Kammback doesn't work on my sedan
 

I made a clear Kammback for my 94 Mazda Protege.
( two pieces of plexi glass joined in the middle )

Unfortunately ( or is it luckily?) my coast down tests confirmed that I should take it off the car because it made it worse!

I have an established coast down section near my house where at a certain
spot cresting a small hill at 59 MPH I pop the car in neutral and about .6 of a mile later I have another landmark ( I call it checkpoint A ) where I consistently get a coasting speed of 39 MPH.

Did 3 runs with the Kammback on and I got:

1st run: 38 MPH
2nd run: 37 MPH
3rd run: 38 MPH

Then I took it off and did a final run and you guessed it,got my 39 MPH without the Kammback on.

Did tuft testing to see what's going on and I got the wiggles on the sides.

The top side appeared to have better flow.

Short video:

http://www.youtube.com/watch?v=9WCmnwAjyus

Did I go too aggressive with the angles? Perhaps I did even though they are
not exceeding 15 degrees on either top or the sides as of my last measuring attempt.

How do you guys post pics within the post? ( not thumbnails )
And how do you embed a youtube video? (not link)
Thanks!
Barna

To me it looks rather short. I know you want to open your trunk, can you make a seem and have the Kammback hinge up with the trunk lid? It also looks like the transition from sides to the top is rather sharp. This might be causing some issues.

What is the angle from the roof to the back of the trunk? you might be making your wake longer with the Kammback (the air leaves at a shorter angle).

A thought on you testing, can you come to a complete stop? Like cost from the hill to where you stop. On my testing I could not make a good comparison until my speed was down to about 10MPH (but the real data was the differences in total coast length).

Ok I see what you mean by the shorter/sharper new angle for the air to leave.

I never measured the angle between the edge of the rear roof and the trunk lid.

I thought my original wake would be as tall as the roof is and as wide as the width of my car but apparently that's not the case.

My "checkpoint B" is another .6 of a mile away ( normally doing 27 MPH ) and after that there is an uphill that I would never make so I could come to a stop but I'm afraid by then the test would be a crawl, easily effected by wind.?
Maybe I will give that a try if traffic allows. Thanks! : )

3dplane -

*IF* my center-of-camber selection is correct, then I would make the claim that the angle is too steep and the airflow is not staying attached :

http://ecomodder.com/forum/member-cf...otege-1994.jpg

I agree with Weather Spotter that a longer kamm-back would be better, but I also think that if you get the angles right, it should still be a net benefit. I think that aerohead had a table of "diminishing returns" of a boat-tail. Obviously a full boat-tail would be best, but pragmatic concerns make it necessary for ecomodders to figure out how far they want to go.

But I'll let the real aero-experts chime in.

CarloSW2

Can you find a back road near you that is flat (or close to it) for about 1 mile. Then set cruse to 55MPH (or the speed limit) go till you hit a known point (drive way, power poll.etc) does not mater what, just so it stays the same. At the mark put it into N and coast to a stop. Mark the spot with some chalk (use one color for A runs, a different one for B). when all the tests are done measure the distance from start to the first mark. then measure from there to each other mark. this will give you ABA data which is way better in terms of useful data.

What slope is your kammback?

-Weather Spotter!

The slope on the top part is about 12 degrees. On the sides even less but
it was challenging to measure because even though the car is boxy looking, it becomes quite curvy when trying to measure angles.

In one of the pix,I made sure to snap the photo with the edge of the roof on the house in the background to be aligned with the top of the Kammback.
Although the front of the car leans down about 1 degree compared to the house in the background.

I am absolutely confident that my test results were true. I get to test my
coast down site every day coming home from work and today I got the same 39 MPH that I always get on a valid run! ( little or no wind, no traffic etc.)

Barna

My aero eyeball says that the angle [12deg] in the picture is to great. although most people on this website us 10 deg as the bench mark I use the slope of 1 in 6 which is about 9.5 deg. you need to keep in mind that this is the maximum slope! So if you are at 12deg. then it is too steep. An other consideration is the lenght to width ratio of the object [car] the drag decreases as the ratio aproches 7 to 1 then increases as the object gets longer. I agree with those who have said that longer is better, and if you reduce the slope to something like 5 or 6 deg and then clear the open trunk, you may find some improvement. When in doubt keep testing...

To measure the angle this is what I did. I took a 4 foot level and had it on the car top, I held it level and measured down off that. I would try a lesser angle (say 5 or 6 degrees) and see what happens.

As for your data, I trust that you are not messing the numbers, I just nodiced in my coast down tests that rolling speed was not a good indicator (we are looking for a very small change 1-4%). Where I drove on the road (crown, side, middle) messed with the rolling speed from moment to moment and run to run. This means that coasting to a stop was the only way I could really tell. I did have a good landmark (intersection) that I would roll through (I had right of way) at about 8 MPH (5 on the runs with out my Kammback, 8-10 on the runs with). any sooner than that and the data had too much variation.

If you can take the hour needed just try the full coast down. If you get the same results than all is good, if not you might have found a better testing method.

Barna,

Man, That is a SWEET lookin Kamm, your fabrication is great on that. In looking at your car's basic design, it is decidedly unaero in the sense it has very little roundness to the roof line, and the transition at the top of the front windscreen to roof is quite abrupt by more modern standards as well as the rear window to roof, That is a cd killer.

I think you're definitely on the right track with the Kamm thought, but you either need to raise the rear edge a bit, or soften the transition from Roof to Kamm by making the shape of the Kamm rounder. You need something here to create a clean detachment for the air going over the roof. I think now your problem is you still have a rather abrupt change in the air from roof to Kamm and it isn't working.

OK, I wrote the above then went and messed around with drawing things over your pics and those of another Protege I found online.....I have to say I'm stumped, it looks all the world to me like you're at 10° on your Kamm, that is how I'd make it too. It could be a bit flatter, but I can't see that making a huge diff. I wonder if you aren't too aggressive on the sides? I don't know how you could modify it, but if you could make the sides less angled in by making the top trailing edge a few inches wider, it may help. I'm thinking you may have hit a "Sweet spot" (as in bad) for creating a large vortex generator. So by making the slope less steep a tad, maybe raise the trailing edge by 1.5" and wider by 3 inches or so, you'll get the clean air release your looking for without some weird pressure differentials creating a pair of horizontal tornados that you're dragging around.

Another thought is your speeds for testing. Is there someplace you can go to get up to 75 or so to start your test with? Since higher speeds really magnify aero problems / benefits, you may get different results starting at higher speeds. (Of course this is coming from a guy in Montana where I'm road kill if I drive 59 MPH where speed limits on our highways are 75 and cops don't even quit chewing their donuts if you pass them at 80)

And to answer your earlier question, seems like you got the You Tube thing worked out. To post pics without uploading them to EcoModder as attachments, I post them on a free web site called TinyPic. Then I copy the little box on there with the "IMG code for Forums & Message Boards" and paste it right into the post.

Like this

http://i49.tinypic.com/wuj96w.jpg

Hint, if you right click any pic then hit "Properties" at the bottom of the pop up menu, it tells you the source of the pic.

Great job again on your Kamm.

And finally (Man, will I ever shut up?) How in the wide wide world of sports did you make the Kamm? I'm considering doing something similar on my 04 Civic. And how much did the plastic cost?

ChazInMT!

Hey! Thanks for the tips on the pics and the complement on my dysfunctional Kammback : )

If anything is wrong with the Kammback itself,then it has to be the sides because that's where the yarn wiggled a lot. On the top they stayed put,straight back.

I could still relax the radius and decrease the angle on the sides but honestly I don't think I'm going to play with it because I think Weather Spotter hit the nail on the head when he pointed out that I might have just gave the airflow a new sharper angle/ larger wake as it leaves the back of the roof and now it has no chance to "touch off" on the trunk!

Sort of like if the trunk area was a boat tail and now it is chopped off!
( Just my speculation )

To answer your questions: The Plexi sheet was a 32"x44" sheet about 2.5mm thick ( .1" ). Cost $33 at Ace hardware. Had it been a longer piece,I believe I could've made it diagonally out of one piece not having to join two pieces.

How I made it:

I started with making a large enough piece of paper using hot glue to attach pieces together and also in two layers so it's not flimsy. If you have cardboard it's probably better!
http://i952.photobucket.com/albums/a...k017Medium.jpg
I taped that to the back of the roof and propped up the back side on the rear window in a fashion that represented my Kammback.
( eyeballing the angle and let it droop down the on the sides )

Then I dug in with my finger nails around the perimeter of the rear glass and followed the outline with a marker so I can cut it out along the line later.

I also made lines with the marker at the radius of the bend so I know where the bend is going to be even when the paper is laying flat.

Chop chop,now I have a paper template. Next,transfer that over to styrofoam insulation sheet that I had laying around.( about 5/8" thick)
This is going to serve as a mold or plug actually for the plexi Kamm.

Cut the styrofoam to the shape and size of the paper template.

Draw the lines of where the bend radius is going to be on to the styrofoam.
( these were not parallel lines in my case because the radius at the roof
end was larger than the radius at the back end)
http://i952.photobucket.com/albums/a...k018Medium.jpg

We are going to have to bend the foam without braking it!
- I used a dremel tool with a cutoff wheel to cut grooves into the inside of the foam along the lines for the radius. ( actually any kind of wheel works,
at 25 k rpm or what ever it turns anything will just melt that foam right out )

Only cut about half way into the thickness of the foam!

Here is a trick! Put tape on the other side of the foam ( outside of bend) opposite of the cut grooves where you removed material to allow bending!

Now you can bend the foam and it will not brake!

http://i952.photobucket.com/albums/a...k019Medium.jpg

Try it on the car to get the idea of how much to bend and to keep it from springing back towards flat, I had my glue gun ready and glued a cross brace piece of foam along the back side ( right above the back window ) to lock the angles in place!

http://i952.photobucket.com/albums/a...k021Medium.jpg

So now we have a foam Kammback! Could actually start coastdown testing with it but it's so white : )
I actually should've tested this before I made the actual plexi Kamm but I guess I just wanted to bend some plexi glass anyways.

-Cut plexi glass according to paper template! ( if you have big enough piece)

-Lay plexi glass over the styrofoam structure and keep it aligned at the edges in the next steps!

http://i952.photobucket.com/albums/a...k022Medium.jpg

- Sneak past wife or girlfriend and unnoticeably borrow her hairdryer!

Actually I have a heat gun but two, or one that puts out some serious heat would work better. Mine was a little wimpy.

- Start heating the sheet where the radius begins on the inside (top of Kamm) Run back and forth along the line of bend until gravity starts bending it down,then continue heating the plexi where it now touches the tape actually that you put on the outside of the foam.

Don't stop moving the heat gun! There will be an ugly spot there!

-Work your way down and keep heating at the new touch point/lines as it
gradually bends conforming to the styrofoam structure.
It will need a little push help bending unless your heat source is blistering

- Repeat other side.

Now go try it on the car with a piece of clear tape and let out an audible "Yaaay"

- Put the hairdryer back where it was taken from : )

Well that's how I made it!

Barna

EDIT! I stuck some images in the story but the pictures were taken after all was done and not step by step as the building process went on. Gives a better idea though!

Nice how to!

Do you have pics of some of those steps?

I would like to add it to the wiki, any objections?

Don't be too discouraged with the results, I am guessing it is an improvement, even though it didn't seem like it. The only way that you are going to screw up your aerodynamics by adding a sweet kammback like that, is if you had such good tail end aerodynamics, that adding a 10-15 degree extension somehow disrupted the separation point... you could test where air is separating from your rear window with another tuft test on your rear window. If it separates above the lowest point of your Kammback, the Kammback is probably helping, if it separates below, it's probably hurting. By looking at your car's tail end, I don't think you have attached flow on your rear glass... it looks too similar to the rear end on my car! The angle looks good, it is a similar angle to what I used on my SedanKamm, and mine was a big gain! Mine was 4' long, and good for a 6% increase in Fuel Economy.
From your video, you have attached flow, which is what you want. This means you do not have to go more shallow on your angle, and could in fact probably go steeper! (I would keep it where it is) In all likelihood, the wiggle you are seeing on the sides is from having mirrors sticking off the side of your car.

Good point about the mirrors causing the movement!

I think that the testing is the culprit (might not be, just my thought).

For fun add 2' to the Kammback using cardboard and run the test a second time.

I assume you mean 2 feet (2') not 2 inches (2"). And yes, that would be a neat experiment!

Weather Spotter!- Thanks! Yes of course you can add it to the wiki!
I made some pictures of the finished pieces and edited it into the post!
It still shows what I'm talking about.
Feel free to modify it if you want to take the jokes out or spelling etc.

Wyatt! - Man you did not play around with the size there! Awesome and it works great too! Very nice smooth transition from roof to Kamm as well!

I will do another tuft test on the car without the Kammback like you suggested! It makes sense to do that!

Thanks to all for the technical tips and good thinking!

Barna

lost you re-attachment
 

3d,if I had to guess what happened,I'd say the the Protege originally had the typical notch-back downwash which would have allowed the sepatated flow from the roof to re-attach at the back of the trunklid.
The Kamm-back is causing an 'overshoot' condition,directing the air out into 'space' where it cannot re-attach,and consequently increasing the wake size and lowering the base-pressure.Both bad.
I like what you did with the plex.I would either extend it further back,or extend the back of the trunklid back and up at a 30-degree angle ( no steeper! ) with a spoiler to give the air a place for re-attachment.
Some cardboard and duct tape should make a showing on your coastdown.

I been thinking about this some more, I wonder if ending the Kamm before the rear end of the car isn't such a good plan, here's why.

If you create lower pressure over the area above the trunk, the higher pressure air along the sides is going to want to fill that void more so than normal, so it will be pulled upwards, and by the time it reaches the downwash from the roof, it will be pushed downwards. This may be creating larger vortices trailing off the rear areas above the tail lamps which would be a real drag.

http://i36.tinypic.com/244qp6u.jpg

Like Aerohed says, by making it longer and making the sides taper in less as well as keeping the sides in place further back, I think this would make the Kamm effective. Also, it seems anything you do to radius the surface will help the shape create a lower Cd too.

your how to looks a lot better with the pictures! thanks for updatting it.

I updated the wiki page with your info:

3dplane's Kammback how to - EcoModder

After digesting this for a few days I have to say yes,that's exactly
what happened what pretty much everybody is describing here: Too short...hm I've heard that somewhere before.:)

If I made a "Priusback" like Wyatt, it would absolutely be an improvement to the
original Protege. At the expense of not being able to open the trunk or having to deal with some kind of hinge system to allow it to be opened.


ChazInMT! May I ask what editor did you use for that swirly arrow on the picture? Great way to explain things!


Weather Spotter! Thanks for adding the procedure to the wiki!

Barna

glad to help, you can add stuff to the wiki just like I can, but I can do it if you want.

Hey Barna,

I'm a goofy wonk for the Draw feature in Microsoft Word. It is quite a powerful little drawing tool once you get the hang of it all. I mostly use the "Autoshapes/Lines/Freestyle" to make a very rough shape, then I right click the feature, select "Edit Points" then right click the points to change them into corner points, then you can make any sort of curve you want out of the lines by manipulating the "guide wings". I can trace a car almost perfectly in like 5 minutes to copy its rough shape. And working with the "Format Autoshape" you can manipulate and measure angles to 1°. (Make the degree "°" by holding ALT and type "0176" (zero one seven six) when you release ALT, Whuuups there it is....) I “Insert Image” to get the picture into the Word Doc that I want to work with. When I’m done, I then convert the Word Doc to a PDF using Adobe, then Save that as a JPG. And Voila, I have a handy modified picture. I crop & resize the JPEG pic using Microsoft Office Picture Manager…..I hope I’ve told you all this and you have a PC not Mac…..Oh well, others may dig this. I keep meaning to sit down and do an extensive “How To” on using the tools I’ve mentioned, maybe in December.

Also, I'm toying with the idea of making a “Prius Back” for my Honda Civic, look for a junkyard trunk lid, then you can get medieval on the junkyard lid mounting whatever you want to it, however you want, and you still have your OEM lid to put back on to normalize your car if you need to. Heck, you could get one that's in awful condition (as long as it seals the trunk area) so you may get a good deal on it. You could easily come up with a quick release and/or hinge system for the Mod to be able to open the trunk if needed.

Charlie

3dplane -

I did (what I hope is) a closer analysis. When I shift the center-of-camber forward, I get a 7.7 degree slope from the edge of the roof to the vertical line denoted by "30" :

http://ecomodder.com/forum/member-cf...speculated.jpg

When I look at the profile of the Protege, I *think* I see a roof that is taller in the back than other cars, most probably for the sake of rear seat headroom. Therefore I think that your stated slope of 12 degrees is too steep.

My Protege image is from tirerack.com, so it is decent but also subject to error. I had to rotate it 1 degree clockwise to make the tires line up.

CarloSW2

if you wanted to do a full cover cam back, would it be possible to make it sectional (3 or more pieces) almost in an accordion style ?
each piece from the roof would overlap the next piece in line as to avoid any parachuting.
this way, you could get the advantages of a full camback and still have the use of your trunk.
see weak mockup below

http://img2.imageshack.us/img2/5456/...k007medium.jpg

Im curious about your results.
plus, what sort of mpg are you seeing with no mods vs with ?
I have a 94 protege, 5 spd, no AC (from the factory) that Im going to be doing some aero mods to as well.

d0sitmatr -

That's really interesting. Conceptually it works, but I suspect the devil is in the implementation details. Would there be a single "hinge point" for all the pieces or multiple hinge points or ?!?!?!?!?

CarloSW2

One thing that's crossed my mind is kammbacking sedans without boattailing them increases the size of the rear area, thus increasing the size of the wake. Hatchbacks can get away with it because their rear ends are already large, and the kamm reduces it. Since airflow still reattaches on the trunk, the "suction cup" isn't as large, even if the air is perfectly clean.

After removing my partial kamm, I couldn't tell if it was even doing much good.

thats a good point, sort of along the lines of how adding a full camper top to a truck takes away from your mileage by creating a larger suction at the tail end.

I went out and looked at my protege real good after my post, and all in all, the sectional kamm wouldnt really be feasible, not with keeping the lines clean and sloping in towards the middle point.

I did think of another method, which would be a quick detachable kamm, if its made out of light weight materials, it wouldnt be too much of a hassle. the only problem I can really see is how to attach it at the roof line and still give you the benefits.

what Id really like to do with mine would be an extreme mod to the rear end, both the window and trunk, which would be to make it into a hatchback such as the 94 escort 5 door HB had. but if I was to go that route, Id just buy an escort 5dr HB :D

Would it make sense to extend the trunk surface? I've seen sedan-shaped drag cars with enormous extensions that continue the line of the trunk. I think the vortex created under the extension might be more beneficial on sedans where you want to maintain easy access to the trunk.

@d0sitmatr! Hey! Before any mods,absolute factory I could hit 53.5 as best MPG and had several tanks of 52+ (summer).
After these mods: Larger wheels (6% GPS confirmed ), holes in wheels blocked off, airdam ,upper grill block, seal gaps at hood,between frt bumper and grill,lights, removed A/C / Power steering belt,..I can't think of any more.
Now I was able to hit...53.75 MPG as a new personal record. A whopping 0.25 MPG increase.
I got to admit that this maybe helped me stay above 53 MPG for 5 tanks in a row as I was trying for a new personal record, but the numbers sure make me look like I was just wasting my time with the mods.

This is also proof to me that the single biggest mod by far is adjusting the nut behind the wheel!
As boxy as the car looks I think we underestimate the aero of it!

Barna

Hey Barno, thanks for the information !

now I know something is wrong with my protege (or the wheel nut is bad...) ;)
there is no way Im coming close to that mileage, and Ive been driving very conservatively in the past 2 weeks.

I'll have to do some looking to see if I can find something.

~Mike

Mine has the 1.8 L single cam engine and I believe this manual trans comes with the tallest ratio,so if you got a different engine in yours,you might also have a "revvier" trans gear ratio so that could waste some fuel for you.

I also live in Florida so that helps a lot!
And obviously I hypermile the crap out of it but I do move along with traffic except my engine spends about 2/3 of the time in the off state,however other commuters can't tell.

(For some reason in the green fuel signature thingy it shows up as a 2000 but it is indeed a 94)

Barna

Im running a 1.8 as well.
but I dont know if the trans is the original as Ive owned the car only for a few months.

Id really like to swap it with the 1.6L from a 323 if I could find one in good condition. with this trans and a smaller engine, Im betting I could see much better miles.

I did find something today when going over the car, there is grease all over one of my rims, so Im guessing I have a blown hub, even though its not making any sounds. that or my CV's are about ready to go out on that side.

It's your CV axle boot busted! Better fix it with a boot kit before it starts clicking on sharp turns!

This,

And this.

It may not look like it, but the airflow is partially re-attaching at the trunk lid. Putting the kammback on there interrupted that airflow, and made a larger drag area. Instead of the airflow going down to trunk level (or close), it is breaking off where the kammback ends. Extending the trunk lip backwards will actually help pull the airflow down and break it off lower.

Also, remember that while 7-12 degrees may be ideal, 20 degrees to a trunk lid is still functional. I'd suggest a flat extension of the trunk lid. It would be interesting to have an adjustable piece, that you could slide out 2-3 inches at a time to try and find the ideal length for your protege.

that is a good thought, and easily attempted :)

http://www.autospectator.com/uploads...onSide_999.jpg

The Bonneville flat spoilers also have vertical plates on the sides.

I took the 10" spoiler I mounted at 30* last week (had it on for a day, and it made things worse) and test-fit it without the 3" ductail spoiler in the way. It actually sits pretty flat, but there is a 1-2* upward angle from the 3/4" lip spoiler I've had on there for quite some time. My trunk does have a slight downward angle, though.

I assume you are talking about your 1-2 foot Kamm. Did you ever do the 4 foot Kamm and see what it would do for you? I started out with a 1 foot extension so my trunk lid could open, and I saw no benefit (although it would have been hard to quantify), when I went to the 4 foot SedanKamm, I saw 6% over no extension.

I haven't gotten to the 4 foot kamm, mostly because of financial constraints. I'd need to get a clear plastic sheet for rear visibility and find suitable supports. It would also eliminate the use of the trunk, which I'm trying to avoid.

With some rough eyeballing, it looks like a 15" flat spoiler on my car would make about a 15* angle from the roof, and the vertical plates on the sides could give the effect of extending the boattailing. The 15" extension is short enough for me to reach the lock on the trunk and open it

I'm going to do some more research on flat spoilers before I cut up more coroplast and test.


Here's some more shots of the Fusion:

http://www.blogcdn.com/www.autoblogg...7/img_7329.jpg

http://www.blogcdn.com/www.autoblogg...7/img_7332.jpg

the one thing you should realize with those wingtails is they are used as both a stabilizer and for downforce to the rear end.
my guess is putting it at a slight angle down from the trunk lid would have more of a beneficial FE than straight out from the tail end ?

d0sitmatr -

In another thread we speculated that the spoiler also "intersects" the boattail :

http://ecomodder.com/forum/member-cf...ces-fusion.jpg

You will see in the above that the "line" looks like it cuts through the vertical fin of the spoiler, but that is because the boattail/teardrop is intersecting the horizontal spoiler. Hope that makes sense.

CarloSW2