Archives

I've been thinking about aerodynamics for a while, so I may as well start a dedicated thread about what I want to do. According to online sources, the '02 Accord has a reasonable drag coefficient of 0.33. It's only natural that this is a concern due to the fact that most of my driving is at highway speeds of roughly 55-65mph. Not only would it be nice to reduce air resistance (and hence engine load) at steady-state, but gains potentially could be compounded by improved NICE-on and FAS abilities. (HS P&G anyone? :) ).

So far I haven't actually done much to the car. I did finally remove the plastic sunroof visor thing just on principle. It's maybe 3' wide by 2" high, and probably does have a small negative impact on aerodynamics. Maybe measureable, maybe not. Grill blocking will probably happen, but I'd prefer to wait until it's not 95F outside with a heat index over 100F. That leaves the most obivious (to me) modification, which is a belly pan.

As I posted at GS last week, I did some reading and came across an SAE paper that deals with the aerodynamics of a production Opel. I can post the details if you like, but for now I'll just cut to the chase. My rough calculations suggest that a small but significant gain could be realized from a well-made belly pan, on the order of a 5% boost in FE. (Some say 10-20% but that sounds awfully optimistic to me.)

At the moment I have 64 square feet worth of 6mm Coroplast sitting in my apartment waiting to be installed. All I need is some time to get my head under the vehicle to figure out just how to attach the stuff. My primary concerns are safety (#1; e.g. no coroplast near hot exhaust parts, nothing that could fly off and pose a hazard to others, etc.) and vehicle integrity (#2; e.g. not drilling holes in metal bits that could invite rust, cause permanent damage to the car that's supposed to last me years, etc.). Fuel economy is the goal, but comes third to the first two objectives. I'm sure you'll understand ;).

I can't say with absolute certainty that I will realize a significant improvement in FE from the belly pan. It should be measureable in theory, but we all know how reality is so good at jumping up and biting us in the rear. Also, there's no perfect way to measure small gains without a wind tunnel. My mileage by technique alone is fairly flat, but a 1-2mpg fluctuation over the same driving conditions and cycle can be expected. I think coast-down testing ad nauseum is the best way to go.

Stay tuned...

I would tend to agree that the 5% is way more in line with reality than the 10-20%, especially on a modern car. Have you been under your Accord? It's probably not all that bad to begin with, especially compared to older cars where they were totally uncaring of what the underbody looked like. Anyway, good luck with it, once you get a chance.

I did stick my head under there and you're right that it doesn't look that bad. It's far cry from a flat plate but at least there aren't many big protrusions to speak of. The rear of the car seems to be the worst part, with the bumper cover hanging down and some other ugly transitions about in line with the rear wheels. Ahead of that it's very rough, but everything is tucked pretty well out of the way.

Hi Tim:

___I wanted to add a few comments before you begin your “Aero-Mod” experience …

___Having the SG-II connected to the Accord these past 3 days, I have been pulling in empirical data as to what an aero mod via drafting may or may not be worth vs. driving technique. What I have seen so far is that even the best of the best Close-in’s (UPS Dualies at ~ 10 feet as an example) traveling at 57 - 60 mph are at best a match to a HS P&G. I am talking so far up the guys under-ride guard you would think we were connected! Under these best of circumstances, the SG-II’s iFCD function will bounce around in the 60 - 80 mpg range on relatively flat terrain with a high 20 mpg consumption rate for an accel during a climb and 120 + range during an almost fuel cut descent. What I am seeing is an ~ 65 - 70 mpg average while in way to tight and if that is not a proxy for a CD of ~ .15 or less, I cannot think of one better? With a HS P&G ranging from 45 - 68 mph (high's and low speed range besides the heavy traffic portions and Warm up AR this morning) 60 + is almost a gimme. The last 2 segments with a not yet calibrated SG-II (I believe she is under-reporting by ~ 7% currently) have allowed a 60.0 and a 62.5 over (2) 95 mile segments respectively. Maybe a max of 15% of those miles were under a close-in or distant, maybe 60% under Surf and Traffic Side which you get anyway, and 25% naked. Along with the usual 20 + miles of Chicago style stop and crawl over those 2 segments …

___What you may want to do is some testing pre- and post Aero-mod using your SG’s TPS. I am running between 13 and 16 for a relatively flat highway cruise under a close-in and 19 - 21 while maintaining similar speeds naked. A HS P&G is usually performed with the TPS in 19 - 23 range (17 - 25 mpg rate) for the Pulse and the Glide is free of course …

___To add … The 7th gen Accord’s NAVI w/ TC was very useful when I first received it for the i and aFCD but now with the advanced techniques, it may as well as a $1,600 boat anchor … Totally and completely useless for a hypermiler imho as the darn thing is always pegged at 50 mpg 90 + percent of the time anyway :( Great NAVI unit though ;)

___Good Luck

___Wayne

Thanks for the insight, Wayne. Using the TPS readout hadn't crossed my mind but that sounds like a good option. Generally I go by the load readout, but I'll call up the TPS readout next to it and see what happens. The real problem IMO is mitigating the effects of the many external variables. Temperature, wind, warmup state, tire pressure, gas formulation...it goes on and on. Now that I think about it, using the ScanGauge instead of timed cost down testing would allow me to do several trials each day at specific locations along my normal route. That would allow me to build a distribution of data that could be compared to a similar distribution of post-mod data. Hey, this is starting to sound like engineering or something :)

Hi Tim:

___Check your PM inbox in a minute or two a sI have some interesting info ... You will really want to ;)

___Good Luck

___Wayne

Hi Tim:

___I was just thinking about something I was considering back when I first picked up my Accord … From what has been posted around the net, the AH’s rear deck spoiler reduced the Accord sedan’s Cd to .29 from .30 all by its lonesome. It looks really slick too. Maybe you can find out if the 6th gen rear spoiler does the same? Not the wing as that is like throwing up a sail but the rear decklid spoiler for that gen Accord if one was ever made? Somebody probably has a Cd number reduction on it with enough snooping.

___A quick post into the Drive Accord forum might make the search a lot easier too?

___Good Luck

___Wayne

That's another great suggestion. Unfortunately, my car came with the the stock "sail" on my trunk lid. I like the way it looks but there's no way it's helping my aerodynamics. (I should post pics of the mounting points, as they look hideous from the perspective of drag.) Removing it involves four bolts as far as I can tell, which leaves four holes that would need to be filled with something.

Anyway, I keep an eye open for 6th gen accords and I have never seen anything other than "naked" trunk lids or the factory wing like mine. So I doubt that such a lip spoiler exists for my specific application in stock form, and the only "factory" option would be trying to adapt a HAH spoiler to my deck lid. That might help or it might not...really depends on how careful you have to be about designing them.

If I understand the aerodynamics right (and I may not) the purpose of that spoiler on the HAH is to "trip" the flow of air as it comes off the trunk lid and produce turbulence, which mixes the wake immediately aft of the car to reduce the size of the lateral vortex, thus reducing the maginitude of the big, nasty low pressure region behind the car. Something I read this morning makes me wonder if there's another device that would serve a similar purpose.

There's a thread on gassavers about aerodynamic vehicle design, and there are a few links to info on vortex generators. So far I've never heard of them being used in place of a spoiler, and there's probably a very good reason for that. Instead they are used to stir up the air flow just forward of a flow separation point, which pushes it farther back on the vehicle and reduces the size of that low pressure region. The application commonly cited is the Mitsubishi EVO, with the generators placed on the roof forward of the rear window. So the question is: would that have a similar effect or would they just make the car look funny? Word is that they would just induce drag and not help anything, but I don't know enough about it to make a good assessment. The only reason I bring vortex generators up at all is because they are relatively easy to fabricate.

Took a few pics. I can shrink and re-upload these if you guys have trouble with them.

Rear profile:
http://www.cleanmpg.com/photos/data/500/Profile.jpg

Rear view:
http://www.cleanmpg.com/photos/data/500/Rear1.jpg

Detail from the front (and slightly high):
http://www.cleanmpg.com/photos/data/500/WingDetailFront.jpg

Detail from the rear:
http://www.cleanmpg.com/photos/data/500/DetailRear.jpg



Mounting Hardware:
http://www.cleanmpg.com/photos/data/500/MountingHardware.jpg

Upon casual inspection it's pretty clear that this is purely cosmetic, as expected. No doubt it causes turbulence, but it is styled to look like a wing that would be used to produce downforce, as opposed to a drag-reducing spoiler. As Wayne pointed out, it probably does catch air a bit like a sail. Look how wide it is where the wing meets the trunk lid! Air shouldn't exactly be moving at free-stream velocity that close to the trunk, but it would be a lot better if the wing weren't shaped to stop it nearly dead in its tracks. Now, I'm sure Honda wouldn't put it on the car if it were that bad, but one could probably do better. It would be worth some testing with and without it.

Just found a pretty good document on drag reduction: (PDF)

http://www.tfd.chalmers.se/~lelo/rvad/reports/rva2002_gr11_contribution-to-drag.pdf

Well, I felt the need to tinker this afternoon so I started playing with that wing. The result: it is now in the trunk rather than on it. I just wanted to see what it would look like and how much cleanup work it would take to make it look whole, and honestly I don't think the result is that bad. The car looks slick, for one thing, and now there's room for a little lip spoiler if I ever get really ambitious and want to either source or (more likely) design and fabricate one of my own. The wing has it's own third brake light, but Honda was kind enough to leave the original intact so that it can be connected with a three second plug swap. The real challenge comes from the fact that a) there are a total of 5 3/8" holes for the wiring and mounting hardware, and b) the paint that was under the mounting points doesn't look so hot. Some good polish and elbow grease ought to take care of the paint, and I've found a novel and reversible solution to those holes.

Home Depot just happened had some 1/4" by 1" carriage bolts, the heads of which completely cover the holes. I also purchased some rubber fender washers that have been cut down to sit nicely under the heads of the carriage bolts (sealing out the rain), and some matching nuts to hold the whole thing in place. I also bought some gloss black Rustoleum for the bolt heads to make them stand out a little less. It's not as nice as a trunk that never had a wing at all, but it's good enough for me. As a bonus the process can be reversed in about 15 minutes if I decide it looks horrid.

Side note: It's just as well that I did this because I finally figured out why I keep finding wet spots in my trunk. The hole where the wires pass through had been sealed very poorly with a dab of black gunk, allowing water to seep in. So there's a bit of surface rust around that hole that I need to take care of before it spreads. Odds are my fender washer/bolt combination will prevent any further leakage if I don't put the wing back.

Pics to come tomorrow since I lost the daylight before breaking out the camera.

hm so i guess in theory, a hatchback (honda fit) back creates more drag than a car with a sloped back like a sedan (toyota yaris). So I got to pick between coasting/fe ability and cargo carrying capability? Tough choice :P

Actually, I think it may turn out to be the other way around. Just look at the Mercedes boxfish thing, with a CD of 0.19. For all intents and purposes, you could scale that up and have a minivan-like vehicle! You can have your cake and eat it too ifmanufacturers are willing to build these things. It's far from impossible.

Pics as promised:

http://www.cleanmpg.com/photos/data/500/Slick2.jpg

http://www.cleanmpg.com/photos/data/500/Slick_Profile.jpg

http://www.cleanmpg.com/photos/data/500/Slick_Rear.jpg

http://www.cleanmpg.com/photos/data/500/Holes_Filled.jpg


The "plugs" for the mounting holes leave a lot to be desired, but I don't think it's all that bad. This is a definite case of function > form. And I'd love to say "wow, that thing was really holding me back on the highway" but I really can't see much of a difference on the ScanGauge. It's likely that the CD is a bit lower, but measurably so? Probably well within the noise.

mmmmmm nice rear end ;P i thought you had those mercedies looking accords and not the NSX inspired one. Too bad it isn't as fast as it looks ;)

hey less noise! more smooth!

hm so i guess in theory, a hatchback (honda fit) back creates more drag than a car with a sloped back like a sedan (toyota yaris). So I got to pick between coasting/fe ability and cargo carrying capability? Tough choice :P

Nah, drag in the rear of the vehicle's probably proportinal to the area of the vehicle where the flow seperates. The Insight's a hatch, but it still has a .25 Cd, better than most sedans. Probably because where the flow seperates, the insight has ~1/2 the area most other hatch's do.

Yeah, mine is the 6th generation, the coupe that looks like a Benz is a 7th gen (2003+). I like 'em both! And if it were as fast as an NSX I don't think I would ever be hitting 40+mpg segments, nevermind tanks!

Anyway, I've been trying to figure out how to rig up my digicam to photograph some tuft testing. Based on the literature research, my bet is that the current rear end shape isn't all that bad for drag due to the very small radius as the lid transitions from near horizontal to near vertical. I guess that makes for a cleaner flow separation than if it were a large radius. However, an article that was linked from gassavers (I'll find it and duplicate the link here shortly) demonstrates graphically that there is still room for improvement due to a low pressure region at the very top of the rear-facing surface. My current thinking is that a simple extension of the horizontal surface, say an inch or two beyond that radius, would do precisely the trick without being terribly dificult to fabricate and affix. My only cosmetic requirement is a shiny black paint job so that it doesn't look completely ridiculous ;).

Naturally, it would be silly to just do it and assume that it it's working without any experimentation. So the other idea that came out of a discussion at gassavers is to set the vehicle up for tuft testing and rig my digital camera up in the rear to take stills or video. (At the moment it looks like it's not capable of video output...very sad.) I'm still pretty sure that coastdown testing would be inconclusive for this kind of drag reduction, but the tuft testing would allow for a visual inference of how the air is separating and where any low pressure regions exist. Improvement to aerodynamics might not be quantifiable, but might be confirmed qualitatively.

**EDIT**
Here's that article: http://www.advantage-cfd.co.uk/Newsletter/RcarV15N3_Aerobytes.pdf
Thanks to MetroMPG over at gassavers for digging this up.

The triangular taillight design actually dates back to the '94 Accord, which was styled to look like the '92 Prelude. Although Mercedes re-popularized the triangle taillights, Honda has used it plenty in the past.

http://www.aerotech.co.nz/products/images/honda/accord94-96Mugen2.sized.jpg

http://www.prelude-fan.de/pics/prelude93d.jpg

Tuft testing is done! Here's the setup, with two tufts on the trailing edge of the trunk lid, one about in the middle, and two right up next to the glass:
http://www.cleanmpg.com/photos/data/500/Setup_528_x_396_.jpg

And two pics taken at 60mph:
http://www.cleanmpg.com/photos/data/500/60mph1_528_x_396_.jpg
http://www.cleanmpg.com/photos/data/500/60mph2_528_x_396_.jpg

To do this, I just fixed tufts to the lid with scotch tape and set my digital camera up on a crude mount near the top of the rear window. The cam was connected to my laptop, which captured images at the maximum allowed frame rate of one ever five seconds. (Not exactly a new camera.)

What does it mean? The tufts pointing consistently to the rear indicate attached flow over the trunk lid at least from the middle right out to the trailing edge. But look at the tufts near the glass. See how they're pointing different directions in the two pics? Detached flow! That indicates a low pressure region, possibly recirculation, and a major contributor to drag on my car.

From a practical standpoint, this means that a lip spoiler would have some value if designed properly. It also means that there's a good chance my factory wing was producing drag since it was sitting right in the air stream. Finally, it means that there would be some value to determining the size of that low pressure region and making an attempt to shrink it. That would require another campaign to image a field of tufts fixed to the rear windscreen.

Just found a pretty good document on drag reduction: (PDF)

http://www.tfd.chalmers.se/~lelo/rvad/reports/rva2002_gr11_contribution-to-drag.pdf

That's a great paper! It gives a good general idea of what areas to tackle first. The first mod I'm going to try is covering my existing wheels. Since I'm going to be using hypermiling techniques (lower speeds, drving without brakes, etc), the brakes should not be overtaxed by the reduced cooling.

I've finally finished and mounted my prototype spoiler, which clearly isn't doing me any good. I need to get out and tuft test like before, but I can get a good idea of the drag properties by trying to coast down the "marginal" hills on my commute, on which I barely maintain or slightly drop speed from my 60mph usual. So far it behaves exactly as before or maybe even a little worse. My suspicion is that a) it doesn't fasten completely to the trunk lid and may be catching a little air (good thing it's bolted down!) and b) it conforms too poorly to the trailing edge of the car, and probably tosses the air around a bit but doesn't really have the smooth shape required to get a clean flow separation. There's also a chance that it's generating big-honkin' vortices right at its edge, which create an even worse low pressure region than before.

I'll tuft tests for the sake of doing it, but it will probably be back to the drawing board on this one. It's as if designing valuable aero mods is supposed to be hard or something ;).