Okay class. Welcome to fluid dynamics 101...
Well, except that I'm not anywhere near qualified to teach such a class, or even TA.
1. Restriction is bad. For this reason small openings and cramped passageways rob power from your fan. If these are to be used, then they'd better be places that you want to remove heat from.
2. Corners are bad. air (or any flowing fluid, for that matter) doesn't like going around corners. Each bend in the air path is equivalent to making the flow longer, thus requiring more power from the fan. The sharper the corner the more the fluid resists going around it.
3. Laminar flow is good. Laminar flow is when the air moves smoothly along a sufrface. This requires the least amount of energy to keep it moving, therefore takes the least amount of power from the fan.
4. Laminar flow is bad. Yes, this is a contradiction. Laminar flow along hot surfaces (which you'd like to be cool) is a bad thing. The nature of the laminar flow keeps the same air against the surface for a longer time, and as we're well aware, air is a good insulator. There are also situations where because of the shape of the air path, having controlled turbulence actually reduces overall drag. (This is the reason for the dimples in a golf ball -- A sphere moving through air has more drag when the air flowing over it is laminar.) If you want to go into detail, then I suggest you start reading aerodynamics textbooks. (Very dry reading, trust me, I know.) Unfortunately the most critical point in the cooling system is the one place that you can't just stick in a bunch of air spoilers. (Ie: between the lamp and the LCD.)
So taking this info into account, long, smooth curves directing the air toward the exhaust, insofar as the light engine design will allow will provide the best return from a given fan. The same would hold true of the intake.
A note about noise. The more power that a fan exerts on the air, the louder it will sound. It is not an increased flow rate that will make the difference, but the amount of power expended. 50cfm from a small 80mm fan running flat out will be much louder than 50cfm from the same fan where the air is beter directed, and has less resistance to flow. Finding that is likely to be a matter of trial and error, unless you have a degree involving a lot of fluid dynamics instruction. (Which I don't. I took a couple of courses because I was interested. They made my head hurt.
As far as fan placement, there are 2 schools of thought.
a) Puller fan. This is the more typical design that we see. A fan is directed to exhaust hot air from an enclosed space. This is the case with most PC fans (though by no means all) and with most of the projectors that I've seen, including commercial ones. The idea here is to place the fan in the area which we expect to be the hottest (Typically up high near the primary source of heat.) and remove that heat. Fresh air can be sucked from the path of least resistance, though any fresh air near the heat source is welcome.
b) Pusher fan. This fan is placed at an intake, and forces air into an enclosure, to find its way out at the exhaust. The idea here is to marginally increase the density of the air, giving it a bit more mass to remove heat with. More importantly this also directs cool air directly at a specific area, which is often a heat sink. This is more typical in the PC world as a CPU fan, though many PC cases now also make use of this idea in the front where the hard drives live. In the case of a projector, this would direct air over the LCD screen and fresnels.
The former has an advantage in simplicity of design. A single fan can draw air from many sources in many locations simply by having an opening in the case. This allows one large powerful fan to effectively cool many components.
The latter has an advantage that even if the integrity of the case is compromised (You opened the lid, didn't you? 🙂 ) the vital parts are still cooled. This allows you a bit more flexibility, even if it requires more fans, and more wiring to deliver power to the fans. It's a trade-off, but then just about any project is.
Well, except that I'm not anywhere near qualified to teach such a class, or even TA.
1. Restriction is bad. For this reason small openings and cramped passageways rob power from your fan. If these are to be used, then they'd better be places that you want to remove heat from.
2. Corners are bad. air (or any flowing fluid, for that matter) doesn't like going around corners. Each bend in the air path is equivalent to making the flow longer, thus requiring more power from the fan. The sharper the corner the more the fluid resists going around it.
3. Laminar flow is good. Laminar flow is when the air moves smoothly along a sufrface. This requires the least amount of energy to keep it moving, therefore takes the least amount of power from the fan.
4. Laminar flow is bad. Yes, this is a contradiction. Laminar flow along hot surfaces (which you'd like to be cool) is a bad thing. The nature of the laminar flow keeps the same air against the surface for a longer time, and as we're well aware, air is a good insulator. There are also situations where because of the shape of the air path, having controlled turbulence actually reduces overall drag. (This is the reason for the dimples in a golf ball -- A sphere moving through air has more drag when the air flowing over it is laminar.) If you want to go into detail, then I suggest you start reading aerodynamics textbooks. (Very dry reading, trust me, I know.) Unfortunately the most critical point in the cooling system is the one place that you can't just stick in a bunch of air spoilers. (Ie: between the lamp and the LCD.)
So taking this info into account, long, smooth curves directing the air toward the exhaust, insofar as the light engine design will allow will provide the best return from a given fan. The same would hold true of the intake.
A note about noise. The more power that a fan exerts on the air, the louder it will sound. It is not an increased flow rate that will make the difference, but the amount of power expended. 50cfm from a small 80mm fan running flat out will be much louder than 50cfm from the same fan where the air is beter directed, and has less resistance to flow. Finding that is likely to be a matter of trial and error, unless you have a degree involving a lot of fluid dynamics instruction. (Which I don't. I took a couple of courses because I was interested. They made my head hurt.
As far as fan placement, there are 2 schools of thought.
a) Puller fan. This is the more typical design that we see. A fan is directed to exhaust hot air from an enclosed space. This is the case with most PC fans (though by no means all) and with most of the projectors that I've seen, including commercial ones. The idea here is to place the fan in the area which we expect to be the hottest (Typically up high near the primary source of heat.) and remove that heat. Fresh air can be sucked from the path of least resistance, though any fresh air near the heat source is welcome.
b) Pusher fan. This fan is placed at an intake, and forces air into an enclosure, to find its way out at the exhaust. The idea here is to marginally increase the density of the air, giving it a bit more mass to remove heat with. More importantly this also directs cool air directly at a specific area, which is often a heat sink. This is more typical in the PC world as a CPU fan, though many PC cases now also make use of this idea in the front where the hard drives live. In the case of a projector, this would direct air over the LCD screen and fresnels.
The former has an advantage in simplicity of design. A single fan can draw air from many sources in many locations simply by having an opening in the case. This allows one large powerful fan to effectively cool many components.
The latter has an advantage that even if the integrity of the case is compromised (You opened the lid, didn't you? 🙂 ) the vital parts are still cooled. This allows you a bit more flexibility, even if it requires more fans, and more wiring to deliver power to the fans. It's a trade-off, but then just about any project is.
SupraGuy said:
Good post nonetheless. Let this 4th year Aerospace engineering student clarify a little.
Poorly designed restriction is bad. A properly designed contraction followed by a diffuser section can actually reduce power required by taking advantage of various quirks of fluid properties, especially where air is concerned.
With that being said, probably nobody around here can design such a section. I know I can't do it and I only know one person who can, my fluid mechanics professor who has had a PhD for twice as long as I've been alive.
Corners also cause massive flow seperation which increases drag exponentially.
Exactly. Streamlines are good. Although I'd doubt highly that a standard PC case fan in a duct to fit it can produce a high enough stream velocity in air to increase the Reynold's number to turbulent levels. The area of the duct is so small and the density so low it would take a significant air velocity to get above the laminar limit.
Again with the low air velocities we're talking I'm not even sure a vortex generator would work. If it did it probably wouldn't work well because you couldn't get the pressure differential high enough.
Good golfball example. Turbulent air "sticks" to the golf ball further downstream than laminar air. Now creating turbulence takes energy but it takes less energy than the laminar flow seperation and the resulting "suction" behind the golf ball.
You also want to keep your air velocity as low as possible everywhere except where you need flow. Fast air is noisy air.
I'm toying with using an air "reservoir" of sorts in the projector I'm designing now. I think by utilizing a short intake duct into a large open space inside the projector and pulling air from that over the LCD and other electronics I can maximize cooling and minimize noise.
I've gotta run some heat transfer numbers to make sure though.
Cool. This might almost have the makings of a sticky post. 😛
Actually, I was aware fo the restriction/diffuser thing, but even trying to think too hard about the theory makes my head hurt. For all practical intents and purposes, restriction is bad.
Headaches are also bad, but that's a different matter.
For your projector... It looks as though the fans are drawing air toward the far side, (Though I can't tell for sure with the other fan.) Personally, I would choose to draw air away from the LCD circuit board to prevent premature failure of the control circuits.
I would place an air intake so that air could enter the projector near the front. Here's an idea... you could put an air filter in the triangle behind your mirror. Cooling air would then enter on either side of the LCD, between it and the fresnels, A seal along top and bottom would force the air to travel the length of the LCD before meeting a curved wall (Or at least a 45 degree baffle) at the opposite end. Like th attatched image...
Actually, I was aware fo the restriction/diffuser thing, but even trying to think too hard about the theory makes my head hurt. For all practical intents and purposes, restriction is bad.
Headaches are also bad, but that's a different matter.
For your projector... It looks as though the fans are drawing air toward the far side, (Though I can't tell for sure with the other fan.) Personally, I would choose to draw air away from the LCD circuit board to prevent premature failure of the control circuits.
I would place an air intake so that air could enter the projector near the front. Here's an idea... you could put an air filter in the triangle behind your mirror. Cooling air would then enter on either side of the LCD, between it and the fresnels, A seal along top and bottom would force the air to travel the length of the LCD before meeting a curved wall (Or at least a 45 degree baffle) at the opposite end. Like th attatched image...
Attachments
You know there is one thing you guys forgot to mention, Pressure. Comp case fans are hopless for having any pressure to push air through a duct efficently at all, hence why you guys are using high power rotary fans when all that is needed is a small low power, high pressure centrivical. Rotary fans also have a dead spot.
Trev🙂
Trev🙂
ace3000_1 said:
The centrifugal is not a bad idea. The design of a centrifugal makes it a little more bulky but I don't see that as a major problem if you designed around it. Maybe a system where it draws intake air over the LCD and blows over the lightbox to an exhaust would work well.
As for the dead spot I don't think thats much of an issue for our purposes. IIRC the dead spot only extends two or three hub diameters axially from the fan as the streamlines re-converge. Unless you've got a big hub I don't know that it would be a problem.
Hehe mine too. I'm more of a flight mechanics guy than an aerodynamics guy myself. What really gets interesting is when you get above Mach 0.3 and start dealing with compressibility effects. Almost everything you thought you knew about air goes out the window hehe.
Its actually smaller. Theres aot of good small high pressure fans around these days 😉. If you wer to compare it to a normal rotary axil fan and its dead spot hub clearence, its acutally quate a fair bit smaller.
How about blowing the cold air over the lcd first then into the light box? we have enough pressure 😉.
Trev🙂
Ace: Centrifugal fans good. I'd acually bought 3 of them at first, but 1 was DOA. I returned all three, also because after looking more closely at them I could see that they weren't really appropriate. I know any of them could have moved more air than the 120mm fans that I'll be using instead, if only the motors were more appropriate for just running some simple voltage to. (Any voltage that I had available would have fried them. 🙁 )
To increase the efficiency of a radial fan (Like most of them are) fins can be used to direct air straight in or out of the fan. The fan moves in a circle, so the air "wants" to follow suit, which reduced the effectiveness of the blades. Forcing it to move straight increases the fan's effectiveness. It makes a more difficult design, but if you need just that little bit more flow, it can do the trick.
To increase the efficiency of a radial fan (Like most of them are) fins can be used to direct air straight in or out of the fan. The fan moves in a circle, so the air "wants" to follow suit, which reduced the effectiveness of the blades. Forcing it to move straight increases the fan's effectiveness. It makes a more difficult design, but if you need just that little bit more flow, it can do the trick.
I have a centrifugal from a photocopier that I plan on using in my projector. Its about 2-3 times the size of the 2 120mm fans I have, but it runs off 120V AC and moves nearly 10 times the air of the two 120 mm fans put together (havn't actually measured the airflow). Oh yeah, the drawback is that its fairly loud. The casing has holes in it, and the pressure is so high inside that it whistles. I'm gonna tape it up and hope for the best, or run it a lower voltage.
ace3000_1 said:
Maybe I just haven't see the right ones then because all the ones I've ever seen have been fairly bulky compared to a standard PC case fan.
That would work as well though I think it might be a little more noisy. You'll have two ducting corners causing flow seperation and turbulence verses one with mine. I was thinking of using the blower itself as one of the corners so it turns the air.
Actually if you put the blower right night to the light box and allow it to draw in over the LCD, you would essentially eliminate both as the first would be a low speed corner.
Of course maybe I'm thinking completely wrong as my cooling experience comes mostly from cooling aircraft instrumentation.
LOL prety much, slightly different deal. I get what your on about though but remember when we have a high pressure airflow we need less air as we have the pressure to push the air through, so less noise. Having air blowing on somthing directly is also cooler then sucking air away from it unless you have heatsinks or such. Also air thats blowing from the fan is more controlable then the air getting sucked into it, well, atleast the pattern.
I dont find air to be that much of a problem with noise, its normally the fan being unballenced and the sound of the motor that causes the noise in our situation, vibration ect. The fans we use dont realy have enough power to create enough turbulent airflow noise, not unless you use a fan thats running at 5000rpm with a 50cfm rating. And to be honest thats miles too much air, all we need is a mere 20cfm over our areas that we intend to cool and its very suficent. Anything more is a complete waste, and will likley give you dust hastles, filter or not.
Trev🙂
- Status
- Not open for further replies.