts of figures in your model.

The 1.9mm thickness of the copper cascade moves it a bit. Slightly worse at low h/flowrates, slightly better at the higher end. This would cause me to re evaluate the values for all the blocks, all are estimates, this is hard data.
Your older prototype. What flowrate? It has a steeper curve. The FOM is a composite value not just describing surface area. It includes impingement and turbulent effects but the relationship is not linear - 55% of the surface area does not lead to an FOM of 55%, for sure. I'm guessing it's cumulative. But, using this value anyway, to be consistant, gives a WW matching block only at very high flowrates. Significantly worse at the low h end.
Using another (higher) FOM for the older block It would be better than the WW at higher flow rates, almost matching, maybe exceeding the Cascade at extremely high rates but at very low flow it is worse.

So sayeth the model. As Bill said "...is dependent on the weighting assigned, ONLY the user can actually define" This probably looks like changing the facts to fit the data. That is missing the point. Or actually that is the point, finding the fit.

Chart:
http://w1.863.telia.com/~u86303493/C.../bollocks2.jpg

Digging my deep hole deeper.

Incoherent

its the pin-hole effect

OK. Got to ask. Pin-hole effect??

<braces for the impending impact>

no, no cheapo gimmes here
actually kinda interesting devices, but of course I've used the phrase in a non-standard manner
(but intelligible none the less)

dig a bit (less effort than the GIGO calcs anyway)

primitive way to look at the sun indirectly? Not a bad analogy to this thread, Bill.

or viewing the firmament from the bottom of a 40 ft. well

Probably optics? Well "interesting devices" suggests camera or glasses.
Abstract. They are very interesting. Otherwise confused. Maybe I am diffraction limited?
I am definitely not looking at sunspots. :confused:
:)

Edit: Cross post

My 2 pennies here.
I would tend to agree with pros here on high innacuracy of discussed theory.
I'd risk a statement that at our level of resources it is cheaper and quicker (as well as much more 'accessible') to make prototypes and by testing them derive simple relationships.
Unless somebody has an access to molecular level 3D flow modelling of liquids there is no point in aking this any further:)
Only handful of peeps are able to even begin to understand all the physics behind it not to mention implement it.

In base plate dept. - anyone ever thought about trapezoid-like base plates? (simplificaion of very hard to machine spherical perfect model)

What I am currently working on is cross breed between foamed carbon thermal devices and copper technology = volume wide coolant to block material heat transfer combined with high velocity of coolant molecules.

I am not quite conviced about the outcome due to fabrication limitations and severe tme constraints but I am of high expectations here :)

P.S. Very interesting thread imho!

I'm surprised no one has thought about inversing the furniture.

Who said no one had?

http://www.ocforums.com/showthread.p...14#post2744514 ;)

First step to something better...

'There will be no pics of #2' he he he, bl00dy high time for that d00de! Get some patents on it first!

Hey, guess you got that covered. Cool to see some pics but I don't blame you for not doing that. I first thought of the idea after looking at how architects designed buildings or bridges to keep a structural entegrity with minimal material. Easier said than done within the limits of a water block is concerned, but it can be done.

Patents are only good if you are willing to spend major money on lawyer's fees to defend your rights.

You will rarely get rich with a patent, but the lawyers get their pockets full.


Best regards

frenchnew