Mitsubishi Lancer Evolution Forums banner

aero: rear diffusers

3931 Views 6 Replies 2 Participants Last post by  rammsteinmatt
Rear Bumper Aerodynamics, Concerning the Implementation of a Rear Diffuser,
Lancer Evolution 8

Identification of Problem
On the factory Lancer Evolution 8 rear bumper, there exists a problem. As air travels under the car it increases above atmospheric pressure. This increase in atmospheric pressure coupled with the decrease below atmospheric pressure of the airflow over the car, leads to a lift force produced on the car. Additionally there is a pressure differential created at the rear of the car from two different air pressure flows coming together. This pressure differential creates pressure drag on the car as a whole. The general shape of the car lends to a large bubble of low-pressure air that is constantly behind the rear of the car when moving. This low pressure behind the car “sucks” the car back while driving similar to how a vacuum uses low pressure to suck items off the floor. This is also another form of pressure drag, which inhibits foreword movement to an extent. The most commonly identified source of drag on a car is the “parachute effect” caused by the rear bumper. The rear bumper sticks down into the natural underbody airflow of the car and due to its shape, that very much like a parachute, it naturally catches air from the air stream, and “parachutes” thereby creating drag.

Identification of a Solution
The solution to these drag problems can be solved or reduced, however. The way to reduce the parachuting effect of the rear bumper would be to place a flat plate under the rear bumper that covers the lip of the rear bumper and through the smooth transition of the plate from behind the axle, to block air from going over the top of the diffuser, to the rear bumper, will create less possibility for the air to get “trapped” on some underbody pieces and create drag as a result.
The low-pressure bubble at the rear of the car cannot be completely eliminated without the use of some radical mechanical devices, but it however can be reduced. The way to eliminate this is to move air into the space, as the air from any other part of the car is at a higher pressure than in the low-pressure bubble. To “blow” air into this area can be accomplished through two different approaches. One approach would be to put an airfoil hanging at the rear bumper that has a steep angle of attack to move air into this area, where the other would be to add fins to the rear diffuser main plate and use the vortices created to move some air into this low-pressure area. The airfoil would move much more air into the low-pressure area, and reduce much more drag than the fins would; however the airfoil produces some ill side effects. The airfoil, since at a very steep angle of attack, would create drag, as drag is a by-product of lift. Additionally since the airfoil would need to be located under the farthest rear part of the car, it would succumb to constant dragging on the ground as the car is driven on inclines, unless the car is sufficiently raised (which would in turn allow more air under the car, which would create more lift and more drag on the car as a whole). Therefore fins will be used because they create very little drag, help to straighten underbody airflow, and can be used to decrease the low-pressure area behind the car.
The pressure differential that is created by airflow over the car and under the car cannot ever be completely eliminated, as air will always be able to get under the car. However you can reduce the amount of air that gets under the car through various other ways, namely lowering the car and the body’s proximity to the ground. The pressure differential that comes back together at the rear of the car creates drag. However the rear diffuser can help that. If the diffuser is placed at a negative angle of attack, so that is still does not catch air from the underbody air stream, one can effectively increase the area under the rear bumper per unit length from the beginning of the diffuser. Through the continuity equation we find that an increase in area leads to a decrease in air stream velocity, and Bernoulli states that an increase in velocity leads to a decrease in underbody airflow pressure. Since the underbody airflow pressure is the highest air pressure of the airstreams around the car, lowering it will decrease the pressure differential, as the two pressures will be closer to one another when they meet at the rear of the car.

Rear Diffuser, and its Implementation of the Solution
The rear diffuser is based off a flat plate with minor modifications to its shape and to accommodate the features of the Lancer Evolution 8, such as the axle and suspension, and the exhaust system. The diffuser uses this flat plate to reduce the drag created by the rear bumper’s “parachute”, and to gradually decrease the air pressure through increasing the area, so as to reduce drag created by the pressure differential at the rear of the car. The implementation of fins on the design will reduce the low-pressure area behind the car while maintaining a very low contribution to total drag on its own. The fins also help to straighten the airflow as it comes across the diffuser, as the air becomes very turbulent as is passes over and under all the irregularities under the car. The “box” design around the exhaust system both serves to straighten air flow with the use of its sides, but also uses the bottom plate of the box as an inverted airfoil to create down force mechanically rather than through pressure like the rest of the diffuser does. The box design was created in the size that it is to accommodate the stock and the different aftermarket exhaust systems available for the Lancer Evolution 8.

Data and Graphs
All runs done in the east and west direction, to account for wind and elevation change, with all windows up

Without Diffuser
Speed During Test [MPH] Drag Force [lbf]
45-40 92.13
65-60 132.22
75-70 204.46

With Diffuser
Speed During Test [MPH] Drag Force [lbf]
45-40 106.85
65-60 141.02
75-70 214.61



With Modified Diffuser (box reduced vertically by 2”, box angle of attack reduced to 6°)
Speed During Test [MPH] Drag Force [lbf]
45-40 95.69
65-60 140.86
75-70 158.88


Data Analysis
Although the diffuser contains all the aforementioned drag-reducing elements, the data shows that the diffuser increased drag. This is due to the fact that the diffuser is creating down force. As stated earlier, drag is a component of down force, since the diffuser is reducing drag from the parts on the factory version of the car, yet showing a net increase in drag, one is lead to believe that down force is causing the calculated drag created by the diffuser, since the drag creating elements on the car have been severely reduced through the implementation of the rear diffuser.

Error Analysis
In the manner that the tests for drag calculation were ran, this involved using the car’s speedometer for a speed change reference and a stopwatch. The error would be involved with the human error associated with starting and stopping a stopwatch and ability to react to a condition. It was calculated that the human’s ability to react was less than one quarter of one second. Since the time values vary, the error accounts for between 0.5% and 7.5%, on average about 4% or between 4 and 8 pounds-force of drag. Additionally, the tests were carried out on different days, therefore one much account for atmospheric changes. It was however calculated that temperature and density changes would account for a difference of less than 1 pound-force of drag between the calculated values and expected values, so those calculations were omitted from the drag calculations for simplicity.
See less See more
1 - 7 of 7 Posts
that is a direct copy/paste from a rear diffuser technical report that i wrote concerning a rear diffuser that i designed, thats why it refers to "my design" alot

basically anyone's diffuser works in this manner: my design, Varis, Vishnu. the HKS kansai and APR piece do not seal off the full rear, so they are kinda cheating for the sake of simplicity. and APR's requires the widebody, however they will be selling one that fits a stockbody in the very near future (remember you heard it here first :thumb: )
Matt(?)

did you produce your diffuser?
or was yours a one off for testing only?

how did you determine drag?
nothere said:
Matt(?)

did you produce your diffuser?
or was yours a one off for testing only?

how did you determine drag?
yes it's matt.

i made my own diffuser, designed, then produced in a CF, E-glass, epoxy conglomeration (for lack of a better word)

well the original intention was to sell them on either socalevo, or though a vendor there. however, school happened, and the project sits idle.

determining drag is relatively simple. drag is a force. newton says F=ma. i can find the mass of my car (with driver and timer) in kg, because metric is easier to work with. i can also find the acceleration (or rather deceleration of my car due to air friction)
how to do that-pick a target speed/range. i did 40-45, 60-65, 70-75 for my tests and documentation. let's make this simple and say we're going to do it for 40-45.
find a road, straight, flat, no traffic whatsoever (because we'll be speeding, slowing, and they will disturb the air - think about how they draft in nascar) go up to 50, put into neutral, the car will nodoubt start to slow have the timer start at 45mph and time to 40 mph (please dont try to do this yourself, you're supposed to be driving). now that we have a delta v (change in velocity) convert to m/s units. and we have a time. well a velocity per time is an acceleration. and ta-da! we have a mass and an acceleration, we can multiply and get a force (in newtons), convert to lbf (pounds). and we have pounds of drag.

this isn't however the whole story. there is also friction in the bearings of the wheels that is contributing to this drag, and the rolling resistance of the tires which also contributes as well. therfore the total drag is equal to:

A + Bv + Cv^2

because things like rolling resistance (moment of inertia stuff) is constant regardless of speed
things like bearking friction vary with speed
and things like aero drag vary with the square of speed (as senn here: D = 1/2(rho)v^2 * Sref)

so solve for A, B, C. and you could theoretically calculate the approx. drag at any speed. or figure out how much aero drag contributes to total drag.

so you need to test stock vs. diffuser. you need to do it at different speeds (and you can make a graph showing trends) and you need to do it both directions on that street you picked out (to account for any windage or slope), and you need to do it abunch of times which will hopefully average and counteract any error you created.


if you dont like this way, because its inprescise, you can get an accelerometer, and that will calculate your acceleration as well. but i dont have one, so this will have to do. [additionally an accelerometer calculates acceleration the same way we do using F=ma and a=v/t, it just calculates it alot more times (instantaneous) than we can do]

some other considerations. air is more dense when it is colder outside. i did my tests in the summer (90+F) and now that its winter (50-60F) my drag numbers may vary by 3 lbf or so

i had pictures of the whole process uploaded and in a progress thread on socal, but they changed software and all the pictures weren't uploaded anymore, and there were atleast 100 of them. on a not fast internet connection, that becomes boring. i can post some specific pictures if you would like.



***mods, im not actively trying to sell a product, i am using a product i designed and developed to demonstrate the purpose of it***
See less See more
do you have inside information on the APR diffuser?
did you carve out a bit of the rear bumper for your design? maybe you could post a picture of it?

no, I was just wondering if you did a field test or inferred through numbers to figure drag.
(can you tell I'm not mathmaticly inclined?)

thanks
Bruce
nothere said:
do you have inside information on the APR diffuser?
did you carve out a bit of the rear bumper for your design? maybe you could post a picture of it?

no, I was just wondering if you did a field test or inferred through numbers to figure drag.
(can you tell I'm not mathmaticly inclined?)

thanks
Bruce
i dont have any info on the APR piece, except that they are making it, it'll fit a stock body evo, and it should be ready to purchase pretty much now.

i leave the full bumper intact (minus a couple holes to bolt it up, so it doesnt come loose - dont worry those holes are a good thing) but it would be more effective if you cut the bumper up a couple inches

i'll post pictures in a minute or so.


also for the record, the evo IX "rear diffuser" is NOT a rear diffuser in the classical sense. because it still parachutes, and there is no account for pressures. but the evo X on the other hand :thumb:
i dont have any good pictures uploaded. i'll take some and upload them tommorow.

but here was the second mounted design


the evolution of the [mounted] designs:
1. outside of bumper (like pictured above) with big box (hung down low to accomidate the real large canister mufflers) - outside, large box design
2. cut box, moved it up 1.5 inches - outside, small box design
3. put inside bumper for a cleaner look - inside, small box
4. re-angled the box for lower angle of attack (less downforce) - inside, modded box design


the diffuser is made in 1 rectangular piece with a box shape, and angle brackets laid into the CF. epoxied, then the diffuser cures, and then is removed from the mold (either clear is added or not). the diffuser is then cut by hand to fit a template. it is ready to go at this stage. its mounted (either inside or outside - owner's preference). box is changed to owner's specs (either leave it alone, raise it, change the angle of attack, cut out a section to accomidate the exhaust, etc...). fins are attached (either the large 4.5" or smaller 2.5") and thats it.

if it is chosen that the box be cut, then the diffuser becomes 3 pieces, with an easy to remove box section that makes it very convenient to change the axle back exhaust. IMO 3 pieces is far easier than the one large piece to take off. when its 1 piece it takes close to 1 hr to remove, and when its 3 pieces it takes probably only 30 mins max
See less See more
1 - 7 of 7 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top