Vehicle Dynamics starting points and design process

[ChristianChalliner]

Z,

Ok, If I read what you've said correctly then the point you are getting at really is that my mindset is incorrect?

I wont argue that you're wrong because maybe you are correct, in which case, how exactly would you prioritise your aims?

One thing I most certainly agree on is that the learning process is not linear and also that I've learned far more from things going wrong than I every have from things going well, it's just a shame that when things go wrong it tends to be rather painful for those involved!

It appears that the main point from everyone is to get a car running as soon as possible, which is great and would be ideal of course but I still have to do a base design which is at least competent before I can even get to that stage, I have taken some of Claude's advice with regards to simple sims rather than going straight to cad so I will get back once I've made some progress on those.

I did create a very basic OptimumLap sim with a 280kg car, the 2014 Michigan autocross course and base friction parameters swept from 1.35 to 1.65 both laterally and longitudinally resulting in the following:

Fy Friction = 1.65, Fx Friction = 1.65, Laptime = 59.93
Fy Friction = 1.65, Fx Friction = 1.35, Laptime = 64.37
Fy Friction = 1.35, Fx Friction = 1.65, Laptime = 61.68

So from base (1.65, 1.65) reducing the Fy friction to 1.35 increased the laptime by 1.92 (roughly +3%), reducing the Fx friction to 1.35 increased the laptime by 4.44 (roughly +7%) so from that basic sim it appears my initial suspicions that lateral events are much more important than longitudinal ones, so a setup biased towards good camber compensation in roll looks the way to go. How much more important seems a little difficult to quantify, to say twice as important seems a gross over simplification of the problem but maybe simple is the way to go.

I wont pretend we have a ergo rig, we don't, so far the driver environment guy has just been using the old car and repositioning the seat, pedals and wheel to get an idea of what most people have as a preference. But we obviously do have 4 wheels and an engine

Jay, I completely agree, everyone has jumped straight into detail design (myself included) which is not the way to go, obviously pulling everyone back to the 'big picture' stage might now take some effort but it can't hurt to try and is probably what is necessary.

Having set parts I actually find to be beneficial, as I have repeatedly told people, what we have are not bad parts, they have been put together as an effective package many, many times so it is not as though we are starting completely from scratch. I also find that knowing you have to use a set part prevents the urge to just go and look for something that is perceived to be 'better'.

As always, thanks for the thoughts

[Ahmad Rezq]

You reached a good point to set a concept: " car which performs very well in corners".
but something to say by fixing engine and tires you will make your live harder especially if you don't have a good power train team you may end with a car which won't be able to run. so consider that.

[Claude Rouelle]

Christian,

You wrote:

"Fy Friction = 1.65, Fx Friction = 1.35, Laptime = 64.37"
"Fy Friction = 1.35, Fx Friction = 1.65, Laptime = 61.68"

But you also wrote "my initial suspicions that lateral events are much more important than longitudinal one"

Did you mix them up?

Claude

[ChristianChalliner]

Claude,

I don't believe I have mixed them up, maybe it is my axis system which is different to yours. I have defined Fy as the longitudinal axis and Fx as the lateral axis as it were. If this still seems wrong to you then perhaps I have made a mistake somewhere.

That does remind me though that I must stop switching between various axis systems. Our CAD program defines x as the lateral axis (left positive if viewed from the rear of the car), z as the longitudinal axis (forward positive) and y as the vertical axis (upward positive) this may seem unconventional (it is not how I would have defined them) but it does mean that even if someone forgets they shouldn't be able to create a part in the wrong orientation (or export points incorrectly).

[MCoach]

Christian, The SAE axis system, which is de facto around automobiles, is Fy is lateral, Fx is longitudinal and Fz is up/down.

https://law.resource.org/pub/us/cfr/...994_004_01.jpg

Z+ is up or down depending on who you ask, but traditionally Z+ is down because it was defined by aeronautical engineers.

Aligning your CAD design software with something like this (or perhaps the ISO designation if more comfortable?) will help standardize everything in design and manufacturing.

[Z]

Christian,

First up,
X = longitudinal (accelerating and braking forces),
Y = lateral (cornering forces),
Z = vertical (gravity and aero forces).

In the whole wide world, this is the only "standard" that is actually quite STANDARD (cf. drive on left/right of road, measure with metric/imperial, etc...).

PLEASE, PLEASE, PLEASE, everyone, try to stick to above!!!

(As for +/- directions, and position of the origin, I won't even try...)
~~~o0o~~~

Back to FS VD...

Question by Christian: "... how exactly would you prioritise your aims?"

Answer from Z (page 1): "To repeat my mantra, at this stage your Team's goal is to build a four-wheeled-chair that can propel itself and the driver at an average speed of 50+ kph, over a distance of 30 km, WITHOUT FALLING APART."

As noted by myself and many others here, the above task is not as simple as most students first think.

Hence you should PRIORITISE A "FIRST DRIVE" DATE that is a good FOUR months before comp, just so you can achieve that "average speed of 50+ kph" at the comp.

Importantly, DO NOT think that you have to have a really good understanding of VD, or of tyres, or of engine-tuning, or of any of those racecar things, BEFORE you start the design. In fact, do not even expect to have a really good understanding of any of those things when you finish the comp.

Instead, you should accept now that much of what you will be doing over the next EIGHT months (yep, time is running out!!!) will be like a blind man searching a darkened room for a black cat ... that isn't there! You, and all the other Team members, just have to put your heads down and work to the above goals.

Namely, DRIVEABLE CAR EARLY MARCH. If diff mounts rip out in mid-March, then work harder until the problem is fixed. Repeat until you can consistently cover those 30 km at average of 50+ kph...
~~~o0o~~~

Some detail stuff below, but more BLUNTNESS now (sorry, but hopefully worth it in the long run... )

I had a quick look at the Aston Uni FS Team website, and it is worse than I imagined.

* The Team has competed at FSUK every (?) year since 2002, so is one of the MOST EXPERIENCED Teams in the UK. Ever.

* The amount of CarbonFibre and CNC-machining in the photos indicates one of the BEST RESOURCED Teams in the UK.

BUT!!!

* No indication of any significant success. Ever.

* No indication of significant knowledge transfer.

* Much useless complexity (eg. glued CF-tube wishbones, +++).

* Atrocious Ackerman on one year's car. And much more...

Bottom line here, is that this Team has a deep CULTURAL problem. Many, many, Teams suffer from this same problem. It is a case of locking yourself into a never-ending loop of underachievement. I have summed this up in the past as "polishing a turd".

M.C. Escher knew this process well. Note how one sub-team is continually INCREASING horsepower, and torsional stiffness, and all the other things that should be increased, while the other sub-team is continually REDUCING mass, etc., etc...



Ahhh, yes..... Non-stop progress, year after year after year...

But, look on the bright side, Christian. You can be the hero who achieves everlasting fame and glory because you freed Team-Aston-FS from this terrible prison. All you have to do is get the car built early!

(Note that this may require some diplomatic negotiations with your Supervisors. I suggest you ask someone more knowledgable than me for advice on this subject. Geoff?)
~~~o0o~~~

Prioritising big-picture VD stuff.

Most of your big decisions have already been made. This means that extensive Simulations are not necessary now, which is good because you can save time there.

The 13" wheels/tyres and four-cylinder-engine make outright victory much harder, but a top-ten at FSUK is still quite achievable. The tube-frame is NO handicap at all. If everything is running well a month before comp, then you might think about some quickly cobbled together aero. (<- DO NOT overthink this, so NO CFD!!!)

For now, your aim is to minimise the junk that is usually piled onto these cars! So SIMPLE suspension, SIMPLE frame, etc. ALL HEAVY PARTS TO BE MOUNTED LOW AND CLOSE TO THE CG. Putting this another way, you want minimum parts count, and squash everything that remains downward and together.

I would aim for minimum wheelbase (+5 to 10 cm to cover for cock-ups, or last minute changes) and driver's feet on front-axle-line. You probably won't manage this, but avoid unnecessary length. A round number of 1.2 metres for F&R track is good enough. "Optimising" these numbers will make very little difference to your overall performance. Wasting time on the optimisation process will give you the same result as last year.

Post concept sketches ASAP, and be prepared for Grumpy-Z to say "NO, NO, no, no, no.... you can do better...". The better the job you do at setting the correct direction now, then the quicker you will reach your destination. Final concept sketches (ie. topologically correct) end of November. Final detailed drawings (ie. with all dimensions), and hence first cutting of metal, by Xmas?
~~~o0o~~~

Some slightly more detailed stuff, though might be helpful to know now.

Here is one of my posts from the "Caster/Camber/Steering shims" thread from back in 2005. This discusses the range needed for these parameters, and their tolerances (I called it "accuracy" back then).

Given that earlier you said,
"Similarly, I was pondering the thought that all systems will have compliance to some extent so maybe analysing and 'tuning' the compliance may be more beneficial to car performance than just saying "well if I make it as stiff as possible it will be negligible" because in reality I'm pretty sure that the compliance can't/wont be negligible.",
it is worth asking yourself why I suggested those particular tolerances.

I can answer that question later if you want, but briefly for now, SOME COMPLIANCES CAN INDEED BE NEGLIGIBLE. (In fact, there have been whole, long-winded, FSAE/FS theses written on just that issue, applied to chassis torsional stiffness.)

It is part of the deep understanding that you might get much later that says that there are some things that you really DO NOT HAVE TO KNOW. (Here, you DO NOT have to know exactly how big the compliance is, just that it is "negligible".)

Z

[Claude Rouelle]

Christian,

Vehicle Dynamics and CAD are not necessarily the same coordinates

But OK now we have erased a communication issue. Good

[ChristianChalliner]

Noted on the axis system, I'll use the SAE definition from now on

Some very quick sketches, yes they are on CAD, why? because I could change the sketch faster than I could draw a new one by hand.

The crossed rectangles represent Springs/Dampers.

http://i1296.photobucket.com/albums/...ps87f4738c.png

Advantages - Simple (aside from creating the longer spring perch), CofG is lower than current arrangement, parts count is lower (compliance reduced as lower number of bearings?) design time potentially reduced as no time spent designing rockers and manufacturing.

Disadvantages - Motion ratios are fixed in design and will probably result in digressive springing (which is a topic for debate, does it matter, doesn't it? I'm not so sure). The new spring perch will be difficult to make and will potentially have to be rather large to prevent buckling.

http://i1296.photobucket.com/albums/...ps36ee34ae.png

Advantages - CofG is low down, could only be lower with a pull rod under the floor (The reason I've not done this is because I'm pretty sure the valving in the damper will be backwards (as in compression is a pull, rebound is a push) and I don't know if we will be able to get the shocks re-valved (yet!)). Other advantages are that the motion ratio can be altered and digressive, linear or rising rate springing can be achieved with the rocker design.

Disadvantages - As you can see, it makes life difficult with the footwell template, again more complexity, more nodes to reinforce, larger forces through the 'pull rod' due to the angle, etc. Probably more!

http://i1296.photobucket.com/albums/...ps7ac0d560.png

This is basically last years arrangement.

Advantages - Similar to above but lower pushrod forces compared to the pull rod.

Disadvantages - Highest CofG of all 3 systems, probably the highest overall weight, complexity,etc

There's probably a lot more I've not mentioned but those are which spring to mind straight away.

Anyway, onto answering some of Z's points:

I know the task isn't a simple one (since last years team failed to achieve it) if anything it's easily the hardest part. The problem is that the project has (and always will be) a third year project so it has to 'fit' with what the university describes as a project which it defines as a full year. So inevitably most students see it as "well I have the full year I can put it off for a while" which is clearly wrong! But you try telling that to someone in October that you've never met and see if they believe you...

We have competed every year (except 2013) yes we've never had any real success, our highest position was 29th in 2009. Yes there is absolutely no knowledge transfer what so ever but we are working on solving that (by keeping proper handover documents and keeping in contact with previous students). Yes for some reason there was an obsession with carbon in the past which resulted in many very heavy cars because of excessive use of billet in the monocoque's due to poor understanding of load paths, etc. With regards to Ackerman I already know Z, it's one of the reasons I want to put the tie rod pickups behind the axle centreline this year.

Looking forward to your thoughts on my sketches and I will have a think about why you suggested those tolerances...

[onemaniac]

It appears that the main point from everyone is to get a car running as soon as possible, which is great and would be ideal of course but I still have to do a base design which is at least competent before I can even get to that stage, I have taken some of Claude's advice with regards to simple sims rather than going straight to cad so I will get back once I've made some progress on those.

Optimum G claims that OptimumLap will simulate lap times with 90% or better accuracy. At first it sounded strange to me because it's a point mass sim, it doesn't consider all that complex suspension geometry, springs and dampers, load transfer etc etc.

BUT, given the fact that Optimum G has confirmed the accuracy through testing, we may say while the 10 input parameters stay the same, the worst design can still be 90% as fast as the best design (within 'acceptable' range I suppose, especially with compliance issue.) IF DRIVEN BY A COMPUTER.

I think the biggest loophole in FSAE is driver training and the second is the stability of the cars. If you look at competition results, the numbers tell us the fastest car around the skidpad barely maintains 1.4G of average lateral acceleration, with wings.
That's how far away FS drivers are from pushing the car to its limit.
The driver needs to be fixed through testing and stability needs to be sorted out through testing as well.

I'm not saying you should neglect details in design and just build something. but my point is that you should aim to spend the same amount of time testing the car and training the driver as you'd spend on design and build.

[ChristianChalliner]

Interesting thoughts onemaniac,

Your point of spending as much time testing as designing/building makes sense, I actually put forward some of the points raised to a few team members today and what I remember most was the chassis guys comment "Yeah I get that but I want to design it right to start with" so you can see the mentality we've had instilled in us here

I also updated the previous post, let me know if you can all actually see the sketches. I may have some more for you tomorrow from the chassis guy too.