White Water type block improvement

Toysrme · May 24, 2003

Welp, we all know it's the best thing going. I thought, what could make it better?

Picturing the block, a half second later it dawned on me the TWO 90 curves to flat surfaces might be hurting it some. If every 90 curve kills the flow by so much, the let's get rid of one!

Why not go back to "Old school" and mount the exit barbs on the sides of the block. Try and fit THAT in your case LoL!

Toysrme · May 24, 2003

Yeah right! How much more perfection dose the block need anyways...

Cathar · May 24, 2003

Because the block has two outlets, and the flow has been split in half prior to reaching those outlets, and pressure drop is proportional to the flow rate squared, the per-90 degree pressure drop is 1/4 of what you would expect, for a total net pressure drop of one half of what you would expect from a single 90-degree outlet. In the grand scheme of things of how the block works, it's not a terribly significant issue. Granted it is definitely a factor, just not a large one.

Toysrme · May 24, 2003

Cool; and the engineer skillfully shows his superiority and fore-thought!

Just curious now. What about a normal larger than core sized chamber at the inlet, tapering to two exits on both ends? That would be some handy Dremel work!

"Granted it is definitely a factor, just not a large one."
Yeah I read that, but I'm still thinking of how the best could become better

What do you think could be done differently? Sory Cather, I can't remember everything from those 5000 page long posts. (well deserved as we all know)

Cathar · May 24, 2003

Toysrme said:
Just curious now. What about a normal larger than core sized chamber at the inlet, tapering to two exits on both ends? That would be some handy Dremel work!

Yeah I read that, but I'm still thinking of how the best could become better

What do you think could be done differently?

The inlet area above the core, if anything, should be smaller and not larger, but this depends on the CPU core you're using. The lateral heat spread from the CPU core barely makes it more than 1mm sideways (across the channels). You see, the point of the design was to minimise copper conduction because the further the heat has to travel through copper, the warmer the CPU will get, and balance that off by cooling the channels in a more efficient fashion than the ability of copper to conduct heat. In doing so, we have effectively minimised the cost for heat to move through the copper to be cooled.

So what can be done differently? Not an awful lot. There may be some better balances of certain elements of the design, but there's not a whole lot left to be gained.

bafbrian · May 24, 2003

I would love to see someone mount a LRWW with barbs on the side.

BTW, I thought you couldn't improve on perfection ?

(J/K)

Giblet Plus! · May 25, 2003

It would be neat with a swiftech-style aluminum top, and the return barbs going out at 45 degree angles.

I doubt that would do much for performance. Maybe .1C better?

Toysrme · May 25, 2003

I didn't get my point accress well.

I mean take the standard middle of it, but from the rectangular center; gradually round the shape to channel the water into the 1/2 barbs. More like a funnel instead of just barbs on the end.

The reason why I don't think this wouldn't hurt (like you said problably wouldn't help at all) is because the heat is concentrated over the core.

LiquidSquirt · May 26, 2003

And with the "fummel " shaped jet nozzle, Cathar could hav added another 20 or 30 bucks for machining/ tooling costs.What you are talking about is not easy #1...
#2 it would negate the effect he was looking for in impingemnt on the WW, as the funnel shape would cause the jet of water to spread outward as it exits....like a waterhose sprayer does when you barely squeeze it. With straight sides on the jet nozzle, you force most of the water to travel straight down where its needed to reduce the boundary layer at the most critical points first.

Toysrme · May 26, 2003

No, it would spread inwards as it exits to get back in 1/2 tubing.

I also said it'd be extreamly hard way up top from the get go. "That would be some handy Dremel work!" Comes to mind...

Also it being easy to make has less to do wit this as you equate. The subject is how to make it better.

Cathar · May 26, 2003

I'm not entirely sure it's possible to do as you suggest Toysrme without significantly changing the height of the block. The block is 12mm tall (without the barbs) and there's not really a whole lot of room to be fiddling with vertical-ranging adjustments. It could be done, but you'd have to make the block taller, which means more copper, and therefore heavier.

Toysrme · May 26, 2003

Ah, That's the kind of input needed

Too bad noone has access to the equipment needed to make molds and pour copper. Wouldn't that be just peachy? Then again I think I could name only a *very* select few that I would let near something over 2000 F. LoL!

Toysrme · May 26, 2003

If noone's ever been in a plant to see how any type metal is made, I highly suggest it! It's very interesting.

My grandfather on dad's side worked his way through USX to become a VP of rail transportations. Dad did two summers at USX during college. Hell... Mom scheduels every piece of pipe made by US Pipe.

I've taken tours of both. It's really neat, but not something I would advise to do on a regular basis.

Cathar · May 27, 2003

Toysrme said:
Too bad noone has access to the equipment needed to make molds and pour copper. Wouldn't that be just peachy? Then again I think I could name only a *very* select few that I would let near something over 2000 F. LoL!

I asked inquired with one of Melbournes premier industry investment casting and moulding company, and one of the few who deal with copper. When I showed them the block they saw the fins and the height of them and commented that there would be a great deal of irregularities as a result of the block. Copper is one of the worst metals to cast with as it traps air-bubbles very readily. Vacuum based centrifugal casting would be required to do it, but that is substantially more expensive than gravity pour style casting.

White Water type block improvement

Toysrme

Member

Toysrme

Member

Cathar

Senior Member

Toysrme

Member

Cathar

Senior Member

bafbrian

Member

Giblet Plus!

Member

Toysrme

Member

LiquidSquirt

Registered

Toysrme

Member

Cathar

Senior Member

Toysrme

Member

Toysrme

Member

Cathar

Senior Member

Similar threads