My micro-channel concept block - Page 5

gone_fishin · 10-11-2002, 09:13 PM

Maybe we should all chip in and get Bill the patch

bigben2k · 10-11-2002, 09:37 PM

Quote:

Originally posted by Cathar
It's okay. I'm not really fussed at all.

I'm surprised that people could read my thread at OCAU and all they get out of it is a friendly caricature of yourself, and miss all the discussion on physical design constraints, machining constraints, thermal transfer, nozzle experimentation, pump flow experimentation, Les's excellent analysis of `h' vs various designs vs flow rate, etc, etc. This list goes on, and yet the lasting impression is a picture of a guy who smokes?

He he...

I just checked in, and find that I'm being referred to! It's all good! The first time I read that OCAU thread, I was amazed to see how much junk was posted that had nothing to do with designing this block of yours. I thought: "surely there must be another thread, where it all started".

I was looking for something new, something that we hadn't discussed here already, but I didn't find it. Les' graphs are truly an excellent piece of work, and I am very grateful that he took the time to produce them.

But like I said, I'll re-read it (I've got my printed copy next to me), see if I can dig up some more info on this "gradient" that Bills refers to, visit UW and maybe try out a few numbers... and if I come up with something new (new here), I'll post it.

Right now, I have to concentrate on exploring this gradient. If I don't get this, then I'm down with the "good enough" crowd, with test-and-go procedures. As BillA pointed out, there's lots of data out there.

As for the caricature, well, it was a free shot at Bill, but I figure that if he can dish it out, he can take it too. It's pretty harmless anyways, and if it lightens up the mood, I'll do it. BillA doesn't have a style: he's blunt and direct. How anyone finds this shocking, or unpredictable, is beyond me.

Yes, that was another shot, but as usual, BillA is correct: fittings can be equated with tubing length. What's not so easy is when the restriction changes size, but it's just a little extra calculation, here and there. Crane 410.

bigben2k · 10-11-2002, 09:37 PM

Quote:

Originally posted by gone_fishin
Maybe we should all chip in and get Bill the patch

I'm in for $20!

BillA · 10-11-2002, 09:59 PM

patch ?
as in for PMS ??
or to cover the shiner ?
shit, Peter will spring for a hunnerd (hes makin' the big bucks as a mfgr)
pretty soon Cathar too

better would be to find me a NI DAQ card (PCI please)

gone_fishin · 10-12-2002, 02:03 AM

Quote:

Originally posted by unregistered
patch ?
as in for PMS ??
or to cover the shiner ?
shit, Peter will spring for a hunnerd (hes makin' the big bucks as a mfgr)
pretty soon Cathar too

better would be to find me a NI DAQ card (PCI please)

A PCI wishlist from BillA, have you been naughty or nice, it's just around the corner ya know

Back to topic of Cathar's thread. Impingement. Air induced, ever contemplated?

GeminiCool · 10-12-2002, 02:10 PM

haha.. Big bucks, that's funny.

I'll make you all a deal, if we ever get to the point where we manufacture, build, and actually sell a few hundred units a month, I'd be happy to begin investing some time, energy, and money into a certified type testing system. I would also be willing to begin looking at low quantity special orders just as a way to say thank you to the grass roots.

For the record. This time Bill is definitely wrong. I had never seen the Fixitt spiral block before introducing my block. I took nothing from him. I learned of Fixitt after my block was finally shown to the public.

-Cathar- wgt your block. Have you looked you considered the consequence of using your blocks at different flow rates? ie, if you have someone who hooks your block in series with another and the flow is restricted, how well does it "deal" with the reduction in flow? Is your overall design goal to produce a block which works very well if sufficient pump flow is generated? Just curious....

Cathar · 10-12-2002, 05:58 PM

Quote:

Originally posted by GeminiCool
-Cathar- wgt your block. Have you looked you considered the consequence of using your blocks at different flow rates? ie, if you have someone who hooks your block in series with another and the flow is restricted, how well does it "deal" with the reduction in flow? Is your overall design goal to produce a block which works very well if sufficient pump flow is generated? Just curious....

I've looked at the Eheim 1048, 1250 and a Pondmaster 4200, at 5.0/6.8/10.0lpm respectively.

Okay, to really cut the flow rates I stuck a small G-clamp onto the hose and tightened it right up while using the Eheim 1250.

The flow rate was cut down to a mere 1.5lpm (0.38gpm). I can't imagine people running with flow rates much lower than this.

I used a Morgan-core Duron (for its extremely high wattage density being notoriously hard to cool) running at 1200MHz/2.15v putting out 81.4W according to the ComputerNerd thermal calculator.

The results (applicable only to my setup):

Eheim 1250 @ 6.8lpm => 21.0/23.4/43.0 (air/water/cpu die)
Eheim 1250 @ 1.5lpm => 21.1/23.3/45.0 (air/water/cpu die)

I think the results speak for themselves. The block performs extremely well even given low flow rates. These results are still better than for a Maze 3 with the Eheim 1250 at 8.0lpm, let alone running a Maze 3 at 1.5lpm.

[Edit: fixed typo - 42.0C for 6.8lpm was meant to be 43.0]

Cathar · 10-13-2002, 08:57 AM

Remounted the block due to a faulty initial mounting and re-ran tests with a very low flow rate.

The flow rate was cut down to 1.0lpm (0.25gpm). I can't imagine people running with flow rates any lower than this.

I used a Morgan-core Duron (for its extremely high wattage density being notoriously hard to cool) running at 1200MHz/2.15v putting out 81.4W according to the ComputerNerd thermal calculator running BurnK7 to generate full thermal load.

The results (applicable only to my setup):

Eheim 1250 @ 6.8lpm => 20.9/23.1/41.0 (air/water/cpu die)
Eheim 1250 @ 1.0lpm => 20.9/22.9/43.0 (air/water/cpu die)

This is somewhat better than what Les's graphs were predicting.

dream caster · 10-13-2002, 10:08 AM

Quote:

Originally posted by Cathar
It's okay. I'm not really fussed at all.

I'm surprised that people could read my thread at OCAU and all they get out of it is a friendly caricature of yourself, and miss all the discussion on physical design constraints, machining constraints, thermal transfer, nozzle experimentation, pump flow experimentation, Les's excellent analysis of `h' vs various designs vs flow rate, etc, etc. This list goes on, and yet the lasting impression is a picture of a guy who smokes?

Myself as having posted in that thread about BillA's caricature i remember more about the block than the caricature but that vision of Bill, as he spoke about himself later, was very clever.

I think forums are about people and not only computer tecnichs; look at two threads in [H]ard forum, [H]ard|Case Gallery and *New* Post your workstation pictures!. The first one is a sticky but it is far less successful than *New* Post your workstation pictures! .
Why?
*New* Post your workstation pictures! is a meeting place, you write comments, question; you can be involved and show yourself there; the other one is a rather cold collection of pictures.

And BillA, as geminicool says

; you are a character, a very noteworthy guy . (Very well respected also).

GeminiCool · 10-13-2002, 10:17 AM

Humm... another interesting test... decreased flow rate can still cool well.

Wow, so thin base and low flow rates CAN cool pretty well... boy now all we need to do is prove that the flow reduction experienced through one reduction is Adequate to APPROXIMATE the system results and the circle will be complete.

Nice Job with the block Cathar, if you do go into production I wish you luck. Perhaps if you need anyone to manufacture here in the US, we could talk.

BillA · 10-13-2002, 10:19 AM

dream caster
an addl 'insight' into your human angle comments (with which I would agree)

Cathar had to e-mail me to explain who the image is actually of
- I don't watch TV (other than the BBC news) and did not have a clue

but the shoe fit, good shot, and appreciated
(all the more so because I did not understand it, which I do to others not infrequently)

DarkEdge · 10-13-2002, 10:33 AM

I really want this micro channel block now.

dream caster · 10-13-2002, 12:53 PM

Quote:

Originally posted by unregistered
dream caster
an addl 'insight' into your human angle comments (with which I would agree)

Cathar had to e-mail me to explain who the image is actually of
- I don't watch TV (other than the BBC news) and did not have a clue

but the shoe fit, good shot, and appreciated
(all the more so because I did not understand it, which I do to others not infrequently)

I do not watch much TV either and don't know whose picture is that but the meaning was clear.

morphling1 · 10-13-2002, 01:22 PM

Hey cathar, I'm a little late to discussion, but I follow your progress. I am thinking of ways to make thiner fins and thiner chanells, maybe even base,and now I finaly got some idea how to make even thiner chanells on my mill, but I need to test this first before making a fool from myself

Anyway the thing with thiner fins and chanells is the the top and how it's designed , I see in your block that the top covers the fins and is basicaly touching them, no problem here if ID of the inlet measures the total width of all chanells and fins, all the chanell are in contact with water flow. But when you test with nozzle like inlet and if you had the same top with no or just very little distance between nozzle discharge those outer chanells won't be realy exposed to water jet. So some rounded cut out on the top, just above the core with nozzle in the centre covering the fins would be better.

Les · 10-13-2002, 01:45 PM

Quote:

Originally posted by Cathar
.................................................. .......................................

This is somewhat better than what Les's graphs were predicting.

As said on OCAu forum ( http://forums.overclockers.com.au/sh...5&pagenumber=4 ) and stated in first post in this thread(when progressing to consider Flow Rate after Conv Coeffs) the model is for "side inlet".
Sure, would expect the model break down when the calculations change from being for Turbulent to Laminar flow. Would not expect an abrupt change in practice even in the "side inlet" case. Probably "flow regime"information should be included in the graphs.

The models are presented to be "shot at" and possibly provoke some contribution/ridicule from Heat Transfer professionals.So far no-one seems to be willing to risk this venture. I am waiting and hoping for correction and enlightenment.

BillA · 10-13-2002, 01:56 PM

I would try Les,
but first I have to find someone to pay me so I can qualify as a "professional"
'till then I'm just another dilettante

hmmm . . .
what if I hooked up Peter with maskedgeek ?
would that qualify as "Heat Transfer" ?

morphling1 · 10-13-2002, 01:58 PM

Les, are any of the programs you use for simulating water block available for public, I would like to play with them a little before I start bothering you

Les · 10-13-2002, 02:00 PM

Quote:

Originally posted by unregistered
I
hmmm . . .
what if I hooked up Peter with maskedgeek ?
would that qualify as "Heat Transfer" ?

Raised a titter, but no comment.

BillA · 10-13-2002, 02:03 PM

aww SHIT !
you mean I got it bassackwards ?

Les · 10-13-2002, 02:23 PM

Quote:

Originally posted by morphling1
Les, are any of the programs you use for simulating water block available for public, I would like to play with them a little before I start bothering you

I use the sub-menus in the "Kryo TESC" portion of the Kryotherm program for the Convection calculations http://www.kryotherm.spb.ru/soft.htm
and Waterloo for the heat spreading http://www.mhtl.uwaterloo.ca/tools.html#

morphling1 · 10-13-2002, 02:33 PM

Thanks, Les for those links.
I will also be trying some nozzle designs calculation from this calculator http://www.coolingzone.com/Content/D...as/fcalc10.htm
h is the same as alpha - coificient for heat transfer in W/m2K right?

Les · 10-13-2002, 03:07 PM

Quote:

Originally posted by morphling1

1)I will also be trying some nozzle designs calculation from this calculator http://www.coolingzone.com/Content/D...as/fcalc10.htm
2)h is the same as alpha - coificient for heat transfer in W/m2K right?

1) Have had only limited success with the nozzle calculator.It only seems to be relevant with nozzle diameters below ~3mm for impingement regions near the die region(~15mm diam?).
Looking more towards "the Billa link" http://www.electronics-cooling.com/h...01_may_a2.html for the answers .
2) That is my understanding.

Cathar · 10-13-2002, 07:28 PM

Quote:

Originally posted by GeminiCool
now all we need to do is prove that the flow reduction experienced through one reduction is Adequate to APPROXIMATE the system results and the circle will be complete.

The flow rate through a system is a function of the sum of its back-pressure resistances versus the pressure of the pump. It doesn't matter at all where or the number of the resistances in the system, the resultant flow rate is the same. It is the flow rate that's important here.

I'm sure BillA and Les will back me up on this one.

BillA · 10-13-2002, 07:41 PM

the beauty of an 'ignore list' is that one only sees what others deem worthy of quoting

Cathar
he is grinding an axe, knowledge and reason have no bearing on the script

the circle referred to is a dead short between . . . .

gone_fishin · 10-13-2002, 10:17 PM

Quote:

Originally posted by Les
1) Have had only limited success with the nozzle calculator.It only seems to be relevant with nozzle diameters below ~3mm for impingement regions near the die region(~15mm diam?).
Looking more towards "the Billa link" http://www.electronics-cooling.com/h...01_may_a2.html for the answers .

Further work on the same problem can be found here.click

10-13-2002, 01:22 PM	#114
morphling1 Cooling Savant Join Date: Sep 2001 Location: Slovenia Posts: 468	Hey cathar, I'm a little late to discussion, but I follow your progress. I am thinking of ways to make thiner fins and thiner chanells, maybe even base,and now I finaly got some idea how to make even thiner chanells on my mill, but I need to test this first before making a fool from myself Anyway the thing with thiner fins and chanells is the the top and how it's designed , I see in your block that the top covers the fins and is basicaly touching them, no problem here if ID of the inlet measures the total width of all chanells and fins, all the chanell are in contact with water flow. But when you test with nozzle like inlet and if you had the same top with no or just very little distance between nozzle discharge those outer chanells won't be realy exposed to water jet. So some rounded cut out on the top, just above the core with nozzle in the centre covering the fins would be better. __________________ [My ftp, with lots of pics, hope to be home page someday

10-13-2002, 01:58 PM	#117
morphling1 Cooling Savant Join Date: Sep 2001 Location: Slovenia Posts: 468	Les, are any of the programs you use for simulating water block available for public, I would like to play with them a little before I start bothering you __________________ [My ftp, with lots of pics, hope to be home page someday

10-13-2002, 02:33 PM	#121
morphling1 Cooling Savant Join Date: Sep 2001 Location: Slovenia Posts: 468	Thanks, Les for those links. I will also be trying some nozzle designs calculation from this calculator http://www.coolingzone.com/Content/D...as/fcalc10.htm h is the same as alpha - coificient for heat transfer in W/m2K right? __________________ [My ftp, with lots of pics, hope to be home page someday

10-11-2002, 09:13 PM	#101
gone_fishin Cooling Savant Join Date: May 2002 Location: Da UP Posts: 517	Maybe we should all chip in and get Bill the patch

10-11-2002, 09:59 PM	#104
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	patch ? as in for PMS ?? or to cover the shiner ? shit, Peter will spring for a hunnerd (hes makin' the big bucks as a mfgr) pretty soon Cathar too better would be to find me a NI DAQ card (PCI please)

10-12-2002, 02:10 PM	#106
GeminiCool Cooling Neophyte Join Date: Mar 2002 Posts: 20	haha.. Big bucks, that's funny. I'll make you all a deal, if we ever get to the point where we manufacture, build, and actually sell a few hundred units a month, I'd be happy to begin investing some time, energy, and money into a certified type testing system. I would also be willing to begin looking at low quantity special orders just as a way to say thank you to the grass roots. For the record. This time Bill is definitely wrong. I had never seen the Fixitt spiral block before introducing my block. I took nothing from him. I learned of Fixitt after my block was finally shown to the public. -Cathar- wgt your block. Have you looked you considered the consequence of using your blocks at different flow rates? ie, if you have someone who hooks your block in series with another and the flow is restricted, how well does it "deal" with the reduction in flow? Is your overall design goal to produce a block which works very well if sufficient pump flow is generated? Just curious....

10-13-2002, 08:57 AM	#108
Cathar Thermophile Join Date: Sep 2002 Location: Melbourne, Australia Posts: 2,538	Remounted the block due to a faulty initial mounting and re-ran tests with a very low flow rate. The flow rate was cut down to 1.0lpm (0.25gpm). I can't imagine people running with flow rates any lower than this. I used a Morgan-core Duron (for its extremely high wattage density being notoriously hard to cool) running at 1200MHz/2.15v putting out 81.4W according to the ComputerNerd thermal calculator running BurnK7 to generate full thermal load. The results (applicable only to my setup): Eheim 1250 @ 6.8lpm => 20.9/23.1/41.0 (air/water/cpu die) Eheim 1250 @ 1.0lpm => 20.9/22.9/43.0 (air/water/cpu die) This is somewhat better than what Les's graphs were predicting.

10-13-2002, 10:17 AM	#110
GeminiCool Cooling Neophyte Join Date: Mar 2002 Posts: 20	Humm... another interesting test... decreased flow rate can still cool well. Wow, so thin base and low flow rates CAN cool pretty well... boy now all we need to do is prove that the flow reduction experienced through one reduction is Adequate to APPROXIMATE the system results and the circle will be complete. Nice Job with the block Cathar, if you do go into production I wish you luck. Perhaps if you need anyone to manufacture here in the US, we could talk.

10-13-2002, 10:19 AM	#111
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	dream caster an addl 'insight' into your human angle comments (with which I would agree) Cathar had to e-mail me to explain who the image is actually of - I don't watch TV (other than the BBC news) and did not have a clue but the shoe fit, good shot, and appreciated (all the more so because I did not understand it, which I do to others not infrequently)

10-13-2002, 10:33 AM	#112
DarkEdge Cooling Savant Join Date: Mar 2001 Location: Sugar Land, Tx Posts: 176	I really want this micro channel block now.

10-13-2002, 01:56 PM	#116
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	I would try Les, but first I have to find someone to pay me so I can qualify as a "professional" 'till then I'm just another dilettante hmmm . . . what if I hooked up Peter with maskedgeek ? would that qualify as "Heat Transfer" ?

10-13-2002, 02:03 PM	#119
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	aww SHIT ! you mean I got it bassackwards ?

10-13-2002, 07:41 PM	#124
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	the beauty of an 'ignore list' is that one only sees what others deem worthy of quoting Cathar he is grinding an axe, knowledge and reason have no bearing on the script the circle referred to is a dead short between . . . .

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)