You should do a comparison between the air flow fans and the static pressure fans. Everyone says that the static pressure fans are better but what I would like to know is how much better they are.
for high fin density and thick radiators the static pressure fans are better. As the name implies, they are build for the high pressure that comes from an obstructed airpath - aka having a rad in the way. it of course heavily depends on your rad-config. But with 30mm and above rads you can't go wrong with pressure fans.
VenoXj1 That is hard to say - cause there are so many things going into this. If you only compare a single 120 rad in open air, than you can easily give a number. But it will greatly change if you have it inside a case. What radiator? what rad-thickness and find-density? How many fans you have/need? The more fans the lower rpm you likely need, and that again has a big impact. Does you case have positive or negative pressure? Just like that you already have to test 200 or more scenarios. but at AnandTech there is an easy to understand pressure vs Flow fan comparison. www.anandtech.com/show/9737/the-corsair-sp140-led-af140-led-sp120-af120-fan-review
I believe the effect causing the piece of paper on the radiator fan to be sucked down is Bernoulli's Principle. It states that as the speed of a fluid (liquid or gas) increases the pressure decreases. So the air under the piece of paper is moving faster because of the fan while the air on top is stagnant. Thus the air on bottom is a low pressure system relative to the air up top. Then the paper is pushed down because of the lift generated by the high pressure pushing down on it trying to equalize with the air on bottom. This is how airplanes fly except it is reversed so the lift generated is pushing up and not down. But it is exactly how high performance cars work, it's known as down force.
+Mike Sico I think the first fan is just putting too much air out for it to happen. I think the radiator lowers the pressure just enough to cause that phenomenon to happen.
I'm not entirely convinced that there will be a lower pressure in this case, as what you're talking about is static air and a moving wing through it. The air may actually have a higher pressure in this case as it is being forced forward from behind. I actually think it's the rotation of the caused by the higher RPM of the fan. The reason it does not occur on the fan without the radiator is that the air is moving sideways quicker, and therefore does not have as tight of a "corkscrew" to suck the end of the paper back down. To put this into the perspective of aircraft again, this is a real thing experienced in single engine propeller aircraft at low speed. When the propeller spins up when the aircraft is at a low speed (typically on the ground), a significant yaw is applied to the aircraft from the propeller wash actually circling around the fuselage of the aircraft and hitting one side of the vertical stabiliser. In the case of the radiator fans here, there is a tight corkscrew of air that is rotating quickly, but not moving as fast horizontally as the fan without the radiator. I could be wrong though.
While I will not be able to built a PC in a while, it is always nice to keep increasing my knowledge base for when I am able to. Keep up the great work!
I just rebuilt my 455 Buick, bored 30 over with Comp cam. The temp was over 200 degrees which makes me extremely uncomfortable. I decided to invest in push fans. Watching your video gave me added assurance I was moving in the right direction. Thank you very much
9:00 - The Bernoulli Principle at work. Increasing fluid velocity decreases the pressure. The fan exhaust is therefore at a lower pressure than the rest of the room, pulling the paper toward it. There's some equilibrium between where the paper is getting pushed away by the air and pulled in by the pressure difference.
I have a very free flowing Pbobya 200 v2 radiator and I wonder, how much of a difference push pull would make when a large surface area of the radiator is not directly in line with my 8 inch fan. Would it then be evermore important to block any air leaks around the fan? Would be interesting to test, not many videos for larger fans as a water cooling topic
Well first off: there is no venturi effect going on behind the fan...my guess is this: you can see that there is no difference on the paper despite higher RPM. This is an indication that the airflow is separated from the fan blades even at low RPM. So right behind the fan is a separation wake. Wakes due to flow separation are always at lower pressure than ambient which is why the paper gets kinda sucked into a certain spot
For me is interesting whole video, because of demonstration of things will help understand why, and what to do in similar, but not exactly same, situations :)
It would be interesting to put a little silicone caulking or something around the fan to decrease the back pressure and force the air through the radiator. Idk if it would make a difference, but it would be interesting. Keep up the good work Jay!
Choosing a radiator with wide bends and maximum thickness is of great importance. High fin count radiators will show wild results with respect to fin shape and paint/finish used. Great video and thanks for making it. I did massive amounts of testing with 6, in loop temperature points, 5 radiators, ~12 fan types, and 3 cases. I found to many variables to call such a simple outcome. Done on 680i & 780i test beds with full mobo cooling and sli configs. Even added RAM cooling lines into the loop.
I would bet that if you matched the angle of the fins of the radiator to the angle of the fan blades, that you would increase total air flow. The question is would that increase or decrease the cooling effect of the radiator.
Yeah because I'm not sure whether I want a 60mm 480 long rad for my Enthoo Primo or an 80 thick. Will prob push pull regardless but how "bad" is it to use 4 only is what I'm curious about.
Jay, not sure if anyone answered your question about the paper being sucked down by the pull configuration: My theory is that in a pull configuration, you are reducing the air density. The radiator resisting, plus the fan sucking, create a vacuum effact. Now obviously not a total vacuum, but pressure between the fan and the radiator is likley measured a pound or two per square inch lower than the atmosphere. This lower density air is then blown away by the fan, which compounds the vacuum effect. So rather than having moving air that has effectively only 1 or 2 psi less measured pressure, it'll be 3-4 psi less meassured pressure. This creates the "sucktion" of the paper down when compared to a standalone fan. Due to the nature of a push configuration, air leaving the radiator will be at an equal or higher density to the surrounding air, ignoring temperature and other variables.
What I appreciate about you Jay, is your commitment to quality content. You provide us videos that are easy to watch, and the watcher leaves having actually learned something. You had good parents.
Awesome video. I've been researching this topic for a current build, and the results are what I need to know. What in infer from the result is this: -Push pull is better across the board. - Push is better across the wider RPM range, if you have to choose. I also think it shows that a less- than-optimal fan can be made to work in a push-pull, and deliver at least comparable results to a higher- end fan in only push or pull (within reason). Good news for the RGBers out there.
the air isnt moving so fast that its pushing the paper up but because the air under is moving faster than the air above the presure bellow the paper is lower. this means the force of the paper being sucked down by the low presure is stronger than the force of the air hitting the paper.
What about the other fan then? Why doesnt this see the same effect. Based on what you are saying both should have responded the same, which they didnt.
Because the force of the air pushing the paper is stronger than the force of the suction from the low presure. I think... ahaha its the same princple as a plane wing :)
+Dan Millard it doesn't matter whether it's the air or the wing (or paper) is moving. All that matters is that there is a velocity difference between the two. Motion is relative between two objects and all that matters is that there is a difference. Staying with the plane example, you can see this with small planes in severe wind. The plane is stationary relative to the ground but there is a difference in velocity between the air and plane and the plane can actually take off. It's also why planes take off into the wind.
+Dan Millard it doesnt have to be moving... the low presure comes from the difference in speed of air move over it, you can test it yourself by hold a piece of paper and blowing over the top of it
The paper is pushed into the airflow because the pressure is slightly lower where the air is flowing. The paper stays in this particular spot because it's in equilibrium between the force of the airflow and the one created by the pressure. I'd say the effect is less noticable on the push config because the airflow is disturbed by the radiator (because it's after the fan)
I think the best setup should be high static pressure fan pushing and high airflow fan pulling. At least I know from my experience with car radiators that push and pull fans have different blade design and fan made for pulling acts absolutely awful while pushing.
I was pondering the results in the same manner as you Jay but I also postulated that there may have been extra heat being generated by the fan motor itself when at full power in addition to the air leakage.
The point at which the fan flap is being sucked towards to is due to pressure difference from the surroundings (higher) and that point (lower) because of delta P across the radiator. If we were to use the bernoulli's equation, assuming datum to be equal (hence potential energy term is out), due to a higher pressure drop, there will be lower velocity of air out (since V1 is constant across both rad and non-rad fans). Thus having a lower velocity of air also accounts for the lower flap of the paper
Best free AV's are Avira and Bitdefender, paid for are Kaspersky and Bitdefender, and these are always in the top 3 of their categories (free, and paid for). Malwarebytes is something you should also have, it's not an AV so won't interfere with your AV, but it's the best tool for malware and it's free.
I see you are testing the effectiveness of push/pull/pushpull on a CPU cooler. Great, but I have a different setup with a tiny ITX case that has liquid cooling with just 1 fan in it. The one fan is the intake fan, the radiator fan, but it is also needs to push air through the entire case, which only has the PSU for outlet fan. In this setup, it's not a matter of how effectively the CPU gets cooled, but rather, how effectively the one fan cools the CPU+Case together. That's why I thought I would need two fans in a push-pull. Your video helps me understand the effect on my CPU temps and air-flow. It shows that if I have the extra room for another fan, push-pull gives me lower temps at lower RPM, and it also ensures good airflow through to the back of the case. I have not yet gauged the effectiveness of 25mm vs slim 15mm fans, but it sems 15mm fans produce less pressure and less cfm, so I'll go with 25mm if they can possibly fit.
Anytime I need an idea for my build I'll always come here no matter how old this video is it's an inspiration and I should be getting to work push and pull...
I don't know why but higher TDP CPUs are releasing less heat? i5 4460 @ 3.4Ghz H110 280mm runs at 75C max i7 3820 @ 4.3Ghz Lucifer V2 air cooler runs at 67C max Both were running in same computer Ambient temperature 26C
My mate said that push/pull is a waste, that it makes no difference to temps. But I have been running push/pull on a 360 rad for my Nvidia 970. Under full load for an hour clocked @ 1670/4055, it never goes above 33 degrees. So I think push/pull works fine. Thanks for the video. It was really helpful.
great video and really usable information. as for the push and pull theory, if i can explain this well it will be understood. being a car guy i know that an engine will breath better with a cold air intake at higher rpm, because there is less restriction. so a pull configuration will cause a vacuum and suck the air through the fins. on the other hand if you super charge an engine..... it will find a leak. if you try to force air through those fins it will cause a positive pressure as there is resistance through the fins and it leaks anywhere it can. air doesn't like to bend or turn, the fins are a straight shot to the vacuum source (fan) in a pull configuration. this is just my theory. sure there is a threshold where push is actually better, so it all depends on how frequently you push your system. if the fans are never at 100% push may be better for your build.
actually no! lmao was no. 1 meaning number 1....srsly this guy 2 years ago made me fall in love of computers....and does a freaking amazing job with these vids we have to admit it!
To explain the full RPM results with static pressure fans on the radiator: Push -> forces air through the radiator Pull -> relies on the air filling the vacuum void in the radiator The reason that the Push configuration performed less than Pull at full RPM is because the fan was attempting to move more air (CFM) than the radiator could freely flow, thus forming a layer of compressed air between the fan and the radiator. The layer of compressed air is what caused the loss in cooling performance. This could be solved by placing an empty spacer ( i.e. an old fan with the internals cut out) between the Push fan and the radiator.
Great video Jay, I myself have also done a lot of fan testing with radiators and I confirm the swirl/interference effect you mentioned when testing the fan with/without rads. It's different per fan though. Anyway, you forgot one thing when mentioning why pull in slower rpm showed worse result than push. It's because the gap between the radiator and the fan itself, since the fan's shroud and or/rad's edge isn't perfect, there's always an air gap between them. Unless you fill that gap with rad covers or wateva (even paper works, just remember to not let them get inside the fan's shroud), you're not going to get all of pull's potential cooling performance. Admittedbly it's not much, but it does, especially at low rpms. Hope to see more videos from you Jay, always great to see these kind of vids. Cheers!
Too bad the #1 argument of pull vs. push is missing: dust not collecting between fan and rad, making it easier to clean it off the fanless surface. This is more meaningful than a 2°C difference ... which is close to margin of error anyway, as we still don't see delta temps on this channel for whatever reason.
Ding, Ding, Ding!!! The winner of the #1 argument of pull vs push goes to Oliver Asus, thank you for your comment today... Is the main reason I use a pull configuration on my AIO...
I would venture a guess that the vortex noted in the pull config is caused by the slow-moving air spending more time being spun around by the blades rather than going through normally. This should be happening to some extent as well between the fan and radiator in the push config. This could also explain why pull performs better at higher RPMs since the air passing through the radiator doesn't have a spin on it.
push/pull with shrouds on both side is the best. it will help remove the dead spots in front of the fan and reduce noise a lot due to less turbulent airflow.
The suction... It's Bernoulli's principle at work (as velocity of a fluid increases, the pressure decreases). The paper is drawn to the point where the static pressure above and below are equal
Keep in mind for a radiator, cooling performance consistency is also very important. Pull has the most consistency and ease of maintenance, as it is easy to clean. Push, and push-pull, are difficult to clean, thus are often cleaned less often, thus for many users, then end up with higher average temperatures in the long run, especially if they are in a home without HVAC and a high end air filter.
@JayzTwoCents , Try to expand this test a bit more by using shrouds. Get Barnacules to 3D print (or buy one or if you're really on a budget and you are using like 120mm fans take some old 120mm fans that are shot/dead/useless and cut out the inner motor and fan blades and the supports so it's just the outer shell and use that as a shroud) a shroud that completely isolates the airflow with the radiator. Probably have to make 2 different shrouds: for placing the fans directly on the radiator and another for placing the fans offset by about 1/2" or so from the radiator. So, hard for me to explain I guess. But basically make a housing so that there's no air or heat leakage from the sides or back-pressure and try Push/Pull/PushPull again and see if it makes a difference. I imagine it would make a few degrees difference because the static-pressure fan doesn't lose some of it's air-flow momentum to air leakage. A shroud could also cut down on ambient-case temperatures as well, if you would like to test that. Oh yeah, I also recommend getting a sheet of very thin silicone or rubber padding to work as a gasket anytime you install fans. It's made a huge difference for me (I've had some vibrating fans recently so I took some gasket-maker I had laying around and some wax paper and basically pressed a layer of gasket maker into a sheet, and then I cut it out to fit my needs. I'm very surprised in the PC D.I.Y. scene and it's lack of fan gaskets.
You should do the same test now. But, the camera be positioned side view directly and not angled view. And there should be some vertical line markers chart(maybe wall paper with marker) beyond that fan to know exactly how much does air pushes the paper to be analyzed properly.
Air flow produces pressure such that it can balance the weight of the sheet. The sheet will certainly drop down if the downward acceleration/net force exists. The 'sucking effect' you named is merely a result of smaller volumetric flow rate (units metre^3/second) due to the fin 'barriers' of the radiator and nothing is sucking the sheet. Correct me if I was wrong. -Cambridge University Engineer's explanation
It's called airflow instability..The fan pulling has 4 small braces that disrupt the airflow. You're still moving the same amount of air. Just not straight out. NASA had this same issue with the Saturn V rockets. Only it was combustion. The flame rotation was not stable.
At a guess I would suggest the paper isn't being pulled back in but pushed back in. No joke intended. The fan's intake is restricted, so the air comes out the far side of the fan radiator combo at a lower pressure than it went in at and hence the surrounding air is flowing back towards this area of low pressure and the paper is resting at* the point of equilibrium between the two. (* not actually at, slightly closer to the fan due to gravity)
wow glad to see this test Im working with a peltier cooling set up and was wondering about the push pull on a 30mm radiator to pull the heat away from the pelt to get the best cooling affect thanks for doing this test even though it was a over a year ago.
JayzTwoCents, you are right; there will be some vortex effect for a Pull config as the higher static pressure fans will cause a venturi effect after stabilizing the air flow. However, the wierd effect you saw earlier was due to the fact the fans are designed to throw air out in a cone rather than in a straight line, unlike the control on the right. This causes the paper to be maintained at an equilibrium position within the vortex currents (you can test this theory using a smoke machine) hence it will push when against the flow when pushed towards the fan and vice-versa. Again, this is just a theory from a mechanical engineer's point of view; I would be happy to correct myself if someone tested this out and found a flaw in the reasoning.
It would be interesting to see your assumption of the "backpressure" verified by using some kind of gasket or sealing to prevent the air from passing out and if using a gasket or sealing with a push-configuration improves that value aswell. Also I'd find it interesting to see values of a push/push- or pull/pull-configuration with/without gasket/sealing.
ThANK YOU JAY!!! This was my biggest probably when pricing out a PC. Im a perfectionist, and even if I'm not spending more, I still want the best performance no matter what. We spend 1-2 grand on these things, obviously we want them to run the best that they can
You should 3d print/design (or get barnacules to print) a vent that goes from the fan to the rad, that you can seal on both ends to prevent air from escaping, then see how much it effects temp.
It's worth tweaking fan speed on the push/pull because the pull isn't going to be able to keep up with the push cfm as pull has to deal with more resistance than push. Solid gold testing though @JayZTwoCents subbed a year ago and catching up on all the stuff I missed!
Just mod your case and put the radiator OUTSIDE far enough from the case, with a good fan under it to push (NOT attached to the radiator, literally UNDER the radiator, at least 5cm) and you are done. I did it with my case, i just had to make some holes with my drill and installed 2 metallic staffs. It works perfectly.
Using 32mm - 38mm thick fan shrouds on both fans in push/pull will help eliminate fan vortex and will drop the temps by a couple of degrees over just push/pull. (Push)>Fan>Shroud>Rad>Shroud>Fan>(Pull) Even just one shroud will help temps (Push) >Fan>Shroud>Rad>Fan>(Pull)
Jay, .... Welllll ... I have a small quandry.... this particular video answered some of my questions - but not all of them .... First .... my setup: RYZEN 7 -2700x Asus ROG Crosshair 7 Hero - wifi 32gb - G Skill Trident Z RGB - DDR4 3200Mhz 16gb sticks ~ x2 EVGA RTX 2080 Ti FTW3 ~ x2 / NV Link ~ 11G-P4-2487-KR Asus ROG Ryujin 360 AIO Asua ROG THOR 1200w Platinum PSU Panteks Enthoo Luxe case Fans: (my final selection based on your reply) Case front : Noctua NF-A14 industrialPPC-2000 IP67 PWM - 31.5 dB/A ~ 182.5 m3/h ~ 4.18 mm H2O *** OR *** Noctua NF-A14 industrialPPC-3000 PWM - 41.3 dB/A ~ 269.3 m3/h ~ 10.52 mm H2O Case rear: Noctua NF-P14s redux 1200 PWM - 19.6 dB/A ~ 110.3 m3/h ~ 1.29 mm H2O ** for positive pressure ** (is that correct?? ~ lower CFM at exhaust??} RYUJIN AIO : Noctua NF-F12 industrialPPC-2000 PWM - 29.7 dB/A ~ 121.8 m3/h ~ 3.94 mm H2O ** installed in top in PUSH ** .....I've been looking at some pretty ARGB fans for the inside of the case ....like the corsair HD / LL / ML / SP Series of 120mm fans ..... they put out around 60-75 CFM at ~1200-1750rpm to use in PUSH and then move the supplied Noctua iPPC fans on the AIO to PULL ...... (should I do that?) OKAY....I DIGRESS... My question is this : In order to maintain POSITIVE pressure in my case .... how should I set up the radiator ??? is my theory sound ??? What GENERALLY type of fan would you use for a push/pull when rad is an exhaust??? Should you always use 2 equal fans or enact a positive pressure setup for a push/pull when its an exhaust ?? JAY ....PLEASE HELP ....thanx ~Cyliss
I'm not sure, but you can be getting a Venturi effect because of the highest part of the radiator, that slows down the air, so it can't travel at the same speed as the fan without radiator. Making the air at the top to gain pressure. I think the results can vary if your radiator is horizontal. The push-pull difference you have, may be just because of the construction of the fans, they might not be designed for high pressure loads
Hey Jay, great video as usual! You should've tried the smoke thing to see how the air flows during each configuration. I believe it would have solved the mistery between the push and pull setups. I personally find your explanation quite convincing though. Keep it up!
i think... because the air is being pulled through the rad fins, the air became lamina, but then hit the fan blade which caused a low pressure in the center of the fan... so then cause the paper was placed off set air was traveling over the back of the paper towards the low pressure at the center of the fan
Excellent testing, thanks! My thoughts are that in a pull config you get a more even airflow through the rad as there's a low pressure void between the fan and fins, and possibly a dead spot in the center of the rad on a push?. Also at high rpm pull fan it's almost beggining to act as a centrifugal fan, it's flinging air outwards instead of flowing axially. (I'd wager you'd get more pressure across the fan itself with the intake blocked vs the outlet). It would be really amazing to see a fog system to see all the cool airflow patterns, nice idea.
Fans are made to blow, not suck. Putting restriction on the inlet side causes surging and stalling as fan blade air requirements exceed supply and you get pressure pulses bouncing off each other as the air moves back forth into the low pressure "vacuum" between the fan and the radiator.
I think the suction you are getting is a result of the resistance of the radiator. Imagine if the radiator was a solid wall instead. No air would be able to come in from that side, and all of the air that the fan blades are pushing out has to come from somewhere. I imagine the air would come from the small gap between the blades, and the cowling. The radiator is not a solid wall, but does create resistance and it would coerce some air to be sucked in through that thin gap between the blades and cowling.
thank you, I actually commented on another of your videos about this subject as I am buying and trying to figure out to do push/pull or push&pull for an "ultra" quiet setup. I still don't know what kind of audio difference I will get but from other readings, the ability to run much lower rpm even with double the fans should still be far lower in dB than a single side fan at faster speeds.
I’m not really sure why, but if your radiator and fan we’re not, itright against one another and you had a standoff with a solid ring gaskets against your radiator to hold your fan blades at least an inch out from your radiator your air flow would increase by 100% on a pull configuration
The big question is cooling versus noise. The results at 12:07 indicate that push and pull at 1600 RPM both do just a little more cooling than push+pull at 1000 RPM. This graph: i.imgur.com/jP01d6Fl.jpg indicates that that particular Corsair fan produces less than 31 dB at 1000 RPM, and less than 40 dB at 1500 RPM. That difference of 9 dB is equivalent to 3 doublings in acoustic energy (e.g., using 8 times as many fans). So a push/pull at 1000 RPM cools nearly as much as push or pull at 1600 RPM, but the acoustic energy at just 1500RPM is already 4 times as much as that produced by 2 fans at 1000 RPM. The results comparing 1600 RPM to 2200 RPM give a similar result. TL/DR: Push/pull wins big-time if your main concern is fan noise. (Will win for energy consumption, too, if that's proportional to acoustic energy produced.)
Although the fans spin at the same speed, the fan on the left has greater resistance pulling air, which reduces its airflow slightly to the paper is not at the same height. Pushing air has no initial resistance, it exhausts through the radiator creating "splash" as you call it! This oscillating air creates additional friction allowing for better cooling in most cases, as it increases the surface area. The key is to pull that oscillating hot air out quickly. By adding a pull fan to the Push config you get the best of both worlds, - it have be further optimised if the fans run a different RPM's.
the suction on the radiator is being magnified by the resistance and the airflow over the paper is a low resistance path for the extra airflow to equalise that huge surface area's negative pressure bubble
the paper os badicaly in equilibrium and has become a ramp for the air because it's blocking the flow of the fans forward direction, to the negative direction air presses on the ramp pf paper at a rate that depends on the dimensions of the radiator and the fan pressure
When the push is alone, it has unrestricted airflow to insert into the radiator, and then the pressure rises inside it, when pull, you do a negative pressure inside the radiator, having an unrestricted exaust, but its actually bottlenecked by the radiator, because the air goes in small blocks in straigh line, instead of a tornado shaped cone. I would like to see if doing some spacing makes a difference, because, if the pull had more space to generate that layer of negative pressure, would be the same. We F1 levels of aero now.
Yes at 1:15 I say fan instead of radiator... that's because I'm dumb.
HI JAY I LOVE YOU LIKE MY MOTHER !
Dope Video xD
:)
Relax there buddy, Lol. Wish people would stop worshipping youtubers.
Nice video Jay. Really enjoyed this informative/testing style vid.
6 years and this video is still a must watch for pc builders. Thanks Jay
7 years now, still holding true. Thanks!
All fan Configuration RPM's
*Control*
1000rpm: 6:07
1600rpm: 6:54
2200rpm: 7:23
*Pull*
1000rpm: 6:07
1600rpm: 6:54
2200rpm: 7:23
*Push*
1000rpm: 9:16
1600rpm: 9:35
2200rpm: 9:57
*Push/Pull*
1000rpm: 10:37
1600rpm: 10:41
2200rpm: 10:58
Temperature Charts: 12:07
Make sure to subscribe for good stuff like this.
Nice! Thank you.
@@Silent_Shadow finally someone commented!!!!
Epic
best off
What I suspected. Thank you
You should do a comparison between the air flow fans and the static pressure fans. Everyone says that the static pressure fans are better but what I would like to know is how much better they are.
for high fin density and thick radiators the static pressure fans are better.
As the name implies, they are build for the high pressure that comes from an obstructed airpath - aka having a rad in the way.
it of course heavily depends on your rad-config. But with 30mm and above rads you can't go wrong with pressure fans.
+ABaumstumpf Sorry but you make no sense here. He KNOWS that static pressure fans are superior in that regard, he just wants to know how much better.
VenoXj1 That is hard to say - cause there are so many things going into this.
If you only compare a single 120 rad in open air, than you can easily give a number.
But it will greatly change if you have it inside a case.
What radiator? what rad-thickness and find-density?
How many fans you have/need?
The more fans the lower rpm you likely need, and that again has a big impact.
Does you case have positive or negative pressure?
Just like that you already have to test 200 or more scenarios.
but at AnandTech there is an easy to understand pressure vs Flow fan comparison.
www.anandtech.com/show/9737/the-corsair-sp140-led-af140-led-sp120-af120-fan-review
ABaumstumpf Are you trolling me right now?
VenoXj1 No?
It is just in general NOT POSSIBLE to give a single answer.
Let me give you a question in return:
How much money does a human have?
I believe the effect causing the piece of paper on the radiator fan to be sucked down is Bernoulli's Principle. It states that as the speed of a fluid (liquid or gas) increases the pressure decreases. So the air under the piece of paper is moving faster because of the fan while the air on top is stagnant. Thus the air on bottom is a low pressure system relative to the air up top. Then the paper is pushed down because of the lift generated by the high pressure pushing down on it trying to equalize with the air on bottom. This is how airplanes fly except it is reversed so the lift generated is pushing up and not down. But it is exactly how high performance cars work, it's known as down force.
Why is the first fan not doing it then? Are the fins creating more pressure?
+Mike Sico I think the first fan is just putting too much air out for it to happen. I think the radiator lowers the pressure just enough to cause that phenomenon to happen.
Yep, this is the correct explanation.
I agree with you!!
I'm not entirely convinced that there will be a lower pressure in this case, as what you're talking about is static air and a moving wing through it. The air may actually have a higher pressure in this case as it is being forced forward from behind. I actually think it's the rotation of the caused by the higher RPM of the fan. The reason it does not occur on the fan without the radiator is that the air is moving sideways quicker, and therefore does not have as tight of a "corkscrew" to suck the end of the paper back down. To put this into the perspective of aircraft again, this is a real thing experienced in single engine propeller aircraft at low speed. When the propeller spins up when the aircraft is at a low speed (typically on the ground), a significant yaw is applied to the aircraft from the propeller wash actually circling around the fuselage of the aircraft and hitting one side of the vertical stabiliser. In the case of the radiator fans here, there is a tight corkscrew of air that is rotating quickly, but not moving as fast horizontally as the fan without the radiator. I could be wrong though.
While I will not be able to built a PC in a while, it is always nice to keep increasing my knowledge base for when I am able to. Keep up the great work!
Did you manage to get your build?
@@Cool-yx7us not yet 🤣
I just rebuilt my 455 Buick, bored 30 over with Comp cam. The temp was over 200 degrees which makes me extremely uncomfortable. I decided to invest in push fans. Watching your video gave me added assurance I was moving in the right direction. Thank you very much
9:00 - The Bernoulli Principle at work. Increasing fluid velocity decreases the pressure. The fan exhaust is therefore at a lower pressure than the rest of the room, pulling the paper toward it. There's some equilibrium between where the paper is getting pushed away by the air and pulled in by the pressure difference.
I have a very free flowing Pbobya 200 v2 radiator and I wonder, how much of a difference push pull would make when a large surface area of the radiator is not directly in line with my 8 inch fan.
Would it then be evermore important to block any air leaks around the fan?
Would be interesting to test, not many videos for larger fans as a water cooling topic
Well first off: there is no venturi effect going on behind the fan...my guess is this: you can see that there is no difference on the paper despite higher RPM. This is an indication that the airflow is separated from the fan blades even at low RPM. So right behind the fan is a separation wake. Wakes due to flow separation are always at lower pressure than ambient which is why the paper gets kinda sucked into a certain spot
11:49 for actual numbers and skip the chit chat padding.
12:07 for all the numbers.
For someone who doesn't know much about this I think watching the whole video could help.
@@testtube173 Obey the Jay
For me is interesting whole video, because of demonstration of things will help understand why, and what to do in similar, but not exactly same, situations :)
It would be interesting to put a little silicone caulking or something around the fan to decrease the back pressure and force the air through the radiator. Idk if it would make a difference, but it would be interesting. Keep up the good work Jay!
Choosing a radiator with wide bends and maximum thickness is of great importance. High fin count radiators will show wild results with respect to fin shape and paint/finish used. Great video and thanks for making it. I did massive amounts of testing with 6, in loop temperature points, 5 radiators, ~12 fan types, and 3 cases. I found to many variables to call such a simple outcome. Done on 680i & 780i test beds with full mobo cooling and sli configs. Even added RAM cooling lines into the loop.
Love the new series thing! The channel is growing so fast!!!
Lol 4 years later
I would bet that if you matched the angle of the fins of the radiator to the angle of the fan blades, that you would increase total air flow. The question is would that increase or decrease the cooling effect of the radiator.
You think you could do this test with 60mm thick and 80 mm thick rads?
I have a 60 and 80mm rad but they are triple rads, I need to get a single 120 and then I can try this again.
Yeah because I'm not sure whether I want a 60mm 480 long rad for my Enthoo Primo or an 80 thick. Will prob push pull regardless but how "bad" is it to use 4 only is what I'm curious about.
The thicker the radiator, the harder it is to get all the air out. Keep that in mind
+Xenteon go with 60mm unless youre doing push pull it wont be able to get enough air through with single fans
+WayStedYou Well yea I will eventually it just costs 100$ verses 200$
Jay, not sure if anyone answered your question about the paper being sucked down by the pull configuration: My theory is that in a pull configuration, you are reducing the air density. The radiator resisting, plus the fan sucking, create a vacuum effact. Now obviously not a total vacuum, but pressure between the fan and the radiator is likley measured a pound or two per square inch lower than the atmosphere. This lower density air is then blown away by the fan, which compounds the vacuum effect. So rather than having moving air that has effectively only 1 or 2 psi less measured pressure, it'll be 3-4 psi less meassured pressure. This creates the "sucktion" of the paper down when compared to a standalone fan. Due to the nature of a push configuration, air leaving the radiator will be at an equal or higher density to the surrounding air, ignoring temperature and other variables.
What I appreciate about you Jay, is your commitment to quality content. You provide us videos that are easy to watch, and the watcher leaves having actually learned something. You had good parents.
Anyone else noticed how much weight Jay has lost over the years???
he did a video on him doing long bike rides to lose weight
you sir are blind and deaf
Of course
Same, its crazy like he's dropped at least a hundred pounds or close to it I'd say
Hard work men hard work that's why
I want to see a video on chassis fan configuration and if it'll make a difference if you change them around. PLEASE MAKE THIS HAPPEN Jay
Thanks for the explanation, 5 years later and still valid!
Awesome video. I've been researching this topic for a current build, and the results are what I need to know.
What in infer from the result is this:
-Push pull is better across the board.
- Push is better across the wider RPM range, if you have to choose.
I also think it shows that a less- than-optimal fan can be made to work in a push-pull, and deliver at least comparable results to a higher- end fan in only push or pull (within reason). Good news for the RGBers out there.
the air isnt moving so fast that its pushing the paper up but because the air under is moving faster than the air above the presure bellow the paper is lower. this means the force of the paper being sucked down by the low presure is stronger than the force of the air hitting the paper.
What about the other fan then? Why doesnt this see the same effect. Based on what you are saying both should have responded the same, which they didnt.
Because the force of the air pushing the paper is stronger than the force of the suction from the low presure. I think... ahaha its the same princple as a plane wing :)
+JayzTwoCents www.explainthatstuff.com/howplaneswork.html
+Dan Millard it doesn't matter whether it's the air or the wing (or paper) is moving. All that matters is that there is a velocity difference between the two. Motion is relative between two objects and all that matters is that there is a difference. Staying with the plane example, you can see this with small planes in severe wind. The plane is stationary relative to the ground but there is a difference in velocity between the air and plane and the plane can actually take off. It's also why planes take off into the wind.
+Dan Millard it doesnt have to be moving... the low presure comes from the difference in speed of air move over it, you can test it yourself by hold a piece of paper and blowing over the top of it
If you're going with single fan on a rad, I always suggest pull just for ease of cleaning when you need to get at the intake side of the rad.
Old post, but this is a VERY good point.
Good to see an update on this topic. I just had to reassemble my PC and was wondering on changing my fan configuration. Really useful info! :)
The paper is pushed into the airflow because the pressure is slightly lower where the air is flowing. The paper stays in this particular spot because it's in equilibrium between the force of the airflow and the one created by the pressure. I'd say the effect is less noticable on the push config because the airflow is disturbed by the radiator (because it's after the fan)
I think the best setup should be high static pressure fan pushing and high airflow fan pulling. At least I know from my experience with car radiators that push and pull fans have different blade design and fan made for pulling acts absolutely awful while pushing.
Who else was waiting for a Norton ad? haha just playing Jay.
I was pondering the results in the same manner as you Jay but I also postulated that there may have been extra heat being generated by the fan motor itself when at full power in addition to the air leakage.
The point at which the fan flap is being sucked towards to is due to pressure difference from the surroundings (higher) and that point (lower) because of delta P across the radiator. If we were to use the bernoulli's equation, assuming datum to be equal (hence potential energy term is out), due to a higher pressure drop, there will be lower velocity of air out (since V1 is constant across both rad and non-rad fans). Thus having a lower velocity of air also accounts for the lower flap of the paper
Thank God, no Norton ad this time around.
it must have ruined your life crybaby
So you don't like Norton what do you think is the best AV then?
+llStarscreamll common sense
Best free AV's are Avira and Bitdefender, paid for are Kaspersky and Bitdefender, and these are always in the top 3 of their categories (free, and paid for). Malwarebytes is something you should also have, it's not an AV so won't interfere with your AV, but it's the best tool for malware and it's free.
Oh boy, all these people suggesting 'common sense' and a free, non real-time scanner....
Have fun with that.
I was really looking forward to the norton ad ;(
@ClownFace1511 did you see LTT featured your build recently :D
Scrivyyy When?
+ClownFace1511 on vessel a couple of days ago :>
Scrivyyy :0 I didnt know. Can you tell me the title of the vid.
+ClownFace1511 think it was the 'which Intel CPU should I buy' video
I see you are testing the effectiveness of push/pull/pushpull on a CPU cooler. Great, but I have a different setup with a tiny ITX case that has liquid cooling with just 1 fan in it. The one fan is the intake fan, the radiator fan, but it is also needs to push air through the entire case, which only has the PSU for outlet fan. In this setup, it's not a matter of how effectively the CPU gets cooled, but rather, how effectively the one fan cools the CPU+Case together. That's why I thought I would need two fans in a push-pull. Your video helps me understand the effect on my CPU temps and air-flow. It shows that if I have the extra room for another fan, push-pull gives me lower temps at lower RPM, and it also ensures good airflow through to the back of the case. I have not yet gauged the effectiveness of 25mm vs slim 15mm fans, but it sems 15mm fans produce less pressure and less cfm, so I'll go with 25mm if they can possibly fit.
Anytime I need an idea for my build I'll always come here no matter how old this video is it's an inspiration and I should be getting to work push and pull...
Push+Pull Master Race.
Corsair H105 240mm
5820k @ 4.3ghz
~35C at idle
~65C max at 100% load with all fans on auto
Pretty good for a 140w TDP CPU
or my 6700k that can't break 60 at 5.0ghz and 1.4v under custom water.
#pushpullmasterrace A10 7850K Antec AIO 4.2ghz Never hitting 40℃... \m/
Cloud Soldier amd actually puts good thermal compound on their cpus's, thats how back in the day my 8150 at 5.2ghz stayed in the 50's
+James Mastroianni IIRC they're soldered!
My 860k at 4.5Ghz 1.51mV is 54c :D on Liqmax 240 ii (about the same as a H105)
I don't know why but higher TDP CPUs are releasing less heat?
i5 4460 @ 3.4Ghz
H110 280mm
runs at 75C max
i7 3820 @ 4.3Ghz
Lucifer V2 air cooler
runs at 67C max
Both were running in same computer
Ambient temperature 26C
Hey Jay is it possible for you to make a video showing the performance difference between the 3GB and the 6GB 1060?
Yes, coming soon
+JayzTwoCents Jayz getting a 1060 3gb iz convirmed.
My mate said that push/pull is a waste, that it makes no difference to temps. But I have been running push/pull on a 360 rad for my Nvidia 970. Under full load for an hour clocked @ 1670/4055, it never goes above 33 degrees. So I think push/pull works fine. Thanks for the video. It was really helpful.
Do u use same speed fan push pull? If i use 2k rpm push and 1.8k rpm for pull is that okey?
great video and really usable information. as for the push and pull theory, if i can explain this well it will be understood. being a car guy i know that an engine will breath better with a cold air intake at higher rpm, because there is less restriction. so a pull configuration will cause a vacuum and suck the air through the fins. on the other hand if you super charge an engine..... it will find a leak. if you try to force air through those fins it will cause a positive pressure as there is resistance through the fins and it leaks anywhere it can. air doesn't like to bend or turn, the fins are a straight shot to the vacuum source (fan) in a pull configuration. this is just my theory. sure there is a threshold where push is actually better, so it all depends on how frequently you push your system. if the fans are never at 100% push may be better for your build.
Wow i'm watching this in 2020 and I didn't realize Skunkworks had Red Fluid! I'm so use to the Green!
good job jay ur no. 1 for me!!!!i always watch ur vids 1st!!!
Thanks Alex!
awwwwwwww
+FookyFooky GMania lmao
actually no! lmao was no. 1 meaning number 1....srsly this guy 2 years ago made me fall in love of computers....and does a freaking amazing job with these vids we have to admit it!
How does it affect noise?
Brought back the Dubstep Kings! Noyce!
Is one 240mm rad better than two 120mm rads
No because surface area is the main factor for cooling
better price/performance ratio if you give it nice fan. 2x120 would cost more, but will have twice the airflow path
what?
240mm rad is basicly just a twice the size 120mm rad so how come apparently they dont perform the same ppl? explain
MoDRun 2x 120 gives you double the airflow area
Gotin Fuklio what u mean double the airflow area? #+# or ## has the same surface area...? same size
120mm+120mm= 240mm
what about Pull Pull?
lol i dont think that would work well lol
+恵勇気 the ek vardar?
+That would be Illegal well if have seen others recommend pull/pull.
You'd probably have very slow moving air, and so high temps.
linus has stated pull/pull or push/push is pointless and i'm inclined to believe him.
i put 4x140mm 3k rpm ippc fans on my AIO.. huge improvement.... huge fan of ippc fans for any job where static pressure matters.
To explain the full RPM results with static pressure fans on the radiator:
Push -> forces air through the radiator
Pull -> relies on the air filling the vacuum void in the radiator
The reason that the Push configuration performed less than Pull at full RPM is because the fan was attempting to move more air (CFM) than the radiator could freely flow, thus forming a layer of compressed air between the fan and the radiator. The layer of compressed air is what caused the loss in cooling performance. This could be solved by placing an empty spacer ( i.e. an old fan with the internals cut out) between the Push fan and the radiator.
So would these results translate over to a cpu heatsink?
pretty much. Heatsinks tend to not be quite as dense though, so similar effect, but not as strong.
yayayya no Norton
That was a one time deal... now rejoice!
this makes me happy.
+JayzTwoCents yasss
Great video Jay, I myself have also done a lot of fan testing with radiators and I confirm the swirl/interference effect you mentioned when testing the fan with/without rads. It's different per fan though.
Anyway, you forgot one thing when mentioning why pull in slower rpm showed worse result than push. It's because the gap between the radiator and the fan itself, since the fan's shroud and or/rad's edge isn't perfect, there's always an air gap between them. Unless you fill that gap with rad covers or wateva (even paper works, just remember to not let them get inside the fan's shroud), you're not going to get all of pull's potential cooling performance. Admittedbly it's not much, but it does, especially at low rpms.
Hope to see more videos from you Jay, always great to see these kind of vids. Cheers!
Love the videos! By far the best PC/gaming material on RUclips hands down. Your hard work shows with the quality of the videos!
You should get nicotine vape and blow through the fans! Just costs a couple of bucks. #YeahiVapeSoWhat #yoloswag
gay
He should put vape liquid in a watercooling loop.
Vape Nation \//\/
No, he should move to Colorado and use a weed bonfire.
Kyle should come by and bring his vape.
Too bad the #1 argument of pull vs. push is missing: dust not collecting between fan and rad, making it easier to clean it off the fanless surface.
This is more meaningful than a 2°C difference ... which is close to margin of error anyway, as we still don't see delta temps on this channel for whatever reason.
*for pull vs. push
Too difficult to calculate ^^
I guess so ;]
The RT is constant. He has a climated room.
Ding, Ding, Ding!!! The winner of the #1 argument of pull vs push goes to Oliver Asus, thank you for your comment today...
Is the main reason I use a pull configuration on my AIO...
I would venture a guess that the vortex noted in the pull config is caused by the slow-moving air spending more time being spun around by the blades rather than going through normally. This should be happening to some extent as well between the fan and radiator in the push config. This could also explain why pull performs better at higher RPMs since the air passing through the radiator doesn't have a spin on it.
push/pull with shrouds on both side is the best. it will help remove the dead spots in front of the fan and reduce noise a lot due to less turbulent airflow.
Jay, I wasn't a FAN of this video
An AIO then?
hAIOooooooo!
Nobody should be running PC case fan above 1000rpm.
The suction... It's Bernoulli's principle at work (as velocity of a fluid increases, the pressure decreases). The paper is drawn to the point where the static pressure above and below are equal
YES on the fan testing! There's not enough good information on the hundreds of options out there. Plus, it seems like it would be right up your alley.
Keep in mind for a radiator, cooling performance consistency is also very important. Pull has the most consistency and ease of maintenance, as it is easy to clean. Push, and push-pull, are difficult to clean, thus are often cleaned less often, thus for many users, then end up with higher average temperatures in the long run, especially if they are in a home without HVAC and a high end air filter.
@JayzTwoCents , Try to expand this test a bit more by using shrouds. Get Barnacules to 3D print (or buy one or if you're really on a budget and you are using like 120mm fans take some old 120mm fans that are shot/dead/useless and cut out the inner motor and fan blades and the supports so it's just the outer shell and use that as a shroud) a shroud that completely isolates the airflow with the radiator. Probably have to make 2 different shrouds: for placing the fans directly on the radiator and another for placing the fans offset by about 1/2" or so from the radiator. So, hard for me to explain I guess. But basically make a housing so that there's no air or heat leakage from the sides or back-pressure and try Push/Pull/PushPull again and see if it makes a difference. I imagine it would make a few degrees difference because the static-pressure fan doesn't lose some of it's air-flow momentum to air leakage. A shroud could also cut down on ambient-case temperatures as well, if you would like to test that.
Oh yeah, I also recommend getting a sheet of very thin silicone or rubber padding to work as a gasket anytime you install fans. It's made a huge difference for me (I've had some vibrating fans recently so I took some gasket-maker I had laying around and some wax paper and basically pressed a layer of gasket maker into a sheet, and then I cut it out to fit my needs. I'm very surprised in the PC D.I.Y. scene and it's lack of fan gaskets.
You should do the same test now. But, the camera be positioned side view directly and not angled view. And there should be some vertical line markers chart(maybe wall paper with marker) beyond that fan to know exactly how much does air pushes the paper to be analyzed properly.
Loves his car so much, has it as wallpaper on his desktop.
I would too, if I owned a car like that.
I would sell it
Air flow produces pressure such that it can balance the weight of the sheet. The sheet will certainly drop down if the downward acceleration/net force exists. The 'sucking effect' you named is merely a result of smaller volumetric flow rate (units metre^3/second) due to the fin 'barriers' of the radiator and nothing is sucking the sheet. Correct me if I was wrong. -Cambridge University Engineer's explanation
Doing a mod on a power supply cooling fan and this is right on point. Thanks and subscribed.
i did NOT expect almost a 10 degree different from push to push pull.
thats insane.
It's called airflow instability..The fan pulling has 4 small braces that disrupt the airflow. You're still moving the same amount of air. Just not straight out. NASA had this same issue with the Saturn V rockets. Only it was combustion. The flame rotation was not stable.
I like the idea of seeing things like this with the fog machine
At a guess I would suggest the paper isn't being pulled back in but pushed back in. No joke intended. The fan's intake is restricted, so the air comes out the far side of the fan radiator combo at a lower pressure than it went in at and hence the surrounding air is flowing back towards this area of low pressure and the paper is resting at* the point of equilibrium between the two. (* not actually at, slightly closer to the fan due to gravity)
2024 still useful af... thank you Jay
wow glad to see this test Im working with a peltier cooling set up and was wondering about the push pull on a 30mm radiator to pull the heat away from the pelt to get the best cooling affect thanks for doing this test even though it was a over a year ago.
JayzTwoCents, you are right; there will be some vortex effect for a Pull config as the higher static pressure fans will cause a venturi effect after stabilizing the air flow. However, the wierd effect you saw earlier was due to the fact the fans are designed to throw air out in a cone rather than in a straight line, unlike the control on the right. This causes the paper to be maintained at an equilibrium position within the vortex currents (you can test this theory using a smoke machine) hence it will push when against the flow when pushed towards the fan and vice-versa. Again, this is just a theory from a mechanical engineer's point of view; I would be happy to correct myself if someone tested this out and found a flaw in the reasoning.
You should've used a rubber gasket between the fan and radiator since there seems to be a 2mm gap between the fan and radiator at 3/4 of the video!
It would be interesting to see your assumption of the "backpressure" verified by using some kind of gasket or sealing to prevent the air from passing out and if using a gasket or sealing with a push-configuration improves that value aswell. Also I'd find it interesting to see values of a push/push- or pull/pull-configuration with/without gasket/sealing.
ThANK YOU JAY!!! This was my biggest probably when pricing out a PC. Im a perfectionist, and even if I'm not spending more, I still want the best performance no matter what. We spend 1-2 grand on these things, obviously we want them to run the best that they can
You should 3d print/design (or get barnacules to print) a vent that goes from the fan to the rad, that you can seal on both ends to prevent air from escaping, then see how much it effects temp.
You can also buy rubber fan gaskets to help seal the fan/fan shroud to a radiator.
Very Nice demonstrations; definitely worth the extra $$ for a Push/Pull config If you've already spent on a liquid cooling setup and radiators.
This video was way more informative and interesting than i expected! Good job! Keep it up!
Also, do one with smoke please, really curious about the backdraft/vortex effect...
It's worth tweaking fan speed on the push/pull because the pull isn't going to be able to keep up with the push cfm as pull has to deal with more resistance than push. Solid gold testing though @JayZTwoCents subbed a year ago and catching up on all the stuff I missed!
Just mod your case and put the radiator OUTSIDE far enough from the case, with a good fan under it to push (NOT attached to the radiator, literally UNDER the radiator, at least 5cm) and you are done. I did it with my case, i just had to make some holes with my drill and installed 2 metallic staffs. It works perfectly.
Using 32mm - 38mm thick fan shrouds on both fans in push/pull will help eliminate fan vortex and will drop the temps by a couple of degrees over just push/pull.
(Push)>Fan>Shroud>Rad>Shroud>Fan>(Pull)
Even just one shroud will help temps
(Push) >Fan>Shroud>Rad>Fan>(Pull)
this was the coolest vid I've ever seen about fans. super informative!
Thx Jay as always... Showing an "Air Flow" fan result with the others while somewhat obvious, would have been interesting to see...
Jay,
.... Welllll ... I have a small quandry.... this particular video answered some of my questions - but not all of them ....
First .... my setup:
RYZEN 7 -2700x
Asus ROG Crosshair 7 Hero - wifi
32gb - G Skill Trident Z RGB - DDR4 3200Mhz 16gb sticks ~ x2
EVGA RTX 2080 Ti FTW3 ~ x2 / NV Link ~ 11G-P4-2487-KR
Asus ROG Ryujin 360 AIO
Asua ROG THOR 1200w Platinum PSU
Panteks Enthoo Luxe case
Fans: (my final selection based on your reply)
Case front :
Noctua NF-A14 industrialPPC-2000 IP67 PWM - 31.5 dB/A ~ 182.5 m3/h ~ 4.18 mm H2O
*** OR ***
Noctua NF-A14 industrialPPC-3000 PWM - 41.3 dB/A ~ 269.3 m3/h ~ 10.52 mm H2O
Case rear:
Noctua NF-P14s redux 1200 PWM - 19.6 dB/A ~ 110.3 m3/h ~ 1.29 mm H2O ** for positive pressure ** (is that correct?? ~ lower CFM at exhaust??}
RYUJIN AIO : Noctua NF-F12 industrialPPC-2000 PWM - 29.7 dB/A ~ 121.8 m3/h ~ 3.94 mm H2O ** installed in top in PUSH **
.....I've been looking at some pretty ARGB fans for the inside of the case ....like the corsair HD / LL / ML / SP Series of 120mm fans .....
they put out around 60-75 CFM at ~1200-1750rpm
to use in PUSH and then move the supplied Noctua iPPC fans on the AIO to PULL ...... (should I do that?)
OKAY....I DIGRESS...
My question is this : In order to maintain POSITIVE pressure in my case ....
how should I set up the radiator ???
is my theory sound ???
What GENERALLY type of fan would you use for a push/pull when rad is an exhaust???
Should you always use 2 equal fans or enact a positive pressure setup for a push/pull when its an exhaust ??
JAY ....PLEASE HELP ....thanx
~Cyliss
Trust Jay to make fan testing methodology look phallic xD
I like the 370z wallpaper on the background though
I'm not sure, but you can be getting a Venturi effect because of the highest part of the radiator, that slows down the air, so it can't travel at the same speed as the fan without radiator. Making the air at the top to gain pressure.
I think the results can vary if your radiator is horizontal.
The push-pull difference you have, may be just because of the construction of the fans, they might not be designed for high pressure loads
Hey Jay, great video as usual! You should've tried the smoke thing to see how the air flows during each configuration. I believe it would have solved the mistery between the push and pull setups. I personally find your explanation quite convincing though. Keep it up!
Great video, still helpful years later 👍
Even if there is a vortex in the pull config, I suspect it makes little difference to airflow.
i think... because the air is being pulled through the rad fins, the air became lamina, but then hit the fan blade which caused a low pressure in the center of the fan... so then cause the paper was placed off set air was traveling over the back of the paper towards the low pressure at the center of the fan
Excellent testing, thanks! My thoughts are that in a pull config you get a more even airflow through the rad as there's a low pressure void between the fan and fins, and possibly a dead spot in the center of the rad on a push?. Also at high rpm pull fan it's almost beggining to act as a centrifugal fan, it's flinging air outwards instead of flowing axially. (I'd wager you'd get more pressure across the fan itself with the intake blocked vs the outlet). It would be really amazing to see a fog system to see all the cool airflow patterns, nice idea.
Fans are made to blow, not suck. Putting restriction on the inlet side causes surging and stalling as fan blade air requirements exceed supply and you get pressure pulses bouncing off each other as the air moves back forth into the low pressure "vacuum" between the fan and the radiator.
I think the suction you are getting is a result of the resistance of the radiator. Imagine if the radiator was a solid wall instead. No air would be able to come in from that side, and all of the air that the fan blades are pushing out has to come from somewhere. I imagine the air would come from the small gap between the blades, and the cowling. The radiator is not a solid wall, but does create resistance and it would coerce some air to be sucked in through that thin gap between the blades and cowling.
thank you, I actually commented on another of your videos about this subject as I am buying and trying to figure out to do push/pull or push&pull for an "ultra" quiet setup. I still don't know what kind of audio difference I will get but from other readings, the ability to run much lower rpm even with double the fans should still be far lower in dB than a single side fan at faster speeds.
I’m not really sure why, but if your radiator and fan we’re not, itright against one another and you had a standoff with a solid ring gaskets against your radiator to hold your fan blades at least an inch out from your radiator your air flow would increase by 100% on a pull configuration
Try it with fan spacers so there is less shrouding from the motors and corners so more of the radiator is actually cooled.
The big question is cooling versus noise. The results at 12:07 indicate that push and pull at 1600 RPM both do just a little more cooling than push+pull at 1000 RPM. This graph: i.imgur.com/jP01d6Fl.jpg indicates that that particular Corsair fan produces less than 31 dB at 1000 RPM, and less than 40 dB at 1500 RPM. That difference of 9 dB is equivalent to 3 doublings in acoustic energy (e.g., using 8 times as many fans). So a push/pull at 1000 RPM cools nearly as much as push or pull at 1600 RPM, but the acoustic energy at just 1500RPM is already 4 times as much as that produced by 2 fans at 1000 RPM. The results comparing 1600 RPM to 2200 RPM give a similar result.
TL/DR: Push/pull wins big-time if your main concern is fan noise. (Will win for energy consumption, too, if that's proportional to acoustic energy produced.)
Although the fans spin at the same speed, the fan on the left has greater resistance pulling air, which reduces its airflow slightly to the paper is not at the same height.
Pushing air has no initial resistance, it exhausts through the radiator creating "splash" as you call it! This oscillating air creates additional friction allowing for better cooling in most cases, as it increases the surface area.
The key is to pull that oscillating hot air out quickly. By adding a pull fan to the Push config you get the best of both worlds, - it have be further optimised if the fans run a different RPM's.
the suction on the radiator is being magnified by the resistance and the airflow over the paper is a low resistance path for the extra airflow to equalise that huge surface area's negative pressure bubble
the paper os badicaly in equilibrium and has become a ramp for the air because it's blocking the flow of the fans forward direction, to the negative direction air presses on the ramp pf paper at a rate that depends on the dimensions of the radiator and the fan pressure
Exactly the kind of video I wanted to see next. Thanks Jay!
When the push is alone, it has unrestricted airflow to insert into the radiator, and then the pressure rises inside it, when pull, you do a negative pressure inside the radiator, having an unrestricted exaust, but its actually bottlenecked by the radiator, because the air goes in small blocks in straigh line, instead of a tornado shaped cone.
I would like to see if doing some spacing makes a difference, because, if the pull had more space to generate that layer of negative pressure, would be the same.
We F1 levels of aero now.