2013 FSAE-Australasia

Thanks all, and thanks Anonymous Aero Judge for chiming in. The input is most appreciated.

AAJ, you have touched upon a point that was raised many times with me across the weekend. I won't mention team names, but approximately 22 of them mentioned at one stage or another across the weekend that they don't have access to the kind of aerodynamic development tools that Monash have. For the sake of these 22 anonymous teams, could I kindly ask you for a few words about:
1. Whether having a big wind tunnel helps your design score
2. What kind of aero development tools and tests would you use if you were building an FSAE car at a non-Monash university

I have a feeling that a team could score really well in Design if they could present a well-developed aero package without the use of a big wind tunnel. I also have a feeling that this is much more achievable than many teams think. And I also have a feeling that teams that call themselves "non-aero" teams might do themselves a lot of good in the design event if they were to present evidence of solid testing program justifying why they DON"T have aero, rather than using the "we don't do aero" argument as a scapegoat.

Apologies Mr AAJ if I am way off the mark with the above, but any feedback would be greatly appreciated. ;)

Cheers all,

BB,

Both aero judges received similar feedback about resource levels. Our parting remarks were almost always to demonstrate that it took better ideas and not better tools to develop a better package. There's stuff that's difficult to conceptualize, stuff that students might just see the links between other systems and aerodynamic influence, there's low-hanging fruit and then there are ideas so obvious, so over-ripe and yet so neglected that they're on the verge of turning into compost. The vast majority of what can be done for aerodynamics at the present stage is nearly compost. They're ideas that are so obvious that they're a trip hazard for good students.

Whilst some teams had a good amount of resources, they were almost universally poorly used relative to any industry standard. Having the ability to one 100M+ element CFD runs means very little, particularly if the methods employed bear little fidelity in critical elements to the real thing, or bear little relevance to real use, or are visualized in manners not extracting truly salient data, or aren't tracked in manners highlighting true gains above noise variables. Or, worst of all, if these '2013' tools are combined with poor research or fundamental skill sets, leading to designs which are born of essentially very finely discretized assessments of flawed ideas.

End result? The vast majority of what's on the 'aero' cars could have been designed in the 70's and 80's. Without wind tunnels. When a pack of 8086's wasn't running much for CFD. There's so much that can be done with far simpler methods, most of which are accessible to all teams. The history of aerodynamics is littered with teams delivering projects on tight resources with smart, simple, accessible and robust testing. Be like this.

If future Monash teams want to use the 'big wind tunnel', I fully support them. Because that wind tunnel, like any other test environment, is just an approximation of the intended practical environment within certain limitations. It isn't absolute. I didn't see a uniformly-understood cognizance of what the limitations of that tunnel actually are relevant to what's being attempted. To reconcile CFD, trackside and experimental (tunnel) project results is an extremely difficult undertaking, one that I think Monash is still developing. Certainly what was presented was very far from complete (which impacted design scores adversely). I support Monash using that tunnel (as I would any other team using it, or a drag strip, or a car park, or whatever) because if they do, they'll have to explain it (and it's relevance, and limitations, and fidelity, and and and) at design judging. Succinctly and unequivocally. Be careful what you're wishing for this Christmas... come next December, you'll have to know it backwards enough to explain it. What resources a team cannot truly stand behind are wasted effort. Don't worry about rocking up next December pitting next to a team that had access to different toys to play with, worry about rocking up next to a team that invested in better thought for what they had. The noise floor between intent and execution in FSAE is so high that a wind tunnel or a cluster of CFD clusters difference between teams isn't going to beat out fundamentally better ideas for some time yet.

The point is to award prizes for smart work, not for a volume of any quality of work. Work accordingly.

Not a single team demonstrated sufficient basics on simple wing theory. I picked on the Gurney flap example as not one team had it employed correctly, because not a single team could explain from base theory how it worked completely, which explained why no single team could design one properly. It's been only a few years that students had ready access to CFD resources capable of 20M+ element runs on reasonable turbulence models. The Gurney flap's been around since 1971. Try mastering 1971 before toying with 2010+ tools. Walk before running. It'll give way to a better design all said whether you've a wind tunnel or a car park at your disposal.

This all said, the bar to pass is understanding, not perfection. Integration in a cohesive design inherently means compromise in some aspects. Demonstrate understanding.

Full credit to the Monash team for chasing me down and embracing critique. Some really robust discussions have followed.

Australians are overrepresented in high-level motorsports aerodynamics, and I'd look forwards to hearing of a few of you (at Monash and other teams) continuing this much.

These discussions here about aero are why I believe that aero is a part of competition that gets weighted far too heavily in the competition. Yes, it is currently all but impossible for a non-aero car (I'm talking no wings or undertrays - Car with suspension and tires) to win any dynamic event other than acceleration and fuel efficiency, but as many have said, this competition ISN'T about racing. Yes you can say that it is rather easy to get CFD and apply your basic fluids classes to make a relatively competent aero package, but coming from a small team, when it comes down to it you need all hands on deck getting your car to competition, and you really can't spare a few members to design and make an aero package that works, and those who you can spare may not necessarily be the most capable making the aero also?

It would seem to me if the design competition is truly about the team members understanding and applying the basics, and knowing why they did everything they did, it shouldn't be a punishment if you decided that investigating an aero package was an inneficient use of resources. Its pretty easy to see from a sim that aero will make your car faster, but if your goal is a cheap, simple, efficient car (in the original spirit of the competition) then I'd have to say lack of aero wouldn't necessarily be a bad thing. If you're using the "weekend warrior" design goal by having all your car cheap and easy to fix in the case of damage (monocoque vs space frame) then you should be able to use the same excuse for not having aero (getting balance correct can be difficult, amateur drivers really won't be able to use it to its full effect, it would add levels of complexity to your package).

Also, with the amount of teams who said they're there to learn rather than to win: I'd rather a team be realistic about their expectations than think they're better than everyone else. If any person involved in this competition truly believes that a small team with no money can win this competition outright, they're delusional. Can they do good? Yes. We did pretty well last year with one of the smallest overall budgets you can get (and a lot of out-of-pocket purchases). But really the gap between the top 5-10 teams in the world and everyone else is really quite large, and the required resources to get that next level higher becomes nearly exponential.

This may be a bit of a rant, but it is.

Quote:

---

Originally Posted by AxelRipper

These discussions here about aero are why I believe that aero is a part of competition that gets weighted far too heavily in the competition. Yes, it is currently all but impossible for a non-aero car (I'm talking no wings or undertrays - Car with suspension and tires) to win any dynamic event other than acceleration and fuel efficiency, but as many have said, this competition ISN'T about racing. Yes you can say that it is rather easy to get CFD and apply your basic fluids classes to make a relatively competent aero package, but coming from a small team, when it comes down to it you need all hands on deck getting your car to competition, and you really can't spare a few members to design and make an aero package that works, and those who you can spare may not necessarily be the most capable making the aero also?

---

Let me reiterate what was said at the end of the competition, as it’s not about mega downforce, wind tunnels and uber resources.

I believe a team can be competitive without a so-called ‘aero package’. I’ll say ‘so-called’ simply because there are a few teams optimising for downforce, but none out there optimising for simply low drag. Which is super low-hanging fruit. Would a super-optimised aerodynamics package obliterate a car not so equipped? Sure, but I’m yet to see such a car. It’s probably beyond student experience to design it, and that’s OK. It’s the nice thing about FSAE, so many ways to skin the proverbial cat. Serious downforce can’t be applied successfully ‘by a few team members’. It’s not a project that can be truly successful without an integrated approach. This is yet to happen in FSAE-A (at least what I saw at competition).

And whilst it’s easy to use CFD, its not easy to use it well – there were very, very few applications at FSAE-A where the bulk of the CFD work was applied in a relevant manner. Yet there were a few cars that had some super basic assessment done, just simple head checks to see it all worked as intended. Very far from CFD. They were favourably assessed (deservedly so).

However whether your car was designed by a large team, a small team, spared people or otherwise, cars do not drive in a vacuum. Effective engine intake paths and cooling paths, nose designs to alleviate front lift are super basic. They’re aerodynamics problems. They’re not an afterthought on any vehicle I’ve worked on – they’re not the sexiest part of development by far – but its basic work to apply the basics, and teams should be assessed on them.

Quote:

---

Originally Posted by AxelRipper

It would seem to me if the design competition is truly about the team members understanding and applying the basics, and knowing why they did everything they did, it shouldn't be a punishment if you decided that investigating an aero package was an inneficient use of resources. Its pretty easy to see from a sim that aero will make your car faster, but if your goal is a cheap, simple, efficient car (in the original spirit of the competition) then I'd have to say lack of aero wouldn't necessarily be a bad thing. If you're using the "weekend warrior" design goal by having all your car cheap and easy to fix in the case of damage (monocoque vs space frame) then you should be able to use the same excuse for not having aero (getting balance correct can be difficult, amateur drivers really won't be able to use it to its full effect, it would add levels of complexity to your package).

---

There’s no reason to have to have a mega-downforce package, but what you’ve offered is a poor set of excuses for not applying any forethought to aerodynamic considerations. Cheap, simple and efficient doesn’t afford a license to be selective about which design elements are embraced at the expense of ignoring others. Teams are free to de-prioritise them but be smart about it. Not knowing how a radiator duct works enough to design one properly – not smart. Runing a high-lift nose design because it ‘looks cool’ – not smart. (Whilst designing pretty things is often an industry requirement, it’s still got to meet performance targets.)

Hey, we had one team that actually had a good excuses for not having any optimisations in lift, drag or cooling – was an electric car, cooling requirements were very low, yet their powertrain afforded them a set of considerations where neither increased lift nor reduced drag would have made significant differences for their points target. ‘Twas a capable explanation from first principles.

Just be capable of explaining how your car makes efficient use of the other main environment it drives in (air – as it drives on road and in air) and teams should be fine. Those that did funky things with it were eligible for a greater points share regardless.

Quote:

---

Originally Posted by Big Bird

Purpose of Design Judging:

The primary purpose of the Design Event judging is to rank and score the teams for the Design Event aspect of the competition....

The "design conflict" we face as designers of the design event, is that of how much time we need to rank and score the team's design effort, (both in terms of the students' knowledge, and its embodiment in the vehicle presented), versus how much time the students want to spend talking about their car...

There is a distinct danger in trying to deliver a Design Event that is all things to all people. I would agree that it is our most prestigious event, and it is a disappointment to me that I was never part of a team that won Design. But just like in vehicle design, if we pile too many hopes / objectives / expectations in the one receptacle, we are going to end up with one Frankenstein’s Monster of a final product. (Note to self – print previous sentence, stick to wall above desk, re-read whenever you are about start writing another forums post …  )

---

Who are your customers? What do they want? That should be on the wall above the desk of every design engineer.

Geoff, I think you're on the right track here, and I agree with your intentions if not your direction. But before I expand on that, a bit of background. I spent the first third of my career as a design engineer designing products for customers: earthmoving equipment for the construction industry, nuclear weapons for the US government and inkjet printheads for consumers. In the 24 years I have been at OSU, my focus has been on designing processes for turning bright young people into work-ready design engineers. I approach that process design the same way I approached product design, starting with: Who are your customers? What do they want?

I just finished my fall term “Introduction to Design” course, with 104 junior level mechanical, manufacturing and industrial engineers. One of the hardest jobs in design engineering is figuring out prioritized customer-driven engineering specifications through which one can measure the “goodness” of a product or process design.

This process is expected of FSAE teams by the design judges, and I would expect it of the FSAE/FS design event organizers as they design the design judging process. Who are the customers, and what do they want? Prioritize those qualitative customer requirements, and develop a measurable set of quantitative engineering requirements by which any proposed design judging processes can be benchmarked against the existing process.

Posting your thoughts and receiving feedback from this community of “customers” is a good start.

Quote:

---

Originally Posted by GTS

"As knowledgable as most design judges are about how to make a car go fast on the track, very few have taken the time to build and validate an FSAE point simulation."

I'm not sure of the point of the comment - whilst quite a few teams had these, very few could actually explain their workings, or explain or justify their limitations.

---

I was responding to Z’s question as to why and how GFR spends the first 5 minutes of design judging on an overview. One of the main things we do in that 5 minutes is describe our vehicle dynamics models and our point simulation, their 6 year evolution, explain and justify their limitations, explain their validation by physical testing, and how we use the them to drive the design of our vehicles. As most design judges have little experience with point sims, we believe this is important information that every design judge evaluating our car needs to know.

Quote:

---

Originally Posted by bob.paasch

I was responding to Z’s question as to why and how GFR spends the first 5 minutes of design judging on an overview. One of the main things we do in that 5 minutes is describe our vehicle dynamics models and our point simulation, their 6 year evolution, explain and justify their limitations, explain their validation by physical testing, and how we use the them to drive the design of our vehicles. As most design judges have little experience with point sims, we believe this is important information that every design judge evaluating our car needs to know.

---

Bob, I understand what you're saying, however:

Most judges are aware of simulation tools for competitive constraints - your statement that most judges have little experience with them is a little arrogant, and,

Whilst most teams at FSAE-A had employed point sims to guide their design choices, only two of them could explain what methods they employed, what assumptions they made, the limitations of the methods. Only one - ONE - could cohesively construct any arguments on the limitations of the directions given by their point sims and how these limitations were handled in design processes (let alone in any validation exercises). Which is poor. As a cool tool unknown has a high propensity of generating garbage results for a variety of reasons most are familiar with.

Teams were still asked to submit design briefs, specs and a video, and those willing to put effort into them were (at least in our area) judged within the context of their project direction, which itself was duly interrogated for rigor and robustness. Done correctly these teams had more than a five minutes of fame to go over how and why their project came to be. I'm not sure why all teams didn't put rigor into it. Possibly it should be made mandatory, as should the judge's reviewing them be made so.

(Apologies for the long-windedness of this post, but I'm going bush for the next week, so wanted to get this out before I forgot it.)
~~~~~o0o~~~~~

CRITICAL THINKING and CRITICISM.
=================================
Clearly my motel bed in Werribee was far too soft, as evidenced by the fairy-floss commentary I wrote earlier. So, time to restore things to their natural balance.

From many discussions I had with students and officials at Werribee, I was astonished to learn that the education system has truly descended further down the S-bend than I ever thought possible. It seems that the current attitude towards "criticism" is that ... most people don't like it! Apparently "critical thinking" is sort of OK, but "criticism" itself is not acceptable. Huh!!!???

It seems that this currently popular codswallop has been brain-washed into at least the last generation of students (X?, Y?, certainly NOT Z!), along with other such nonsense, such as;
"There are no wrong answers..."
"All solutions are equally valid..."
"It is all about choices, and you are free to make whatever choices you wish..."

Aaaaaarrrghhh!!!!!!!!!
~~~~~o0o~~~~~

I feel that I should give you, dear readers, a bit of my background, so that you can better appreciate my views.

Shortly after WWII, Old Mr & Mrs Z (my parents) escaped from behind the Iron Curtain. This was a time when most choices were VERY BAD CHOICES, with "terminal" consequences. Nevertheless, by the application of the right sort of critical thinking, they made it to a central European refugee camp. After a rather cold winter there, wearing newspapers under their one set of clothes, Old Mr Z decided to head for the peacefulness, warmth, and tranquility of Kenya Colony, East Africa, then part of the very civilised British Empire.

After a blissful half-decade or so, Little Z entered the scene. Unfortunately, he (I) did so right in the middle of the Mau Mau Uprising. So Mother Z had to give birth to Baby Z, with, quite literally, a Colt-45 strapped to her thigh. This was another period when critical thinking was of utmost importance, and most choices were bad ones.

Of course, I didn't experience the above events directly, but along with my parent's stories of those times, I did spend enough of my formative years going "on safari" in the Serengeti Plains to learn some important facts of life. One such fact is that if Young Grasseater thinks that he has "the right" to drink at the waterhole while Mr Meateater is there, even though Old Grasseater criticises such thinking, then Young Grasseater's contribution to the gene pool is very quickly eliminated.

Furthermore, on the Serengeti Plains a Grasseater aiming for "a Podium", or "Top 5", or similar in the running races, is making A BAD CHOICE. Any lower than First Place has you inside someone else's stomach. In fact, looking at the great herds of Grasseaters you see roughly one-in-ten who have a large flap of meat hanging off their rump (it looks a bit like an open fuel-filler-flap on a car), often with four long claw marks leading to it. These Grasseaters have actually WON their last running race, but only just! Their close finish means they have received some mild criticism from Mr Meateater, and they will certainly try harder next time.

And Mr Meateater is also a little bit hungrier now, so will be applying a generous dose of self-criticism to his flawed technique. "Should I have used more power? Or perhaps reacted more quickly...?" Fortunately for Mr Meateater, the hungrier he gets, the better his self-criticism and critical thinking skills become. That is how it works in the real world. Very educational. You should all try to see it some time...

I would like to waffle on a lot more about these things. But briefly for now, the oft-repeated notion that "... in Engineering there are many different, but EQUALLY VALID, ways of solving the same problem..." is a load of nonsense. It is a well known principle of ecology that if two different species enter a newly opened ecological niche, then, in the long run, only one species will come to dominate that niche. There is some sublety here, but for a given niche, one species is always FITTER than the other.

By analogy, for a given set of FSAE Rules, one car concept is always fitter than the others. All solutions are NOT equally valid, and most answers are quite definitely WRONG.
~~~~~o0o~~~~~

The above approach to "criticism" (ie. eating the criticised one) works well in the real, natural, world. But how do we helpfully criticise each other in our civilised, artificial, world?

The best example I can think of is the period from roughly 600 BC to 200 BC in Ancient Greece. Not surprisingly, the words "critic", "critical", "criticism", "critique", "criterion", and even "crisis", all stem from Greek kritikos = able to discern, from krinein = to decide/judge. IMO, at least half (the more fundamental half) of our present knowledge of maths, science, technology, medicine, political systems like democracy, +++, was all developed in this very short period of about four centuries.

How were so many new concepts developed in such a short time?

Most likely because the national pastime back then was debating stuff, any stuff, in the Agoras (ie. the Forums). One person would get up and present their views, such as Aristarchus in 3rd century BC arguing that the world is a sphere spinning on its axis, and, together with the other Planets, it circles the Sun at the centre of the Solar system, with the fixed stars much further away. Then other critics would get up and, in a very civilised way, criticise the begeezus out of such nonsense. After much toing-and-froing, the spectators, probably eating the ancient equivalent of pop-corn, would draw their own conclusions as to whose was the most reasonable argument.

Worth noting that shortly after Aristarchus, Eratosthenes measured the diameter of the Earth (quite accurately), and also the distance from Earth to Sun. Then, slowly, slowy, the debate was stifled. Eventually people had to believe whatever the high priests told them was the truth, because the high priests were, well, Very Important People! So for the next ~two thousand years the Earth was flat, and at the centre of the Universe, and you had to do whatever the important people told you to do...

Also worth noting, in Antiquity (and also in Medieval times), the education system consisted of studying the Seven Liberal Arts. These were split into two divisions, with the Trivium (the easy, or "trivial" subjects) of Grammar, Rhetoric, and Logic being studied first. These trivial subjects had to be learnt first, because the student had to be capable of correctly composing arguments (= G), delivering those arguments in a convincing manner (= R), and forming arguments with no obvious fallacies (= L), in order to competently argue their particular hypotheses, or to criticise others. Sadly, these trivial skills seem to be increasingly ignored in today's education system.

To summarise the above for FSAE, and also for society in general, CRITICISM IS GOOD FOR YOU!

Criticism is simply someone else's critical thinking coming towards you. It is educational. You are free to, and SHOULD, apply your own critical thinking skills to other people's criticism. You also can, and should, spend as much time as is feasible applying self-criticism to your own work. To repeat, criticism, critical thinking, and self-criticism are all educational. In the long run they get you to a better place.

Now it is true that in the current world-of-plenty anyone can do just fine, even if they are totally incompetent. There is an over-abundance of food, and very few real threats. But when times become more competitive, as they inevitably will (maybe, if you are lucky, still a few generations away), then the inability to think critically will have you inside someone else's stomach, or whatever is the more "civilised" equivalent.
~~~~~o0o~~~~~

(More next post - can't fit it all in...)

Z

CRITICAL THINKING and CRITICISM (continued...).
=================================
Getting back to Werribee, while I was wondering around the hall many individual students asked me for comments on their car, or their particular sub-system (or else I just offered the comments without being asked :)). I hope that these students apply their critical thinking skills to my ramblings, so they can decide whether to reject them, or consider them further.

However, only one team asked me, on an official Team-Management basis, to give an overall critique of their car. Perhaps the adrenalin left in my system from watching an exciting weekend of competition affected me, but I spent maybe 20+ minutes pointing out everything I considered wrong with their car. It is worth noting that this team were not the Dead-enders, or the Tail-enders, or even the Pretenders. They were one of the Contenders.

Now if this team got nothing more out of their half-hour of ducking my spray than a chance to exercise THEIR critical thinking skills on MY criticism, then they will be a little better for it. If only one of the many criticisms I made leads to one slightly improved part on their car, then they are another little bit ahead. And if my criticisms in some way help them SELF-CRITICISE other parts of the car that I didn't mention, then yet another little bit ahead...

To sum-up this rant (and as a last reminder to the above team ;)), the most important point I wanted to make is that you can learn far more about good Engineering by STUDYING NATURE, than you will ever learn from a textbook, or in a lecture theatre. (Well, unless you have a particularly good teacher, which the above team definitely does!)

In Nature almost everything is curved and tapered, rather than being properly straight, as is demanded in one of the FSAE Rules. In Nature things are almost always mounted in single-shear, and cantilevered, so that they have those horrible bending loads detested by all the Experts. Aaack!!! But DO NOT BE AFRAID to design such Natural structures, even though the Very Important People tell you that you mustn't.

Because I am quite sure that if the current crop of FSAE cars were running around in the Natural world, then most of them would very quickly end up in a more Natural creature's stomach. And that, to repeat, is the most effective form of criticism! :)

Z

Z, you got your voice back. The real Z voice, the raspy one with the barbs and the jagged edges. I was beginning to think we were being duped by a benevolent impersonator.

May I offer some criticism of your critique of critical thinking, as critically I’ve uncovered a critical flaw in the critical argument of your critique. I hope you are OK with criticism.

The nature of the complex ecosystem that we live in is that there ARE multiple winners. If the rules of the game are complex enough, that is. then good ol’ mother nature just dumps as many mutations on the ground as she can see fit and lets them play dice with each other, to see which mutations comes out on top. The winners are the ones who still retain a place on the planet after the game has gone on for some time.
In the game of life we have mammals and marsupials and insects and bacteria and the Kardashians and fungus. And much much more. Each life form has its own “survival trick”, but also its own weakness. Some have camouflage, some are fierce hunters, some survive by mass reproduction. Humans have the ability to co-operate and communicate better than other animals, so we can outwit our furry rivals. Our weakness – well we aren’t really all that threatening without our tools and weapons.

The Lion is big and powerful and fast. Therein lies the lions strength, but also its weakness. Power = energy / time, so we can give the lion power by enabling it to burn off lots of energy in a short time frame. The cost? It has to stop and rest for long periods of time. Why didn’t mother nature give it never-ending power? The super-lions fuel needs would exhaust the food available to eat – and thus it would eat its way to its own extinction. Nature is all about balance…
Nature needs its direct conflicts to deliver us the variety of life forms we have here on earth. The mighty lion can take down the buffalo or the deer or any of the other herbivorous meat packages, and eat mightily. But the lion can be taken down by the tiniest bacteria or virus. Spoiler strategies make life interesting, and stop “runaway successes”.

(As an aside, I personally think we did ourselves a disservice when we changed fuel scoring from outright fuel used, to “fuel efficiency”, being a combined measure of track speed and fuel used. We already had plenty of events awarding points for on-track speed, and the Fuel event offered a neat little spoiler opportunity for a team to build a car out of a lawnmower motor, tootle it around on Sunday drive pace, and win the event very nicely. As a pointscoring event, I really liked that we had the fuel opportunity and the cost opportunity for alternative strategies to come in from left field to steal points from the teams who focussed on being fast. But, alas, the teams and the supporters of the strategy of “go as fast as you can at all times” kicked up enough of a fuss that the organizers changed the Fuel Economy rules to a rather contrived “fuel per lap speed” formula…)

FSAE has inherent conflicts – mass vs stiffness, power versus fuel usage, downforce vs drag, etc etc. From these conflicts (if and only if the rules are balanced) come the different strategies leading to different concepts. The powerful car that suffers a little for fuel usage. The very light car that suffers a little for cost and maybe also stiffness. The low powered car that loses some lap time but gains a fuel economy trophy. The narrow car that kicks goals in slaloms, but may risk rolling in long steady state corners. Some are meat-eaters, some are herbivores, but all added in together we end up with lots of variety and a healthy competition.

In summary – umm I can’t remember. Something about multiple winners. Yeah, take a look on the roads outside. Have we converged on one design of vehicle for all our wants? No, because there are many different customers and customer needs. Same in FSAE – there are enough competing requirements that there very well be multiple equal strategies, but some focussing on cost and manufacturability, and others on carbon fibre and track speed.

Hope all that made sense. Time for bed. And Merry Christmas to all!!

Quote:

---

Originally Posted by Big Bird

...could I kindly ask you for a few words about:
1. Whether having a big wind tunnel helps your design score

---

(Cue Voice of Doom and Reason)... NO. It does not.

Any test environment is simply an approximation of the real world with (hopefully) measured noise and (ideally) known limitations. The real world (the track) has bugger-all limitations (as it's what you're trying to simulate) and a lot of noise.

More speed from useful downforce = f(results in test environment +/- noise)

Simply put your results need to be relevant and need to exceed the noise floor of the method used. I didn't see any teams at FSAE-A talking about their gains for 2013 in terms of being beyond a tested (noise) threshold for gains on track - which is super important as FSAE-A drivers are not professionals (for the most part) and the point simulators used are exceptionally crude in linking high-level aerodynamic performance data (CD.A, CL.A were thrown about a lot) to actual track gains.

Monash's tunnel will simulate a few bits of the car well, it will simulate a few poorly. If the team - any team - can explain this in competition and talk well of how their test plan was mindful of the limitations of their resources, then good work. For the level of development shown and the limitations of the tunnel being used, a team using a drag strip (or any straight bit of road), good speed sensors, some calibrated Idiot Lighting in the cockpit, some load-measuring sensors (get creative here) and a yaw probe would be no less disadvantaged. There would be a considerably higher noise floor to account for, much of which could be deconstructed and acquired. But there'd be a lot more in transferable results too, things Monash's tunnel cannot readily test with any accuracy owing to its inherent limitations would be particularly easy to develop this way.

Quote:

---

Originally Posted by Big Bird

2. What kind of aero development tools and tests would you use if you were building an FSAE car at a non-Monash university

---

A bit like this:

If min(KnowsBernoulliItsNotAPastaSauce(ProspectiveStu dents), UnderstandsWingTheory(ProspectiveStudents), CanWorkAManometerFrom1stPrinciples(ProspectiveStud ents), GetsDeLavalNozzle(ProspectiveStudents), CognisantOfGroundEffect(ProspectiveStudents), UnderstandsWingTheory(ProspectiveStudents), IOwnWhatASlotGapDoes(ProspectiveStudents), ReadUpOnKinematicVehicleLoadTransferBasics(Prospec tiveStudents), UpWithBluffBodyBasics(ProspectiveStudents), StraightlineCLMeansLittle(ProspectiveStudents), MyCoolingPathsAreTight(ProspectiveStudents), MinPressureDropsOnMyWatch(ProspectiveStudents), AeroNoAfterthough(ProspectiveStudents), DefenderOfTheLoadPath(ProspectiveStudents), abs(FearOfMATLAB(ProspectiveStudents)-1), FreeThinking(ProspectiveStudents), TinkeringSpirit(ProspectiveStudents))=1

THEN

You'd need to acquire velocity, load and wind conditions.

For speed use GPS. It's more accurate at lower cost than most will ever manage any other way. And it's cheap, really cheap, Sparkfun cheap. Don't turn up with MoTeC (or whatever DAQ package a team uses) plots if you have no idea what smoothing, averaging, etc MoTeC puts in their software on what's probably the single most critical variable in any aerodynamic breakdown. I really mean that. Small errors mean big differences here, so own that data path. Back it up with pressure-based speed if you're good. More later.

Load can be done a number of ways. Yes, uber linear shock pots are fantastic to have. Rotary pots or encoders repurposed from another life are the same thing with different noise/accuracy profiles. If it's below what you can live with, good. Pressure sensors are similarly useful if you know what flow structures you're looking for and what trends to analyze. They're useless without speed, so get that right first. Usable sensors are under $10, seriously. You'll need an aerodynamic speed measurement to backup your GPS if you want to get serious and use static pressure too. Research why, or risk having an aero judge next year (me, if I'm invited back) ask you why you didn't, and what it means.

Yaw sensors don't need to be expensive. Make your own, there's no need to get fancy, under $100 in parts and most universities have small wind tunnels enough to calibrate these things in. I'd mark anyone down turning up with a rapid prototyped version of this. Or take a weather station with and log wind performance sync'd with the car. Or triangulate from online data. Pros and cons of each.

Make sure it's all super-tight and super-tough. Acceptable safety factors in aerodynamic loads for your level are not 1 or 'umm'. If your rear wing delivers 60kg downforce, a 60kg person should be able to do a light tapdance on it and break neither themselves nor the car, and anything reading suspension displacement should be able to play back whatever Riverdance went on at full fidelity. If you can't transfer loads, don't bother, you're just adding useless mass and other complications. Pick your battles in build - not all aero components are easy to manufacture, but there are many ways to buy in wing sections and other components at low or otherwise reasonable cost - if you can't do it strong in-house, buy it strong. Don't short-change the design intent at the last hurdle, that's just silly.

Have a big think about the other stuff you need to acquire. There are a few more parameters needed to cut assumptions and limit noise. Or make assumptions and live with noise, choice is yours.

Then debug your DAQ - which you should do for anything acquiring a signal on a car that you hope to do anything with anyway. You'll need 1 part Understanding of Sampling Intent, 1 part Nyquist criteria, 1 part Calibrated Multimeter, 1 part Cathode Ray Oscilloscope or digital substitute, 1 part Understand of Signal Theory (or did we all sleep through Control Systems) and 2 parts Engineering Head. When your electrical system is under proper load and your DAQ is spitting out reliable, noise-minimised and credible data, you're good to go. Take records of this process, design judges will like you for it. A lot. To the point of thinking that you're exceptional. Seriously exceptional.

Then you need stuff to test. Don't waste time moving winglets in 2mm steps, this is not F1. Run significant changes. Take notes. Take setup logs. Talk to drivers and don't tell them what you're doing, just ask them what they feel. Your drivers are not Rudi Uhlenhaut. Work with your team to build a dialogue of 'this feeling means that factor which we can effect with these systems'. Pack multiple configurations. Don't argue over which idea is best, more good ideas are great and are stuff to test. Just find people to build them. Then work out what flow structures were at play.

And on that last point, flow viz EVERYTHING. There are good reference papers on understanding what you'll see deposited on an aerodynamic surface. Find them, print them, etch them into your heads. Use this to understand flow structures. Take many, many photos. Then think and plot and plan for a month and hit the track again. A handful of good flow viz photos >>>> wind tunnel and/or CFD development. For what happens in the product's natural habitat overrules all else.

There is plenty of information a Google search away on methods of testing on track for reliable aerodynamic data, on statistical methods to reduce variance, on relevant flow structures and systems, the lot.

There is also plenty around on what you'd pack to go testing with. Again, nothing expensive.

Done properly this approach would yield a product that would absolutely shred a car developed with a wind tunnel that can't simulate rolling road or rotating wheel effects... which is not to say the Monash team does so. The sticker on the underside of the rear wing might have mentioned something about a wind tunnel, but the talking heads talked considerably broader. As should we all strive to.

Who knows, BB, might be happy to partake in some sort of pizza night for those interested to talk methods and shop on the topic.

ELSEIF CantTestUntilCarBuiltThenItsTooLate=1

2D methods. Best practice in design. Build in some adjustment. Build test stands on sleds and drive them around. Get creative. Review the best in the field and conduct a decent lit review. Using your head is free, and thankfully not an option of you want to do well in design.

Be very wary of CFD. Saw a few simulations in the 60M+ element range but didn't see a thing about mesh quality (it's not just y+), any level of accuracy quoted beyond % (which isn't much beyond a headline number). The going rate for an industry-grade mesh for a car as complex a form as an FSAE car should be enough to suggest that it's beyond the skill range of any student in the competition. No slight on any of you (I hope at least a few of you go on to study more CFD in postgrad - its a fascinating field). Meshing is CRITICAL. Then wall functions are critical, then turbulence model is critical, and so on and so forth. And the majority of these runs weren't in relevant scenarios, and none were with CFD models of sufficient accuracy to justify the intended resolution of the method employed. A 60M-element run of a CAD model that's not quite the real deal in a vehicle position that's not quite a relevant deal, is, put sharply, simply breaking crap up into 60 million parts and then attempting to draw inferences from it. That's not to say these efforts are wasted or that there weren't some very sharp operators at FSAE-A (there were), just to say this: put more thought into how you use it, and why, as there's a good gap here that can either be closed to make for simpler runs, or more accurate runs, or both, and that it's a development from what's being done presently that will increase the fidelity and usefulness of this method.

ELSE

Don't bug me.

END (!)