5th generation civic hatchback - improving aerodynamics

[AndrewJ]

And here are the mirror delete pics, used a couple of round blind-spot mirrors that I got for $1.45 at the local autoparts store for interior side-mirror duty.

So here's the deletion, I'm kinda in the middle of getting the hardware situation ironed out, need to get that shiny stuff painted black.



And the gory details of the (new) mirror and it's mounting bits

[AndrewJ]

So I actually got my lazy butt off of the couch today and went out and did some coastdown testing.
Here be the results.

Methodology:
using Iwilltrys' instructable "Measure the drag coefficient of your car"
I used a camera mounted on a tripod in the backseat to record the data points. I used my remote shutter control to take pictures of the speedometer every 10 seconds as indicated by my stopwatch.

Frontal area was measured using this technique but with photoshop's histogram tool instead of GIMP.

The Raw Data:

Vehicle Weight (with occupant) 2300lbs/1043.26kg

Frontal Area 1.768 m^2

Test run-----V1-----V2----V3----V4-----V5----V6
0 sec---------55-----55----55----55-----55----55 mph
10 sec-------49.5---51-----51----50----49.5--50 mph
20 sec-------44-----45.5---45----44----43----43 mph
30 sec-------38-----40-----40----38----37----38 mph
40 sec-------33-----35-----35----33----32----33 mph
50 sec-------29-----30------------28----27----27 mph
60 sec-------25--------------------23----22----23 mph
70 sec------------------------------------19----19.5 mph

The Results

Well, here's where it gets tricky. I used the solver function in the excel spreadsheet, but it keeps giving me really strange numbers, impossible numbers. So I tried my best to eyeball the numbers to get all of the data points to line up, here's what I got.

--------Solver-------Me eyeballing it
Cd------0.062-----------0.20
Crr-----0.0218----------0.018

Perhaps someone who has done this before can lend a hand?

[IWillTry]

Hi AndrewJ,
MetroMPG contacted me through the Instructables site and asked if I would comment here since I created the spreadsheet you're using. Unfortunately, the lower your Cd is, the less accurate the spreadsheet is since the contribution of Crr dominates the results. I punched in your numbers and the data comes out almost linear which means there isn't enough change in drag force during the test to accurately isolate the contribution of Cd from that of Crr.

The easiest way to improve results would be to determine Crr from a different test and enter that value directly. Then run solver again, changing only the Cd value to match the model to the data.

To measure Crr directly you can use a spring scale (50lb capacity should be enough). Measure the force required to pull your car at a constant speed (a walking pace) on level ground. Crr = Frr/M/g where Frr is the force in Newtons (1lb = 4.45 Newtons), M is vehicle mass in kg, and g is 9.81. Measure the force to pull the car forward in one direction, then turn the car around and measure the force pulling forward in the other direction. Use the average force to calculate Crr. Drive the car for a while first to warm up the drivetrain.

Hope that helps.

[IWillTry]

One additional comment (hopefully not too critical ) now that I've read all the other posts. You said in a prior post that the air dam would increase your CdA, but would decrease your Cd. I just wanted to clarify that CdA is what you should actually be trying to decrease (not Cd). Drag force is proportional to CdA.
Cheers and good luck.

[AndrewJ]

Thanks for clearing that up! Well, looks like I have more testing to do, and a spring-scale to locate. As for the CdA/Cd thing, I basically have no idea what I'm talking about , that's why I have all the other smart folks on here to keep me straightened out.

[AndrewJ]

Well, maybe I don't need a spring scale after all. Perhaps a humble bathroom scale will do the trick if I push the car from behind. In that case, I'll have the Crr isolated by the end of the day!

Can't believe my 2 brain cells that are awake thought that up so quick. Alternately, I can't believe it took so long to think of that.

[MetroMPG]

Originally Posted by IWillTry

You said in a prior post that the air dam would increase your CdA, but would decrease your Cd. I just wanted to clarify that CdA is what you should actually be trying to decrease (not Cd). Drag force is proportional to CdA.

Good catch, IWillTry.

And thanks for signing up to clarify the coastdown testing with your spreadsheet also.

I don't know if you have a stat counter at your Instructable on this subject, but if you do, I bet it's seen a lot of traffic in the past week.

[IWillTry]

A bathroom scale may or may not work. Some scales will only work in a horizontal orientation. They also can be difficult to balance on a car bumper (depending on the car). If you do use a bathroom scale, be sure to "zero" the scale while it is in a vertical position to calibrate out the weight of the scale platform. Good luck.

[AndrewJ]

So, using the bathroom scale trick (zeroed at vertical) I found that my car takes 35lbs of force to keep it moving on a level surface.

*a note to people who try this method*
smoothing out your stride seems to be an important part of getting an accurate measurement. Just pushing normally produces quite an oscillation of force, making the measurement "jump around"
So focus on the dial of the scale and smooth out your power delivery to keep the dial steady.

So, 35lbs = 160.2 newtons, so on an so fourth. Voila! Crr for my car is 0.0169

Plug it in to Excel, set the solver to solve for Cd (cell F41 only) and it spits out a Cd of 0.194

Hmm, still seems low to me. If the "stock" car is 0.33 and I managed to knock that down to less than 0.20 with a mirror delete, airdam/radiator block , and rear skirts? Seems unlikely.

I think I'm gonna take all the aero goodies off the car tomorrow and run another set of coastdown tests in "stock" form. That way I'll have some idea of how accurate the 0.194 actually is.

[MetroMPG]

Interesting update. *applaudsyourtenacity*

[AndrewJ]

Originally Posted by AndrewJ

Sure, this will increase the CdA of my car, but it will also take care of two (or three) nasty areas of aerodynamic drag, thereby decreasing my Cd.

What I *should* have said was:
"Sure, this will increase the frontal area (A) of my car, but it will also take care of two (or three) nasty areas of aerodynamic drag (Cd), thereby decreasing my coefficient of drag (CdA).

Basically the coefficent of drag is CdA which is comprised of the Cd multiplied by A. See wikipedia for further detail.

So the big idea behind airdams is that they will improve your aerodynamics IF you can improve your Cd without adding to much A. If you tip the balance the other direction, by adding a lot of frontal area without streamlining the shape much then you will actually be hurting your aerodynamics.

This is why an airdam like mine would actually be a bad idea for basjoos's CX. He has a full belly pan, so adding an airdam wouldn't improve his Cd any, and it would add a bit to his A, thereby raising his CdA.

As for the original post, "You have to hear what I'm thinking."

[AndrewJ]

I went out to do some more testing today and while I didn't get to do all of the testing I had planned to do, the intrigue continues.

I did 2 of 6 runs of coastdown testing in "stock" form to verify that the Excel spreadsheet was giving me decent numbers. By comparing the Cd that the spreadsheet gave me with Honda's Cd of 0.33 for a stock 5th generation Civic hatchback, I can see how far off (if any) the numbers in the spreadsheet are.

Now, I only got to do 2 runs, one in each direction, before I had to quit due to the sudden appearance of a fierce crosswind. I didn't want my data corrupted by any wind, so I just used those two sets of data to plug into Excel.

I know, I know... two points of data isn't much, in fact, it's probably safe to consider it flawed.

But then again...

Guess what Excel came up with based on my 2 data points?

Cd = 0.33

Grrr....more testing....damn wind.


Given the windyness situation I headed on over to my "calibrated" hill for a little session of A-B-A. The hill is usually windy, but today it was on the leeward side of the wind, so it was unusually calm on the downhill side.

I brought the car to 55mph at the top of the hill, and at the same point in all 3 runs, cut the engine reset the trip odo. I then kept track of speed and at what distance on the trip odometer those speeds were reached. The runs ended when I coasted back down to 55mph at the bottom of the hill.

---------------------------A(stock)----B(aero)---A(stock)
Initial speed-------------55mph-----55mph-----55mph
Slowed to----------------52mph-----55mph-----52mph
Miles@lowest speed-----0.375------0-0.4------0.375
Top speed---------------66mph-----71mph-----65mph
Miles@top speed--------1.1---------1.05-------1.05
Miles@55mph (end)-----1.35-------1.45-------1.25

Also, on the "B" run, 66mph was reached at 0.75 miles.

Looks like this aerodynamics stuff might actually work

[SVOboy]

Haha, interesting results,

Stick with it though, I'm glad you have the patience to be so methodical about it,

[MetroMPG]

and while I didn't get to do all of the testing I had planned to do,

Welcome to the club!

Originally Posted by AndrewJ

Looks like this aerodynamics stuff might actually work

Nice work. I envy you people and your winter time testing weather.

[basjoos]

In my opinion, an airdam is just a band-aid to use if you have a aerodynamically dirty underside you want to divert air away from. In a max FE application, we're not overly concerned about generating added downforce, which is the other main purpose of an air dam. On my car, I have a totally smooth underside and the only "airdams" on my car are the two air diverters (wheel spoilers) that sit ahead of my front wheels to split the airflow around the wheels. This split airflow is maintained via the double side skirts to the rear wheels and then ideally (future project) should have a small boattail fin behind each rear wheel to reduce its eddy. If you look at the Loremo and the Daihatsu UFEIII, they have airflow splitters in front of each of their front wheels and the UFEIII has a small boattail behind each rear wheel. If you insist on going with an airdam, you can make it a variable height airdam that is deepest in front of each wheel and highest in the section between the wheels. Some airdam-equipped Audis and BMW's go this route.

[AndrewJ]

I agree that the airdam isn't optimal FE-wise. However, I'm finding that it is a really nice thing to have as part of a work-in-progress car.
I will be modifying the airdam further as I continue to work on the car, eventually resulting in it being just a grille block/wheel spoiler type thing.
But for now, while I won't be able to do a full bellypan anytime soon, it's helping to staunch the aero losses.

I like the mini-boattail idea behind the wheels, I may have to work that into my car somehow. I've been thinking of trying sideskirts next, maybe work them into the mix there?

And after re-reading IWillTrys post I think I'll have to look for a new place to do my next round of coastdown testing. It looks like the tests might be more accurate if I start out with a higher top speed, where the areo-drag is more pronounced.
My current testing circuit is a 50mph speed limit, and right next to the Highway patrol headquarters. Probably not wise to speed

basjoos: How did you get your Cd numbers? Any tips?

[newtonsfirstlaw]

Originally Posted by AndrewJ

I agree that the airdam isn't optimal FE-wise.

The full belly pan/undertray thing is not that hard to do. You need a couple 8*4 sheets of 0.6mm aluminum, probably some sort of heat resistant padding to stick in the middle somewhere. Jack stands, your regular jack, a friend, a drill, screws with big heads, some metal shears and about 5 hours is all you will need.

[AndrewJ]

One thing that I feel hasn't been discussed too much yet about airdams vs. bellypans is the point of diminishing returns.
So far we know (at least in theory) that a bellypan is better because of the lower frontal area, and also the probable decrease in Cd from all the hanging bits on the underside.

But what we don't know is how much better a belly pan is. Has anyone seen any data on this so far? The most relevant thing I've found so far has been this refrence (PDF file) to a 1958 study
"Hoerner 1958 discusses the underbody as the biggest single avoidable
aerodynamic drag component and gives an old example where drag goes
from 0.3 for a smooth underbody to 0.6 for an open and rough as
customary car!"

Still no word on how much an airdam would knock off that 0.6

So, I hope eventually when I do get myself a jack, some jackstands and a new (working) drill I can get a belly pan together to test it out and get some numbers.

This list of things to do is getting quite long.

[MetroMPG]

Originally Posted by AndrewJ

One thing that I feel hasn't been discussed too much yet about airdams vs. bellypans is the point of diminishing returns.

Well, one thing Phil Knox has said in terms of the lowest extension of an air dam is that it shouldn't go lower than the lowest hanging suspension/underbody bit, or else you're just adding frontal area at the expense of Cd.

That said, your tests so far, and Coyote X's, seem to suggest that more can in fact be better.

But what we don't know is how much better a belly pan is. Has anyone seen any data on this so far? The most relevant thing I've found so far has been this refrence (PDF file) to a 1958 study

I'd be wary of relying on 1958 data. Cars back then would have had MUCH worse undersides (basically being body-on-frame truck like construction) than today.

Here's a couple more modern references:

• Ford's hydrogen fuel cell Bonneville racer was lowered and had a full undertray added. They claimed a 0.01 reduction in Cd with the pan

• Audi improved the Cd of the A4 with a patchwork of panels applied to the underbody. The sum difference was Cd -0.019