Is 42mpg too low a goal?
 

Greetings everybody.

I just made a 2900 mile trip out to visit friends in Montana. I've been here a few weeks, and I'm here for a few more. While websurfing I stumbled across the Ecomodder forum, and it got me thinking about all that gas I'll be using on the way home. So I read lots and lots of posts until I got an idea of what I'd be capable of, and comfortable with, doing.

My initial question: if I can get a bit over 30 mpg on the highway without any modifications at all, is it out of the question to expect a 10 mpg improvement from some modest aero mods?

Car is a FWD Subaru Legacy sedan. It turns 18 in October (of 2009). The good news is that I (mostly) don't care how it looks at this point (as you can tell by the pop-riveted sheet metal patches in five places that, at the time, saved me $200 in replacement fenders and who knows how much bodyshop payments).

The bad news is it's 17 years old (so no OBDII). The engine's got some wear on it (though not as much as could be, at about 189k miles), it's an automatic transmission, the speedometer sticks and it has no cruise control.

So, please let me show off my progress so far (modest as it will be; no full boat-tails in the near future, and even if I had a manual, engine-off pulse-and-glide is more effort than I feel like spending while driving).

Here's Car in as good an "original" picture as I can find.
http://i78.photobucket.com/albums/j1...g?t=1244354933
Go pop-rivets! I've had a comment from someone that it reminded them of something from Mad Max.

My first thought was, what's the easiest and most reversible mod I could do? After reading plenty of threads on this forum, I decided on a door mirror delete. PA apparently also has that "at least one mirror that can see behind the car" law, so I went for it.

You can see the mirrors are pretty big. Using two different calculation methods, I found that each mirror has a frontal area of about 40 square inches. Folding them in (which the stock mirrors do, glad Subaru realized they're huge) reduces that to about 16 square inches. Good, but not as good as it can be.

Removing the mirrors entirely reduces the frontal area of Car by 3%. Three. Percent. That brings the CdA from 6.81 to 6.61, and that's assuming no reduction in drag.

Here's a comparison pic of my three mirroring options.
http://i78.photobucket.com/albums/j1...g?t=1244380130

They came out very easily, and I'm able to put them back in whenever I want. (which is good for A-B-A testing that will come; also, I believe I need to put them back to pass inspection)

I bent up some sheet metal (to keep with the theme, of course), and with a couple of bolts, nuts and washers, there were no more exposed holes. Also, due to the way I angled the metal in some areas (and folded the edge over in others), there's not a sharp corner to be felt, even when the door's open.
http://i78.photobucket.com/albums/j1...g?t=1244351968
(I won't bother showing a picture of the driver's side here, as it's just a mirror image of this one. Get it? Mirror image? *cheesy grin*)

So what next? Thinking about it, I decided on the mod that would be the easiest (in my opinion) to do while having the least chance of degrading performance, something sure-fire that at the very least would not hurt anything. Rear wheel skirts.

Following the example of several people on this forum, I went with the aluminum bar support at the bottom.
http://i78.photobucket.com/albums/j1...g?t=1244378706
I kept at least an inch of clearance from the tires, figuring that a little too much clearance would be better than too little.

(by the way, don't make the same mistake I did, figuring that steel bars would be stronger for the same dimension of bar, and cost a bit less; I returned them because it was far too difficult to work with)

After I had the support bar and brackets in, I used a piece of cardboard to get a rough outline of the fenders, then traced that onto the galvanized sheet metal I so love and cut them out with tin snips.

For each one, I folded over a bit at the bottom for the smooth, non-dangerous edge (using channel-lock pliers to get to 90 degrees, a flathead screwdriver to get it to an acute angle, then a hammer to get it flat), and drilled the holes in the support bar.

I lined up the slightly oversized sheet metal where I wanted it, leaving about half an inch overlap around the whole curve, then marked where to drill the two center holes in the sheet metal with a pencil while an assistant held it in place for me. Drilled those two holes, then used washers, bolts and nuts to fasten it to the support bar in just those two places.

I then used the pliers to bend the edge of the metal 90 degrees to fit the curve, going an inch at a time, bending then checking the fit, unbending and rebending when necessary, and trimming the little bit I needed to with the tin snips.

I took the skirt off, further folded and hammered the whole edge, put it back in place with the two bolts, marked where to drill the two top holes and the remaining front hole on the support bar, got them drilled, then put the fender securely in place and drilled the final, rear hole that goes into the plastic bumper at the bottom.

I'm happy with how they turned out. As with the mirrors, there's not a sharp edge to be felt. They sit right inside the edge of the fenders, and they're quite sturdy, no wobble at all.
http://i78.photobucket.com/albums/j1...g?t=1244351474
http://i78.photobucket.com/albums/j1...g?t=1244355309
http://i78.photobucket.com/albums/j1...g?t=1244355333
The bottom of each skirt undulates a little, and the trailing curve at the bottom is a little severe for aero, but on the whole it has to be doing more good than harm, I think.

Next up, simple to do and likely an improvement: partial grille block.

You may have noticed I had Car sporting some nice cardboard-and-duct-tape in the last pic; I proof-of-concepted a grille block before I worked on the skirts, but I got the skirts complete while the cardboard was still on, so I posted them first.

Car has quite a bit of area open in the front, some of it completely useless.
http://i78.photobucket.com/albums/j1...g?t=1244349525

After a tip from RobertSmalls, and also looking at other posts on here, I knew that the top openings were the ones to block off completely. I did a quick patch job with cardboard and plenty of duct tape.
http://i78.photobucket.com/albums/j1...g?t=1244351742
http://i78.photobucket.com/albums/j1...g?t=1244351874
After sustained highway driving on a warm, sunny summer's day, I was stopped 30 seconds at a light on an off-ramp before the radiator fans kicked in (and of course the thermostat needle didn't budge), so I'm fairly confident I left enough of the bottom vent open (basically all of it: what I covered on the bottom was the non-functional parts).

I'm definitely going to block the middle and top radiator vents, and it will be pretty easy to do (hook the top over the back of that plastic grille, screw the bottom into the license plate mount (my resident state of PA doesn't require a front plate)). I do have a bit of a question about the bottom, though.

The bottom lip angles in in the front, so the center vent is the furthest out.
http://i78.photobucket.com/albums/j1...umper-side.jpg
The tempting way to do it is the way I have it in the cardboard mockup, just follow the slant from the upper lip to the bottom. (I'll also duct the inner sides of the opening when I do that, unlike the cardboard)

I worry that this will funnel air under the car, though I still think a tapered angle pushing air down is better than the useless parachutes that the edges of the bottom vent are right now.

Another option I'm thinking of is a short, vertical air dam extending from the furthest-point-forward part of the bumper, down level with the bottom lip, and ducting a radiator-wide, one-quarter-high-as-the-radiator opening in to the radiator. It'd be short enough to not even need support bars, just screw it to plastic bumper.

Still another option is doing a full air dam, extending vertically from the tip of the bumper down to the level of the lowest protrusion of the underbody. I worry about the lack of clearance with this option, though.

Car has pretty high ground clearance for a sedan, I seem to remember reading somewhere on here that high-ground-clearance cars should avoid big air dams?

Any and all feedback welcome. I've got about three weeks before the return almost-cross-country trip.

Mods look awesome; I really like the look.
Make sure you do all the easy stuff, like maxinflating your tires, etc. How fast did you drive on the way out? Slowing down by 5 mph will be huge.

Thanks! I take pride in Car's "intentionally unfinished" look on the patches and mods. If I'm not going to take the time and money to make everything look stock, it might as well look so bad it's good. *grin* I really don't want to try matching Car's color, and I think a not-quite-matching-and-likely-to-fade brown would look worse than bare sheet metal.

One of the reasons I got my current tires was that they had max 44 psi, so that's what I'm running them at.

A problem I've noticed though is that the centerline of the front tires seem to be wearing down a bit more than either the outside or inside edges. I've seen other people say that they've never experienced uneven or increased wear, but maybe that's because their tires are more narrow? Stock tires on Car are 185/70R14. Next time I need tires I'll be getting 175/70R14 HTR T4's. (I'm having trouble finding narrower tires than that for 14" rims that are still all-seasons and -also- don't sacrifice tire radius by a decreased height/width ratio) A centimeter narrower and 7psi higher than my current tires.

I tried to stick around 60mph on the way out, when I could. I-44, though, in Missouri into Oklahoma, is insane. I'm not exaggerating when I say that traffic even in the slow lane on that highway tends to move about 85mph, at least at the time of day I went through. And even when it wasn't congested, it seemed like people tended to drive beside each other, not planning ahead at all for slower traffic in the slower lane, so that I constantly had cars and pickups riding up on my bumper. I sped up more than a little there, but still didn't suffer much in fuel economy according to the fill-ups.

I'm thinking the gas through Indiana and Missouri is just better for high mileage. Or maybe they've already switched to summer gas, whereas PA hasn't. Who knows, I'm not about to take samples of gasoline to have it analyzed. :)

Istas -

I really like the workmanship. I may copy your side view mirror metal look. The rivets make me think of unpainted aluminum planes from the 1930's, Spirit of St. Louis and all that. I also understand the "intentionally unfinished" look. It's kind of an anti-cool cool statement.

CarloSW2

Thank you for the compliment! I never thought of that comparison before. And I'm glad to see I'm not the only one who appreciates that look. (I get ribbing from my one friend all the time)

Hello,

Is the undercarriage in the engine area lower than the bottom of the lower grill? Maybe you can integrate a chin spoiler/air dam under there? It might help a smidgen to angle the upper block back to the headlights, eliminating the side gusset triangle; forming a bit of a "point" on the upper part of the nose?

What is your tire pressure? I would try a foam gasket all along the front edge of the hood to seal the gap.

The oil pan is a bit lower than the lowest lip of the bumper. (I have a skid plate I made for the car out of some thick stainless steel, but I didn't put it on for this trip :/ ) Parts of the undercarriage are too. It's hard to get a good picture right now as the driveway I'm using is gravel and woodchip, but I'll try to get some good pics detailing the underside tomorrow.

And hmm, yeah, a chin spoiler/air dam/whatnot would alleviate some of the clearance issues I was worrying about, but still give some effect. I do hesitate to do anything along those lines, though, without having an MPG gauge to see if what I'm trying helps or hurts. I would also like to get some lower spoilers just in front of the front tires to shove the air off to the outside of the car.

I'm not really sure what you mean by angling the upper block back to the headlights. I can post a pic without the cardboard on at all if you want to sketch something quick and easy in Paint, if you wish. I'm interested in what you mean.

I have tire pressure at max sidewall of 44 psi, and am experiencing a bit of uneven wear (center of front tires wearing out faster than edges, characteristic, I believe, of too much pressure for the application).

The foam gasket is a good idea, I'm planning on getting that for the front and side gaps of the hood. I'm also going to get some clear caulk and seal up the gaps between the bottoms of the headlights and the bumper.

I'm going to do tuft testing as soon as we get a rain-free day to do so (there's a stationary front stalled over us)

Your car looks great! I like the look as well, as long as it works, who cares what it looks like ;)

You should definitely get some instrumentation. You might not be able to use a Scangauge, but as long as the engine is fuel injected, you'll be able to use an MPGuino, which is just as good for older cars.

It looks like your car has a CD of .33 (source), which is pretty good, and with your mods, you've definitely lowered it. So, aerodynamically, I think you're doing great, now with some instrumentation, I think you'll be able to notice even greater improvements.

This is pretty good work, but I don't think you can expect a 10mpg increase if you're not willing to do some EOC and other techniques. You can get away with it on the highway if you are willing to at least put it in neutral instead. Any combination of the mods I recently did gave me about a 10mpg increase from my previous norm, but I also pay very close attention to techniques while I'm driving. You might see a 5mpg or so increase, not to be a naysayer. More power to you if you can get an increase of 10. I don't know how well you can do with an AWD car either (pretty much all subies except for a few like the justy are AWD, right?) I'd suggest doing what I may end up doing, which is making a little wiper spoiler so the wipers don't create any drag - I'm currently running with only the driver side wiper, which I'm perfectly comfortable with.

As for the mirror thing, you may want to opt for the driver's side, folded in setting. Many state laws require that mirror there, but don't say anything explicit about it being folded in, so if you get stopped by smokey just apologize and fold it back out. You could even make a nice little fairing for when it's folded in like MetroMPG has (although I believe subie mirrors fold in more gracefully):
http://metrompg.com/posts/photos/mir...ded-detail.jpg

McTimson: yeah, I want to pick up an MPGuino, but one thing worries me about it; my speedometer sticks sporadically. I'm not sure if this problem is at the instrument panel, or if it's at the sensor (wherever it is). If the problem is with the sensor, getting an MPGuino won't do me much good for testing mods, and its effectiveness will be reduced for altering driving style.

Yup, I found that 0.33 as well, through wikipedia, through a link on that energy/mpg calculator that's on this site. It's not too bad to begin with, definitely, but there's also room for lots of improvement (as with most cars). Thanks for the encouragement.

Wonderboy: well, I'm still going to try for a 10mpg improvement. :) I am willing to put it in neutral for long hills, although that means kicking it out of torque converter lock. I do pay a lot of attention to that, and have gotten to the point where I know when it will kick in. I'm also trying to drive with load recently, too, which is not always easy because of traffic, also not easy because of the automatic transmission, and I'm already going at, or slightly below, the speed limit (which is of course slower than 99% of people on the highway).

For AWD, I think it was 1993 that Subaru started putting that on all their vehicles. Mine's FWD, so at least I don't have the extra weight and drivetrain losses of AWD.

*nods* I am planning on the wiper spoiler/deflector. I don't quite want to take the wiper off, wouldn't feel comfortable losing visibility for that much of the windshield when it's raining/snowing. I'll be doing tuft testing on the hood and windshield (more from curiosity than anything at this point), as well as the rear window and trunk, so I'll be able to do that again after I put the wiper deflector on and see what's changed.

As for the door mirror, as I read the PA law, the only thing I can find on it is "at least one mirror that can see directly behind the car for 200 feet", as long as it's not obstructed or broken, and I don't think other states can enforce their state-specific laws like that for out-of-state vehicles. However, it is something I'm still a bit nervous about it. Even if I can't get a ticket for it, I don't like the idea of getting pulled over and hassled.

You are correct, fortunately for me: Subaru mirrors, at least on this car, do fold in quite forgivingly. Observe.
http://i78.photobucket.com/albums/j1...g?t=1244474446

So perhaps that's what I'll do, just as insurance against PITA fuzz. 16 square inches for the driver side mirror, hugging close to the car, is still better than having 24 square inches in addition to that, sticking way out.

The speedometer is probably mechanically driven by a cable from the transmission. Most cars have a separate VSS electrical signal used for things like cruise control - I doubt that the VSS signal is 'sticking', because it's rare for an electrical signal to stick like that.

Ahh, that is very encouraging to know. I was worried the electrical signal was sent to the dashboard needle, and that the electrical sensor was faulty. But from what you said, that doesn't make a whole lot of sense.

I'll try to get an MPGuino to me before the 2500 mile trip home. Though I still have the fun of getting to puzzle out the wiring. >.< Thank you very much for the information.

I guess that would mean I might have success installing a cruise control too, though I'm not sure I'd have use for it other than quantifying aero mods.

I wonder if there's a way to scavenge and retrofit a power mirror folding mechanism from a car that has it. It would be cool to have mirrors that stay folded until you signal in that direction, and then have them automatically unfold.

That would be a cool mod indeed. They'd have to move pretty quickly to be useful, though, and I would think it would be good to check the mirror before putting the signal on. Still a cool idea. And it would be nice to have some way to fold and unfold the mirror without needing to reach outside, especially in cold weather. Maybe a separate switch? Hand crank? Lever?

I just went out and took pics of the underbody relative to the bottom of the front bumper, as NeilBlanchard was asking. The oil pan is lower than the front lip, but lower still by a bit are the two exhaust headers, and the thing I use for the rear mounting point of the skidplate.
http://i78.photobucket.com/albums/j1...g?t=1244490312
It's hard to say just how much further down from the lip the car parts are, but I'm going to guesstimate two or three inches lower.

So I'm thinking, an air dam that extends down about 2.5 inches from the bottom of the bottom lip of the bumper, all the way around from one wheel well to the other, and then a partial underbody tray that goes from the bottom edge of that to the skidplate mounting points at the lower-back of the engine.

What does anyone think of that idea?

And I think I discovered some more cause of the road noise I experience.
http://i78.photobucket.com/albums/j1...g?t=1244490178
It looks like the bottom of the car just before the front wheel wells actually encourages the flow to go up into the wheel well. C'mon, Subaru, yer killin' me! This is not a race car, there doesn't need to be that much airflow into the wheel well!

I'll plan out some front-of-wheel fairings (splitter? spoiler?) to go with the air dam, as well.

10-mpg
 

Istas,I tossed your numbers around on the calculator.A jump from 30-mpg to say 40-mpg would be a 33% increase.If say,your mpg were at 55-mph,then to get an extra 10-mpg would require a 66.6% drag reduction if you were going to use streamlining exclusively.That would require dropping the Subarus drag coefficient down to Cd0.11.If you can imagine GM's Sunraycer of 1987,that's what we're talking about.------- Low rolling-resistance tires would help remedy some of the challenge.------- I like the mods you've done.Consider a bi-wing,or tri-wing rear deck to clean up some of the turbulence behind the roof and extend it out even with the length of the rear bumper.Ford's German Merkur Xr4Ti will give you some idea of a multi-vane spoiler.And use Search EcoModder to find Project: permanent Kamm back for Metro to find a template to gauge angles and curvature for the spoiler.

Aerohead: Thanks for bringing that to my attention.

You're using this calculator, here? Aerodynamic & rolling resistance, power & MPG calculator - EcoModder.com

I put my car's Cd (0.33) and weight (2830, plus 200 pounds for the driver) in there, along with the original area (20.63 sq ft), as well as the rolling resistance measurement I got last week (0.0103, I'll be re-doing this to check), then dropped the engine efficiency to .21 to bring the numbers more in line with my experienced fuel mileage. (derived by fill-the-tank-and-calculate, but better than nothing)

Using that as a baseline, it spits out 35.96 mpg at 60mph (my preferred compromise highway speed). When I drop the area to 20.13 (which is what it is after passenger mirror removal and driver mirror fold-in), and plug in a Cd of 0.23, the calculator spits out a fuel economy of 45.71 mpg at 60 mph.

I know all of these numbers are very rough, but a Cd of 0.23 is a lot less intimidating than 0.11, heh heh.

I've also taken the air conditioner belt off, and will, when next I need tires, be switching from 185cm-wide 44psi-max tires to 175cm-wide 51psi-max tires.

I'm likely going to pick up some airtabs and see how that helps the airflow down and over the rear window and trunk.

I've also pondered just riveting a flat sheet-metal spoiler onto the back of the trunk, to help airflow separation.

That Merkur looks interesting, I think if I was going to try that route I might just opt for a full Kammback instead. Which I might do, but that would be a daunting task for me at this point.

calculator
 

Istas,I didn't use the calculator,just ran numbers for aero.Cd 0.23 is realistic goal as I hit this with the CRX.I hope you've seen basjoos AeroCivic,he's at Cd0.17(wickedly delicious!).I did a thousand dollars in wheels and LLR Michelins and saw absolutely no improvement in MPG.I also did the AC belt delete with no reward.The airtabs might help you,as they were developed for cars like yours.I don't have any first hand experience with them.The flat sheet extending back would have to help.The Kammback with all compound surfaces are the most complicated to form,so I thought the multi-wing spoiler might be more doable.I don't understand the mention of engine efficiency at 0.21,can you help out there?------ Of course the stickys are a good source for all the mods we know can contribute to better mpg,so indulge yourself there if you haven't already.The Legacy is a great car.I've got a lot of time in a wagon model and I think it will be a great project for you.If you can get to 45-mpg it would be a real coup!Keep us posted.

Well, the default engine efficiency that the calculator lists is 22%. I don't have any idea on what my engine efficiency is (or the drivetrain efficiency, or 'parasitic overhead'), but when I put in the other numbers I have a rough idea for (weight, Crr, Cd, A), the gas mileage seems to be a bit higher than what I find I get. So I adjusted the engine efficiency down a tad to make the estimates more in line with my observations.

I'm also assuming some loss due to inefficient driving tactics, as I don't yet have an MPG readout to help train and guide my foot on the gas pedal. I'm intimidated by working with Car's wiring, but I'm going to get an MPGuino and try it out.

I have read basjoos's Aerocivic thread and website, and I find it amazing. I never would have thought somebody could get away with doing something like that to a vehicle. His efforts are instruction and inspiration wrapped together. (so is AndrewJ's boattail, also very nice looking)

Hmm, you did standardized, instrument-backed testing and the AC belt removal didn't help at all? In that case I think I'll put it back on. (I hear that having it not rotate for an extended time can harm the compressor)

I'm still a bit leery about doing the aerotabs without having a way to quantify testing them, but I'm going to just take a leap and assume (until I can test them) that having just enough of them to reattach airflow to my rear window will do more good than harm.

The flat sheetmetal ledge spoiler will look interesting, and be easy to do. Your words are encouragement. A multi-spoiler would certainly be interesting.

If I do decide to go for a boattail, at least I have this advantage: the back edge of my car is already lower than the back edge of a hatchback.

I look over those two stickies every once in a while, re-evaluating what I'm willing to do as my boldness waxes and wanes. I better order the rollers for the front wheel skirts now, so I have them in time to work with them before the trip home.

I also am working on getting some CFD software working on my computer. I have a start, but it's tough diving into programs that are so complex and powerful... Overload of information to deal with. It would be excellent to test the trunk spoiler(s).

Do the MPGuino first. I've not heard positive results on the forums from aerotabs, but if they did work for you, it would be such a subtle change that you probably wouldn't be able to tell just from filling up. Doing a grill block is a no-brainer, but having that instant readout is important to quantify the smaller mods you'll probably have to do to get to 42 mpg.

Hi,

Icksnay on the AirTabs-ay -- I doubt they will help, and they probably would hurt your FE. I've tried them, and it was a waste of a fair bit of money...

Here's my sketch of the grill block wedge I'm suggesting -- it's similar to mine:
http://img.photobucket.com/albums/v7...lWithLines.png

Some more air may flow to the sides, though probably the difference is subtle. It also would not block (the little bit of) light from the inside edge of the headlights. The blue lines are foam gasket or tape or caulking to seal the air gap -- the front edge of the hood is the more important -- the air flow velocity there is higher and keeping it out from under the hood is important.

Here's my Coroplast grill block:
http://img.photobucket.com/albums/v7...d/DSC_2366.jpg
Your Legacy has less curve to the sides (along the brown line on my drawing) so shedding air to the sides may be less important than on my xA.

LOL. Make sure you draw the teeth on the finished product. Makes the car look like a hammerhead shark. :D

FYI, that is called tessellation! :p

Hi, Istas: impressive the way you've jumped in with both feet!

FYI, these two "controlled as possible" A-B-A instrumented tests of airtab vortex generators on 4-door sedans showed no significant change in fuel economy:

• http://ecomodder.com/forum/showthrea...olla-2390.html
  tested at 90 km/h (~55 mph)

• http://ecomodder.com/forum/showthrea...ting-7864.html
  tested at 75 mph

One of these should bump you up to the 40's mpg:

http://cruisenews.net/whacked/versa/02V2009/2.jpg

Metro MPG: Thank you!

Yeah, I've read both those articles, but both of those were also done on cars that already had a smoother transition to the rear window, and the window at a more gentle angle. I think both the 2006 Camry and the 2003 Prius are significantly better in this attribute than Car, a 1992 Legacy:
http://i78.photobucket.com/albums/j1...g?t=1244512315
I'm thinking that maybe, maybe, it might be more effective on cars with steeper back windows. However, I won't even think of trying the airtabs until I have an MPGuino installed and calibrated. And maybe when I do, since they didn't work, I can borrow those loners that were floating around? *hint hint*

Thank you for discouraging me from just buying them because I could.

Clev: Hahaha! Yeah, because the first thing people think of when they think fuel efficiency is an upside-down hammerhead shark. *grin*

Neil: Very slick job on your grille block. That's really smooth. And yeah, I can see how it definitely benefits your car, with the swept headlights. It's definitely an interesting idea, and I will keep it in mind, but two things: one, it takes away being able to use the front license plate holes as mounting points and it'd be a lot more complex to secure it by just using the grille alone (somehow), and two, it takes more finesse to shape it the way you have it. I'm going to go with the... hmm, I'm paraphrasing here, but I know I read it somewhere on this forum, "first, do it. Then do it better". I'll make it the simple way, then possibly refine later. I'll definitely keep your shape in mind when I do.

And yeah, I'll be using foam tape or something similar under the front and side edges of the hood, and clear caulk around the edges of the headlights (and possibly the grille blocks).

Thanks to all of you.

skylark: Wow! I hadn't seen that job before. That's your car and mod, right? Is there a thread showing how you built it? Have you tuft-tested the top of it to see where flow rejoins it? Can you still access the trunk?

That looks like a viable option for me (I especially like the way it uses the stock rear window, which will aid in keeping it clean in the wintertime rather than having a homemade plastic viewport, and probably annoy Smokey less). Only thing is I would loathe loss of trunk space. I guess it wouldn't be too hard to just start it after the trunk, or make a panel over the trunk raise up with the trunk lid.

See, I would've never considered that. Never would have thought of a boat-tail that didn't start at the rear edge of the roof. Awesome design.

Edit add: Ahh, it's in your sig. Didn't even look before. *reads*

Tuft testing!
 

I likes me some testing with tufts.

The documenting didn't go as smoothly as planned (I do not have video of mirror out, but I do have pictures of mirror out), but I did some tuft testing.

Main goal of the tufting was to see if the flow, once separated from the roof or upper rear window, re-attached at all at the back of the trunk (I had figured it separates from the rear window, and it does)

Another goal was to document the turbulence effects of removing the side-view mirror.

Tufts!
http://i78.photobucket.com/albums/j1...g?t=1244604635
I like tufts.

Mmm, laminar. Nothing out of the ordinary, just wanted to see what it looked like. I do plan on putting a deflector just in front of the wipers.
http://i78.photobucket.com/albums/j1...g?t=1244605152

Rear window. As I thought, total separation. Even the top row of tufts wasn't aiming the right way.
http://i78.photobucket.com/albums/j1...g?t=1244604638


I could use some interpretation help with the trunk. It looks to me like the airflow tentatively rejoins the trunk a bit, but that it's not smooth at all.
http://i78.photobucket.com/albums/j1...g?t=1244604637
That second to last row was swinging left and right, both sorta towards the middle, but not aiming forward (towards the front of the car). I'm not sure if that last row is pushed off the trunk, or if it's being pulled out by the turbulent airflow behind the trunk.

Now, for the side. Mirror in "normal" position.
http://i78.photobucket.com/albums/j1...g?t=1244604649
http://i78.photobucket.com/albums/j1...g?t=1244604650
http://i78.photobucket.com/albums/j1...g?t=1244604652

Lots of movement on all three rows. I guess a 40-square-inch-front-area mirror will do that.

Second, folded in.
http://i78.photobucket.com/albums/j1...g?t=1244604657
Not as bad, but still turbulent down the whole side. Bottom row was calmer than the middle.

Last, mirror off completely.
http://i78.photobucket.com/albums/j1...g?t=1244604659
Mmmm, two smooth rows. The top row, though, is still angled up. The angle gets progressively less, it's barely that way in the back window. That top row's wiggling, but the other two are pretty smooth.

I guess the a-pillar's creating a sort of vortex just behind it against the glass, and that it's pulled along next to the upper windows, so it's angling the tufts up that way. Kinda nifty, actually. This would be worth testing some turbulator tape on the A-pillar, another time.

I didn't see much difference on the rear window or trunk with regard to the different mirror configs.

I have two videos, the mirror off and the mirror folded, neither very high quality and both from inside the car. I've never really posted to Youtube before, but I might try it, maybe. Not just now though. If the quality I got were higher I would be more inclined, but I don't have a video camera, so we went with what we had.

rear spoiler
 

Istas,I made a thumbnail sketch off the computer screen of the back of your roof/rear deck.From the top of the backlite header to the rear edge of the trunklid,I'm measuring about 18-degrees.That's a bit steep for the air to follow and I don't believe it is re-attaching.You might make a simple rear spoiler which angles up at no steeper than a 30-degree angle ( as measured from the horizon) projecting up and to the rear until it intersects an imaginary line down from the roof at no more than say 13-degrees.It would be a cheap and dirty way to get your flow re-attached without adding adding any significant length,your turbulent wake will be reduced and it should show up at the gas station.

aerohead: This is why I was thinking those airtabs might actually be a benefit in this situation. The two reports on airtabs that MetroMPG pointed me to were done on an '06 Camry and an '03 Prius, both smoother cars than mine. But even if I do try them (and really, given the design of the aluminum vortex generators, it shouldn't be too difficult to mock up some of my own on the cheap to get at least a rough guideline), I'm going to wait until I have my MPGuino installed and calibrated before I even think about testing VGs.

(I don't trust that picture either, even I'm not sure if it's level, and at any rate I get the feeling my camera (which is more than a few years old) warps pictures.)

Also, getting cfd software running will help too.

A rear spoiler off the edge of the trunk seems like it'd be really easy to do. Also, I get the impression that adding that sharp transition at the trailing edge of other parts of the car would help as well? like on the sides of the rear bumpers? (it's rounded there too)

Mmm, gorgeous car. Just a few comments:

On the 1995+ Legacy, there is no speedo cable. Since very little else changed from 1994 to 1995, I'm guessing you don't have one either. I actually have a spare instrument cluster from a 1995, but I don't think it'll do you any good.

For CFD software, if you're willing to pay thousands, SolidWorks Cosmos FloWorks (not FloWorks Express, which is pretty much a teaser) is easy to use, but it doesn't have as many features as I'd like. Sure, you can do simulated wind tunnel testing, draw streamlines over your car and all that. It can report the force exerted on a single surface by the flow. But it wasn't able to tell me the Cd of the model I drew, and I really couldn't get it to quanify the difference between my model with and without mirrors or wheel skirts.

OpenFOAM seems like a respected open-source CFD toolbox, but you'd better be a computer programmer to get it to run. You'll also need to be running Linux, in a VM or otherwise.

What speed did you do the tuft testing at?

The speedometer will sometimes be responsive showing changes in velocity, but sometimes it will stick, then suddenly drop (or rise) 10 or 20 mph. McTimson said earlier he thinks it'd be weird for an electrical signal to be sticking (though I'm thinking the instrument panel might be at fault in that case), but I really don't know one way or the other what's in there. However, cruise control was an option on first gen Legacys, so whether it's cable driven or electric, I'm -hoping- there's a VSS somewhere that will give a consistent, reliable signal. Confident enough that I'm going to be getting a pre-assembled 'Guino.

I have Ansys CFX 11.0 installed on this computer, as well as SolidWorks 2007 for 3D modeling. There's still so much I need to learn in both, though, that I haven't even yet figured out how to put anything I create in Solidworks into Ansys for testing. Haven't given up yet though. (encouraging that there's an Ansys menu option right in Solidworks).

The tuft testing was done around 60-70 mph (speed limit is 70). It was hard to remain consistent because it's really the only road for traffic in the area, so I had to keep moving fast enough to not get passed in a passing zone (which would then mess up airflow in front), but there were also cars pulling out in front too. It wasn't the most controlled of tests, but I now know some things I didn't have any idea about before.

that car in maine bent in half a decade ago. :confused:
what is up with the brown scoobs anyway.best give it the mismatched paint before no paint at all, the ecu could fry in a thunderstorm (resistor R1 goes first).

anyhoo, do you know this cars mechanics, is it the only subaru in history with the backwards firing order? (it came in that generation legacy). 189k miles with an egr hooked up best be letting that engine get as much air as possible....I am quite serious. keep a window open on the long run. In fact this car is disturbing to think about. Go find a loyale and weld the rust out ..you'd be better off buddy. :thumbup:

Tell you what, bgd73, you buy me a car with decent power that still gets over 30 mpg on the highway, that'll last me 4.5 years and 94,000 miles, and I'll drive that when this Legacy quits on me.

I see you said something about aluminum VG's. I could not find any info on the aluminum ones ( I am tired its almost 1am). I went looking for the ones on the Mitsu EVO. They are a totally different shape than the ones most people try on here. I would think this is a better shape because Mitsubishi actually owns a wind tunnel or two as opposed to the company who makes those other ones. LOL I have been wrong before though LOL. I think the back of your cabin has a similar angle to the EVO.

Took me a while to find them. They may not be aluminum, I may have gotten that incorrect. However, they are metal, and very simply constructed:

Fuel Savers

I remember reading about the evo's VGs, that they worked when as tall as the boundary layer and angled 15 degrees off from the direction of airflow, or something like that. If I recall correctly from the comparison I read (I've read too many pages over the past couple of weeks to remember where I saw it), the metal "fuelsavers" are inferior to the airtabs in their actual effectiveness... But still, both the "fuelsavers" and the mitsubishi VGs are a simple enough shape. It wouldn't hurt for me to mock up some cheap ones and tape them to my roof to see if I can manage to make something that does anything at all.

However, I will not even bother with this until I get my 'Guino installed and calibrated.

Yeah, they do have a rather steep window and short trunk, don't they?


I got the tuft testing videos up on youtube, but for some reason the framerate when uploaded is abysmal. Coupled with the bad image quality I don't even want to link them.

Just ordered an MPGuino. Here's hoping I have a VSS wire, hah hah.

Oh yeah I remember seeing those in another post. Good Idea to wait for instrumentation so you can document, document, document, compare and document. You are really researching this I see. Keep up the good work. I would tuft the rear window when you try the different styles of VG if possible.

keep up the good work
Dean

Of course, QLTD. Tufting's fun. :) Thanks for the encouragement.