About the air filter test shown on this site, please consider the following:
A "high flow" air filter may produce a significant reduction in the pressure drop across the filter on one car and may not produce much of an effect on another. It is my opinion that they probably do very little for most cars, but to determine that they don't do anything on all cars may be inaccurate. Here is why. A lot of cars use similar sized panel filters. Example - my
4.0 Aurora V8 and 3.8 series II V6 (Impala) use the same size panel. I think GM's 3.1 is about the same size panel filter as well.
However, not all cars "breath" the same. Clearly, the much higher revving larger
displacement 4.0 V8 will be pulling more air (volume) at the top end than the other engines mentioned. Now, please be aware that the "losses" experienced through the filter are a function of the air velocity to an exponent
(probably squared or close to it ^2). So as the approach velocity increases, the losses across the filter increase EXPONENTIALLY. Also
consider that with most cars, the entire filter area is NOT utilized due to the design of the air box. Most often, a SMALL FRACTION of the filter area is used thus increasing the velocity through the filter. This is done to minimize resonations in the intake and keep operation quiet for the general public's desires. Sure, a full flowing box that uses the entire filter area could be designed that uses double wall construction for sound insulation and the approach and exit flow paths could be extended for a smooth expansion of the flow approach and exit to truly utilize the flow area but that would be expensive if not impractical for design. So, for some cars, V^2 may not amount to much even under top rpms at WOT, but on others is may be significantly more.
Additionally, the pressure drop across the filter is also a function of some constant, or "K" to multiply the V^2 portion of the loss equation. From the simplest demonstrations (yes K&N demo) it can be seen that for a constant flow rate across the same filter area, the K&N type media has a lower loss coefficient - a lower "K" value to be multiplied by the V^2 portion of the equation. So, with some cars (probably most or many IMHO) a moderate or lower approach velocity squared (^2) multiplied by the higher "K" coefficient of a paper filter will not be much more than that for a free flow media. But it is entirely possible that with some cars, the higher approach velocity squared is significant and multiplying it by the lower coefficient may help.
Also, (just an additional thought - not really sure on this) as the velocity through the filter media itself actually increases, the "K" loss coefficient might itself change. Thus giving more advantage to a free flowing media. I see this in practice with losses through various elements. The loss coeficient will be accurate only for a given flow velocity range. Higher velocities require a different coefficient.
Therefore, I submit to you that while many cars may not perform or "breath" enough to create a high intake velocity and experience the steep portion of the pressure drop curve, some probably do and could benefit from a free flowing media - due to the reasons explained above.
The tests shown on this site used a manometer to measure the pressure drop and I suspect that the tubing used was long tubing to extend from the engine compartment to the cabin - my guess. If so, that could easily be too long for a manometer to function accurately. It's in the engineering books. Perhaps the compressible nature of the air in the tube oven an extended length could dampen and skew the results. I also question
the effect of under-hood temperatures on the manometer tube/air. Also, the placement of the manometer in the air box could be critical. I think the air flow in the box could be quite turbulent and placing the tube in an "eddy" location could skew results.
In short, even if this test was performed flawlessly, it is not likely representative of all cars in any way.
Like the guy that did these tests, I'm an engineer. I have a professional engineers license in Illinois and Indiana and have practiced for 14 years. Not to be nasty, but I would have also known better (even if not an engineer) that tests performed on a filter to test filtering efficiency (as shown on this site) must be done by weight under strictly controlled
conditions with the weight of the dirt fed to the filter known and the weight captured known. I'd
never offer a visual test as evidence of anything at all. No, I don't work for an air filter company or an automotive company. I work for a
consulting engineering firm and do a lot of municipal infrastructure work.
By the way, I've used a K&N type filter for a couple of years. I've sent several oil samples in for analysis and the dirt and wear numbers have been extremely low, so I'm pretty confident that my intake filtering is fine. Also, I've dyno'd (back to back same day) a K&N on an Aurora on several different visits and always saw a consistent +6 to +7 HP increase at the wheels
- every time consistently. Admittedly it was top end only and probably proves that if my car had a little less displacement or revved a bit lower (not red-line at 6200 but say 5600), maybe I'd have seen little gain at all. Again, some cars may see a gain on the top end, many others won't see anything.
I'm just afraid that there is a lot of well-meaning experimenting going on without being aware of the myriad of variables. Some possibly inaccurate and definitely broad based conclusions are being made as well.
A "high flow" air filter may produce a significant reduction in the pressure drop across the filter on one car and may not produce much of an effect on another. It is my opinion that they probably do very little for most cars, but to determine that they don't do anything on all cars may be inaccurate. Here is why. A lot of cars use similar sized panel filters. Example - my
4.0 Aurora V8 and 3.8 series II V6 (Impala) use the same size panel. I think GM's 3.1 is about the same size panel filter as well.
However, not all cars "breath" the same. Clearly, the much higher revving larger
displacement 4.0 V8 will be pulling more air (volume) at the top end than the other engines mentioned. Now, please be aware that the "losses" experienced through the filter are a function of the air velocity to an exponent
(probably squared or close to it ^2). So as the approach velocity increases, the losses across the filter increase EXPONENTIALLY. Also
consider that with most cars, the entire filter area is NOT utilized due to the design of the air box. Most often, a SMALL FRACTION of the filter area is used thus increasing the velocity through the filter. This is done to minimize resonations in the intake and keep operation quiet for the general public's desires. Sure, a full flowing box that uses the entire filter area could be designed that uses double wall construction for sound insulation and the approach and exit flow paths could be extended for a smooth expansion of the flow approach and exit to truly utilize the flow area but that would be expensive if not impractical for design. So, for some cars, V^2 may not amount to much even under top rpms at WOT, but on others is may be significantly more.
Additionally, the pressure drop across the filter is also a function of some constant, or "K" to multiply the V^2 portion of the loss equation. From the simplest demonstrations (yes K&N demo) it can be seen that for a constant flow rate across the same filter area, the K&N type media has a lower loss coefficient - a lower "K" value to be multiplied by the V^2 portion of the equation. So, with some cars (probably most or many IMHO) a moderate or lower approach velocity squared (^2) multiplied by the higher "K" coefficient of a paper filter will not be much more than that for a free flow media. But it is entirely possible that with some cars, the higher approach velocity squared is significant and multiplying it by the lower coefficient may help.
Also, (just an additional thought - not really sure on this) as the velocity through the filter media itself actually increases, the "K" loss coefficient might itself change. Thus giving more advantage to a free flowing media. I see this in practice with losses through various elements. The loss coeficient will be accurate only for a given flow velocity range. Higher velocities require a different coefficient.
Therefore, I submit to you that while many cars may not perform or "breath" enough to create a high intake velocity and experience the steep portion of the pressure drop curve, some probably do and could benefit from a free flowing media - due to the reasons explained above.
The tests shown on this site used a manometer to measure the pressure drop and I suspect that the tubing used was long tubing to extend from the engine compartment to the cabin - my guess. If so, that could easily be too long for a manometer to function accurately. It's in the engineering books. Perhaps the compressible nature of the air in the tube oven an extended length could dampen and skew the results. I also question
the effect of under-hood temperatures on the manometer tube/air. Also, the placement of the manometer in the air box could be critical. I think the air flow in the box could be quite turbulent and placing the tube in an "eddy" location could skew results.
In short, even if this test was performed flawlessly, it is not likely representative of all cars in any way.
Like the guy that did these tests, I'm an engineer. I have a professional engineers license in Illinois and Indiana and have practiced for 14 years. Not to be nasty, but I would have also known better (even if not an engineer) that tests performed on a filter to test filtering efficiency (as shown on this site) must be done by weight under strictly controlled
conditions with the weight of the dirt fed to the filter known and the weight captured known. I'd
never offer a visual test as evidence of anything at all. No, I don't work for an air filter company or an automotive company. I work for a
consulting engineering firm and do a lot of municipal infrastructure work.
By the way, I've used a K&N type filter for a couple of years. I've sent several oil samples in for analysis and the dirt and wear numbers have been extremely low, so I'm pretty confident that my intake filtering is fine. Also, I've dyno'd (back to back same day) a K&N on an Aurora on several different visits and always saw a consistent +6 to +7 HP increase at the wheels
- every time consistently. Admittedly it was top end only and probably proves that if my car had a little less displacement or revved a bit lower (not red-line at 6200 but say 5600), maybe I'd have seen little gain at all. Again, some cars may see a gain on the top end, many others won't see anything.
I'm just afraid that there is a lot of well-meaning experimenting going on without being aware of the myriad of variables. Some possibly inaccurate and definitely broad based conclusions are being made as well.
