design idea
 

http://perso.wanadoo.fr/miladiou123/0.jpg
http://perso.wanadoo.fr/miladiou123/10.jpg
http://perso.wanadoo.fr/miladiou123/20.jpg

this is based on the cascade but with oblique cups, arranged in concentrique circles.
i'd love to see the result of a milled model, if some are interested let me know what are your results.

Interesting concept. Splayed holes. Can't wait to see the finished result and numbers.

it will take few days, the more "advanced" tool i have is a drill :cry:

I had an idea for a block similar to this one - in cooincidence with Incoherants recent block;

Hemispherical dome; however, smaller than his (down to 10-15mm diameter, i believe his is larger) with surrounding jets at 90 deg...

Nice work :)

A puzzling idea... Positioning the jet nozzles is going to be tricky, but this idea has potential. I'm just not sure that it's worth all the hassle though: the results may not be better.

The Cascade design has straight up holes not just for simplicity, but because the resulting "fins" on the baseplate are steady throughout the core area. It may be possible to optimize this "fin" pattern in a radial pattern here, but careful attention to the geometry of the heat source would be required. Also, hole size versus resulting fin thickness is going to be critical.

The only advantage I see is that since the fin is effectively longer, it increases the surface area of the fin, which may help heat transfer at that point, but that's just secondary: the primary cooling is the jet inpingement at the bottom of the cup, and that doesn't change here.


I'll wait for results!

If you cut a horizontal cross section of the crater you get something similar but with more holes. At first sight I had a nightmarish thought that pHaestus had got his hacksaw out again.

the idea behind this design is to try to manage the hot spots between the cups, found in the cascade,

http://perso.wanadoo.fr/miladiou123/alt.jpg

the jet would strike the wall of the cup creating a first impinging effect less efficient that the cascade one (but on a thicker copper height), on an other hand the velocity of the water would probably less affected by stricking the oblique wall, allowing a more efficient second effect.
The first impinging effect would be on the top of the potential hot spot...
the evacuation of the water from the cup would probably be easier ...
maybe a kind of washing machine effect ? (the european washing machines...) ??


this is more some kind of intuitive feeling that theory, but i would be curious to have the opinion of the theorical coolers on this point

Yeah, none of us are versed enough in fluid dynamics, we're all grasping at straws her.

My first instinct, when I see that graph is that the first "hit" is going to spread out and reduce the size of the second inpingement, but this is a cylindrical hole, so the spread might not be so bad, plus, the second inpingement is really nicely located.

The problem is that the first inpingment isn't right up against the baseplate, but the second inpingement is nicely located.

I don't think it'll beat a Cascade, but you might be able to match the performance. Anyone else wanna place bets? :p

i'd love to have a cascade to compare with... :D

I have posted a design based on a very same idea a long time ago...the only difference was in the pattern of the holes...it had not been welcome at the time, though...:)


http://forums.procooling.com/vbb/sho...8&page=2&pp=25

was it working?
i feel that it could be important that the cups converge toward a central point.
i'm finishing the leak testing, some results very soon... :)

Well, I've never actually machined it, although I have had a G-code somewhere...there was a deficit of 5-axis mills at the time...:)

Your thinking that it could be important that the cups converge toward a central point is definitely worth considering. Of course, we will never know whether there would be any difference between the two patterns.

ok, the block is not leaking...yet
some pics of the product before further lapping.

http://perso.wanadoo.fr/miladiou123/b10.jpg
http://perso.wanadoo.fr/miladiou123/b20.jpg
http://perso.wanadoo.fr/miladiou123/b30.jpg

just glued my fingers few times with super glue putting the small tubes in the middle plate :p

not very nice (finishing work to do) but every tube enter his cup (that was a challenge...)

Nice work.

Nice work!

Cathar, you must be the trailblazer for other D.I.Y. waterblocks. It seems that everytime I see a person with a new Idea for a waterblock in the forums, it is in comparison to the cascade, and hoping to be better than, or it looks just like the Cascade with a slight difference.

Can I get your autograph?

ok, the testing of the block is in progress...

i'm using an athlon xp-m 2500 at 2700 Mhz @ 1.95v (1.9 real) on an abit nf7-s v2
my pump is a viaAqua 1300. the thermal paste i use is artic silver 5 (not yet settled).
i read the water temp with a cheap aquarium thermometer (accuracy is probably ~+/- 1 degre). for the cpu temp i use the w83627hf:2 sensor with mbprobe.

to summarize nothing very sensitive or accurate....

after running burnK7 for 1 hour my delta temp is 15 degres

the only block i have to compare with is that

http://perso.wanadoo.fr/miladiou123/cas0.jpg


this block is far from being a good copy of cathar's cascade (cup repartition is not perfectly regular, rounded bottom of the cups, too much space between the cups....), and i was reading a delta temp of 17 degres with the test conditions described above (but with settled AS5).

due to the "high quality" of my test bed :rolleyes: i don't know if the difference is significative, but i like it....


another point is that if the block, i'm comparing with is not as well realized as a cascade, the achievement of this one is also far from being perfect (easily seen on the first post) and could be improved using a drill press (to consider the use of advanced tools... ;) )