28 replies to this topic

[TimW]

Hello,

A friend of mine suggested I jump on here and ask some questions.

I am in the process of designing the suspension for a own built vehicle and I was hoping I could get some feedback.

I know its not going to be perfect, but I want to ensure I am not making any howling blunders in what I am doing.

Weight of the car is going to be between 350~400kg (I hope)

Mid mounted R1 engine with chain drive

 

 

For suspension.

Double a-arm suspension.

Currently mounts are parallel in side view, but not in top view (due to chassis design, clearances and aesthetics).

Custom uprights

 

Current thinking is:

RC Fr 30mm Rr 65mm

Caster Fr 5.5deg Rr 0

Scrub Fr 15mm, Rr 0mm

KPI Fr 7deg Rr 8.7deg

Camber gain 0.5deg per 25mm of bump travel

 

Weight distr is very approx 35/65.

 

No plans for roll bars at the moment, but if I did they would be approx 20% of vertical stiffness.

 

I was originally designing it so the RC distance to body stayed the same through bump, but camber gain was approx 0.9deg per 25mm. I thought that was excessive.

So now I have to revert to having some RC movement to get the camber gain where I think it should be.

Some feedback / advice would be great.

I am not expecting it to be per perfect, but make sure it is not a mess.

 

I am currently using an older version of Lotus Shark to assist with the number crunching (if it ever stops crashing on me).

[TimW]

Can share the summary file, not easy to put up here.

[Fat Boy]

Start with goals. Right now you've basically thrown poo on a wall to see what sticks. When evaluating FSAE projects, I ask them to explain how they arrived at the basics. Tires, engine, wheelbase, tracks, CG position & aero (if any) pretty much define the car. Once defined, you can work towards the middle.

[Greg Locock]

It is very hard to design a suspension if you don't know much about the tires. And you haven't even described the usage - is this an off road buggy? A hill climber? circuit racer? superkart ? (well probably not)

[TimW]

Sorry it was a bit late when I posted this, tried to upload pictures to give you a better idea. After much failure I gave up and hence left a lot of info out...

 

https://flic.kr/p/2iWkGwX

 

https://flic.kr/p/2iWgoVo

 

So basically, it is a 2 seater exo skelecton vehicle. Vast majority of use will be on the road with very limited track, maybe 5% if that (but who knows after I build it).

Tyre wise, 195/50 R15, something like a Toyo – Proxes R1R. A bit of fun really.

 

I want to car to be fun to drive, good agility. Due to it not having any electric devices I need to maintain some stability and predictability.

What other information would you like to see? Been doing it for so long kinda forget how to explain it.

 

Here is a link to summary PDF,

 

https://drive.google...iew?usp=sharing

 

I understand this question is very difficult and opens so many issues, even more questions etc.

Main thing is I would like someone to have a look and let me know if I am somewhat going in the right direction. Once again, that can be so subjective.

 

Please note, the damper positions haven't been resolved yet.

CG height I am working on, will definitely be low (a good engineering term.......) Doing that in CAD now.

No aero.

Edited by TimW, 02 May 2020 - 03:23.

[gruntguru]

65% is a lot of rear weight and I suspect it may turn out to be less than that with driver. Less still with passenger. It will be easier to tune at 60% or less. Have you looked at Malcolm Oastler's video in the other thread? You should find it reassuring.

 

Neil Roshier is building/has built something similar to yours although single seat and track focused. He is sure to have some tips.

 

I assume you have sorted the steering column/angles/etc?

[Ben]

Why those RC heights? Unless you know what TLLTD you need for handling then why arbitrarily fix them now? Which brings me to the WD. Why would you chose 35/65, what tyres do you have and why did they lead you to that WD?

 

Why no roll bars? Again seems strange to arbitrarily make that call without being clear on what TLLTD you need vs. the heave rates you need for ride. If they're not compatible you may need bars (at least one end) to hit both targets.

 

BTW if I was designing a completely from scratch vehicle I'd do a soft twist mode suspension - longitudinal Z-bars are probably the simplest way of packaging it. Then you avoid the high warp stiffness that stiff ARBs give you without needing to run soft in roll. Maybe a bit full on if you've already committed to a design though...

 

Impossible to have an opinion on the scrub, KPI and camber gain without tyre data. But I will say zero scrub radius at the rear seems strange. With compliance it's likely to go negative and on top of that you have pneumatic scrub - the tyre CP will move inboard under cornering load. So for zero net scrub radius at max gLat for example you're likely to need a non-zero value, and maybe more than 15mm at the front. Not to say you will, but you might, and it's worth thinking about

 

Ben

[NRoshier]

Ah, let me state that I plead almost total ignorance of suspension design. As a layperson and not an engineer, I have opinions on certain things, but very little practical/technical knowledge - I simply ask questions of people more clever than me. Tim is also familiar with the Social Climber. Also a pure track focus allows for certain liberties that I suspect you do not want to ignore in a road car.

Edited by NRoshier, 02 May 2020 - 13:46.

[Greg Locock]

Designing from a clean sheet of paper is good fun, but for a road car it is often a good idea to start with some sort of image car and then ask what you'd like to fix. The much derided Montego had the best target setting process I've ever come across - build an Audi A4 (I think it was) but for X% less.

 

This meant all of our targets were defined. Similarly for the Territory, after much hemming and hawing, they decided they liked the X5, so we had to beat it objectively and sort out a couple of its strange handling traits. Then they added a new process for steering targets at the last minute. Ha. And of course we couldn't use the BMW tires so the tire was being developed in parallel to the steering cal.

 

35/65 weight distribution is going to make getting your linear range  understeer budget, essential for a road car, rather difficult, unless you have big rear tires.

 

Here's some understeer budgets, note that they all rely on some weight distribution understeer. Try any of these links they might work

 

http://www.mediafire...mmary.xlsx/file

 

http://www.mediafire...               

 

http://www.mediafire...rf_summary.xlsx

Edited by Greg Locock, 03 May 2020 - 02:39.

[NRoshier]

The bundorf files were interesting to look at, but I have to admit that yet again I did not know all of the terminology used.

[Greg Locock]

Weight/tires steer/fy camber/fy SAT/ steer SAT camber roll steer roll camber RBAT CC US

Weight/tires - tire load sensitivity of the tires

steer/fy - compliance steer

camber/fy - camber thrust

SAT/ steer - self aligning torque

SAT camber 

roll steer

roll camber

RBAT rigid body effect of aligning torque. basically the SAT is applied to the whole car and tends to spin it round.

CC - cornering compliance

US

[NRoshier]

Thank you, that helps a lot ... now I know I'm probably not going to make sense of much of that and know what is important without a lot of work! Note I'm reflecting on my own lack of knowledge.

Edited by NRoshier, 07 May 2020 - 00:32.

[Greg Locock]

I don't think this matters as much for circuit cars, to be honest. And until you know your tires the whole thing is unreliable. Obviously we have to design cars that someone might put any old tire on, so can't rely too much on fine tuning around them. BMW used to, they've now done a 180 on that. I have worked on programs that were designed around the tire.

 

The problem is that measuring steer/fy is not impossible, but it isn't something you can do on a Saturday morning in your garage either. Hot tip if you do is to try with Fy at different locations on the contact patch, that way you can work out the SAT steer effect just using force,  typically we use 0 and 30mm, as that's the range of pneumatic trail we typically see in handling events.

[NRoshier]

How much would it vary between different tyres within the same tyre size: not putting 20" wheels on a Territory, but say changing from the design tyre to the top cost variant from the same manufacturer?

[Greg Locock]

It has always baffled me that bigname brand is happy to still use bigname on the side of the tire, when it is now produced on repurposed ancient machinery in a  third world country using a somewhat unknown compound. Inevitably some of these cheap old tires find their way back into first world markets and must cause significant loss of customer satisfaction.

 

The easy way for me to quantify is to compare a 75%ile tire against a 25%ile tire. Quite a sporty 75%ile tire with a 40 AR profile cornering stiffness @ rated load 2627 N/deg @rated load-1 kN 2381

 

25%ile, cornering stiffness @ rated load 2354 N/deg @rated load-1 kN 2132

 

So in this case the cornering stiffness is about 11% less for the bad tire, but the tire load sensitivity isn't much different. That's a statistical model, not an actual example from a manufacturers range. Because the TLS doesn't change much then there wouldn't be that much difference in understeer, in this case.

[SGM]

Although I am not an engineer it seems to me that you are on the right track. Like others have said when building a car you really start with the tyres you are going to use and work backwards from there. This will more or less dictate your ground clearance, and from there you can work out suspension travel and then RC etc. 

 

For a car that is going to be used 95% on the road I wouldn't get too caught up in all the minute detail of engineering design. Remember a Lotus Seven is a relatively simple 1950's design but that doesn't diminish the fact it is still a fun and fast road car. Don't try to push the design envelope, be conservative and use proven and simple designs and parameters and you will be fine. Build some adjustability into the car i.e. sway bars/ camber so that you can tune things to your liking. To me, the way your A arms are angled towards the chassis make for unnecessary complications, but if you are happy you can make them work, go for it.

 

Never forget that whatever you design has to be made. If you are building entirely yourself this is not such a big issue but if you are paying someone, make sure your design can be easily manufactured first. Sometimes small changes can make a huge difference in cost. 

 

Good luck with your project!

[Kelpiecross]

Sorry it was a bit late when I posted this, tried to upload pictures to give you a better idea. After much failure I gave up and hence left a lot of info out...

 

https://flic.kr/p/2iWkGwX

 

https://flic.kr/p/2iWgoVo

 

So basically, it is a 2 seater exo skelecton vehicle. Vast majority of use will be on the road with very limited track, maybe 5% if that (but who knows after I build it).

Tyre wise, 195/50 R15, something like a Toyo – Proxes R1R. A bit of fun really.

 

I want to car to be fun to drive, good agility. Due to it not having any electric devices I need to maintain some stability and predictability.

What other information would you like to see? Been doing it for so long kinda forget how to explain it.

 

Here is a link to summary PDF,

 

https://drive.google...iew?usp=sharing

 

I understand this question is very difficult and opens so many issues, even more questions etc.

Main thing is I would like someone to have a look and let me know if I am somewhat going in the right direction. Once again, that can be so subjective.

 

Please note, the damper positions haven't been resolved yet.

CG height I am working on, will definitely be low (a good engineering term.......) Doing that in CAD now.

No aero.

 

  Why does the proposed car have to look like an Ariel Atom.   The chassis on an Atom looks (to me) very odd indeed - more "multi-tubular"  rather than "space-frame".   Is there any actual engineering advantage  in having those  strange  curved  chassis members?   It  would seem to me that a more conventional space-frame   approach with straight  tubes etc.  would be simpler.   There is no question that the AA  performs well  -  maybe there is some point in the unusual (almost grotesque) chassis design?     

[kikiturbo2]

Ariel Atom's chassis is design first, engineering second.. Nice car tho..

[NotAPineapple]

While I don't thnk that the Ariel Atom is super optimised structurally, if you look at the output of a topological optimisation for stiffness the result is always a type of curved spaceframe shape.

Edited by NotAPineapple, 10 June 2020 - 06:32.

[Kelpiecross]

While I don't thnk that the Ariel Atom is super optimised structurally, if you look at the output of a topological optimisation for stiffness the result is always a type of curved spaceframe shape.

 

  "Curved spaceframe"  sounds like a self-contradiction  - can you give an example of a "curved spaceframe" that is better than a "straight spaceframe'? 

[Greg Locock]

Sydney Harbour Bridge may provide some inspiration. I don't know how much the Atom's chassis tubes weigh, but they are pretty small beer in the overall scheme of things, so a bit of inefficiency there is not a deal breaker.

[Kelpiecross]

Sydney Harbour Bridge may provide some inspiration. I don't know how much the Atom's chassis tubes weigh, but they are pretty small beer in the overall scheme of things, so a bit of inefficiency there is not a deal breaker.

 

 I see your point about the "curved" Sydney Harbour Bridge  - but - the truss is made up of many straight elements - none are curved themselves. 

[gruntguru]

Yeah, lots of straight elements combining to form an arch. The loads are all applied at joins.

 

Curves are great for carrying distributed loads - arches, pressure vessels etc.

[desmo]

Yeah, but does the traditional curved bend of a bicycle fork have any engineering basis?

[gruntguru]

Springy?

[desmo]

With the offset designed into the fork crown putting the force off the centerline of the tubes, straight tubes should be able to act as springs. Indeed, straight fork tubes have largely supplanted traditional curved ones, but I wonder how the two approaches differ in an engineering sense—if there actually are any significant differences.

[carlt]

The Atom design seems t me to be a very good compromise.

 

The chassis needs to be narrow in front to accommodate steering angles of front wheels.

 

wide enough centre for 2 passengers (modern 'wide' humans)

 

narrowing at rear for relatively equal front/rear track and wider? rear tyres.

 

That's the general shape, then ease/cost of production follows that shape as 2x curved tubes each side, tied with triangulations.

 

The Atom is 'only' a refinement of Derek Buckler's 1947 design which became the Buckler MK5 http://www.fordspeci...uk/buckler.html

[gruntguru]

With the offset designed into the fork crown putting the force off the centerline of the tubes, straight tubes should be able to act as springs. Indeed, straight fork tubes have largely supplanted traditional curved ones, but I wonder how the two approaches differ in an engineering sense—if there actually are any significant differences.

I got nuthin.

 

I know my CF bike frame has straight forks - bent at the crown as you say.

Edited by gruntguru, 17 June 2020 - 02:18.

[Fat Boy]

With the offset designed into the fork crown putting the force off the centerline of the tubes, straight tubes should be able to act as springs. Indeed, straight fork tubes have largely supplanted traditional curved ones, but I wonder how the two approaches differ in an engineering sense—if there actually are any significant differences.

You bet it's significant. It has a lot to do with the feel difference among American/English/Continental bikes of old. They're all special. I think you mostly see straight blade forks today because the designers can adjust the stiffness with the layup and it's a cleaner load path.