posted on September 22, 2013 13:31
Project S2000: Part 19 - Ram Air!!! Plus Hot Air Testing!
Khiem Dinh is an engineer for Honeywell Turbo Technologies at the time of this writing. All statements and opinions expressed by Khiem Dinh are solely those of Khiem Dinh and not reflective of Honeywell Turbo Technologies.
Did I mention there would be more hood hacking? Why yes I did. Somewhere along the other 18 parts of this project, I noticed the stock air box looked relatively well sealed along with having a location ideal for ram air. Ram air is used on practically every sport bike you can buy to coax as much power as possible from the engine. Even cars such as the Corvette Z06 and Dodge Viper use ram air. Lastly, the last time I took the S2000 up Angeles Crest Highway, I got pulled over for not having a front plate. During the
interrogation friendly banter with the police officer, he asked if I had an intake to which I replied no (FYI, he was digging the StopTech brakes). So, keeping the stock intake deters further police inquiry. There’s nothing illegal (I don’t think…) about chopping up the hood and adding a ram air NACA duct (plus, this is my track-only hood), so that’s what I did.
The duct in the hood of the Viper seals up against the air box providing a ram air effect.
The principle of ram air is simple: it converts the dynamic pressure of air at a high velocity into a higher than ambient static pressure. Basically, it supercharges your engine by increasing the air pressure in the intake manifold. The maximum pressure that is achievable is a simple function of velocity. The higher the velocity, the more dynamic pressure there is to convert to static pressure. Using Bernoulli’s equation, dynamic pressure is equal to 0.5 * air density * velocity^2. As total pressure along a streamline stays constant, the change in static pressure is equal to the change in dynamic pressure. So when you bring the velocity down to 0 (hence, the term stagnation pressure), you have maximum static pressure.
Using Bernoulli’s equation gives us this curve of the stagnation pressure vs. vehicle speed.
Now that we know how ram air works, it was time to figure out a way to do it. I decided to design my own NACA duct to feed the snorkel of the stock air box. There are some NACA ducts you can buy on the market, but they were too narrow or too long or not deep enough. Also, none of them would line up perfectly because of the curve of the hood. So, I designed one specifically to work for the S2000. There is actually science behind the shape of a NACA duct to draw in as much air as possible. I had no clue as to where to start in designing one, but a Google search netted a free spreadsheet someone else had already developed. All I had to do was enter my desired length, width, and depth and the spreadsheet cranked out all the dimensions I needed to design my duct. The tricky part in designing the duct was to make it match the curves of the hood.
I began by using C.A.D to locate the snorkel relative to the hood.
Monday, September 23, 2013 3:36 AM
Monday, September 23, 2013 6:01 AM
I am wondering the affect of a branch intake resonator on the new improved intake. I see the standard resonator as part of the lower box,but can't help wondering if a second resonator inline would help with the oscillations that your logging.
Monday, September 23, 2013 7:06 AM
Just as a thought, I know there's articles on here about doing composites. If you cut some of the bracing out on the underside of the hood, you could have epoxied the printed NACA duct and possibly printed hood vents to the underside of the hood, gotten everything nice and flush-ish with a die grinder or something, and pulled molds for a DIY carbon/glass/whatever hood. But no arguing with results, and I really do like the printed bits on this, especially given I have a 3d printer at my disposal.
Monday, September 23, 2013 8:31 AM
Would the rough texture of the NACA duct affect the efficiency?
Monday, September 23, 2013 9:48 AM
Oh yeah, also do some research on NACA ducts; long story short is that the make air that would otherwise be parallel to a surface go into a duct at all, but they aren't nearly as efficient as a scoop that projects into the airstream. If you had pressure transducers it might be interesting to block off the exit from the duct completely and see what kind of pressure it's generating; I bet that your air box is sealed well enough and you're just not seeing all the theoretical stagnation pressure due to the inlet design.
Might be fun printing a little scoop in a compatible shape to try an A/B if you were printing yourself instead of sending it out.
Monday, September 23, 2013 10:03 AM
Dan, you are absolutely correct on the purpose and effectiveness of a NACA duct. Considering how much flak the hoods vents got, can you imagine what people would say about a scoop? ;)
Monday, September 23, 2013 10:29 AM
Bah, but Science!
You're giving me a lot of impetus to build a bigger 3d printer, I'll have you know. ;) Since I'm building an S13 science experiment anyway...
Monday, September 23, 2013 10:57 AM
Here's another thought: the scoop isn't at a flat part of the hood. It's at the very front, which is typically a high pressure zone. Were it midway back, I could see it being a lot less effective.
Monday, September 23, 2013 1:51 PM
Another great article Khiem. Of course, if you had called me beforehand, I'd have told you to expect the bigger gain from air temp reduction ;)
It looks like you might have enough thickness in the scoop plastic to switch to a countersunk head on the bolts... if you do the front ones at least, it should cut a decent bit of turbulence across the duct and make it more ideal.
Monday, September 23, 2013 3:17 PM
I would have been interesting to monitor the A/F ratio since you are speed density.
Monday, September 23, 2013 7:46 PM
Nice work, great writeup as always.
I may not want to hack up my hood for now, but I'm interested in buying your brake ducts though :)
Monday, September 23, 2013 10:22 PM
No A/F, but I did look at my datalogs again. Without the duct, air temp = 54degC, timing = 26-27 deg, map = 95.1kPa, injector = 10.1ms. With duct, air temp = 46-47degC, timing = 28 deg, map = 96.1kPa, injector = 10.2ms. Running the numbers, the A/F stayed virtually the same (of course, more air so more fuel and more power), and the lower intake temp gained 1-2 degree of timing in the map.
@klch, I priced out some tooling. Basically, I'd have to sell at least 100 sets to break even. Yeah... I don't think that's going to happen and I don't have that kind of money to float for tooling.
Wednesday, September 25, 2013 11:06 AM
@Wrecked, I think it might actually help a little bit in keeping the flow attached to the lower surface of the duct.
So, NACA ducts are not really suited for a ram air application as you loose some of the pressure, but it works out decent here because of the relatively steep slope of the hood. So the air has a pretty straight shot into the snorkel. It probably would have been more effective if I would have done a rectangular duct instead of the trianguler NACA shape, but that would have required cutting a lot of the hood closer to the edge possibly significantly compromising strength. What I had to cut to fit this duct only required cutting through a little bit of the double layer of material.
Wednesday, September 25, 2013 1:32 PM
Probably has less of an impact on drag than a projecting scoop though, which may be a decent tradeoff.
Man, where's the official MotoIQ wind tunnel when you need it? ;)
Wednesday, September 25, 2013 1:42 PM
Could it be that the oscillations you saw were aliased frequencies of valvetrain oscillations? I wonder if they could be manipulated with resonance to get a little more torque out of the motor.
Wednesday, September 25, 2013 5:15 PM
For those of you who aspire to play with NACA ducts for experiments like this, cheap thermoformed NACA ducts with hose outlets are readily available from a lot of sources, including racecar supply companies. They may not be able to take full engine heat, so be careful of placement. Khiem's FDM duct isn't exactly great for that either though, but again, proper placement....
Wednesday, September 25, 2013 8:06 PM
I have the Mugen air box on my 06 DC5 RSX. I wonder if making a duct in the stock grill will help draw in cool air and give a slight ram air effect. I don't know if the area in the grill is a high or low pressure zone. I'm not sure if air isn't just hitting the pointy front bumper and mostly sailing passed the upper grill and over the hood.
The Mugen bumper has an intake duct along with a solid upper grill.However it is very expensive and draws too much attention to the car.
Think it might be worthwhile to cut the grill and fashion a duct?
Thursday, September 26, 2013 11:10 PM
I looked up the Mugen air box. So that snorkel for it is not compatible with the stock from bumper? Are you using the stock snorkel for the stock airbox? That should be grabbing air from a cold place.
I wouldn't cut up the front grill, but maybe just run 2"-3" neoprene ducting like that used for brake ducting and zip tie it to the front grill and run the other end to the Mugen air box.
Friday, September 27, 2013 12:15 AM
The stock intake snorkel makes a 90 degree turn towards the battery and draws in air here right next it to. The Mugen snorkel points right over the radiator and into the Mugen bumper's opening. What I've seen people do is cut out the top part of the grill since you can't even see it with the hood closed, and take out the length of weather strip along the opening. This isn't my car but the picture makes it very clear:
What I would want to do is seal the end of the snorkel with foam to seal up against the front grill. That's IF this would work to draw fresh air (maybe even ram air) in, it would be a very clean invisible way to do it. I guess without testing I won't find out if air is even making it to that area.
Saturday, September 28, 2013 7:24 PM
That's the wrong use of a NACA duct. It is intended to not get air IN, rather, to get are out. You're not getting any sort of ram air effect with that setup. A NACA duct should be positioned 180 degrees from how you have it. It's intended to create a vacuum. Way to be a ricer.
Saturday, September 28, 2013 8:26 PM
They're used as both low drag inlets and outlets
Saturday, September 28, 2013 9:04 PM
@Miznitic.... so, I have data showing it both increased pressure and reduced intake temperatures which both served to increase power. I also showed four different ways it improved power. That's rice? Please help me out with your definition of rice. I'm curious to hear your answer.
If you're going to make a negative comment, you better back it up. In this case, not only are you being rude, but you're also wrong which makes you look like a double jackass.
"A NACA duct, also sometimes called a NACA scoop or NACA inlet, is a common form of low-drag air inlet design, originally developed by the U.S. National Advisory Committee for Aeronautics (NACA), the precursor to NASA, in 1945.
When properly implemented, a NACA duct allows air to flow into an internal duct, often for cooling purposes, with a minimal disturbance to the flow. The design was originally called a "submerged inlet", since it consists of a shallow ramp with curved walls recessed into the exposed surface of a streamlined body, such as an aircraft. It is especially favored in racing car design."
I'd like an apology. Otherwise, go back to whatever hole you crawled out of.
Saturday, September 28, 2013 9:20 PM
Also, the original NACA reports proposed using them as air exits, but a later one found that they were actually crap for that - the air exiting out of them if you run them reversed is about perpendicular to the air stream, and a simple ramp exit is better.
Saturday, September 28, 2013 10:44 PM
Using Wikipedia as a cite? Really?
Saturday, September 28, 2013 10:58 PM
They're used as inlets on airplanes and cars all the time. Check out the Ferrari F40 for example on a car. Also look at the side window inlets on racing cars. They usually use naca ducts there too.
Saturday, September 28, 2013 11:14 PM
Sorry... Are there really people arguing about whether NACA ducts are proper to use as inlets....? Have we run out of legitimate things to argue about?
Saturday, September 28, 2013 11:34 PM
http://naca.central.cranfield.ac.uk/reports/1948/naca-rm-a8a20.pdf http://naca.central.cranfield.ac.uk/reports/1948/naca-rm-a7i30.pdf and http://naca.central.cranfield.ac.uk/reports/1945/naca-acr-5i20.pdf for non-wiki sources then. When they talk about submerged air inlets or entrances on the opening page of all of those... look, really? Can we put this to bed yet?
Sunday, September 29, 2013 2:37 AM
Miznitic, you have nothing to counter my hard evidence taken from my datalogs? Do you have nothing to form a reasonable counter argument? Here's a citation for you:
An Experimental Investigation of NACA Submerged-Duct Entrance"
Frick, Charles W.
Davis, Wallace F.
Mossman, Emmet A.
May 23, 1945
Organization Source: NASA Ames Research Center; Moffett Field, CA, United States
"The results of an investigation of submerged-duct entrances are presented. It is shown that this type of entrance possesses the following characteristics: 1) very high-critical-compressibility speeds throughout the range of high-speed inlet velocity ratios; 2) very low pressure losses for the air entering the duct at all inlet-velocity ratios; and, 3) low external drag. These characteristics are obtained by the proper shaping of the contour of the upstream approach to the submerged inlets and by proper alignment of the duct lip. Design data are presented and the application of these data to a specific high-speed fighter-airplane design is discussed."
I was being relatively nice before. Now I'm just going to call you out for what you are which is a freakin idiot who apparently lacks the ability to read, analyze data, and do any type of logical reasoning.
All you can come back with is, "Using Wikipedia as a cite? Really?"
Actually, I'm saying is that all you REALLY have? If it's so ricer, then why did it increase intake manifold pressure? If it's so ricer, why did it decrease intake manifold temperature? If it's so ricer, why did it increase power?
Really? Is that all you have?
Either man up and apologize as all the evidence points to you being wrong, or do us all a favor and go back to the cave you came from.
Sunday, September 29, 2013 9:11 AM
Miznitic must be trollin'.
Friendly local CFD engineer here again. NACA ducts are used as intakes because their shape causes the boundary layer to pull air in along with it by inducing vortex shear. Because of this, they tend to have downwind vortices as well if there's any resistance to the air being sent through the intake at the bottom of the duct (if the air going in gets in a traffic jam, so to speak, they generate more drag outside the car). Remember vortices are not your friend for reducing drag, they're your friend for _trading_ drag. If you have a blunt body with a sudden drop off to its trailing area (like the back window of an Evo for example), you may want to generate some vortices there to get some more energy into the air stream. Doing so allows the air stream to follow an inverse pressure gradient longer (doing so takes energy, remember), causing the air to make a smaller hole behind the blunt body, thus reducing form drag by increasing a little skin drag. It's a careful balance to make sure you end up ahead, and even then it's going to depend heavily on intended velocity. Like most things in aerodynamics, they don't have a hard absolute use, just places where you can, in the right conditions, exploit them in a way that works in your favor. But I digress.
So NACA ducts have got good and bad to them. An intake along a surface where the air would otherwise have no reason to get into the car (like a trailing edge or an edge parallel with flow) that feeds into the cabin to give the driver some sweet sweet face breeze, and they're your ticket. Flowing directly into an intake trumpet, and there are probably better options.
So, as Dr. Dinh has explained in the comments, there are a number of facets about his intake situation that make a NACA duct rear its bad sides over its good, namely the duct being in a place where air would likely be drawn in of its own volition, and the fact that a car's air intake is inherently a flow-limited area (the engine will draw air in from the duct plenum when it's good and ready, and not before). So, now we've created external vortices instead of intake scoop form drag. What happens then?
Engineer's favorite answer: IT DEPENDS! Let's devolve into conjecture from here. The intake is cut along a surface that will make a positive pressure gradient for the air flowing outside the car. We're adding vortices to this from the corners of the NACA duct, probably. Now you've got air following along a surface with even more energy on tap to follow that surface, which we've traded a little surface drag from in the shape of those vortices. What's downstream of the NACA duct? Hood vents! So we send this energized air ramping up this vent to grab some sweet air, but it has more energy to follow around the backside, so the passenger side duct doesn't create as much stream disturbance. At that point, the effects of the NACA duct are probably largely muted, unless you're driving over, say, 100mph and off throttle, where the airbox will be highly pressurized with little airflow.
In conclusion, Dr. Dinh didn't make a mistake, he made a trade. He could probably increase airbox pressure with a different shape of intake, or could possibly get the same performance from a different intake location, perhaps up closer to the actual throttle body, if desired. But what he did was far from flat out wrong. NACA ducts make great intakes for certain applications. This application would be football-akin to sending in a tight end to do a wide out. Can it work? Hell yeah! Will it be your absolute best option? That needs further investigation to answer correctly, but chances are that it isn't your best option, no.
Sunday, September 29, 2013 7:38 PM
@cycleflight, great info! And I only have a MS, so no "Dr." :) Though, maybe a Dr. of half-baked ideas; some better baked than others.
Yup, I think the evidence of the pressure gain only being about half of the theoretical shows this setup is not ideal. A simple rectangular opening would most likely work better from a ram air perspective. Given a clean sheet start for an entire hood (as in molding a completely new hood), I would do a rectangular opening. However, in modifying the stock hood, I think it would put the edge of the opening too close to the edge of the stock hood compromising strength.
Very interesting about the vortex generation. I don't have a good feel for trying to visual them in my head. As for the example of using them in a situation such as the steeply raked rear windshield on the Evo 8/9... Well, I'm planning on getting an OEM style hardtop which also has a steeply raked rear windshield and half thought about adding the vortex generators. I read the paper Mitsubishi put out on it and it was pretty interesting. Always fun to see the theory learned in class applied to a real-world case.
So while the ram air performance was only so-so, it did a decent job as a cold-air source. And as you stated, it probably compromised my hood venting/cooling efforts, but it seems the overall cooling system efficiency is pretty good still.
I'm sure some people have questioned why I do such hair-brained things. Well, you can theorize and simulate until you're blue in the face. At some point, you just have to do it and see how well something works. Sometimes it works pretty well, sometimes not quite as well. Haha. But we learn something either way!
Monday, September 30, 2013 8:33 PM
Speaking as someone who screwed up some otherwise good, long-NLA 12A rotor housings by milling them for spark plugs that were never intended for them... making a NACA duct as an air inlet isn't really hair-brained. ;)
Thursday, October 03, 2013 3:43 PM
I think people overuse the word "ricer" without knowing what it really is.
That's just plain stupid.
Aside from that, I've learned a great deal is information from reading spdracerut and cycleflight's comments. :)
Thursday, October 03, 2013 3:44 PM
Wednesday, October 09, 2013 12:03 AM
I'm glad I finally read this, nice idea and implementation. That being said, I believer there is some smallroom for improvement. Currently you are achieving ~1000 Pa at 125 mph, with a theoretical max of ~2000 Pa. This represents a dynamic pressure recovery of 50%, which is pretty good considering the fact that most NACA inlets max out around 80%. In my opinion, the weak points of the current duct a directly coupled with the installation technique. First, you have a fairly large gap between the NACA inlet and the inlet to the airbox; second, the bump around the inlet is likely having an adverse effect, potentially generating additional loss within the inlet. My suggestion would be to make the airbox inlet and NACA inlet one continuous component, which you would then shut the hood down on top of. You could then seal the inlet-to-hood interface with some of the foam you have around the radiator. This would reduce the leakage of air around the intake, thus reducing the velocity ratio (V_inlet/V_freestream) and thus improving the efficiency. This would likely get you around 10-20%, but the remaining benefits are much harder to acquire though. They will likely require a change in the surface roughness of the material and a redesign of the duct divergence rate and upper lip shape. Ultimately you have achieved most of the effect and with some additional work you could produce a cleaner and more effective design, but ultimately you will miss the 80% mark unless you get an aerodynamicist to tweak the current inlet design.
Thursday, October 10, 2013 10:06 PM
Kyle, thanks for the insight. I was thinking of adding a foam strip to the top of the snorkel part of the air box to help seal against the top of the hood. Check out the very last picture; I did tape up the gap between the hood and the snorkel effectively sealing the top part. I did not try to seal the sides though between the duct and snorkel. I glanced at the datalogs and it did seem to help a tiny bit, but the ambient conditions were so different that I didn't feel it was a good comparison.
Friday, October 11, 2013 7:35 AM
I see it now that you point it out. At this point I would fix the remaining leakage and then call it a day. At this point it is likely that no one change is going to close the remaining 30%. Rather it will be a combination of changes, each worth 5% or less. In my opinion the remaining 30% is not worth it, you've already achieved a really good inlet temperature drop and an increase in pressure, the remaining 30% pressure recovery (0.5% air density) will take much more work than its worth...