Vehicle Dynamics starting points and design process

[Z]

MCoach,

It is true that making adjustable parts can be quite easy, such as your example of stacked washers, perhaps at each end of a toe-link to adjust bump-steer.

But the point I think Christian was making, and that I strongly agree with, is that it is not the MAKING of adjustable parts that takes the most time, but the TESTING of them. Especially as the number of adjustments increase, and the combinations and permutations skyrocket. And then you get lost...

Funnily enough, I reckon it is teams in the middle of the ladder that probably need the most adjustable parts. Teams at the top have already figured-out most of this stuff, so they just make parts with settings they know will work.

But the goal of teams at the bottom of the ladder is to "build a car that can drive 30 km at an average speed of 50 kph". If they can manage this before comp, then there is most to be gained by simply getting the drivers familiar with the car (ie. lots of seat time), regardless of the car's details.

Then, if still some time left, small ergo fixes (move the shifter a bit, make throttle smoother...), multiple tyre pressures to test, multiple static-toe settings, brake-balance, spring-rates, ... and the platinum-plated Drexler ("You mean we've had it for five years, and no one has ever adjusted it!?", which means resetting all the spring-rates again, etc., etc... There are a lot of adjustments that can be made on a very simple car.

If the team is successful (ie. finish mid-field), then next year they can start with mega-testing of last year's car, but retro-fitted with multi-adjustable-everything. Which is sort of the end of Christian's quote above.

Z

(PS. Or they can say "Engine, frame, suspension all look OK, so time to dive into this aero thing!".)

[ChristianChalliner]

I'm short of time tonight but will get back tomorrow with more! In essence Z is spot on in terms of how he has interpreted my post, my feeling is that we are much more likely to get good returns from running many, many laps with fixed settings to get our drivers up to a reasonable pace both in understanding the car and the kind of courses they will be running on. As such, this is not the time to be making adjustments since it's 5-15 minutes per adjustment which is 5-15 minutes less running and practice for the driver! At this kind of level I'm firmly of the belief that 'improving the bracket which connects the wheel to the pedals' is the most important thing we can do.

Will get back with more tomorrow!

Christian

Edit: if the beams go well this year, next year I'm pushing for someone to do an un-sprung tray connecting the front and rear beams.

[ChristianChalliner]

Finally got some time to write something not in a rush!

To start here are a few more pictures as to how the design has progressed, there are some issues with it. For example, the front roll hoop definitely needs to be raised for the cockpit template so I am already aware of that. With the accurate masses of all known components plus the masses of designed components with applied materials in SolidWorks the overall weight at this stage is 149kg of course this is still missing a radiator, exhaust, intake, diff hanger, etc.







The mentioned issue about the steering wheel and front roll hoop is a definite concern and one that we've been working on. In essence it's driven by the desire to limit the number of tubes in the chassis, we didn't want to have to put in another roll hoop for the steering wheel since we already have the current one where the front shocks mount. It's possible to move it slightly rearward but then the shocks are inclined rearwards, I'm not sure how much of a problem this is, or if it is even a problem?

As Z pointed out the frame doesn't have to have high torsional stiffness since the beams are doing all of the work (aside from feeding in the arm loads and the SD forces) which is why we're not targeting high torsional stiffness but more an aim of reduced overall compliance. It was always our intention to have the floor as a 0.5mm sheet of steel welded in but we hadn't considered the reinforcement with spot welds so we will give that a try.

The steering is doing my head in, it's super super easy to run a setup which has reverse ackermann or even parallel steer but trying to run a setup with pro ackermann where there isn't interference with the beam, frame, shocks, basically anything.

There are carbon arms in those images currently but those are there because they are someones project, the first run of the car will be with steel arms and the car will probably run with steel arms at the competition but we can't stop people developing their own projects as they are what people are actually graded on rather than the car performance.

The pedal box and hand clutch I think needs reconsidering since the column does make it difficult to press the brake pedal in its current configuration and it's not really ideal to push the brake pedal further to the right.

At least the team is getting along better now and we're starting to get parts to actually work together now which was a big problem earlier on however, my frustrations with upper management are getting worse, basically because they seemingly have the whole process backwards, they're intent on having a chassis designed and built soon but I don't see the point because why on earth does someone want the part which everything bolts to designed first! Surely all the sub systems need to be designed first and then the chassis designed around those but that just falls on deaf ears.

In short, internal politics ruins everything, of course, these are only my opinions and don't represent those of the rest of the team/university, etc.

[mech5496]

Good job so far! I believe it would be useful to add actual masses to all components to keep track of your CoG position and yaw inertia at the design stage.

The pedal box and hand clutch I think needs reconsidering since the column does make it difficult to press the brake pedal in its current configuration and it's not really ideal to push the brake pedal further to the right.

RUN A HAND CLUTCH! It is easier to package, lighter, provides better ergonomics, requires less parts to be manufactured, it is way cheaper. The easiest way is to add a tube near FRH and bolt a motorbike clutch handle directly on it. It is really easy to add a simple mechanical linkage to actuate the shifter as well.

[Adam Farabaugh]

I think everyone seems to race to the goal of building the chassis, which is not a bad idea if you have everything designed before you start building it! You can look at some teams' facebook pages etc. and see that they are welding their chassis, which to me means that they must have already had their design freeze. Or are they really starting to weld without being 99% sure everything fits? I know I wouldn't want to angle-grind our frame apart...

[MCoach]

If you're starting off fresh, parallel steer will get you a long way in this competition.


Taking a second look at your chassis the proportions seem a little odd to me. Is your front roll hoop waaay out in front of your driver or is your wheelbase really long?
In regards to that, there is a maximum distance that you'll need to adhere to for the distance from the steering wheel to the front roll hoop.

If you don't have a model of a 95% hooman, I'd suggest making Percy a solid model with some width to represent a normal driver. That may help decide some more of the chassis width with regard to the shoulder, hip, and thigh area of the car. Other than that, model the templates, I'd hate to see you roll into tech checking to see if they fit for the first time.

[Z]

Christian,

... we didn't want to have to put in another roll hoop for the steering wheel since we already have the current one where the front shocks mount. It's possible to move it slightly rearward but then the shocks are inclined rearwards, I'm not sure how much of a problem this is, or if it is even a problem?

NO problem. In fact, it is an ADVANTAGE (ie. "Mass Distribution Rule 1. All major masses to be mounted low down and close to CG."). I reckon the centreline of FRH can easily be ~0.2 m behind the front-axle-line (or more). With the top-SD-mount coming a bit forward from the FRH, and the lower-SD-mount welded to the back of the beam, the SD's final slope in side-view is just right.

(Edit: Also, having the SD slope from-wheelprint-to-CG helps better react all the major wheelprint forces. See below.)
~o0o~

More importantly, think about the structural performance of your front-beam-axle and its locating links. Consider side-view when under HEAVY braking. The bottom links are under GREATER (X-axis) tension than the actual wheelprint Fxs, and the top links are under smallish compression. BUT, because the bottom links are at 45 degrees in plan-view, their tension loads go up another 40+%. Right now I would not get into that car if it had (student!!!) glued CF-tube links!

But even more concerning is that these large bottom-link loads are reacted at the centre of the beam. So, at the very least, you should design the beam to be adequately strong and stiff against this LARGE bending load.
~o0o~

I don't want to sound as if I am harping on about this, but in your current layout I see a much better fit for the "Model-T" style beam.

This could have its main "apex-BJ" about half-way between your FRH and MRH, fitted to a lateral floor cross-member there (which is braced with the floor-sheet). Your diagonal SIS tube(s) would now be wide "V"s (in side-view) that meet this floor cross-member. The floor from MRH to the apex-BJ would be horizontal, for lower CG of driver+car, then slope upwards to clear the beam.

Importantly here, the largest front wheelprint loads of outer-wheel-braking+cornering are now fed directly through the beam's diagonal torque-arms, into the apex-BJ, and to the most central, and hopefully strongest, part of the car.

Lateral location of this beam can be via a P&S as in the "Twin Beam-Wing" sketch. Or via something like your bottom-link-wishbone, but with a Ball-in-Tube joint at the axle. Or, better yet, reverse your bottom-link-wishbone, and have the B-T joint under the front-bulkhead (so the bottom-link-wishbone connects to beam out near the wheels).
~o0o~

The steering is doing my head in, ...

Keep at it. It WILL work!

BGB with crown-gear ABOVE pinion. Rearward facing Pitman-Arm (as in your image). Angle of PA-to-Toe-Link (ie. along their "centrelines") less than 90 degrees. FORWARD facing Steer-Arms. Angle of TL-to-SA (along centrelines) considerably less than 90 degrees (say <70?). PA a bit shorter than SA.

Moving the spring-damper rearwards, as above, will help with clearance issues.
~o0o~

...my frustrations with upper management are getting worse,
... whole process backwards, they're intent on having a chassis designed and built soon ...
... internal politics ruins everything ...

I suggest that whenever the above "chassis first" ideas are raised in meetings, you raise a small 4" angle-grinder, and say, with TERRIFIED look on your face and tone of voice taken from that "Jaws" movie,

"Weeee're gonna need a BIGGER ANGLE-GRINDER!!!"

Seriously, keep reminding them of all the EXTRA cutting-and-butting that will INEVITABLY have to be done if they follow that course. Even when things go really well, there is always some rework. But try to keep cool...

Z

[ChristianChalliner]

Mech - You don't need to convince me, I've been told "you go away, you come back with something that works and if we like it we'll do it" so that's easy enough with the bike hand clutch like you say.

Adam - Yup, well that's what happened last year, the amount of brackets and bars which were added in post paint was ridiculous.

MCoach - The front roll hoop really is that far forward, luckily we do have a person who fits the 95th pretty much identically so we're using him to make sure it seems reasonable along with the Percy template, it looks odd in the CAD but it actually works ok even if the steering wheel is right at the edge of the 250mm envelope.

Z - I see what you're saying and to be honest... I want to change it as for exactly what you're saying I realised I'd backed myself into a compromised design but I fought so hard to get the design accepted I fear that if I propose a new idea now I'll just be told to get lost. I'll work with the chassis guy though as an offshoot project to see if we can get a model T style beam in (I think it would work well as the anti-sub belts need a floor brace anyway so they can go in the same place as the beams ball joint). If I can get one in and with steering which works then I'll put the idea forward, my main concern is how I would make a model T style beam? It's easy to bend some tube to the shape I've got currently but it's a different story to start flattening tubes and still argue it's easy to fabricate! Also, I understand how the P&S works from a conceptual point as a lateral restraint but how does it account for the fore/aft movement of the beam whilst retaining pre-load as in your sketch?

Rearward facing pitman arm with forward upright pickups was the only way to get pro-ackermann steer when I was messing around with drawings, maybe I missed something at the time but all other configs ended up with parallel or reverse ackermann unless silly configurations were ran. Why do I want the crown gear above the pinion? Springloading gears together when the gears mesh in a conventional manner is pretty simple but I'm not sure how I'd achieve that through a 90 degree turn, unless springloading the crown gear against the casing would be enough? I don't know, I'm probably completely wrong!

Ahh if only that would scare them! I try and keep calm about it all but sometimes it's just impossible when people come out with the absurd, the idea that a bevel box and pitman arm is somehow more complex than a R&P is mystifying to me. But it is what it is and I just have to get on with it

Oh actually, just something that came into my head, I was considering purchasing an in-line bias adjuster to place in the line running to the rear brakes and removing the bias bar. My logic on this was that given that we have the same caliper size front and rear and the same master cylinder we are never going to run out of rear brakes. So reducing the pressure to them wouldn't be a problem and would simplify the pedal box, it would also mean that probably for the first time ever the bias adjustment would actually work! My only concerns are will the adjuster have enough range to be of use? I was considering that someone as a development of our car could place some in-line pressure gauges and find out what master cylinder size they actually wanted for the rear rather than just running the same as the front.

Christian

[MCoach]

One thing that the beam axle does lend itself to is making unsprung aero really easy to do compared to most other suspension types.
...You know, food for thought when you get to that point.

[ChristianChalliner]

Oh I know, I've already said that if time permits it or at the very least next year someone should do a tray connecting both beams together so we can have a full length un-sprung aero tray. You know what I was told? "Yeah, lets just get this working first"