PDA

View Full Version : Moveable Rear Wing


ffreitas
2011.04.13, 11:28 PM
This is a "proof of concept" of a Formula One style moveable rear wing for the Mini-Z MR-03.

http://4.bp.blogspot.com/-DnAXJR7J-8Y/TaZ5JdvZzUI/AAAAAAAAAIY/7DADjmFXelM/s1600/2011-04-13+22-41-06.362.jpg

http://1.bp.blogspot.com/-nZkuc4Bofks/TaZ5MKLkjDI/AAAAAAAAAIc/x0P5dnwvnHk/s1600/2011-04-13+22-41-41.845.jpg

A video of it can be found here: http://haagringnews.blogspot.com/2011/04/mini-z-mr-03-moveable-rear-wing-proof.html

As I said, it is just a "proof of concept" showing that it can be done. It took me more time to shoot the video than to do it...

color01
2011.04.14, 08:24 AM
Just making a wing movable is easy as sin, what he's done is definitely not new. The real trouble is to make a functional, useful rear wing that is not only steering dependent (imagine what happens if you make a correction... the downforce suddenly changes!! :eek:) but also speed and power dependent. The true ideal solution is probably going to be gyro based, but someone smart is going to have to write the software and produce a usable chip before anything like this is practical.

If you know anybody, I have two, maybe three control algorithms I think could be used to dictate the action of the rear wing. This is something I've been trying to figure out since 2009 and have not made significant headway on because none of the control algorithms are truly perfect, they cannot meld "perfectly" with the driver's driving style and expectations unless all parameters are customizable, which is unacceptable because most people would not be able to successfully tune so many hard-to-define parameters (probably at least 20).

hrdrvr
2011.04.14, 08:37 AM
Well, if it was more like the F1 wing, where it was just used to flatten out the wing to reduce drag down the straght, it could be used with a third channel. That would be pretty cool, and teamed with the new Tiki Tiki mode would be just like racing F1 with DRS and KERS :D

The steering controlled wing, Im not sure sure is viable. We will never know 'til it is track tested. :D

JuniorWKR
2011.04.14, 09:01 AM
wouldnt you want it to work opposite?... i would think that when the car was going straight you would want the flap up to give rear bite then when you were to hit the turn the flap should go down so you can put more force to the front and free the rear up for rotation...

great execution... i like what brian pointed out... that gyro based would be amazing... escecially if it was a diffuser that was attached to the bottom of the motor pod...

hrdrvr
2011.04.14, 09:02 AM
I think the idea is to have down force on the car while cornering (increasing grip), then reduce drag going down the straights.

JuniorWKR
2011.04.14, 09:18 AM
i guess...

when i set up my cars i want them to be planted and stable going straight... why i love the side shock setup so much... i wanna be able to come outta a turn know my car is pointing in the right direction and mash the throttle... i then want my car to break free off throttle so i can dive into the turn and rotate through as fast as i can...

ffreitas
2011.04.14, 06:55 PM
Hi there,

Thanks for all the comments on the moveable rear wing. As I said, the idea was just to make a "prrof of concept", that is, a very rough implementation just to kick off the discussion.

I have to confess that my implementation was so rough that it could never see a real track, it would just not work. At the same time, it was really simple, it took me 15 min to put it together. I guess with some more thought and time someone (with a lot more knowledge and expertise than me) could come up with a functional model.

I myself still have doubts on the actual aero effects that such a wing would have, particularly if the purpose is to increase down force and grip in the slower portion of the track (turns). I guess, and I'm no expert on the matter, that that's when the aerodynamic effects, if any, would be even weaker. I guess that is another discussion though.

Anyways, thanks for your feedback.

Fernando

ruf
2011.04.14, 07:17 PM
Very cool Fernando! Reminiscent of the Porsche 908 hill climb cars! IIRC, those had independent flaps actuated by steering almost like airplane ailerons.

arch2b
2011.04.14, 08:21 PM
Even if it could be made functional and effective, one hit in the back and your investment is gone. Sounds great, looks nice but I'm not sure of the feasibility of producing something that would be durable enough for mini-z racing.

Great topic though! Really interesting stuff.

JuniorWKR
2011.04.15, 10:52 AM
thats why i thought if it was o the pod it would be better... it could go up when on throttle and lay down when off throttle or vice versa depending on how you like your car to be setup... and it will be well protected..

hrdrvr
2011.04.15, 03:08 PM
Junior, just to clarify, I don't think this is how I would use it either, but rather how I understood its intentions...aside form the fact of just being "proof of concept" :D

On another note, most of us mere mortals don't set our cars up like you do :D I keep mine as stable as possible, even on the infield. I give up time because of that to drivers with more sensitive set ups, but I tend to drive a lot cleaner with a car set up a bit more "dumb".

arch2b
2011.04.15, 05:16 PM
yep, understood it's a proof of concept project at the moment. providing it works as intended, getting a durable working product to hang off the back end of a mini-z may be just as much a hurdle:p i love the science and ingenuity of the project! i always like to learn new things and you guys never cease to amaze me in that regard.:cool:

cosmicsoul
2011.04.15, 05:30 PM
"Dumb" is definately a proof of concept I can relate to.

I have definately customized all the paremeters of this concept.:eek:

If anyone is interested in the appropriate algorithims to attain maximum effectiveness of this concept shoot me a PM;).

SabRC
2011.04.17, 08:47 AM
Heres an adjustable f1 wing in action.

http://www.youtube.com/watch?v=qliL5A2fqOc

As you can see to reduce drag and produce a cleaner airflow on the straights the top wing levels out and in the turns returns to apply down force (in some turns a car can hit 5gs so you want the rear end planted). Also in in f1 after returning to normal position from a straight it almost acts as a drag brake and slightly reduces pressure on the brake calipers and rotors.

ffreitas
2011.04.28, 08:02 PM
I have been playing with a CFD - Computational Fluid Dynamics software package to try to see if the moveable rear wing would have any impact on the car aerodynamics.

http://2.bp.blogspot.com/-iX3dbg0jTds/Tbn4txYhxXI/AAAAAAAAAI0/GLfHjKs31Uc/s1600/20110428+Mazda+787b+1901.jpg

This is what I found so far: higher pressure at the front, lower pressure at the rear.

I'm still struggling with the software, so these results most likely don't mean much. If anything, it is at least a cool color scheme for a custom body...

<Update 01 May 2011> The results above are wrong, it happens that the software is not fully taking into account the contours of the body, due to the geometry of the 3D model being used. I should have more meaningful results soon.

ffreitas
2011.05.03, 02:05 AM
http://4.bp.blogspot.com/-rabwex4PXco/Tb-SJkgbhKI/AAAAAAAAAI8/JiCyxhDMFik/s1600/20110503+Mazda+787b+0001+Scaled+Faceted+Body+With+ Angled+Rear+Wing+Cp+U+P+Drag+DForce.jpg
In order to get results from the software I'm using I had to start from scratch and significantly simplify the geometry of the car 3D model. After doing that, I got the results above.

Now that the simulation is running I guess it may be possible to add features to the model and also test for different car speeds, body and wing configurations.

I have some more info here: http://haagringnews.blogspot.com/2011/05/mini-z-visit-to-virtual-wind-tunnel.html.

color01
2011.05.03, 05:38 AM
So far, for a speed of 10 m/s, the results indicate a drag force of 0.84 N and a down force of 0.1 N.

If I am not mistaken, this is the equivalent of masses of 8.4 g and 10 g.

You made a typo on the first number, since you divide 0.84 and 0.10N by 9.8N/kg, which gives you 0.086kg and 0.01kg, and then you multiply to get grams, so your simulation's actually saying you get 86g worth of drag force and 10g worth of downforce. I think these are high numbers as well, especially the drag force. I don't doubt that 10g worth of downforce can be achieved, but 86g worth of drag would probably be holding the cars back from achieving a top speed anywhere close to 10m/s = 22.3mph or 36km/h. Yet I'm sure we have Mod cars that will handily reach that speed so they are either pushing more current/power than I thought, or the drag number needs to be re-calculated.

Do you have ground, or do you have space between the car model and the ground? A simulation like this cannot be handled as if the car's floating, having the ground there should change things up due to boundary layer formation, more sciencey stuff, etc...

ffreitas
2011.05.03, 07:44 AM
You made a typo on the first number, since you divide 0.84 and 0.10N by 9.8N/kg, which gives you 0.086kg and 0.01kg, and then you multiply to get grams, so your simulation's actually saying you get 86g worth of drag force and 10g worth of downforce. I think these are high numbers as well, especially the drag force. I don't doubt that 10g worth of downforce can be achieved, but 86g worth of drag would probably be holding the cars back from achieving a top speed anywhere close to 10m/s = 22.3mph or 36km/h. Yet I'm sure we have Mod cars that will handily reach that speed so they are either pushing more current/power than I thought, or the drag number needs to be re-calculated.

Do you have ground, or do you have space between the car model and the ground? A simulation like this cannot be handled as if the car's floating, having the ground there should change things up due to boundary layer formation, more sciencey stuff, etc...

You are right about the typo (already fixed). The correct number for the drag force was 0.084 N, which, as you pointed out, corresponds to 8.6 g.

About the ground, the model is positioned 4 mm above the road/ground, and moves at 10 m/s relatively to the road and to the air surrounding it. The car model floor (so far) is completely flat though, which is not completely accurate. The body height of 4 mm is also relatively high, I used it to make sure that I would be able to see the at least an indication of some of the effects of air flowing underneath the car, in spite of the low mesh resolution I used in the simulation.

I guess as I feel more confident about the results I'm getting it may be worth trying different parameters and working more on the body features, particularly the undebody. I do not expect to get accurate absolute results, but it may be interesting to compare results relative to each other, as a directional indication the impact of aerodynamics of the car.

ffreitas
2011.05.16, 12:31 AM
I was able to run the simulation on a more detailed, although still rough, model. This time instead of a solid block (like the previous model) I considered the body as a shell and included the main parts of the chassis.

http://1.bp.blogspot.com/-Nq0ITIDTPok/TdBfxUaNIxI/AAAAAAAAAJE/ssbdD27M2-E/s1600/20110515+Mini-Z+MR03+Mazda+787b+0000+10+ms-1+Perspective+Cp.jpg

Results are still questionable, as any non-validated CFD results are. These come with the additional caveat of having been produced by an amateur rookie practitioner, so any advice from the experts is much appreciated.

I posted more pictures here: http://haagringnews.blogspot.com/2011/05/virtual-wind-tunnel-update.html

ruf
2011.05.16, 01:12 AM
The latest model looks pretty close to what I'm used to seeing. The way that the F1 DRS works is that it essentially vents the high pressure air on the wing. The position of your proof of concept wing is located in a relatively low pressure zone, but still could be moderately effective even with turbulent air. I'm interested to see what the model would look like with your lower wing element? If you could reduce that low pressure ridge at the rear wheels with a turbulator, you might get a better effect on the lower wing element.

ffreitas
2011.05.16, 01:32 AM
The latest model looks pretty close to what I'm used to seeing. The way that the F1 DRS works is that it essentially vents the high pressure air on the wing. The position of your proof of concept wing is located in a relatively low pressure zone, but still could be moderately effective even with turbulent air. I'm interested to see what the model would look like with your lower wing element? If you could reduce that low pressure ridge at the rear wheels with a turbulator, you might get a better effect on the lower wing element.

The "lower wing element" shown in the proof of concept was there just because I didn't want to ruin my Mazda 787 body! :) Never thought of actually implementing it there, for all the reasons exposed in some of the previous posts (it would not resist even a minor accident).

I'm planning on experimenting with different spoiler angles (current model has a 45 degree angle) at the more typical "high" position. Changing the model geometry is not as easy as I originally assumed as this definition is really on the upstream part of the simulation workflow.

ffreitas
2011.05.22, 09:45 PM
I concluded some additional simulations, comparing drag and down forces for three different rear wing configurations: no wing, flat wing, and 45 degree wing.

The results were the following:

- No rear wing: 6.2 gf of drag, 7.8 gf of lift
- Flat rear wing: 6.6 gf of drag, 0.5 gf of lift
- 45 degree rear wing: 9.0 gf of drag, 6.2 gf of down force

All simulations were done considering the car "racing" at 10 m/s (36 km/h).

Detailed results and pictures can be found here: http://haagringnews.blogspot.com/2011/05/virtual-wind-tunnel-no-wing-vs-flat.html

color01
2011.05.23, 01:54 AM
Wow -- great info! Could you also do an analysis for 15 and 30-degree angles? I think those are the most common angles that people will run a flat spoiler.

Do you also plan to try simulating a scoop spoiler in the future? For RC applications scoop spoilers generally are more efficient than flat spoilers, I'd like to see if your software can validate that empirical result. :)

EMU
2011.05.23, 01:56 AM
Thanks for the info :)

ffreitas
2011.05.27, 10:05 PM
Results for 15 and 30 degree angle rear wing are in.

For a 15 degree angle rear wing, at 10 m/s (36 km/h), the simulation indicates 7.7 gf of drag and 4.6 gf of down force.

For the 30 degree angle rear wing, also at 10 m/s (36 km/h), the simulation indicates 8.9 gf of drag and 8.0 gf of down force.

Details and colorful pictures here: http://haagringnews.blogspot.com/2011/05/virtual-wind-tunnel-15-degree-wing.html

ffreitas
2011.05.30, 12:07 AM
I completed the series of CFD simulations, with "no wing" and for 0 - 15 - 30 - 45 - 60 degree rear wing.

Results are:

Configuration: drag, down force
No Rear Wing: 6.2, -7.8 (lift)
0 Degree: 6.6, -0.5 (lift)
15 Degree: 7.7, 4.6
30 Degree: 8.9, 8.0
45 Degree: 10.1, 8.5
60 Degree: 10.6, 9.3

All results are in gf, all simulations considered the car "racing" at 10 m/s (36 km/h).

Results for the 45 degree wing are different than the ones previously posted because I had to change the geometry of the wing (increasing it a bit) to make it comparable to the other configurations tested.

A summary of these results with links to the detailed analysis and pictures can be found here: http://haagringnews.blogspot.com/2011/05/virtual-wind-tunnel-update_28.html

Below is a series of pictures showing the pressure coefficient (Cp) on the rear wing for each configuration (no wing, 0, 15, 30, 45, and 60 degree). Blue areas indicate low pressure, red areas indicate high pressure.

http://1.bp.blogspot.com/-oitYjWaFMBo/TeMW_k5jEEI/AAAAAAAAARs/ddEWx_Yld2U/s1600/Wing+Series.jpg

I also did the simulations considering other speeds: 2.5 m/s, 5.0 m/s, 7.5 m/s, in addition to the original simulations at 10m/s. The conclusion was that, as expected, both drag and down forces vary proportinally to the speed squared.

Charts showing the relationship between drag and down forces and speed, as recorded by the different simulation scenarios, can be found here: http://haagringnews.blogspot.com/2011/05/virtual-wind-tunnel-forces-vs-speed.html

All results come from simulations, none have actually been validated on a real wind tunnel.

color01
2011.05.30, 02:48 AM
Those results mirror a lot of our real-world testing, so although we can't validate your exact testing it at least gives a clue that it's on the right track. Thanks for the good work! :)

ruf
2011.05.30, 03:11 AM
Ok, someone explain to me the low pressure zone occurring at the top edge of the wing element? Turbulence from the rolloff?

ffreitas
2011.05.30, 05:53 PM
Ok, someone explain to me the low pressure zone occurring at the top edge of the wing element? Turbulence from the rolloff?

I was hoping that the experts (you guys) could help interpreting and explaining the results :). I'm way out of my league here...

I believe what you are referring to is just the transition from the high pressure region above the spoiler to the low pressure region underneath it.

This transition region can be better seen on the cross section view (shown below for the 45 degree configuration):

http://4.bp.blogspot.com/-Z7GzHNdYU1g/TeQRN0VbrCI/AAAAAAAAARw/uohijfOVSNI/s1600/20110523+Mini-Z+MR03+Mazda+787b+Multi-Angle+Rear+Wing+Build+08+45+Degree+Physics+4+Resul ts+05+10+ms-1+Cross+Section.jpg

Digitalis West
2011.05.31, 03:22 AM
The results look relatively normal to me. Compressible flow dynamics can appear strange if you are not thinking about it in the right mindset. A few things to remember about these results....

The lower pressure at the trailing edge of the wing has two causes... that are actually the same cause... but I digress... The air at the back of the wing is not being "held in" by the wing and as this air on top of the wing expands and speeds up to rejoin the flow from under the wing, the pressure is lowered by the Bernoulli effect.

Total downforce is a result of the low pressure under the wing added to the high pressure above the wing. For a wing in clear air, the effect of the low pressure below the wing is often stronger than increased pressure on top of the wing.

color01
2011.05.31, 06:42 AM
That last picture seems to make sense to me, it's just an overlap near the junction between the high- and low-pressure zones.

However, I wanted to point out that it looks like you're already modeling a rudimentary scoop wing for us, based on the endplates and the way you've implemented the 45 degree angle. Thank you very much! I'm definitely keeping this in mind the next time I build a spoiler.

ffreitas
2011.06.02, 11:45 PM
That's it! By popular demand, here are the latest results, including the simulation of a scoop wing, a gurney flap, and a "broken" flat wing.

Configuration: Drag (gf), Down Force (gf)
No Rear Wing: 6.2, - 7.8 (lift)
Broken Wing: 6.4, - 3.8 (lift)
0 Degree (flat) Rear Wing: 6.6, - 0.5 (lift)
15 Degree Rear Wing: 7.7, 4.6
Gurney Flap: 7.2, 5.7
30 Degree Rear Wing: 8.9, 8.0
45 Degree Rear Wing: 10.1, 8.5
60 Degree Rear Wing: 10.6, 9.3
Scoop Wing: 11.7, 13.5

Complete results, with lots of colorful pictures, some analysis charts, and even the streamlines around the car are posted here: Virtual Wind Tunnel Update (http://haagringnews.blogspot.com/2011/06/virtual-wind-tunnel-june-02-2011-update.html)

color01
2011.06.03, 12:13 AM
Thanks so much for your work! Great insights in this thread.

So, is a Gurney flap better than a 15 degree flat wing? :) We should find out experimentally...

EMU
2011.06.03, 03:38 AM
Thanks for adding the additional parameters to the test :) Good to see all of this data... If you have any extra time, could you try moving the location of the wing around a little (forward, up/down) and see how it affects the downforce/drag... I have always wondered how much of a difference that would make. I know where the wing is placed changes the leverage of the force on the body/wheels, but how much does it affect the overall force on the wing...

Rune
2011.06.03, 04:53 AM
Great follow up!:)
Thanks for your work! It is really interesting. I join in with EMU, and would really like to see more tests done as to wing location.:)

ruf
2011.06.03, 08:32 PM
Awesome work! I really "see" streamlines better than pressure zones, though the combination of both is really helping me visualize the effect of different wing geometries.

It's really fun for me to see a lot of the theories behind the scoop wing played out in simulation! What it boils down to is that for a similar sized wing, a scoop wing has a little more drag, and a LOT more downforce compared to a chord angle change. This is due to 2 phenomena:

- the scoop is grabbing "cleaner" air above the car. This air will "stick" to the wing surface and follow it down into the main chord area to create a very high pressure zone.

- the underside of the scoop evacuates the "dirty" turbulent air directly behind the rear windshield which is resultant of the air curling off the roof and meeting the air curling in from the side windows/c-pillar. This should speed up the airflow under the wing creating a low pressure zone under the wing instead of behind it. This creates awesome downforce, and smoother air behind the wing. You can see that the streamline under the wing doesn't curl back on itself, which would indicate high turbulence. This means reduced drag and more efficiency.

http://3.bp.blogspot.com/-1wun6K8uwXU/TebAfe6NhuI/AAAAAAAAATg/KAKDlzUUDoI/s1600/20110519+Mini-Z+MR03+Mazda+787b+0000+Scoop+Wing+Build+04+Physics +02+Results+07+Cross+Section.jpg
http://4.bp.blogspot.com/-CUtYC1ak2Ws/TehQzGaHKNI/AAAAAAAAAUU/TCDfmH_ahFo/s1600/20110523+Mini-Z+MR03+Mazda+787b+0000+Scoop+Wing+Build+04+Physics +02+Results+11+10+ms-1+Streamlines+U.jpg

Best of both worlds! If you need less drag, then trim it down a little! ;)

Thanks again ffreitas!

ffreitas
2011.06.04, 02:15 PM
I apologize but it may not be possible to test additional configurations as you guys suggest, at least for a while: the free trial of the CFD software expired!

I did have the time to produce some glitchy but cool videos that may help see what happens with the air flow around the car. I posted them in the following links:

Vertical Plane Streamlines Video (http://haagringnews.blogspot.com/2011/06/virtual-wind-tunnel-vertical-plane.html)

"3D" Streamlines Video (http://haagringnews.blogspot.com/2011/06/virtual-wind-tunnel-3d-streamlines.html)

Particle Tracks Video (http://haagringnews.blogspot.com/2011/06/virtual-wind-tunnel-particle-tracks.html)

ruf
2011.06.04, 06:02 PM
Rats! Have you tried a "free" software like Symscape/Caedium?

http://www.symscape.com/
http://www.symscape.com/blog/how-much-does-cfd-cost
http://www.symscape.com/product/professional
http://www.symscape.com/product/comparison

ffreitas
2011.06.04, 06:35 PM
Yes, that's what I used, why? Have you used it before?

ffreitas
2011.07.29, 06:30 PM
A summary of our Mini-Z CFD study was featured at the Symscape's home page. Symscape is the publisher of Caedium, the software used to perform the analysis.

CHeck it out at: http://www.symscape.com

It's at the bottom of the page, but it's there...

ruf
2011.08.03, 06:15 AM
The posted it on f******k! Sweet! Nice work!

color01
2011.08.03, 11:52 AM
Didn't even see that much, but congrats on being featured! Should I find myself with 30 less busy days ahead of me, I'll be sure to continue your work. :)

ffreitas
2011.08.05, 03:01 PM
If you guys were willing to do more tests what would you try? Additional wing configurations and positions? Rear/front diffusers? Flat underbody? Front splitters? Difference among several different bodies?