Aero packages and race points

[Lorenzo Pessa]

Continuing the discussion started in the topic "New Aero Packages for 2011 Rules"



Originally posted by sbrenaman:
The only way to properly justify running aero (and anything else, really) on a car is a laptime sim that takes into consideration the change in NSM weight distribution F/R as a function of speed. Then, do this for accel, skidpad, autox, endurance, and fuel economy, and figure out what gets you the most points.

Not the droids you were looking for? http://fsae.com/groupee_common/emoticons/icon_biggrin.gif



Originally posted by Lorenzo Pessa:
At least you need to account also the cost event.



Originally posted by Scott Wordley:
Lorenzo, I disagree.
In our experience the actual (perceived?) cost of the car does not correlate strongly with the points a team scores for that portion of the cost event. Have a look over the results from any competition and you will see this is the case.

In terms of a competition point sim we find it is easiest to assume every car scores full points for all static events. That way you get an indication of the maximum point scoring potential of the car.
[...]




Originally posted by Lorenzo Pessa:
[...]
The 100 cost points are partitioned between cost of the car (40 pts), accuracy&...&visual inspection (40 pts), real case (20 pts).

Aero package influences directly only the cost of the car.
Taking data from the last FSAE-Italy I add the cost of aero packages to the cost of my team car (ET3).

I have done a rough estimate of the costs (conservative! so the resulting costs are less than real, I hope).

wings: 1600$
undertray: 900$

[Is someone laughing now? :-) ]

the cost of the car points are
ET3 : 16.064 pts
ET3+wings: 12.112 pts
ET3+underbody: 13.739 pts
ET3+wings+undertray: 10.187 pts

There is a difference from 4 to 6 points.

Aero doesn't affect other static events.



Originally posted by Matthew Bell:
Lorenzo,
Remember that the total points score is 1000 points. So the cost difference makes up roughly 0.5% of a team's score.

That being said, I'm not 100% with Scott on this one either. I see the point of "eliminating" the static events from the event simulation when you're trying to decide "do we do wings or not?". But I think you have to keep in mind that you better be prepared for the static events when you get there - it's a lot easier to lose points in design than gain them because you overestimated your skills and resources.

This is all part of the game though, and topics like this could probably dominate the forum if we all stated our opinions.
[...]

[Lorenzo Pessa]

Originally posted by Matthew Bell:
Lorenzo,
Remember that the total points score is 1000 points. So the cost difference makes up roughly 0.5% of a team's score.


Let's do some quick calculations.

In the thesis of Karl Jensen "Aerodynamic Undertray Design for Formula SAE" from GFR is stated that with that undertray you gain an improvement of 1% in lap times.

I assume my car (ET3) has an 1% of improvement with an undertray. I focus on autocross and endurance at FSAE-Italy 2010.
h t t p ://w w w .ata.it/upload/allegati_submenu_fata/175/fsaeitalyresultsclass1-2010_finale_eventi-dinamici.pdf

Endurance (min time: 1204.07 s; max time: 1745.9 s)
ET3 time (not adjusted): 1242.35 s
ET3 "undertray" time: 1229.93 s
penalty: 58 s
ET3 "undertray" adj time: 1287.93 s

ET3 endurance score: 240,354 pts
ET3 "undertray" endurance score: 247,548 pts
variation: +7.194 pts


Autocross (min time: 47.777 s; max time: 59.721 s)
ET3 time: 50.825 s (no cones)
ET3 "undertray" time: 50.317 s

ET3 autocross score: 107.271 pts
ET3 "undertray" autocross score: 114.032 pts
variation: +6.761 pts

autocross + endurance= + 13.955 pts


car cost= - 2.328 pts


I'm not confident to use 1% to make calculation on skid-pad.


Looking at overall results, the difference from #2 to #5 is about 30 pts.

Which is the influence of cost over the overall score?

It would be interesting to know why GFR decided to not run the undertray in 2010.

[bob.paasch]

Originally posted by Lorenzo Pessa:
It would be interesting to know why GFR decided to not run the undertray in 2010.

Lorenzo:

In 2010, GFR only built one undertray. The one undertray we built was somewhat overweight due to the CF material used. We had planned to build a second, lighter undertray, but ran out of time.

We did not run the undertray at Michigan because of limited test time with the undertray on the car.

We did have the undertray on the US car at California. The autocross and endurance tracks at FSAE California are typically fairly fast, and aero cars tend to do well there. If you look at the autocross results, the top 6 cars were all running some type of aero.

In discussions among the team right after California, we decided that the logistics of shipping the undertray to Europe and installing it on the German car in the three weeks between California and Silverstone would be very difficult. We could have installed the undertray between Silverstone and Hockenheim, but the FSG tracks tend to be pretty tight, and our belief was that aero would not have the potential there that it had at tracks like California. The undertray stayed in the US.

This year we expect to have more test time with aero, and hope to have it on at least one of the cars (C or E) in Europe. We hope to be able to answer Scott's question about the effectiveness of aero on a 50 HP 160 kg car. http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Lightsun]

I really like to see your opinions about aero packages and race points http://fsae.com/groupee_common/emoticons/icon_wink.gif

First of all, I completely agree with sbrenaman. How can you decide if an aero package will make your car faster without a laptime simulator? A laptime simulator should be a MUST in a top team.
Yes, I understand that it is a lot of work. But ask your faculty advisor for a new team member, with the only purpose of developing a laptime simulator with MATLAB or C++ or whatever code you like. The results should be impressive.

Second, I agree with Scott Wordley. The Cost Report is not only about the cost of your car. Prepare better your "Real Case" to get some extra points lost with the cost of the aero package http://fsae.com/groupee_common/emoticons/icon_wink.gif
But what is even more important, the three static events are affected by the aero package.
Design judges are humans, with emotions. A car with a good aero package, and a couple of students explaining the reason why they decided to run a car with wings (with data from their own laptime simulator), and good CFD simulations (I understand that only a few universities have access to wind tunnels) will give many points in the design event.
Moreover, I don't know what the Business Event judges look for, but I imagine that it is easier to sell a racecar with wings only for its appearance. It looks faster, and more agressive... just paint it in red and you have a Ferrari F1 http://fsae.com/groupee_common/emoticons/icon_wink.gif

Finally, this summer, having lunch (or dinner, I cannot remember it) with Mr. Claude Rouelle, I asked him for its opinion about aerodynamics in FSAE. We speak about different things, but he finally asked me "What do you think about movable wings"? Maybe in 2012 you will see a car winning FSAE events with them...

And just my opinion as CFD engineer about different softwares (for teams which are not using CFD);
Don't use CosmoFloworks (Solidworks CFD), it is a waste of time.
I don't recommend using OpenFOAM; I developed a WTCC for Seat Sport aero package with it, and it was a nightmare.
StarCCM+ is excellent! Maybe it is not free, but I'm sure that your university can afford it (I don't know the current price, but a license for a university should not cost more than 1000$). With a little experience, you could start simulating in a few days! And it is used in many F1 teams... so it is interesting to learn a software widely used in F1!

Perhaps I will run some examples in StarCCM+ and upload them to have more data about downforce...

Hope it helps!

[nowhere fast]

Originally posted by Lightsun:
A laptime simulator should be a MUST in a top team.
Yes, I understand that it is a lot of work.

I agree that a lap time simulator is immensely valuable but disagree that building one takes a lot of work. With lap sims (most sims actually) it is best to start simple and iteratively build in refinement.
I built a Matlab based lap time sim recently to help with conceptual design of a motorbike I'm working on; and it only took a couple of hours here and there after work to get something useful. From my experience I would say that a good lap sim could be built in one weekend.

[MalcolmG]

Originally posted by Lightsun:
Design judges are humans, with emotions. A car with a good aero package, and a couple of students explaining the reason why they decided to run a car with wings (with data from their own laptime simulator), and good CFD simulations (I understand that only a few universities have access to wind tunnels) will give many points in the design event.
Moreover, I don't know what the Business Event judges look for, but I imagine that it is easier to sell a racecar with wings only for its appearance. It looks faster, and more agressive... just paint it in red and you have a Ferrari F1 http://fsae.com/groupee_common/emoticons/icon_wink.gif

Disagree on both points there. I've never seen any evidence to suggest cars with wings do better in design or presentation. I'd almost suggest the opposite is true for design event.

[Luke Phersson]

The only way wings can make you lose points in Design Event is if the team doesn't justify them sufficiently. It’s the same for any major system on the car really.

I also wouldn't say wings will gain you more points in Design - it really depends on how well the team has utilised the resources available to them and how well they communicate this. Basically just cost(resource)/benefit relationships.
If team x gets a better cost/benefit from implementing Aero than they would if they focused on increasing power/any other system - then it is justified. This is where your lapsims come into play (as well as knowing the resources available to the team), we actually have a 'competition sim' which incorporates all dynamic events.

Interesting point; do those teams utilising lapsims take the design path that gives them the greatest possible points out of 1000? We focus on the greatest delta between the top competitors, ie a lightweight single (RMIT, Delft, GFR), a powerful four (UWA, Munich, Stuttgart), a light narrow car (Auckland) and our 2010 four with large amounts of downforce. It’s interesting to see how the point’s distribution changes when track parameters are varied, giving you an idea on your design sensitivity.

[Scott Wordley]

I was hoping to start a topic based purely on Lap sims and points sims, without any specific reference to aero, but it seems Lorenzo has beaten me to it!

On the topic of Cost event...

Most performance features on the car will incur a cost penalty, whether they be wings, undertrays, turbos, different engines, carbon tubs etc. But if you look at the cost event results from most competitions you will see that teams with lots of these supposedly high performance and added cost features do not always end up at the back of the field.

I think this is generally because they are able to minimise the costs they report by doing extra work to justify it, and using creative accounting techniques which is a product of the team's experience. These teams generally also do better on the question answer based part of the event and so can claw back most of the points lost.

The thing is, you cannot easily assess the added costs of parts you have never run before. So given the small amount of points difference it would make I would suggest you don't worry too much about it, particularly as it will severely slow down your iteration time and complicate the process, particularly at the start when you are looking at very broad effects.

If we were to do a study looking into the effects of increased downforce and drag on our car due to bigger wings, there is no way I would want to try and estimate what the cost differential would be for each increment. Same goes for the weight, ie we would estimate the max weight added but not bother to tweak it slightly for every single design point. 2 variables (or dimensions) are complicated enough to look at simultaneously, you need to draw a line some where.

Once you have found the ball park you want to operate in only then would you go firming up the details and looking at the lower order point scoring factors. But in general we leave cost completely out of the equation.

[Lorenzo Pessa]

Thank you Bob for your answer!

About cost event I would to specify that calculation are about only the "cost of the car" part of cost event that is based only on the cost of the car.
Everyone is free to choose if account it or not.

About moveable wings, the solution used in F1 to increase the gap and change the flap angle isn't the only one.
I red on a report on FSAE-A that Sophia use a spring mounted moving rear wing. Other solution are possible. Why are we waiting for 2012? http://fsae.com/groupee_common/emoticons/icon_smile.gif

About SIM, in Pisa the more advanced sim we have is a realtime Simulink based simulator. It is not FSAE dedicated but a Formula SAE car model is implemented. The main problem is to find the time to use it! Other simpler models are in use.

[Lightsun]

Lorenzo: Good point, really interesting! Maybe you remember where did you read this? Or maybe you can share this report? And yes, you don't have to wait until 2012 for movable wings, of course! http://fsae.com/groupee_common/emoticons/icon_wink.gif
In F1, they move the flap to reduce drag. But by movable wings, I mean front & rear wings, moving while braking, cornering and accelerating... To have less drag or more downforce, depending on each situation. Perhaps this can lead to smaller wings, cheaper and with a little bit less weight (despite having to add the weight and cost of the device to move the wings)... Although it would be interesting to see how the teams manage to install a device to move the wings and control it. In any case, it would be really interesting engineering!

In fact, to know the potential of them, some CFD simulations must be done and some test with a laptime simulator (as nowhere fast said, the lapsim can be done with a weekend to have some initial values, although I would prefer a laptime sim more advanced). I will try to make some CFD simulations in my free time to have some drag&downforce data to discuss...

MalcolmG: You didn't understand me. A good aero package doesn't mean necessary wings, it can be an undertray+diffuser... And perhaps you don't see cars with wings doing better in design because 1- It is not correctly understood/justified by the students (a lapsim is essential to justify them) or 2- The overall of the car is not as good as the aero package they have developed. Moreover, I think there are not enough teams running wings and the importance of the wings in the design event is not big enough to make a statistical analysis and obtain conclusions of correlation between wings/design event score.

Anyway, I have to tell you, if a team runs a car with an aero package (with wings or with only a diffuser, or well developed sidepods), the CFD calculations are OK, and they demonstrate with a lap time simulator that the car with the aero package is better, and it is correctly implemented, they will obtain some extra points! Good engineering exercises are always welcome!http://fsae.com/groupee_common/emoticons/icon_wink.gif
If you find a team of 30 students, and they say: no, we didn't make any CFD calculations because aerodynamics are not important in FSAE, in my opinion, they will lose some points! I don't buy it, I'm sorry!
However, if a new team of 15 members, on their first year of FSAE, explains that they don't have the resources/manpower to work on aerodynamics, and they explain that the potential of an aero package is not big enough, because they have read some documents about the aerodynamics on FSAE cars and they know downforce levels generated and the consequences on lap time, I would say: Excellent job!Congratulations!

About the business plan, it was only a joke! http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Lorenzo Pessa]

Originally posted by Lightsun:
Lorenzo: Good point, really interesting! Maybe you remember where did you read this? Or maybe you can share this report?


Sorry, You are right!
Google give it me again
h t t p:/ /w ww.fapm.com.au/LinkClick.aspx?fileticket=lH3OVuGhwnQ%3D&tabid=37
It is at the "Formula SAE-A Technical Review" by Pat Clarke.


Originally posted by Lightsun:
Perhaps this can lead to smaller wings, cheaper and with a little bit less weight (despite having to add the weight and cost of the device to move the wings)... Although it would be interesting to see how the teams manage to install a device to move the wings and control it. In any case, it would be really interesting engineering!


Maybe with actuators connected with ECU.
An intelligent system like the advanced aircraft stability control... cool!

Or an old fashioned Chaparall style third pedal mechanically connected to the wing. simple, efficient, cheap.

[Lightsun]

Lorenzo, thank you very much for the link!

It is very interesting to look at aircraft technology about moving wings, they really know how to move them! http://fsae.com/groupee_common/emoticons/icon_wink.gif In fact, I sometimes read aircraft design books, not only for fun, but also to have new ideas! I think that fighter designers are masters in ergonomics (although it is not my subject), so aircraft technology is not only about expensive technology!

A third pedal mechanically connected to the wing... Or an F-Duct!haha just kidding!

[theTTshark]

Instead of Lapsims you could always do it the old fashion way. Wing tunnels (full or even scale) and actual driving! http://fsae.com/groupee_common/emoticons/icon_biggrin.gif Drive a car with and without wings and you'll feel the difference. Then go buy some simple data recording devices and do some real tests. Granted time may be short before SAE East or even West, but you have a lot of time over the summer.

[sbrenaman]

Why would you waste your time designing, building, and testing wings, if you didn't know that they would benefit the car?

[Thrainer]

Originally posted by Lorenzo Pessa:
...
ET3 endurance score: 240,354 pts
ET3 "undertray" endurance score: 247,548 pts
variation: +7.194 pts
...

Hi,

has anybody analized the effects of aerodynamic devices on Economy and Efficiency score? For the events with FSAE rules, you should see how many points you loose. For events under FSG rules (effiency score based on consumption and lap time), you might be able to find an area with positive effects.

How is the topic of energy consumption treated at Monash and other aero teams?
Has any team built aero with the intention to reduce air drag (compared to a car without aero)?

Regards,
Thomas

[Demon Of Speed]

Originally posted by nowhere fast:

I agree that a lap time simulator is immensely valuable but disagree that building one takes a lot of work. With lap sims (most sims actually) it is best to start simple and iteratively build in refinement.
I built a Matlab based lap time sim recently to help with conceptual design of a motorbike I'm working on; and it only took a couple of hours here and there after work to get something useful. From my experience I would say that a good lap sim could be built in one weekend.

I have no idea what you are are accounting for, but just to make an acceleration simulator took me like 6 months (and it isn't very user friendly). To add breaking to that shouldn't be that bad though.

I started on a "simple" model which was going to not look at line or corner transients (which was because I was having such large issues with transients) and after around many hours and not getting very far I gave up. I don't have detailed enough tire models, to be able to accelerate account for the suspension kinematics that effect weight transfer. Not to even mention how road surface will effect (besides just grip) both the tire and damper oscillation, which can intermittently effect grip levels. Well on the road surface I remember about banking, hills, valley, and everything that can happen, even well steady state.

Well looking at transients a two degree of freedom model dosen't even come close to what you need to have to look at the vehicle well in transients. I think I had calculated around a 57 degree of freedom model to be able to know what the car was doing at all points. The model had I solved it would have been able to do ss conditions and save me time. After solving a couple of 3 degree of freedom models for a class and taking most of a day to derive the equations and solve them I choose that the model was simply to complicated.

Next we have "line". Sure one can start with just making the car follow a single line, but that is not what really happens on a course. This really doesn't do you much good. So you then have inputs in the tack to add width and calculate out the line based on keeping the turning radius's as large as possible, which isn't that bad computationally (takes about a day so it will automatically do it for any track configuration). Then you remember back to the road surface and have to desired how hills and valleys and banking effect the line (which is integrated into how the car is). Then you can add in the transients of the vehicle on how it effects the line (which is also effected by how the car is).

The question I have for people who have a "lap sim" what do you feel is the minimum level needed to be able to get enough accuracy when writing a simulator (I know it is subjective, but would like opinions). So I know I am way off topic, and I will fix the spelling and grammatical errors later.

[Luke Phersson]

Thrainer,

We've done both on-track and theoretical (lap/comp sim) studies on the effects downforce and drag play on fuel usage. Interestingly, CL has a greater effect on fuel consumption than CD (roughly 2/3 from CL, 1/3 from CD) due to the higher average velocities you can achieve around the lap - you can essentially react/use more energy from the engine due to the higher grip levels.

In terms of the lapsim, it’s relatively easy to implement fuel consumption through changes in kinetic energy throughout the ‘lap’ – just assume and tune some thermal efficiencies to give you similar quantities you’ve found through physical testing. All this has been covered quite well in Geoff’s posts.
Acknowledging fuel consumption is essential to figure out the best compromises to make between target CL and CD – it’s possible to reach a crossover point when you’re getting faster lap times, but scoring less overall points due to fuel consumption. This makes heavily adjustable aero useful where you can run a lot more downforce for skid-pad and autocross then back it off for the endurance event (Did anyone notice our rear wing the last few years in skid-pad?).
I’m not sure why teams would go into analysing (to a reasonable degree) their cars aerodynamics purely to reduce drag, where for the same amount of resources and effort they could get more points from downforce?


Demon of Speed,

The level of complexity really depends on how many people you have at your disposal to work on this, and how competent they are. You can get incredibly useful information from a very simple lapsim (Wheelbase, track, CG, average mu, and power), see a lot of Geoff’s posts about that subject. A simple lapsim with a lot of assumptions that works is significantly more valuable than one that explains exactly what the car is doing everywhere and doesn’t work. Once you begin to complicate the sim with transients and detailed tyre models you lose the ability to quickly iterate design choices which is the whole point of the lapsim, although if you could incorporate those into the model successfully – great!

[murpia]

Originally posted by Lorenzo Pessa:
Or an old fashioned Chaparall style third pedal mechanically connected to the wing. simple, efficient, cheap.
Why bother with a 3rd pedal? Connect it to the throttle pedal above say 60% travel:

0% Throttle = need downforce (braking or cornering)
1%->60% Throttle = need downforce (accelerating on the traction limit)
60%->100% Throttle = need low drag (engine limited acceleration in a straight line)

Regards, Ian

[theTTshark]

Originally posted by sbrenaman:
Why would you waste your time designing, building, and testing wings, if you didn't know that they would benefit the car?

Haha I agree you need to have some sort of means of justifying even starting the process, but I think that sometimes we forget the way engineers used to do it (physical tests). When Dan Gurney first stuck a Gurney Flap on one of his car's, he didn't really know if it would work. He didn't have simulations, but he had laptimes. All I'm merely trying to say is that we shouldn't focus just on one way of justifying a design. The complex nature of air flow around our cars is sometimes best dealt with in actual testing. I'm just merely pointing out a different way of thinking and approaching the problem, not saying that your's is wrong. Because you're right, you need to justify it, but also no one has really spent any time about the reality aspect of testing.

[andyman61]

Has anyone (or does anyone know of a team) done a full evaluation of an aerodynamics package from a points perspective and made the decision NOT to run aero? I can fully understand not having the resources (manpower, design tools, or manufacturing) to do it, but has anyone had all those things available and actively made the choice not to run full aero because they felt it would be worth fewer points?

[Lorenzo Pessa]

A really short report of my MS thesis.
I'm sorry I can't be more precise. I started immediatly to work and I have really few free time.

5 different aero configuration studied.

1. basic: no aero car [basic; on the geometry of my team 2010 car; 280 kg with driver]
2. full aero: front & rear wing, underbody with tunnels [F&R_W&UB]
3. front & rear wing [F&R_W]
4. rear wing and underbody with tunnels [R_W&UB]
5. underbody with tunnels [UB]

the wings are simply shaped on a two element Eppler 420 wing profile with a C_lift of -2.97 (not so much but I had no time to optimize it!)
the rear wing package is 570 * 1350 * 630 mm and 4.2 kg
the front wing package is 470 * 1250 * 180 mm and 2.4 kg
the underbody weights 3.6 kg

radiator simulated as porous media without heat exchange.
tested at 10 m/s
roll 2°, yaw 0°
front wheels in a right turn
wings mounted on frame

2.6 M polyhedral cells


CFD Results (coefficient * frontal area)

config: [basic; F&R_W&UB; F&R_W; R_W&UB; UB]

Drag: [0.65; 1.04; 1.05; 1.03; 0.59]
Lift: [0.01; -1.82; -1.79; -1.25; -0.16]

Efficiency (-Lift/Drag) [-0.02; 1.75; 1.71; 1.21; 0.28]

Radiator inlet mass flow drop (basic is the reference)

mass flow drop left radiator (%): [0; -56; -50; -7; -14]
mass flow drop right radiator (%): [0; -48; -33; -1; -9]



Cost of aero components (with a rough extimation)

Front wing 570$
Rear wing 1290$
Underbody 930$

The event points are calculated from the results of FSAE-Italy 2010

Lowest cost event (basic as reference)

Point variation: [0; -6.43; -4.5; -5.26; -2.35]


The skid-pad event is simulated. A neutral setup is used for all configurations.
Because the friction coefficient of the skidpad track is unknown, three different results are calculated. Skidpad event position 9# (5.176s), 14# (5.303s), 28# (5.638s).
It would be better to vary the friction coefficient, again... I had no more time.
Points evaluated on FSAE-Italy 2010 results (Tmin=5,050s).
the setup of R_W&UB is far from optimal so don't thrust the result of this configuration. The others are not perfect but quite good.

09#; point variation: [0; 3.31; 3.86; -2.79; -.15]
14#; point variation: [0; 3.18; 3.68; -2.64; -.15]
28#; point variation: [0; 2.81; 3.23; -2.36; -.12]


Final variation (skidpad+cost)

09#; point variation: [0; -3.12; -0.64; -8.05; -2.50]
14#; point variation: [0; -3.25; -0.82; -7.90; -2.50]
28#; point variation: [0; -3.62; -1.27; -7.62; -2.47]


The remaining dynamic events have to be evaluated with a dynamic car sim.

You are free to ask for more detail.

[Scott Wordley]

Very interesting data, great work.

A few comments:

Was this a symmetry model, ie only half the car? I would guess not if the front wheels have been turned? If its full car, then you are getting very light on mesh, with only 2.6M cells, but you make do with what you have computer wise. Would perhaps be better to make your model symmetrical and only analyse half?

The numbers you quote look pretty reasonable. Just to confirm, you have given the CL.A and CD.A right? We measured around 3.3 (CL.A) and 1.7(CD.A) for our 2003 car which had bigger more aggressive wings than what you simulated.

I was surprised by how little downforce the undertray made on its own. Judging by the amount of teams running undertrays and no wings, you would think they would work a lot better. Perhaps they just provide a good talking point for the aero judge in design? http://fsae.com/groupee_common/emoticons/icon_wink.gif

Your radiator cooling flow drop was very interesting, and in line with our experiences. You need to move them right back near the rear wheel if you want any reasonable flow.

Your cost are way high, from our experience anyway. In reality our full aero package costs about half that to build, and using the FSAE costing formulas maybe half that again (for the cost event). So if you used those costs for the point analysis you will be penalising wings quite heavily, as seems to be the case.

For skidpan do you take into account weight and CG Hieght changes due to the various components?

Even still, I think the results are very clear.
No one should put wings or diffusers on their cars... please!

I might ask one of our guys to run these numbers through our sim and see what they get by way of comparison.

Cheers,

Scott

[Lorenzo Pessa]

Originally posted by Scott Wordley:
> Was this a symmetry model, ie only half the car? I would guess not if the front wheels have been turned? If
> its full car, then you are getting very light on mesh, with only 2.6M cells, but you make do with what you
> have computer wise. Would perhaps be better to make your model symmetrical and only analyse half?

I used a complete car with turned front wheel, 2° of roll and 0° of yaw.
I prefer to work on a turning car because everyone in the world make the CFD of the car on the straight. I need downforce in a turn so I made the CFD on a turn.
In a more reasonable way I worked on a "method": how to study a FSAE aero car? I tried to setting up a general case linked with car dynamic simulation. I was also corious about flux asymmetry under the car.
Thinking about an example, it was easier for me to work on skid-pad. The dynamic simulation need a very "simple" simulator so it was easy for our sim guy to find time to run it for me.

2.6 M of cells because I could use an "old" 8 core pc in my lab without making queue for the everytime busy university clusters. When I started I had no idea how many sim I have to run (I started with 2 cases and I finished with 5).


> The numbers you quote look pretty reasonable. Just to confirm, you have given the CL.A and CD.A right? We
> measured around 3.3 (CL.A) and 1.7(CD.A) for our 2003 car which had bigger more aggressive wings than what
> you simulated.

yes: CL.A and CD.A

> I was surprised by how little downforce the undertray made on its own. Judging by the amount of teams running
> undertrays and no wings, you would think they would work a lot better. Perhaps they just provide a good
> talking point for the aero judge in design? http://fsae.com/groupee_common/emoticons/icon_wink.gif

My undertray could be designed better! I had to design it around a frame not optimal for mounting an undertray and I had no time to optimise it. I'm not sure it work properly.

> Your cost are way high, from our experience anyway. In reality our full aero package costs about half that to
> build, and using the FSAE costing formulas maybe half that again (for the cost event). So if you used those
> costs for the point analysis you will be penalising wings quite heavily, as seems to be the case.

I used FSAE costs indirectly. I started from 2009 and 2010 costification then I made up a simplified cost model.

> For skidpan do you take into account weight and CG Hieght changes due to the various components?

Yes.

> Even still, I think the results are very clear.
> No one should put wings or diffusers on their cars... please!

I hope someone in my team will work on aero package next year.

> I might ask one of our guys to run these numbers through our sim and see what they get by way of comparison.

It would be very interesting!

[StevenWebb]

what car and wing weights were you working from?

also, what were the skidpad times? i'm getting a 4.59 points advantage for config 3 in skidpad times wer 5.107 for base car, 4.995 for config 3(210kg bas car, 12kg added for wings in config 3)

FWIW our sim has constant tyre coefficient with load so will always show an over-advantage of wings

FWIW2 config 3 showed 4.59 points in skidpad, and a 23.53 point advantage in all dynamic events (+4.59 skidpad,-1.72 accel,+18 autoX,+20.45 endurance,-17.68 fuel) assumed 60kw car

[Lorenzo Pessa]

Originally posted by StevenWebb:
what car and wing weights were you working from?

also, what were the skidpad times? i'm getting a 4.59 points advantage for config 3 in skidpad times wer 5.107 for base car, 4.995 for config 3(210kg bas car, 12kg added for wings in config 3)

FWIW our sim has constant tyre coefficient with load so will always show an over-advantage of wings

FWIW2 config 3 showed 4.59 points in skidpad, and a 23.53 point advantage in all dynamic events (+4.59 skidpad,-1.72 accel,+18 autoX,+20.45 endurance,-17.68 fuel) assumed 60kw car

Hi Steve,
I update my post with data
The sim used a costant with load friction coefficient.
Try with 50kw :-)

Thank you for this comparison!