View Full Version : square duct or circle ?
adamtekh
11-24-01, 08:29 AM
what works best in a duct ? square/ rectangle , or a tube shape
Circles seem to work best , usually, but one of mine is a big square so that I can get max flow in a tight spot.
Circle
maximum air volume in smallest space plus all your fans are circle!! ;) :p
Not a square or a circle but a OCTAGON for a hole on the case. Not with equal side but more like a square with the corners cut(shape of the fan area that moves the air). The duct should not just tapper into a smaller opening but should have the shape of a velocity duct.
what's a velocity duct? I would have thought that a circle would be marginally the best since it has less surface area per volume and fewer place that are likely to resist airflow. But I don't have all that much to base it on.
nihili
Originally posted by nihili
what's a velocity duct?I meant a velocity stack. Lets say you want a 120mm fan blowing on a HSF that was designed for a 60mm fan. First you will need 2 funnels with the narrow sides facing each other joined by a cyclinder & that's it. The airflow will be cleaned up, no loss of CFM due to back pressure, & no center dead spots. Just like air trumpets on individual race intake manifolds on cars. If anyone has taken a close look on fan design you will notice that on both side of the inside actually taper in not a 90° angle.
EDIT - no loss of CFM due to LESS back pressure
adamtekh
11-24-01, 02:02 PM
draw pictures please :beer:
Originally posted by sonny
I meant a velocity stack. Lets say you want a 120mm fan blowing on a HSF that was designed for a 60mm fan. First you will need 2 funnels with the narrow sides facing each other joined by a cyclinder & that's it. The airflow will be cleaned up, no loss of CFM due to back pressure, & no center dead spots. Just like air trumpets on individual race intake manifolds on cars. If anyone has taken a close look on fan design you will notice that on both side of the inside actually taper in not a 90° angle.
I got ya, but that would create a lot more resistance than what it is worth! I still think a 120MM circle laying over the HS with the 120 just open over the rest of the mobo.... that way the excess can cool other components and supply an intake for the case...
my .02c
Just to remind you guys this was done in paint in less than a minute so be happy with it 'cause it aint gonna get much better with a mouse:D
jbell - The reason this was designed was that it would give better flow rates than an open design. The 120mm-60mm example was to give people an idea of it's capabilities in a realworld environment. I got one inside my rig that helps concentrate the air & keep the flow & CFM & also increase the pressure since the air follows a uniform path even with the smaller exit. The flaired ends are the secret. They help accelerate the air. If this would cause more resistance then F1 cars would not have them.
Doh! I didn't even make the association, but its another example of automotive technology in the overclocking world: Velocity Stacks have been used on carbs for 30+ years.
Originally posted by rogerdugans
Doh! I didn't even make the association, but its another example of automotive technology in the overclocking world: Velocity Stacks have been used on carbs for 30+ years. Don't forget that even the F1 cars with their dual stream injection(normally a spray pattern) still have them on top of the cyclinder heads even if the stacks are only 42mm tall instead of just an open path from its airbox.
When used on cars, what is providing the airflow, and what is the approximate cross section at the widest and narrowest spots?
nihili
Originally posted by nihili
When used on cars, what is providing the airflow, and what is the approximate cross section at the widest and narrowest spots?
nihili Is that the smart guy with a beard? The movement of the vehicle provides airflow. Depends on stacks for a two barrel or 4 barrel carb. Outlet is the size of carb throat, inlet roughly tulip shaped. Now you should of known that!:beer:
Originally posted by mw521
Is that the smart guy with a beard? The movement of the vehicle provides airflow. Depends on stacks for a two barrel or 4 barrel carb. Outlet is the size of carb throat, inlet roughly tulip shaped. Now you should of known that!:beer: That's roughly "trumpet" shaped!
Well that's sort of what I figured, but I'm no mechanic. And if we're talking about somehting designed to take in air at more than 30 mph and shove it down into a carb, I'm not sure how applicable it is to computers. The narrow neck design will increase flow resistance. In the carburater (now there's a word I don't know how to spell) case, you have massive pressure both from the movement of the vehicle and also from the exhausting of the cylinders. If I'm understanding right (and I'm not sure that I am) it's the fact that the system generates extremely high pressures that allow these things to work.
But in a computer we're talking very low pressure. Notice that the static pressure of these fans is measure in tenths of inches of water. That means most fans couldn't suck water up through a normal straw, not even close. So while narrowing the neck may not make a difference to air hurtling into the opening at race car speeds, it seems that it would for your typical case fan.
What am I missing?
nihili
Originally posted by nihili
Well that's sort of what I figured, but I'm no mechanic. And if we're talking about somehting designed to take in air at more than 30 mph and shove it down into a carb, I'm not sure how applicable it is to computers. The narrow neck design will increase flow resistance. In the carburater (now there's a word I don't know how to spell) case, you have massive pressure both from the movement of the vehicle and also from the exhausting of the cylinders. If I'm understanding right (and I'm not sure that I am) it's the fact that the system generates extremely high pressures that allow these things to work.
But in a computer we're talking very low pressure. Notice that the static pressure of these fans is measure in tenths of inches of water. That means most fans couldn't suck water up through a normal straw, not even close. So while narrowing the neck may not make a difference to air hurtling into the opening at race car speeds, it seems that it would for your typical case fan.
What am I missing? The only thing you are missing is the ability to spell carb. (notice that little shortcut?, I can't spell it either!) The principle behind these things is that air is pulled/pushed through a larger "cone" into a smaller one, increasing the velocity of the air. In autos, most air intake=mucho bueno! Hence the auto world name(and others) Velocity stacks!
nihili :beer:
Dang, messed that post up. Got it all in quotes!
I figured out the quote things.
I'm still not sure this velocity stack is a good idea in your typical computer. I'd be all for it if we had a fan that could push a couple feet of water for static pressure. But given the low pressure systems we're dealing with, it would seem unadvisable to pass the air through any narrower opening than you absolutely had to.
I don't have any hard data to back this up. Maybe it would work even in a low pressure system. But the disparity in pressure between the two types of systems makes me a bit skeptical.
nihili
Originally posted by nihili
I figured out the quote things.
I'm still not sure this velocity stack is a good idea in your typical computer. I'd be all for it if we had a fan that could push a couple feet of water for static pressure. But given the low pressure systems we're dealing with, it would seem unadvisable to pass the air through any narrower opening than you absolutely had to.
I don't have any hard data to back this up. Maybe it would work even in a low pressure system. But the disparity in pressure between the two types of systems makes me a bit skeptical.
nihili
You are good with figures, take a 120, 80, etc. fan. Then figure cfm in and cfm out of a "velocity stack" make it say, inlet, same size as fan, outlet maybe 3/4ths or so smaller. This will keep you busy and out of trouble for a day or so!:beer:
Originally posted by mw521
You are good with figures, take a 120, 80, etc. fan. Then figure cfm in and cfm out of a "velocity stack" make it say, inlet, same size as fan, outlet maybe 3/4ths or so smaller. This will keep you busy and out of trouble for a day or so!:beer:
You'll never get more cfm out of a velocity stack (once you adjust for barometric pressure) than the CFM of the highest rated fan on the stack. Now, I can figure out the Velocity of the air under the assumption that the fan isn't slowed by the back pressure.
If the outlet is 3/4 the size of the inlet, then it's cross sectional area is about 56 percent of the inlet's. Hence the velocity of the air would have to be 1.8 times faster in order to maintain the same CFM of airflow (I'm discounting changes in barometric pressure. But there are two very important points to be made.
1. Although the velocity of the air increases, the amount of air does not. Air cooling depends primarily on the amount of air, not on it's speed. Simply increasing the speed of the air won't do anything.
2. THe assumption that the fan is able to push the same amount of air through the narrowed opening as it can without any such narrowing is almost certainly wrong given the low pressure characteristics of computer fans. Hence you'll actually be pushing less air through the sink than without the velocity stack. Thus you will have poorer cooling, not better cooling.
I do see the point of using a fan that's bigger than your heatsink and necking it down to the size of the sink. An 80mm fan can push more air than a 60mm fan. So even after accounting for the loss of airflow from reducing the channel from 80 to 60mm, you'll probably still end up with a net gain in the amount of air flowing over the sink.
What I don't see is why the channel should be narrowed to less than 60mm and then flared back out to 60. Can you explain this to me?
nihili
Dang, I'm not even gonna quote all that! I said measue cfm in my reply, I meant velocity. All this is going to do is increase the velocity of the air ie: say you are catching a breeze in the open , fells nice huh? Now say you are catching that same breeze from an opening between say 2 rocks. Feels much cooler huh? This all depends on if your heatsink has the right feelings though.
Wait, don't say it! Cooling the skin is evaporative cooling, now if that cpu was sweating!
adamtekh
11-24-01, 11:50 PM
umm yeah , i dont need all that science mombo jobo , just tell me whats better to duct it down to my 6cu + , i wana get rid of that loud fan. parents geting annoyed :mad:
Originally posted by adamtekh
umm yeah , i dont need all that science mombo jobo , just tell me whats better to duct it down to my 6cu + , i wana get rid of that loud fan. parents geting annoyed :mad: This is what makes the Forums a hotbed of great ideas:DOriginally posted by mw521 The movement of the vehicle provides airflow. This is incorrect regarding the way a velocity stack works. What you are thinking of is what is called RAM AIR INDUCTION. Which only works at high pressure & at a certain vehicle speed & the effects are not at a constant with vehicle speed, you get to a point when you are travelling fast enought that there becomes back pressure & HP drops off instead of increasing with speed. An engine generates its own negative pressure to suck air in.
nihili - Do you remember the last time we talked about ducts?
1. CFM & Flow remains constant but area is decreased. So what variable increases?
2. This is the beauty of the Stack. The flared exit actually accelerates the air from the minimal "slow down" that the center creates bringing it back to normal even with the smaller area.
Originally posted by nihili
What I don't see is why the channel should be narrowed to less than 60mm and then flared back out to 60. Can you explain this to me?Those were examples and you must look beyond the numbers as airflow is as much of an Art as it is a Science. For one thing I did not say that the size of the opening towards the 60mm Fan sized HSF should also be 60mm. That would be a total failure as you might have thought. Here is another drawing to try to better explain it to you all. The box below the stack is the HSF. Flaring the exit out is what speeds up the airflow at the smaller end.
I'm still not getting it. Why not just neck down to a 60mm opening connected directly to the fan? What function does the outlet flare serve? You say that it speeds up the air. But by all accounts I can think of, the air should slow down as the tube widens. It will speed up in terms of velocity as it goes through the narrow point, but then slow down at the exit flare. I can see why you might get some benefit in terms of the homogeneity of the air pattern, and that would be of benefit in a combustion situation.
As for your question about the variables. It depends on what you mean by flow. I take it we're assuming that the out put from the fan is constant; it outputs X cfm at a given pressure. WHen you push that through a narrower opening you have to either increase the velocity or the pressure in order to maintain the flow rate in terms of mass per minute.
2. This is the beauty of the Stack. The flared exit actually accelerates the air from the minimal "slow down" that the center creates bringing it back to normal even with the smaller area.
I'm sorry, but I can't figure out what you're saying here. If you're talking about velocity, then as I said above, it seems to me that the velocity would decrease as the flare widens. The only exception to this would be if the flare opened in to a very low pressure area. In that case the sudden pressure drop might result in an increase in velocity. But in such a case it would do so even without the flare.
However, velocity is not the important factor in air cooling a computer. Instead the important variable is mass per minute. And I can't see any way for a velocity stack to improve the mass per minute variable.
Sorry to be so dense about this. I'll have a look through our previous discussions and see if I can puzzle it out.
nihili
Originally posted by adamtekh
umm yeah , i dont need all that science mombo jobo , just tell me whats better to duct it down to my 6cu + , i wana get rid of that loud fan. parents geting annoyed :mad:
The short answer is that it won't make a lot of difference. Round is probably a bit better than square. Smooth insides are probably a bit better than rough. However my duct is made out of accordion fold dryer ducting just because that was the easiest. Some day when I get extra time I'll build a better duct, but I doubt it will make more than 1 or 2 degrees difference.
nihili
Originally posted by sonny
Just to remind you guys this was done in paint in less than a minute so be happy with it 'cause it aint gonna get much better with a mouse:D
jbell - The reason this was designed was that it would give better flow rates than an open design. The 120mm-60mm example was to give people an idea of it's capabilities in a realworld environment. I got one inside my rig that helps concentrate the air & keep the flow & CFM & also increase the pressure since the air follows a uniform path even with the smaller exit. The flaired ends are the secret. They help accelerate the air. If this would cause more resistance then F1 cars would not have them.
In that case - you should try applying what I learn when I built my fist model rocket... and when I made a simple modification to my truck intake system
Add ridges to the inside.. that are not in a straight line but rather help to spin the air... by spinning it you create a cyclone effect that will ''RAM'' the air onto the HS and help even further your air compression attemps... also by spinning the air this way - which should be in the same direction as the fans blades - to reduce friction will actually straighten the air out even more... definitly worth a try!
***edit***
When spinning the air - you actually do a compression thing to it - you get a greater volume of air and it holds in place a little longer??...that should give a greater ideal effect on the HS
Originally posted by nihili
I'm still not getting it. Sorry to be so dense about this. I'll have a look through our previous discussions and see if I can puzzle it out. I think the best way is to give it a try. I think both of us understand Bernoulli's Law. Now What I mean by airflow is pressure+velocity. Since velocity is affected by the size of the duct decreasing it will increase velocity. Then the problem of decreased pressure comes into play as the size decreases so the exit then flares out to maintain it but not to the extent that the flare is as big as the intake side. Another thing that the flare contributes to is cleaning up of the air path so there are no "eddies" at the exit to pull the air away from the center. If you attach the duct directly on top of the HSF you can experience some back pressure that will hinder cooling, less air thru the HS.
Our last discussion about ducts was a couple of months ago concerning the 120mm 127CFM fan I ordered to duct cold air to my heatsink. Pressure was a bigger concern than total CFM. We were not able to continue that topic.
Originally posted by jbell
In that case - you should try applying what I learn when I built my fist model rocket... and when I made a simple modification to my truck intake system
Add ridges to the inside.. that are not in a straight line but rather help to spin the air... by spinning it you create a cyclone effect that will ''RAM'' the air onto the HS and help even further your air compression attemps... also by spinning the air this way - which should be in the same direction as the fans blades - to reduce friction will actually straighten the air out even more... definitly worth a try!
***edit***
When spinning the air - you actually do a compression thing to it - you get a greater volume of air and it holds in place a little longer??...that should give a greater ideal effect on the HS Laminar Flow is the first goal concerning ducts. If there is a strong enough fan to compress the air with minimal noise & size(120mm MAX) this would be something I would give a good try.
Originally posted by sonny
Laminar Flow is the first goal concerning ducts. If there is a strong enough fan to compress the air with minimal noise & size(120mm MAX) this would be something I would give a good try.
A side case mounted 120mm would work well.... and insulate the tube from the fan to the HS to keep the case temps off the pipe - so you have even slightly cooler air running to the HS
Originally posted by jbell
A side case mounted 120mm would work well.... and insulate the tube from the fan to the HS to keep the case temps off the pipe - so you have even slightly cooler air running to the HS I guess your talking about your "compressor tube". You probably do not have to go thru the hassle of insulating it 'cause the time air spend idle inside the tube, most probably made of plastic which does not conduct heat well, is none existent. Give it a try & tell us how it pans out.:beer:
adamtekh
11-25-01, 02:20 PM
i made a square one :eh?:
Great! Let us know how it works out.
nihili
Originally posted by sonny
I guess your talking about your "compressor tube". You probably do not have to go thru the hassle of insulating it 'cause the time air spend idle inside the tube, most probably made of plastic which does not conduct heat well, is none existent. Give it a try & tell us how it pans out.:beer:
Where do I get a case mod kit to mount 120mm fans? I want one for directly above my CPU and one down by my PCI slots and one above my PSU for exhuast...
I got mine from Bart (http://www.1coolpc.com/). Not a kit but you can get some screws & filters from him.
Originally posted by sonny
I got mine from Bart (http://www.1coolpc.com/). Not a kit but you can get some screws & filters from him.
Thanks!
Any recomendations on 120mm??? I need a high CFM for exhuast/HS Duct
and a quiet not neccassrily high CFM intake for PCI slots....
Get the high output one he sells. There is really just one choice there. 127-131CFM rating for his 120mm. It's a YS-Tech.
Originally posted by sonny
Get the high output one he sells. There is really just one choice there. 127-131CFM rating for his 120mm. It's a YS-Tech.
damn thats huge!! LOL
I looked at your link and the links on this site - where do I get mod kits to install fans and cutting holes and stuff!
William
11-25-01, 04:21 PM
as far as I know and have read adapting an 80mm fan to 60 won't yield results as good as a delta until you slap on (guess what!) a delta 80mm fan. It goes back to that backpressure thing. I am with Nihili on that cone thing, that would only decrease the velocity at the end.
Back to the original question, i don't think it will really matter.
Oh, and to see an interesting airflow design, go to the cooling section on the front page and check out Hoots Cohesive Aircooler or Hoot Chute.
adamtekh
11-25-01, 05:30 PM
well i think this will cool better , and be LOTS quieter
Hey guys,
I tried both types of ducts. 1 square cardboard duct and 1 Med. size Tim Hortons cup. I cut both ducts to be the same in height so that the distance in air flow would be the same. I used an 80mm fan as the primary air intake to the hsf 60mm fan. What I found was that Tim Hortons cup duct was able to keep temps at a more stable level but not necessarily at lower temps. Why is that?
My guess would be less turbulence.
nihili
Originally posted by nihili
Great! Let us know how it works out.
nihili After all this debating, you get out the cardboard and some tape! Well, if it works, it works!
adamtekh
11-26-01, 04:42 PM
:cool: :cool: :cool: :cool: :cool:
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.