Static pressure or airflow fans?

[Y33]

I'm looking to add an intake fan to my case (NZXT S340) and have my eyes on the Corsair AF140 and Corsair SP140 fans. One is airflow optimised and the other static pressure optimised.

Which one should I get? There are no obstructions like drive cages in the case but there will be a dust filter in front of the fan.

[Blackhole890]

intaking, i'd get pressure fans because dust filters so you'd be making sure that the air will enter properly

[wzrd]

3 minutes ago, ylst said:

I'm looking to add an intake fan to my case (NZXT S340) and have my eyes on the Corsair AF140 and Corsair SP140 fans. One is airflow optimised and the other static pressure optimised.

Which one should I get? There are no obstructions like drive cages in the case but there will be a dust filter in front of the fan.

Get the AF140s. Dust filter doesn't provide significant air resistance compared to a radiator - they are designed to provide as little air resistance as possible, and will run quieter at a given RPM because of their blade design. You can also run them at lower RPMs and get the same air flow, so your system will be quieter.

[Y33]

2 minutes ago, wzrd said:

Get the AF140s. Dust filter doesn't provide significant air resistance compared to a radiator - they are designed to provide as little air resistance as possible, and will run quieter at a given RPM because of their blade design. You can also run them at lower RPMs and get the same air flow, so your system will be quieter.

Ah great! I like my system quiet so that's good to hear. Thanks.

[wzrd]

Just now, ylst said:

Ah great! I like my system quiet so that's good to hear. Thanks.

There is specifically a "quiet edition" of the AF140s I believe. I would assume that would be the one to go for in that situation.

[Y33]

Just now, wzrd said:

There is specifically a "quiet edition" of the AF140s I believe. I would assume that would be the one to go for in that situation.

Yes and those are on sale as well so that's nice. Still having doubts on whether I should get a one or two pack. Already have 2 fans in my system so getting another pair would be a bit of an overkill I think.

[wzrd]

Just now, ylst said:

Yes and those are on sale as well so that's nice. Still having doubts on whether I should get a one or two pack. Already have 2 fans in my system so getting another pair would be a bit of an overkill I think.

The more fans you have, the slower you can run them, so the less noise in theory. Whether or not it is worth it in terms of cooling depends on the heat load of the system you have installed.

[WoodenMarker]

23 minutes ago, wzrd said:

Get the AF140s. Dust filter doesn't provide significant air resistance compared to a radiator - they are designed to provide as little air resistance as possible, and will run quieter at a given RPM because of their blade design. You can also run them at lower RPMs and get the same air flow, so your system will be quieter.

With the front of the S340, a bit more static pressure helps to pull air in. SP140's would perform a bit better in this case.

 

@ylst Neither are the best fans you can use though. Why these fans in particular?

[Y33]

18 minutes ago, WoodenMarker said:

With the front of the S340, a bit more static pressure helps to pull air in. SP140's would perform a bit better in this case.

 

@ylst Neither are the best fans you can use though. Why these fans in particular?

Well, a combination of price, Corsair being a brand I know and trust and looks (white LED's) led me to these. What would your suggestion be?

[wzrd]

41 minutes ago, WoodenMarker said:

With the front of the S340, a bit more static pressure helps to pull air in. SP140's would perform a bit better in this case.

 

@ylst Neither are the best fans you can use though. Why these fans in particular?

@WoodenMarker Actually, that is a common misconception when it comes to dust filters.

 

Assuming that the fan filter has a weight of 20g (overestimate, as it is the weight of an entire silverstone fan filter, including the metal frame), and a thickness of 2mm (again an overestimate), and assuming that the material of the filter made up half of the area of the filter (again a massive overestimate, worst case scenario), it would provide an air resistance of only 2mN across the entire fan, even at a grossly overestimated air flow speed of 50m/s.

 

Given that an AF140 has a static pressure rating of 8.23N/m according to the Corsair website, and therefore of approx. 420N across the fan radius, the dust filter provides a resistance of 0.002/420 or 0.00047619% of its static pressure rating.

 

If we apply the same calculation to an SP140 fan, with a rating of 11.47N/m, and therefore approx. 585N across the fan radius, it provides a resistance of 0.002/585 or 0.00034188% of its static pressure rating.

 

In summary, dust filters provide air flow restriction that is almost negligible. In further summary, I need to get outside more.

 

EDIT: Not trying to be confrontational, just pointing out a common incorrect assumption. What you're saying makes sense from a logical perspective, but the thickness of dust filters compared to, say, a radiator means the restriction is actually minimal.

 

EDIT AGAIN: Should also point out that, in reality, the resistance is actually probably much less, as I hugely overestimated all of the figures.

[Y33]

@wzrd @WoodenMarker might also be referring to the way the S340 intakes air. Since there aren't really any vents in front of the case it uses a slit on top and below to suck in air. It's spaced about 3cm away from the filter. (see image)

I currently do not have any intake fans and still feel air being sucked in so I didn't think this would be much of an issue (hence it not being in the original post).

[wzrd]

Just now, ylst said:

@wzrd @WoodenMarker might also be referring to the way the S340 intakes air. Since there aren't really any vents in front of the case it uses a slit on top and below to suck in air. It's spaced about 3cm away from the filter. (see image)

I currently do not have any intake fans and still feel air being sucked in so I didn't think this would be much of an issue (hence it not being in the original post).

The 3cm gap in front of the filter is even less of a factor actually. The relative volume of air available to the fan, as well as the suction exerted over the slits grossly exceeds what the fan would be capable of moving, i.e. the fan would need to be running at an insane amount of RPM to exceed the volume of air available to it; the airflow capability of the fan itself is still going to be the limiting factor.

[WoodenMarker]

36 minutes ago, wzrd said:

@WoodenMarker Actually, that is a common misconception when it comes to dust filters.

Assuming that the fan filter has a weight of 20g (overestimate, as it is the weight of an entire silverstone fan filter, including the metal frame), and a thickness of 2mm (again an overestimate), and assuming that the material of the filter made up half of the area of the filter (again a massive overestimate, worst case scenario), it would provide an air resistance of only 2mN across the entire fan, even at a grossly overestimated air flow speed of 50m/s.

Given that an AF140 has a static pressure rating of 8.23N/m according to the Corsair website, and therefore of approx. 420N across the fan radius, the dust filter provides a resistance of 0.002/420 or 0.00047619% of its static pressure rating.

If we apply the same calculation to an SP140 fan, with a rating of 11.47N/m, and therefore approx. 585N across the fan radius, it provides a resistance of 0.002/585 or 0.00034188% of its static pressure rating.

In summary, dust filters provide air flow restriction that is almost negligible. In further summary, I need to get outside more.

EDIT: Not trying to be confrontational, just pointing out a common incorrect assumption. What you're saying makes sense from a logical perspective, but the thickness of dust filters compared to, say, a radiator means the restriction is actually minimal.

Dust filters make a significant difference in airflow. It's not abnormal to lose >50% airflow when using them.

Here's an example of how even relatively open fan grills inhibit airflow: https://www.pugetsystems.com/labs/articles/Effects-of-Grill-Patterns-on-Fan-Performance-Noise-107/

Airflow isn't so simple that one can look at the thickness of an object and know how it would interact with a fan against it.

25 minutes ago, ylst said:

I currently do not have any intake fans and still feel air being sucked in so I didn't think this would be much of an issue (hence it not being in the original post).

From experience with a variety of fans, I found that more static pressure oriented fans performed better at the front of the S340. For example, I found that PH-F140SP's performed better than AF140 QE's when matching loudness. NF-P14S has a similar balance of airflow/static pressure to PH-F140SP and were a bit quieter though. 

Along with the results of various other fans, I extrapolated that something along the lines of the NF-P14S or PH-F140SP was a good representation of optimal airflow/static pressure for the front of the S340. This is far from comprehensive but it's an educated guess. 

22 minutes ago, wzrd said:

The 3cm gap in front of the filter is even less of a factor actually. The relative volume of air available to the fan, as well as the suction exerted over the slits grossly exceeds what the fan would be capable of moving, i.e. the fan would need to be running at an insane amount of RPM to exceed the volume of air available to it; the airflow capability of the fan itself is still going to be the limiting factor.

Any restriction like the front panel of the S340 causing airflow to change direction or a dust filter will can inhibit airflow. The size of the openings at the top and bottom aren't the only factors that inhibit performance.

37 minutes ago, ylst said:

Well, a combination of price, Corsair being a brand I know and trust and looks (white LED's) led me to these. What would your suggestion be?

Brand loyalty is one thing but are you unwillingly to explore other options?

What's your budget? Are white leds necessary? I personally prefer and recommend going with an led strip if you want case lighting.

[Y33]

@WoodenMarker No I'm completely open to other options if there's something better out there. My budget is in between 20 - 30 euros / USD.

I'm not necessarily looking for a fan with LED's, I like subtle lighting in my case and thought of them as a simple way of doing that (and given that other plug and play solutions like NZXT's Hue+ are really expensive for some lights). But again, if there's a better alternative then I'd be more than happy to look into it.

[wzrd]

23 minutes ago, WoodenMarker said:

 

Dust filters make a significant difference in airflow. It's not abnormal to lose >50% airflow when using them.

Here's an example of how even relatively open fan grills inhibit airflow: https://www.pugetsystems.com/labs/articles/Effects-of-Grill-Patterns-on-Fan-Performance-Noise-107/

Airflow isn't so simple that one can look at the thickness of an object and know how it would interact with a fan against it.

@WoodenMarker The drops measured were too small to have a practical effect, and fall inside (way inside) the margin for error of a handheld anemometer in an uncontrolled environment.

 

Actually you are incorrect, air resistance is determined by the shape of an object, its density, the air density (constant), the area of the object, and the speed at which air moves over it. All of these factors were accounted for in the calculation I provided, and actually hugely exaggerated in a lot of areas, and the realistic values would be much smaller, in all likelihood.

 

23 minutes ago, WoodenMarker said:

Any restriction like the front panel of the S340 causing airflow to change direction or a dust filter will can inhibit airflow. The size of the openings at the top and bottom aren't the only factors that inhibit performance.

It isn't affecting the direction of the airflow, because the effective suction radius of the fan is nowhere near 3cm. After that, the panel may as well not be there in terms of measurable effect.

 

For the purposes of calculation, we can look at it as a two-stage process, whereby the fan will pull air from within the 3cm gap. This causes negative pressure, pulling air into the gap from outside the case. The only way this could affect the fan, is if the air flow into the case from outside is less than the volume of air the fan pulls from the 3cm gap around it, which it most definitely isn't. I can provide calculations to that effect too.

 

23 minutes ago, WoodenMarker said:

From experience with a variety of fans, I found that more static pressure oriented fans performed better at the front of the S340. For example, I found that PH-F140SP's performed better than AF140 QE's when matching loudness. NF-P14S has a similar balance of airflow/static pressure to PH-F140SP and were a bit quieter though. 

Along with the results of various other fans, I extrapolated that something along the lines of the NF-P14S or PH-F140SP was a good representation of optimal airflow/static pressure for the front of the S340. This is far from comprehensive but it's an educated guess. 

Those are two completely different brands of fans. You can't compare static pressure optimised fans from one brand to air flow optimised fans of another, it's an apples to oranges comparison. A better quality static pressure fan will probably perform better than a lesser quality airflow fan, because of how little a factor it actually is in this scenario. I'm not saying Corsair fans are better than any other brand - far from it. I just used their AF and SP fans for consistency. What I'm demonstrating is that, all other factors being equal, that a dust filter, and/or the enclosed space of an S340.

 

There are way more factors to consider when comparing across different brands, and there are definitely better brands than Corsair out there. I don't doubt that the quality of the fan is a significant factor, but the addition of a dust filter and/or the enclosed space of the S340 is not one.

[WoodenMarker]

1 minute ago, ylst said:

@WoodenMarker No I'm completely open to other options if there's something better out there. My budget is in between 20 - 30 euros / USD.

I'm not necessarily looking for a fan with LED's, I like subtle lighting in my case and thought of them as a simple way of doing that (and given that other plug and play solutions like NZXT's Hue+ are really expensive for some lights). But again, if there's a better alternative then I'd be more than happy to look into it.

Are you just using the stock case fans and planning to add a single intake?

If that's the case, a single led intake might look a bit weird.

[Y33]

3 minutes ago, WoodenMarker said:

Are you just using the stock case fans and planning to add a single intake?

If that's the case, a single led intake might look a bit weird.

Yes, I'm using NZXT's stock fans on the rear and top exhausts and want to add an intake fan.

As for the lighting, it might indeed make more sense to go for a LED strip. Anything you could recommend?

[Y33]

@wzrd@WoodenMarker What alternative would you suggest to the Corsair ones then?

[wzrd]

Just now, ylst said:

@wzrd@WoodenMarker What alternative would you suggest to the Corsair ones then?

I use almost exclusively EK-Vardar fans, because the fans in my builds are always radiator-mounted. I have never tested air flow fans.

 

@WoodenMarker will have a better recommendation than me.

[WoodenMarker]

27 minutes ago, ylst said:

Yes, I'm using NZXT's stock fans on the rear and top exhausts and want to add an intake fan.

As for the lighting, it might indeed make more sense to go for a LED strip. Anything you could recommend?

If you're fine with the stock fans, A Silent 14 would fit right in: https://tweakers.net/pricewatch/326668/gelid-solutions-silent-14pwm-140mm.html

PW2's are quieter if that's what you plan to replace the stock fans in the future: https://tweakers.net/pricewatch/365691/be-quiet!-pure-wings-2-140mm.html

NZXT's sleeved led strip is alright if you don't mind a cool-ish white: https://tweakers.net/pricewatch/361785/nzxt-sleeved-led-kit-1-meter-wit.html

It has 3 settings and can be dimmed if you want more subtle lighting.

34 minutes ago, wzrd said:

@WoodenMarker The drops measured were too small to have a practical effect, and fall inside (way inside) the margin for error of a handheld anemometer in an uncontrolled environment.

Actually you are incorrect, air resistance is determined by the shape of an object, its density, the air density (constant), the area of the object, and the speed at which air moves over it. All of these factors were accounted for in the calculation I provided, and actually hugely exaggerated in a lot of areas, and the realistic values would be much smaller, in all likelihood.

 

It isn't affecting the direction of the airflow, because the effective suction radius of the fan is nowhere near 3cm. After that, the panel may as well not be there in terms of measurable effect.

For the purposes of calculation, we can look at it as a two-stage process, whereby the fan will pull air from within the 3cm gap. This causes negative pressure, pulling air into the gap from outside the case. The only way this could affect the fan, is if the air flow into the case from outside is less than the volume of air the fan pulls from the 3cm gap around it, which it most definitely isn't. I can provide calculations to that effect too.

 

Those are two completely different brands of fans. You can't compare static pressure optimised fans from one brand to air flow optimised fans of another, it's an apples to oranges comparison. A better quality static pressure fan will probably perform better than a lesser quality airflow fan, because of how little a factor it actually is in this scenario. I'm not saying Corsair fans are better than any other brand - far from it. I just used their AF and SP fans for consistency. What I'm demonstrating is that, all other factors being equal, that a dust filter, and/or the enclosed space of an S340.

There are way more factors to consider when comparing across different brands, and there are definitely better brands than Corsair out there. I don't doubt that the quality of the fan is a significant factor, but the addition of a dust filter and/or the enclosed space of the S340 is not one.

That seems to be a really large margin if error if that's the case. The results reflect what I've seen using different meshes. 

Here's another example: http://www.xtremesystems.org/forums/showthread.php?209801-Some-Wire-Mesh-Filter-Fan-Grill-Testing

There's a filter vs. no filter test here: https://www.reddit.com/r/buildapc/comments/2c8oum/discussion_optimizing_case_airflow_10/

It's 1-2c difference but that's relatively significant when it comes to case fans.

 

It's not a particularly popular topic so I'm having trouble finding some results I remember seeing.

 

I don't know the math behind determining performance but there are some principles of airflow that I've read up on.

The front panel of the S340 very restrictive but it does impede the airflow to some extent.

 

I didn't notice that I didn't include it in the post but I acknowledge that they aren't the same brand and aren't that similar. Even fans from the same brand can't necessarily be compared as if all that's changed is the frame and blade design. Like I said, it's an educated guess.

[wzrd]

16 minutes ago, WoodenMarker said:

If you're fine with the stock fans, A Silent 14 would fit right in: https://www.alternate.nl/Gelid-Solutions/Silent-14-(FN-SX14-10)-Case-fan/html/product/1220866?lk=9301

PW2's are quieter if that's what you plan to replace the stock fans in the future: https://www.alternate.nl/be-quiet!/Pure-Wings-2-140mm-Case-fan-/html/product/1094161?lk=9301

That seems to be a really large margin if error if that's the case. The results reflect what I've seen using different meshes. 

Here's another example: http://www.xtremesystems.org/forums/showthread.php?209801-Some-Wire-Mesh-Filter-Fan-Grill-Testing

There's a filter vs. no filter test here: https://www.reddit.com/r/buildapc/comments/2c8oum/discussion_optimizing_case_airflow_10/

It's 1-2c difference but that's relatively significant when it comes to case fans.

 

It's not a particularly popular topic so I'm having trouble finding some results I remember seeing.

 

I don't know the math behind determining performance but there are some principles of airflow that I've read up on.

The front panel of the S340 very restrictive but it does impede the airflow to some extent.

 

I didn't notice that I didn't include it in the post but I acknowledge that they aren't the same brand and aren't that similar. Even fans from the same brand can't necessarily be compared as if all that's changed is the frame and blade design. Like I said, it's an educated guess.

It is. Most handheld anemometers derive the airflow by measuring the RPM, and then calculate the airflow based on the known resistance of the installed bearing. But these characteristics change over time and with different ambient conditions. They're not really sensitive enough to properly measure the relatively small airflow of a case fan at all, their margin for error is fine for larger environments, but it's way too big for this scenario. Also, if you look at the positioning of the anemometer in the first link, for example, you'll see that its positioned at the edge of the fan. This is where most of the airflow is without a dust filter. Adding a filter disperses the air, but doesn't necessarily reduce flow rate, so that positioning is hugely biased towards an unfiltered fan.

 

It will impede airflow throughout the case in general, yes, because it will restrict the rate at which the pressure inside the case can cycle. However, it won't affect the front fans ability to move air, which is an important distinction - it will affect cooling, but is not cause to use a static pressure optimised fan. I will explain using calculations as best I can in a separate post if you wish.

 

I agree, I was only illustrating the negligible degree of resistance dust filters provide.

 

What you're saying does make sense from an educated guess perspective, and would make perfect sense on a larger scale, if air was moving at enough speed. But the rate of air movement in a case isn't large enough for these type of factors to have a real world effect. I'm not saying they don't have any effect, but as I illustrated in my earlier calculations, the effect is just nowhere near enough to have a performance application.

 

EDIT: To clarify point about anemometer: +/-10CFM (for example) is nothing when measuring a gale force wind, but is nearly 20% when measuring from case fans.

[Y33]

13 minutes ago, WoodenMarker said:

If you're fine with the stock fans, A Silent 14 would fit right in: https://tweakers.net/pricewatch/326668/gelid-solutions-silent-14pwm-140mm.html

PW2's are quieter if that's what you plan to replace the stock fans in the future: https://tweakers.net/pricewatch/365691/be-quiet!-pure-wings-2-140mm.html

NZXT's sleeved led strip is alright if you don't mind a cool-ish white: https://tweakers.net/pricewatch/361785/nzxt-sleeved-led-kit-1-meter-wit.html

It has 3 settings and can be dimmed if you want more subtle lighting.

Thanks for your suggestions! (and them linking to Dutch sites!) I think the PW2 and that led strip could be a really great combo, especially for the price!

 

And thanks to both you and @wzrd for the info on this topic.

[WoodenMarker]

48 minutes ago, wzrd said:

It is. Most handheld anemometers derive the airflow by measuring the RPM, and then calculate the airflow based on the known resistance of the installed bearing. But these characteristics change over time and with different ambient conditions. They're not really sensitive enough to properly measure the relatively small airflow of a case fan at all, their margin for error is fine for larger environments, but it's way too big for this scenario. Also, if you look at the positioning of the anemometer in the first link, for example, you'll see that its positioned at the edge of the fan. This is where most of the airflow is without a dust filter. Adding a filter disperses the air, but doesn't necessarily reduce flow rate, so that positioning is hugely biased towards an unfiltered fan.

 

It will impede airflow throughout the case in general, yes, because it will restrict the rate at which the pressure inside the case can cycle. However, it won't affect the front fans ability to move air, which is an important distinction - it will affect cooling, but is not cause to use a static pressure optimised fan. I will explain using calculations as best I can in a separate post if you wish.

I agree, I was only illustrating the negligible degree of resistance dust filters provide.

What you're saying does make sense from an educated guess perspective, and would make perfect sense on a larger scale, if air was moving at enough speed. But the rate of air movement in a case isn't large enough for these type of factors to have a real world effect. I'm not saying they don't have any effect, but as I illustrated in my earlier calculations, the effect is just nowhere near enough to have a performance application.

Yeah, they're not optimal but I'm not sure what else to point to. The first link from XtremeSystems is a bit less elegant that Puget's with an adapter to the anemometer.

 

All fans have some degree of static pressure and airflow and designing a fan for static pressure will usually mean a sacrifice to airflow and vice-versa. The best fan is usually the one that has as much airflow as possible with enough static pressure.

What's considered a static pressure fan is relative to other fans available. What I'm observing is that a certain amount of static pressure is optimal relative to the airflow provided for the use case of intake at the front of the S340. In the case of the AF140, I don't think it provides enough static pressure to take advantage of its airflow to perform better than the SP140.

 

Other people and I find the difference in performance when using a dust filter pretty apparent. How is it negligible?

Although you provided the math showing a resistance of 0.00047619%, it doesn't match up with actual results. I'd appreciate an explanation as to why the weight of a fan filter matters and an example of how much restriction a theoretical filter with the same numbers where the material covers 99% of the fan. I have a feeling the result would demonstrate an obvious disconnect between the math and actual performance.

 

I don't see how the speed of air makes it not matter. The AF140 and SP140 are examples are otherwise identical fans performing differently mainly due to the design lending towards more airflow vs static pressure.

Depending on what you consider negligible, you could boil it down to 'case fans don't matter'. To some extent, this is true but there's definitely a measurable difference in performance and acoustics between case fans. 

[wzrd]

2 hours ago, WoodenMarker said:

Yeah, they're not optimal but I'm not sure what else to point to. The first link from XtremeSystems is a bit less elegant that Puget's with an adapter to the anemometer.

 

All fans have some degree of static pressure and airflow and designing a fan for static pressure will usually mean a sacrifice to airflow and vice-versa. The best fan is usually the one that has as much airflow as possible with enough static pressure.

What's considered a static pressure fan is relative to other fans available. What I'm observing is that a certain amount of static pressure is optimal relative to the airflow provided for the use case of intake at the front of the S340. In the case of the AF140, I don't think it provides enough static pressure to take advantage of its airflow to perform better than the SP140.

Other people and I find the difference in performance when using a dust filter pretty apparent. How is it negligible?

Although you provided the math showing a resistance of 0.00047619%, it doesn't match up with actual results. I'd appreciate an explanation as to why the weight of a fan filter matters and an example of how much restriction a theoretical filter with the same numbers where the material covers 99% of the fan. I have a feeling the result would demonstrate an obvious disconnect between the math and actual performance.

 

I don't see how the speed of air makes it not matter. The AF140 and SP140 are examples are otherwise identical fans performing differently mainly due to the design lending towards more airflow vs static pressure.

 

Depending on what you consider negligible, you could boil it down to 'case fans don't matter'. To some extent, this is true but there's definitely a measurable difference in performance and acoustics between case fans. 

@WoodenMarker But even that uses a derivation and not an actual measurement, and so is still inherently flawed.

 

Well what's considered a 'static pressure fan' is completely up to the manufacturers, and so obviously varies hugely as there is no prescribed standard. Static pressure is the ability to overcome resistance, airflow is the ability to move a given weight of air, irrespective of resistance. They are not related, i.e. the AF140 does not need a given amount of static pressure to take advantage of its airflow capability. What I think you are trying to say is that it won't have enough air to feed it, which is simply not the case in the S340, though it may easily be the case in a different chassis. I can provide calculations to this effect if necessary, too.

 

It is negligible because it doesn't affect the air flow to a significant enough degree to effect performance, I will include calculations to this effect below also. I am considering the fan itself. How the panel effects the cooling of the entire case is a totally separate issue.

 

The weight of the fan filter matters as it is directly related to the density of the filter, which is a determinant of air resistance. Without going into too much detail, air resistance is partly determined by the density, as the malleability effects the reactive force it can exert on the air (for any physicists that may be reading, I realise this is a very simplistic explanation).

 

In the example I provided, the filter itself already covers 100% 140mm fan. For my calculations, I assumed the mesh was made up of 50% material, with 50% of the surface area being space for air to pass through. In reality, the material covers far less area than the space between it, but just for a worst case scenario I went as high as 50%. 

 

I said that computer case fans don't move anywhere near enough air to be restricted by something like the S340s front panel. However, if they were moving more air (a LOT more), this could eventually become the bottleneck, in theory.

 

What I consider negligible is anything that does not have an impact on performance that lies outside of the margin for error, as is the case in any scientific data. 

 

I never said that, and never disputed that there is a difference between case fans. I said that putting a dust filter over a fan does not change the performance of the fan itself, not that there is no discrepancy between different fans.

 

CALCULATIONS:

For this calculation, I will use the data provided by Puget that you referenced earlier.

 

Volume of air in main chamber of NZXT S340 = 38.48 L (had to use data for entire case as I don't have one to measure)

Unrestricted airflow of fan: 46.7 CFM = 22.04 L/s

Restricted airflow according to Puget: 41.8 CFM = 19.73 L/s 

 

For the record I am dubious of their results, but the net effect is so negligible that I will use them anyway, and it will make no difference.

 

Lets assume we have a CPU cooler with 1m² of surface area (which is insane), and it is made entirely of pure copper (best case scenario), and the CPU is at a crazy 90C, and thermal conductivity to the cooler is perfect (best case scenario), so the entire cooler is 90C, and we're at an ambient temperature of 20C (nice and cool). And, best case scenario again, the fan is intaking air at exactly ambient temperature.

 

Assuming the copper is pure, we're going to be dissipating heat at a rate of 401W/m/K. Air has a specific latent heat of approximately 555J/kg/K, meaning that it takes 555 joules to change the temperature of 1kg of air by 1C. For every 1C carried away from the cooler to the air, the temperature of the cooler itself drops by 1C, and so does the CPU (would not really happen, but is the absolute best case scenario, the thermal paste of the future ). Assuming air flow is perfect over the CPU cooler, which is perfectly square shaped for convenience, meaning we travel across the hypotenuse (which would then be route 2), giving it as long as possible to transfer heat.

 

So we are moving 22.04L/s of air without the mesh, weighing 26.2276g. Assuming all of this moves over the cooler (perfect direct flow, highly unrealistic), it would take (555x0.26276=145.83J) 145.83 joules to change the temperature by 1C. Given that we are dissipating at a rate of 401J/s, in that second in which we pass 22.04L of air through the cooler, we should change its temperature by 401/145.83 = 2.75C (give me this cooler over a dark rock any day).

 

If we apply the same calculations to our impeded fan, our volume of air weighs only 23.4787g, needing 130.31 joules to change its temperature. 

 

EDIT: For clarity, the higher this joule value, the more heat energy it can absorb before changing temperature and therefore dissipate heat faster, as rate of heat dissipation is directly related to the temperature difference between the two materials.

 

130/145 is approximately a 10% drop in performance. Put in perspective, this will affect CPU temperatures by about 0.35C at 90C and perfect heat transfer from CPU to heatsink, which is physically impossible.

 

In summary, even using the numbers provided by Puget, even with a measurable air flow difference, that does not come along with any significant decrease in cooling capability, even with the best possible heatsink imaginable. In reality, there will be efficiency loss between the CPU die and IHS, IHS and cooler, the coldplate and heatpipes, heat pipes and fins, etc., and as a result the actual dissipation of heat from the CPU core would be way less, so the difference in cooling would be way less than I demonstrated.

[WoodenMarker]

14 minutes ago, wzrd said:

But even that uses a derivation and not an actual measurement, and so is still inherently flawed.

 the AF140 does not need a given amount of static pressure to take advantage of its airflow capability. 

 

It is negligible because it doesn't affect the air flow to a significant enough degree to effect performance, I will include calculations to this effect below also. I am considering the fan itself. How the panel effects the cooling of the entire case is a totally separate issue.

 

The weight of the fan filter matters as it is directly related to the density of the filter, which is a determinant of air resistance. Without going into too much detail, air resistance is partly determined by the density, as the malleability effects the reactive force it can exert on the air (for any physicists that may be reading, I realise this is a very simplistic explanation).

 

CALCULATIONS:

For this calculation, I will use the data provided by Puget that you referenced earlier.

 

Volume of air in main chamber of NZXT S340 = 38.48 L (had to use data for entire case as I don't have one to measure)

Unrestricted airflow of fan: 46.7 CFM = 22.04 L/s

Restricted airflow according to Puget: 41.8 CFM = 19.73 L/s 

 

Lets assume we have a CPU cooler with 1m² of surface area (which is insane), and it is made entirely of pure copper (best case scenario), and the CPU is at a crazy 90C, and thermal conductivity to the cooler is perfect (best case scenario), so the entire cooler is 90C, and we're at an ambient temperature of 20C (nice and cool). And, best case scenario again, the fan is intaking air at exactly ambient temperature.

 

So we are moving 22.04L/s of air without the mesh, weighing 26.2276g. Assuming all of this moves over the cooler (perfect direct flow, highly unrealistic), it would take (555x0.26276=145.83J) 145.83 joules to change the temperature by 1C. Given that we are dissipating at a rate of 401J/s, in that second in which we pass 22.04L of air through the cooler, we should change its temperature by 401/145.83 = 2.75C (give me this cooler over a dark rock any day).

What's a derivation and not an actual measurement?

The airflow spec is in respect to zero static pressure. Any restriction will lower that amount of airflow and any fan with zero static pressure would fail to push air past any restriction.

 

I asked earlier but you didn't provide the numbers. I was referring to your calculation here:

3 hours ago, wzrd said:

assuming that the material of the filter made up half of the area of the filter 

Given that an AF140 has a static pressure rating of 8.23N/m according to the Corsair website, and therefore of approx. 420N across the fan radius, the dust filter provides a resistance of 0.002/420 or 0.00047619% of its static pressure rating.

You assumed that the filter covered half of the fan and only provided a resistance of 0.00047619% 

Let's say 99% of the fan was covered by the filter's material. How much resistance would that provide using similar calculations?

 

The dust filter is pretty much static and not malleable since it's mounted to the case. It's not moving out of the way for air.

 

Regarding your calculation here:

The results from Puget are from fan grills. Dust filters are more restrictive. The heatsink itself would also inhibit airflow.

1m² isn't insane. The 212 Evo for example has a bit over 1.1m² of fin surface.

Copper isn't necessarily a better scenario. A heatsink at copper at 90c contains more heat than an aluminum heatsink of the same size at 90c.

Seems extraneous and I don't get it.

What does that 2.75c mean? It's not like the cpu doesn't stop producing heat.