The Physics Involved with Intake Design
Registered User
Joined: Oct 2000
Posts: 7,134
Likes: 3
From: Columbia, MD
I'm using the filter from (I believe) MP Tuning...sold by timg (?). I never took the time to fully look it over, but if memory serves, it's a fine mesh screen (about the size of a typical screen door) surrounding a gauze filter. Supposedly it can be cleaned by blowing compressed air form the inside out, but I can't say I've ever done anything more than spray it with water while still on the pipe. I just wonder how much stuff it's letting in...
Hey, that's my day job...I'm allowed weekends off 
Oroginally posted by tenblade2001
You have a shift light to make 1ST!!!!!
You have a shift light to make 1ST!!!!!
Registered User
Joined: Oct 2002
Posts: 1,871
Likes: 0
From: Scottsdale
The company I work for always warns people about spraying water on the filter. As a matter of fact when I am in front of customers, I am usually pounding on the table to make this point.
First, you can actually drive the particulate into the filter media by direct impingement of the water on the particulate on the surface of the filter.
Second, water has a tendency to wick into the filter media - it's a surface tension effect - and carry particulate with it into the pore structure.
The effect of both of these is to reduce the permeability of the filter media.
I would recommend that you use clean (oil free) compressed air from the inside to clean the filter. Another way to clean the filter would be to force clean (filtered with a filter that has a smaller pore size than the air filter) water through the filter media from inside the element. Both these approaches will do a pretty good job of cleaning the filter, but the water wash does it better.
Just make sure to dry the filter well before using it.
First, you can actually drive the particulate into the filter media by direct impingement of the water on the particulate on the surface of the filter.
Second, water has a tendency to wick into the filter media - it's a surface tension effect - and carry particulate with it into the pore structure.
The effect of both of these is to reduce the permeability of the filter media.
I would recommend that you use clean (oil free) compressed air from the inside to clean the filter. Another way to clean the filter would be to force clean (filtered with a filter that has a smaller pore size than the air filter) water through the filter media from inside the element. Both these approaches will do a pretty good job of cleaning the filter, but the water wash does it better.
Just make sure to dry the filter well before using it.
Registered User
Joined: Oct 2002
Posts: 1,871
Likes: 0
From: Scottsdale
OK I said I would do some checking when I got to the office, so here we go. With the following assumptions:
inlet air temp =150 F
WOT should pull 9000 liters/min (correct me if I am wrong) at 9000 rpm (9000 rpm*2 liter engine*1intake/2 revolutions)
ID of inlet - 3"
The velocity of air in the inlet pipe is 107 ft/sec. This is a pretty good velocity - process piping typically is designed for 30 - 100 ft/sec velocity, but you could certainly go much higher at the expense of pressure drop.
Using this velocity, and the viscosity and density at 150 F, I get a Reynolds number of 128,000 which is WAY turbulent. The laminar to turbulent transition occurs at a Reynolds number of about 2100. Reynolds number is directly proportional to velocity, so it will vary accordingly with RPM (I think). I'm not sure how air flow varies with RPM, but it should be according to the above formula (at least in my simple mind).
The dp of 1 ft of pipe at these conditions is:
(4fl/d)*v^2/2*Gc*density (I have assumed incompressible flow for simplicity)
Gc is a constant that is 32.174 with a bunch of dimensions like pounds mass, pounds force, seconds and feet.
Sorry I am old school and have to use English units just cause my brain works that way.
So (4*(f)*1 ft/.25 ft)*107^2/2*(32.174)*.0651
f (worst case) is about .0075
density is .0651 lb/cu. ft.
so I get about 0.27 inches water guage dp per foot of inlet pipe. 0.27 inches w.g. is about .01 psi.
To compare, a 90 degree pipe bend (3") has a pressure drop of about .016 psi at these same conditions.
I would guess that the pressure drop though the filter at WOT is on the order of about 0.3 psi or about 9 in w.g. - That is just a guess though, one would need the filter surface area, and permeability to calculate the dp, neither of which I have.
Don't know if this gives anyone any insight into the inlet design, but it was kind of fun for me - These are thngs I have been meaning to calculate on the S2000 for awhile - so now I have done it.
One question on my mind at this point is then when you open the OEM air box, there is a divider right in front (to the left as you stand in front of the car) of the filter, with a fairly small slot at the top to let air pass from the inlet chamber to the filter chamber. Does anyone have any idea why they put that divider there? I don't think it is for mechanical reasons (i.e., support) I think they did it for some other reason.
Come to think of it- that box probably sees pretty high suction during WOT, so maybe it is for mechanical reasons. It just seems that it would be pretty restrictive for air flowing from the inlet chamber to the filter.
Well enough mental masturbation for the day, I really do have work that I need to do.
inlet air temp =150 F
WOT should pull 9000 liters/min (correct me if I am wrong) at 9000 rpm (9000 rpm*2 liter engine*1intake/2 revolutions)
ID of inlet - 3"
The velocity of air in the inlet pipe is 107 ft/sec. This is a pretty good velocity - process piping typically is designed for 30 - 100 ft/sec velocity, but you could certainly go much higher at the expense of pressure drop.
Using this velocity, and the viscosity and density at 150 F, I get a Reynolds number of 128,000 which is WAY turbulent. The laminar to turbulent transition occurs at a Reynolds number of about 2100. Reynolds number is directly proportional to velocity, so it will vary accordingly with RPM (I think). I'm not sure how air flow varies with RPM, but it should be according to the above formula (at least in my simple mind).
The dp of 1 ft of pipe at these conditions is:
(4fl/d)*v^2/2*Gc*density (I have assumed incompressible flow for simplicity)
Gc is a constant that is 32.174 with a bunch of dimensions like pounds mass, pounds force, seconds and feet.
Sorry I am old school and have to use English units just cause my brain works that way.
So (4*(f)*1 ft/.25 ft)*107^2/2*(32.174)*.0651
f (worst case) is about .0075
density is .0651 lb/cu. ft.
so I get about 0.27 inches water guage dp per foot of inlet pipe. 0.27 inches w.g. is about .01 psi.
To compare, a 90 degree pipe bend (3") has a pressure drop of about .016 psi at these same conditions.
I would guess that the pressure drop though the filter at WOT is on the order of about 0.3 psi or about 9 in w.g. - That is just a guess though, one would need the filter surface area, and permeability to calculate the dp, neither of which I have.
Don't know if this gives anyone any insight into the inlet design, but it was kind of fun for me - These are thngs I have been meaning to calculate on the S2000 for awhile - so now I have done it.
One question on my mind at this point is then when you open the OEM air box, there is a divider right in front (to the left as you stand in front of the car) of the filter, with a fairly small slot at the top to let air pass from the inlet chamber to the filter chamber. Does anyone have any idea why they put that divider there? I don't think it is for mechanical reasons (i.e., support) I think they did it for some other reason.
Come to think of it- that box probably sees pretty high suction during WOT, so maybe it is for mechanical reasons. It just seems that it would be pretty restrictive for air flowing from the inlet chamber to the filter.
Well enough mental masturbation for the day, I really do have work that I need to do.
Registered User
Joined: Feb 2001
Posts: 2,895
Likes: 1
From: You wish
Couple real world reference points for you folks to ponder.
1. We did extensive testing on Slick Rick's S2000 in developing the "ideal" intake length for his combination. This consisted of basically straight section of 3" piping (with a small bend to clear obstacles) to which we attached various additional lengths of piping topped off by a theoretically ideal air horn entry. As we made the tubes shorter and shorter, there was a natural increase in the primary resonant frequency and the corresponding torque peak. Overlaying the dyno graphs was quite interesting as it basically looked like a bunch of sine wave peaks about 90 degrees out of phase.
Now, this result was not unexpected, and without giving too much away, we were trying to get the primary resonant peak to move high enough to help us enter VTEC as early as possible and broaden the torque curve. However, once we got the primary peak high enough, we began to notice a significant drop in torque throughout the 6500-7500 rpm range and a less significant, but still repeatable drop near the power peak. IOW, there were other order effects of resonance that were impacting performance at higher rpms - they are not as noticeable, but they do have an impact. So, barring a sliding tube intake pipe, what to do? You'll have to resolve that yourselves.
2. In datalogging intake manifold pressure, we can often spot when an engine is intake restricted, whether it be by filter clogging, insufficient intake pipe flow, or too small an intake manifold. We can also see the resonant peaks caused by intake pipe effects. These peaks will usually be in the vicinity of 0.2-04 psi depending on various factors. The typical S2000 will also show an increase in manifold vacuum as revs climb (decrease in pressure). Between 2000 and 7000 rpm, the vacuum increase is only about 0.1 psi. Between 7000 and 8500 rpm the vacuum will climb another 0.2 psi or more depending on filter condition, intake, etc. (aftermarket intakes do show a little less restriction when new than the stock airbox). With heavily modified engines (read headwork, cams, etc.) we've seen vacuum increase by as much as 0.5-0.7 psi at high rpm while retaining the stock intake manifold - clearly there is a restriction there at high power levels when normally aspirated.
3. Box type intakes (Comptech for example) do not show noticeably more restriction when placed on stock engines. But on modified engines they begin to cause problems. On heavily modified B-series engines, we've seen pressure drops cut in half when switching from an Icebox style intake to a more traditional open filter pipe setup (still drawing cool air of course).
UL
1. We did extensive testing on Slick Rick's S2000 in developing the "ideal" intake length for his combination. This consisted of basically straight section of 3" piping (with a small bend to clear obstacles) to which we attached various additional lengths of piping topped off by a theoretically ideal air horn entry. As we made the tubes shorter and shorter, there was a natural increase in the primary resonant frequency and the corresponding torque peak. Overlaying the dyno graphs was quite interesting as it basically looked like a bunch of sine wave peaks about 90 degrees out of phase.
Now, this result was not unexpected, and without giving too much away, we were trying to get the primary resonant peak to move high enough to help us enter VTEC as early as possible and broaden the torque curve. However, once we got the primary peak high enough, we began to notice a significant drop in torque throughout the 6500-7500 rpm range and a less significant, but still repeatable drop near the power peak. IOW, there were other order effects of resonance that were impacting performance at higher rpms - they are not as noticeable, but they do have an impact. So, barring a sliding tube intake pipe, what to do? You'll have to resolve that yourselves.
2. In datalogging intake manifold pressure, we can often spot when an engine is intake restricted, whether it be by filter clogging, insufficient intake pipe flow, or too small an intake manifold. We can also see the resonant peaks caused by intake pipe effects. These peaks will usually be in the vicinity of 0.2-04 psi depending on various factors. The typical S2000 will also show an increase in manifold vacuum as revs climb (decrease in pressure). Between 2000 and 7000 rpm, the vacuum increase is only about 0.1 psi. Between 7000 and 8500 rpm the vacuum will climb another 0.2 psi or more depending on filter condition, intake, etc. (aftermarket intakes do show a little less restriction when new than the stock airbox). With heavily modified engines (read headwork, cams, etc.) we've seen vacuum increase by as much as 0.5-0.7 psi at high rpm while retaining the stock intake manifold - clearly there is a restriction there at high power levels when normally aspirated.
3. Box type intakes (Comptech for example) do not show noticeably more restriction when placed on stock engines. But on modified engines they begin to cause problems. On heavily modified B-series engines, we've seen pressure drops cut in half when switching from an Icebox style intake to a more traditional open filter pipe setup (still drawing cool air of course).
UL
Registered User
Joined: Oct 2002
Posts: 1,871
Likes: 0
From: Scottsdale
Now, this result was not unexpected, and without giving too much away, we were trying to get the primary resonant peak to move high enough to help us enter VTEC as early as possible and broaden the torque curve. However, once we got the primary peak high enough, we began to notice a significant drop in torque throughout the 6500-7500 rpm range and a less significant, but still repeatable drop near the power peak. IOW, there were other order effects of resonance that were impacting performance at higher rpms - they are not as noticeable, but they do have an impact. So, barring a sliding tube intake pipe, what to do? You'll have to resolve that yourselves.
Do you have any idea where the restriction is in the oem airbox. I would speculate that it is in one of two places. 1. The bend in the runner just behind the airbox. 2. The shield/stiffener in front of the filter in the airbox. In addition, I'm not totally happy with the airbox inlet - I think it is pretty small for the air flows which can be generated.
Registered User
Joined: Feb 2001
Posts: 2,895
Likes: 1
From: You wish
I have my suspicions about box style intakes. In theory, there is nothing wrong with them. But in practice, for space savings, etc. manufacturers tend to wrap them around the filter like a glove on a hand. This tends to disturb airflow. Ever noticed how the dirt on a stock S2000 intake filter tends to accumulate in certain areas? Ideally you'd need a much larger box to provide plenty of free space around the pipe inlet (a good airhorn please). And the larger the box, the smaller the pressure variations with each cylinder intake event. There is a reason why motorcycle manufacturers keep striving to place bigger and bigger airboxes in their bikes. Hell, some 600cc bikes these days are running 12-14 liter airboxes (or even bigger). In the meantime, the S2000 intake plenum displaces well under 2 liters, and the airbox can't be much bigger than 10-12 itself.
UL
UL
Registered User
Joined: Oct 2002
Posts: 1,871
Likes: 0
From: Scottsdale
has anyone removed the shield/stiffener in the OEM airbox? Have they noticed any change (increase or decrease) in performance?
Interesting point about air box size - now I have to think about that. The only way to know what is going on is to do computer modeling or some kind of visual flow modeling (i.e., build a model out of clear plastic and watch what happens).
Gee if we just had $25000 to get someone to do a CFD model in Fluent. (3D flow modeling program)
Interesting point about air box size - now I have to think about that. The only way to know what is going on is to do computer modeling or some kind of visual flow modeling (i.e., build a model out of clear plastic and watch what happens).
Gee if we just had $25000 to get someone to do a CFD model in Fluent. (3D flow modeling program)
