MAF Airflow Straightener Testing

[superskaterxes]

so ive been studying alot of my logs and i cant seem to get a nice smooth repeatable AFR at WOT. It differers by .1-.2 and no matter how many slight MAF curve tweaks i make its never really consistent. i have chalked this up to the 3.25" MAF housing as its significantly bigger then stock and the added area prob causes a little more turbulence with a sensor that was designed for something smaller.




so i did some googling of airflow straighteners and i came across this:

Honeycomb for MAF

this guy originally made them for PC tower fans but i guess he has an F-Body and stumbled upon an HPtuners forum of people discussing it here.

100mm Maf kickin my Butt!!! - HP Tuners Bulletin Board



here are a few picts from the thread:

here is the resulting scan before and after installing the screen Before: After: now the MAF Low table is only using 1 maybe two cells for idle. my air/fuel raiti seems more stable as well.

Here is some info they posted about how honeycomb is used in windtunnels to help make flow more laminar

Wind Tunnels - Honeycomb

The restriction of the material seems to be somewhat beneficial. Incoming air is a fluid it will take the easiest path. Which will have a normal distribution within the system like the natural flow image below. The honeycomb cell placed in the location of the previous easiest path restricts it a %, but instead of losing that flow volume, it is displaced to a previous difficult path cell. Making a more consistant flow across the entire system. I noticed this playing with a few dozen smoke bombs. Here the system is open and the smoke lazily takes the course to the fan output. With the cells in place it is all business, the air sweeps in fast, goes directly down. One of the wind tunnel sites I have read said once you "straighten" the air and remove all the tiny eddies, and the distribution is even. It is much easier to control. For corners and such.

So i bought the 1/8" 6:1 ratio in the aluminum and plastic versions to do my testing with. heres the plastic and the aluminum was seen above






will update this thread with testing when i get them, they were a total of like $19 shipped and i ordered the 96mm version and i will just cut them down. im not really worried about it getting sucked into the turbo because A) the MAF is right in front of it and B) its a pretty large piece. i may even do some before/after vacuum measurements on the inlet to see if there is really any flow restriction from this thing. i would bet the benefits far out weigh the disadvantages.

[Alpha]

Hopefully SURE addresses these issues when they release they're large mouth intake. I'm confident that they will just like one the Aero's. Nice work!

[SSinstaller]

If I ever try a blow through setup I'll have to order some...

[LuczOr]

Dang that is some serious air straightening. The smoke pictures are crazy.

[BlueStreak]

Looks great.

That meth setup is badass as well.

EDIT: Didn't PTP make a ghetto fabbed airflow straightener way back using mesh screen? I guess John was on the right path just not quite there.

[superskaterxes]

Originally Posted by BlueStreak Looks great. That meth setup is badass as well. EDIT: Didn't PTP make a ghetto fabbed airflow straightener way back using mesh screen? I guess John was on the right path just not quite there.

im not so sure you would want 2 nozzles facing each other like that, something tells me the water droplets would hit each other and form bigger ones lol

and yes he did but his wasent very good. theres actually alot of stuff in that link above for the wind tunnel that also talks about a screen and how it changes airflow.

[86azms3]

this is what I was trying to tell that deuchebag at ptp long long ago when he introduced his maf screen. Honeycomb>>>>shitty overpriced wireing fucksauce.

Glad you know this as well.

[rfinkle2]

Originally Posted by 86azms3 fucksauce.

?fucksauce?fucksauce

[Lex]

How much airflow restriction are you getting using that honeybcomb?

Is it truly the MAF instability that's causing your AFR changes?

[86azms3]

Originally Posted by rfinkle2 ?fucksauce?fucksauce

Yeah fucksauce it was.

I wouldn't see this as much of a restriction. But you definately have the right idea. You have the proper length channels to create that nice laminar boundary layer in each channel.

I would have to look it up, but to create nice laminar flow you need something like a channel length of 4x time the diameter of the channel. I would have to research that. I drank all my knowledge of engineering away.

[Monotonous ONE]

needs moar maf screen

[06Speed6]

The honeycomb will reduce the effective diameter of the maf by a small percentage so you will need to recalibrate, but overall I am a fan of them on larger diameter maf setups. You just cant get a good maf reading when they are 3"+, especially at or around idle.

[superskaterxes]

Originally Posted by Lex How much airflow restriction are you getting using that honeybcomb? Is it truly the MAF instability that's causing your AFR changes?

lol did you read OP? i havent even gotten these yet to test haha

Originally Posted by 06Speed6 The honeycomb will reduce the effective diameter of the maf by a small percentage so you will need to recalibrate, but overall I am a fan of them on larger diameter maf setups. You just cant get a good maf reading when they are 3"+, especially at or around idle.

the aluminum is .001" thick and i dont know how much area it takes up but when you think about it its not very much compared to the overall area. the pictures can be very deceiving to the eye with so much going on.

[06Speed6]

True, but most of the blockage will come from the outside edge where it hits the pipe so a little flow disruption there can end up being a big percentage of flow because most of the area of the crossection of the pipe is in the outside edge.

Just make sure it meets the pipe as cleanly as possible and it shouldnt be a big deal.

[superskaterxes]

Originally Posted by 06Speed6 True, but most of the blockage will come from the outside edge where it hits the pipe so a little flow disruption there can end up being a big percentage of flow because most of the area of the crossection of the pipe is in the outside edge. Just make sure it meets the pipe as cleanly as possible and it shouldnt be a big deal.

the plastic version looks like it does ALOT cleaner job of keeping this intact then the aluminum version. im still only at about 4.5V max on my curve so a little decrease isent going to hurt me.

[mason]

you could tack weld or braze the aluminum one in place so you know it will sit put?

[LuczOr]

Originally Posted by 06Speed6 True, but most of the blockage will come from the outside edge where it hits the pipe so a little flow disruption there can end up being a big percentage of flow because most of the area of the crossection of the pipe is in the outside edge. Just make sure it meets the pipe as cleanly as possible and it shouldnt be a big deal.

I'm going to preface this post by saying that i am the wind tunnel engineer at freightliner trucks.

I would say that the edges are the area where you should be least concerned. The maf is located center of flow so the edge effects wouldn't really make it to the maf. Especially if the straightener is close to the maf inlet. That's really the whole point, to eliminate turbulent flow characteristics, which are mostly caused by boundry layer development.

However, the turbulent region in the intake tube is probably most likely caused by the non uniform intake material (aka cone shaped filter going to tube shaped maf housing). You can imagine the not so smooth course that the air has to take to get to the maf.

Basically if your maf is close enough to your straightener, you should not see any turbulent effects. In the wind tunnel, we have a 1ft thick straightener for a 40x40ft intake and we can't measure any turbulent region at the straightener outlet. To give you an idea of turbulent boundry layer growth, we see a 10 in deep boundry layer at the bumper of the truck which is 100 ft away from the straightener outlet. This is all at 60 mph of course, and the boundry layer growth will be much greater at inlet air speeds.

But that's still at the walls of the maf housing.

Cliff notes: do it, it will work.

[06Speed6]

I love picking an engineers brain...

@lucz0r would that boundry layer around the inner circumfrence of the tube, say the outermost 20-30% of the ID have a much lower average flow velocity than the innermost 20-30% of the tube?

If that was the case then the crushed part of the straightener around the outer edge would have almost no negative impact on the flow.

Also you mentioned that your wind tunnel uses a 1' thick straightener, for a tube with a 3-4"ID and intake velocities less than 450mph, is there an ideal thickness for the straightener?

And last could the stem that holds the maf electronics be creating its own boundry layer with a thickness that increases in proportion to the intake velocity? If thats the case then maybe it would be a good idea to build the straightener around the maf stem such that maybe the boundry layer will be distributed across the entire ID.

Thanks for posting in the thread, it just got alot more interesting.

[LuczOr]

Originally Posted by 06Speed6 I love picking an engineers brain... @lucz0r would that boundry layer around the inner circumfrence of the tube, say the outermost 20-30% of the ID have a much lower average flow velocity than the innermost 20-30% of the tube? If that was the case then the crushed part of the straightener around the outer edge would have almost no negative impact on the flow. Also you mentioned that your wind tunnel uses a 1' thick straightener, for a tube with a 3-4"ID and intake velocities less than 450mph, is there an ideal thickness for the straightener? And last could the stem that holds the maf electronics be creating its own boundry layer with a thickness that increases in proportion to the intake velocity? If thats the case then maybe it would be a good idea to build the straightener around the maf stem such that maybe the boundry layer will be distributed across the entire ID. Thanks for posting in the thread, it just got alot more interesting.

In the case of an air straightener, the boundry layer is eliminated at the walls when dealing with internal flow. You can think of it this way: since all the air must pass through holes that are all the same size, it all experiences the same amount of drag. However if the air straightener is crushed, the air at the crushed areas will experience more drag. HOWEVER it may also be the case that due to the way the air must "turn the corner" to get from the cone filter to the maf housing, it may accelerate, negating this effect.

Either way this is the area that has the smallest effect on the maf in the center and I would say it is of little or no concern.

As far as the straightener length, the only way to really know the optimal setting is by testing. It might be the case that a 1in straightener is fine or you might need it to be longer. It is a function of the size of the holes in the straightener and the length. The smaller the holes, the shorter it needs to be, BUT the restriction will be greater with smaller holes (essentially, more flat plate flow. Google flow over flat plates for more info, this is essentially how the boundry layer behaves as well). There are equations that govern these characteristics, but they are only really used in air straightener development. I can run some quick calcs when i have time and post some results later.

As for the maf extension, it is located in the same plane as the inlet for the maf so its only effects would be flow deflection, but I think it's safe to assume that the designers of the maf have accounted for any of these effects.

The thing with aero is that there is always more shit going on. Scary to think this is just scratching the surface.

FUCK that was a beast to type on my phone.

[superskaterxes]

jesus that was on your phone!?!?! haha

if you look at the original thread they go over the ratio of length to hole size. i went with a 6:1 ratio with 1/8" holes.

luczor, r u saying the honeycomb should be as close to the maf as possible? i think theres 2-3 inches between the MAF and end of the housing so i have about that much play in where to place it. testing will certainly help but we may just find it smooths out the air so well it doesn't matter one way or the other.

[06Speed6]

Thanks @lucz0r ! I knew I was going to get some good info.

+1 on the spacing question

[laxplayermjd]

guinea pigs ftw!

[driver311]

this is the exact reason i sold my cpe big maf intake. Mine was the same way. Car just didnt run perfect all the time. Im holding out for Mark to send me the sure 3.25 intake coming soon. But your on to something good for this particular intake. Im hoping that marks intake works as good as the current one.

[Ziggo]

Its important to note that at high flow rates, the reynolds number in even a big maf car is very high, indicating that the flow is a fully developed turbulent flow. I did the calcs awhile back when I was proving why the "straightener" in MSCAIs are for improving maf stability around idle, not at WOT. I can dig up the spreadsheet if needed.

Airflow "straighteners" would be more appropriately called turbulence inducers as the reduce the effective diameter of the pipe in the vicinity of the MAF. Those airflow straighteners in windtunnels are for recircing windtunnels, which typically have low flow velocities (ie laminar flow)

[lengineering diatribe]
MAF sensors are based upon the amount of cooling a resistive wire wire receives. The amount of cooling is directly related to the mass of air passing by the wire, but only is valid if the flow remains turbulent. If the flow is slow enough, is starts to fall out of a turbulent flow near idle, causing eddies to develop and the equations that relate the cooling of the wire to the amount of massflow to no longer be valid. This screws royally with the fuel trims. Larger pipes lower the flow velocity, causing it to drop out of the fully turbulent range earlier. The MSCAI has a larger MAF diameter than stock and the bend before the MAF tends to organize the flow making the problem worse, thus requiring the "straightener"
[/engineering diatribe]

[superskaterxes]

Originally Posted by Ziggo Its important to note that at high flow rates, the reynolds number in even a big maf car is very high, indicating that the flow is a fully developed turbulent flow. I did the calcs awhile back when I was proving why the "straightener" in MSCAIs are for improving maf stability around idle, not at WOT. I can dig up the spreadsheet if needed. Airflow "straighteners" would be more appropriately called turbulence inducers as the reduce the effective diameter of the pipe in the vicinity of the MAF. Those airflow straighteners in windtunnels are for recircing windtunnels, which typically have low flow velocities (ie laminar flow) [lengineering diatribe] MAF sensors are based upon the amount of cooling a resistive wire wire receives. The amount of cooling is directly related to the mass of air passing by the wire, but only is valid if the flow remains turbulent. If the flow is slow enough, is starts to fall out of a turbulent flow near idle, causing eddies to develop and the equations that relate the cooling of the wire to the amount of massflow to no longer be valid. This screws royally with the fuel trims. Larger pipes lower the flow velocity, causing it to drop out of the fully turbulent range earlier. The MSCAI has a larger MAF diameter than stock and the bend before the MAF tends to organize the flow making the problem worse, thus requiring the "straightener" [/engineering diatribe]

im going to have to severely disagree with you here with no fluids/mechanical background purely on the fact that most automotive manufacturers use honeycomb straighteners in stock applications.....


i fully plan to prove you wrong lol

[Ziggo]

It allows you to add margin to the top end so that you have a big enough maf dia to ensure that it wont max out the sensor in cold weather without sacrificing low end maf stability.

[Fobio]

with the cp-e big maf, I have been able to extract more top-end power relative to other similarly modded cars, but I have been chasing idle AFR's for a few MAF calc's and LTFT's been anywhere from 0 - +8 without changing maps or hardware.

I have 2 solutions in mind:

1. a straightener like this to help "average" out the flow so the MAF sensor sees a similar flow as it is flowing nearer to the centre of the housing.

or

2. Find a replacement MAF that has a longer housing, so that it's sits closer to the centre of the MAF housing.

$19 sounds a hell lot cheaper, but I wonder how much it affects peak g/s and where the added material becomes an obstruction to flow.

[LuczOr]

Originally Posted by Ziggo Its important to note that at high flow rates, the reynolds number in even a big maf car is very high, indicating that the flow is a fully developed turbulent flow. I did the calcs awhile back when I was proving why the "straightener" in MSCAIs are for improving maf stability around idle, not at WOT. I can dig up the spreadsheet if needed. Airflow "straighteners" would be more appropriately called turbulence inducers as the reduce the effective diameter of the pipe in the vicinity of the MAF. Those airflow straighteners in windtunnels are for recircing windtunnels, which typically have low flow velocities (ie laminar flow) [lengineering diatribe] MAF sensors are based upon the amount of cooling a resistive wire wire receives. The amount of cooling is directly related to the mass of air passing by the wire, but only is valid if the flow remains turbulent. If the flow is slow enough, is starts to fall out of a turbulent flow near idle, causing eddies to develop and the equations that relate the cooling of the wire to the amount of massflow to no longer be valid. This screws royally with the fuel trims. Larger pipes lower the flow velocity, causing it to drop out of the fully turbulent range earlier. The MSCAI has a larger MAF diameter than stock and the bend before the MAF tends to organize the flow making the problem worse, thus requiring the "straightener" [/engineering diatribe]

Ziggo is right about the Reynolds number being really high. It is a low density fluid in a high speed flow characteristic. However, the air straightener reduces the Reynolds number. It decreases hydraulic diameter of the pipe, which if I remember correctly is in the top of the Re equation.

As for the maf using turbulent air to cool the wire, I'm pretty sure they are talking about the region immediately around the wire itself. I do not think this is a concern for our maf since the wire is not located directly in the free stream, although I don't exactly remember where that wire is... I just remember that plastic extender with the wire "in there somewhere."

As far as air straighteners in wind tunnels, we use an open loop system as opposed to closed loop (recirculating) and still utilize an air straightener for obtaining laminar flow and eliminating boundary layer.

[Ziggo]

The wire is located in the free stream, its inside that plastic extender. The plastic on either side is to help lower the effective diameter of the pipe around the maf also. The whole basis for a MAF operating is that wire sitting in the free stream. If it gets even crud on it, it wont work very well and thats when you hit it with the maf cleaner. As for the air straighteners eliminating boundry flow, they do this by INTRODUCING TURBULENCE. Its why golf balls have dimples, its why race cars use diffusers, and why ricers put those stupid little triangles on their roof (even if they don't know it) Turbulence cuts the boundry layer and reduces drag.

The characteristic length that goes on the top of the RE equation is a minor player compared to the velocity term that also goes on top when talking about pipes. Velocity increases as a pipe gets smaller much faster than the pipe diameter drives it down. I have linked the spreadsheet below, check the math if you don't believe me, change the pipe diameter around and see how it changes the "RE" column. Note the temperature above the columns is the IAT temperature.

If the RE is less than 2300 you are getting laminar flow. If it is above 4000 you are getting fully turbulent flow. The in between is like the eddies behind a paddle in the water, which can be organized by bends in the pipe making the MAF freak out.

https://docs.google.com/spreadsheet/...0Tms2b2c#gid=0

[06Speed6]

In this episode of Engineer Wars...

Sorry I just had to, I am very seriously watching this thread to see what you guys come up with.

[LuczOr]

Ziggo is once again correct on all accounts. However I think he is over looking some small points when it comes to the Reynolds Number. He is absolutely correct that the velocity grows much more rapidly as the pipe diameter decreases. This is due to the fact that the velocity of the air is dependent on a Velocity*Area constant, ergo, as the area decreases, the velocity must increase proportionately, and since the area of a circular cross section is pi*r^2, the velocity must increase at a squared rate as compared to the decrease in radius.

However in the case of a straightener, the area does not decrease very much since the material of the straightener is extremely thin. BUT the effective diameter (characteristic length) is greatly reduced from the diameter of your big fat maf housing to the size of the holes in the straightener. But Ziggo makes a good point, this alone will not drop Re below 2500 and achieve laminar flow. Not to mention the fact that Ziggo did not account for the dynamic pressure drop due to suction (which I would be quite surprised to see since it's almost impossible to calculate) and the consequent drop in dynamic viscosity.

But, once again, I feel Ziggo has overlooked some small details. In the case of turbulent flow, there is a rule of thumb that explains how long it takes for internal flow to become fully developed, which is 10x the pipe diameter. What this means, is that the effects of the straightener will not disappear for 32.5in.

The effect that the straightener has is to cut down the size of turbulent vorticies in turbulent flow. The smaller vorticies will have less pressure and density difference within them and will tend to neutralize since the they are moving in parallel. Boundary layer effects will not affect the flow in the center till 10 diameters.

Ziggo, no disrespect. in fact, I'm impressed that you put together the spreadsheet.

Hope what I'm saying makes sense.

[Ziggo]

I don't really get your point. My point was that these things do not affect MAF stability at WOT, they improve it around idle. There are no "eddies" in a fully turbulent flow to make the MAF freak out, and you can see that the RE numbers are nowhere near the transitional borders at WOT like MAF flows.

I am not trying to engineer the next delta rocket here, and I don't believe these equations are 100% accurate or that I have accounted for all effects, but the results are very clear and the basic physics all point in the same direction. At 100g/s RE values are in the order of 125K, and reducing the effective diameter in a pipe increases turbulence. (increases RE)

The MAF straighteners will tend to flatten out the velocity profile in the pipe, is this what you are talking about? I don't see the velocity profile affecting MAF stability as while the profile would throw off the sensor, it does it in a repeatable way that could be accounted for by trimming the sensor with the MAF cal tables.

[laxplayermjd]

this thread makes my head hurt

[06Speed6]

Aaaaand... I feel dumb.

But continue because maybe at some point I will be back on the bus.

[86azms3]

I think the problem is we are confusing laminar flow with an evenly distributed velocity profile. And we need to realize that we need a fully developed flow to be able to achieve a consistent reading.



So who’s ready to bust out some navier stokes equations? Joking fuck that short hand heuristics>>fucking theory any day.

And looking back at some of my engineering books, it turns out turbulent flow is not such a bad thing. Really it is a good thing since we are creating a uniform velocity profile. In laminar profile the greatest velocity is at the center of the pipe decreasing exponentially across the radial axis. Use navier stokes equations if you really want to find the exact profile. But the picture below shows the velocity profile as we digress from laminar flow to turbulent. As we become more and more turbulent we see that the velocity profile becomes a straight line where at any point the velocity will equal the average.

So then if we already have turbulent flow as shown by Ziggo, then why do we need an air straightener?

This is because flow is not necessary fully developed turbulent flow. So any inconsistencies (ie. bends, bungs in the pipe, any non-conformity) will cause disruptions. These disruptions cause eddies and other irregularities. So now we have to ask how much time will it take for these irregularities to go away. In fully turbulent systems

The entrance length can be expressed with the dimensionless entrance length number expressed as

El = le / d (1)
where
El = Entrance Length Number
le = length to fully developed velocity profile (m, ft)
d = tube or duct diameter (m, ft)

We want to find le , we already know the diameter of the tube. So we need to calculate El the entrance length. The equation is found below.

Elturbulent = 4.4 Re^1/6 (2)

So say we are just cruising at low flows well say we have a Re of 5000 then plugging and chugging we get an El of 18.2

Now to get the le (length for fully developed flow) we use equation 1 and assume a diameter of 3” tube

18.2=le/(3”)
Solving le=54.6”

So to get a uniform flow before the maf you would need a pipe that is 54.6” long. And this pipe has to be straight with no obstructions. So obviously that is a little too long and unreasonable. So how bout we throw an air straightener and reduce the effective diameter. We’ll use the same Reynolds number assuming it is constant up unto the straightener.

Then solving for the effective diameter is. We’ll assume the diameter of one of those cells is one tenth of an inch

18.2=le/(.1”)

le then equals 1.82” so then the length of those small channels will need to be 1.82” long for fully developed flow

And really it doesn’t need to be 1.82” I’m sure if we looked at the actual equations instead of heuristics we would find that even a length of 1” will probably get you like 95% fully developed flow. No doubt that the equation is exponential.

[Ziggo]

Excellent, now I am on board. I never thought about the velocity profile or entrance length till I started questioning myself yesterday and you just proved my suspicions. The more turbulent the flow, the longer the entrance length will be, and while a honeycomb will increase the turbulence, it will also significantly decrease the entrance length.

In summary: The flow in the pipes is turbulent and should remain so.
MAF straighteners have 2 benefits: they increase turbulence at low flow rates to keep the flow turbulent at low loads, and decrease the entrance length at all loads to remove the effect of bends before the MAF.

Now we just need 06speed6 to prove the theory. I will add an entrance length column to my spreadsheet so he can figure out his spacing.

Fuck Navier-stokes
---
- Zigatapatalka

[superskaterxes]

^^^^^

<<<<

will be doing the testing lol

[Ziggo]

I get you speed6 guys all mixed up. Its what you get for all looking the same.

[LuczOr]

86azms3 said exactly what I was trying to say... In a much more concise and understandable way LOL. I was just trying to get at the fact that I think a flat velocity profile and fully developed flow would make maf readings more consistent.

Word up on testing, I wanna see what effects this really has or if we are all just blowing smoke out of our air straighteners.

[RichieRichness]

this is a good idea...I too am seeing .1 to .2 AFR fluctuation.