First all the usual caveats; I make no claim to be an expert, I’m not a formally trained engineer, I’m just someone who enjoys tinkering with mechanical things. The following notes on upright design can be commented on, criticised, used for inspiration or discarded out-of-hand. However I will take no responsibility for other people’s construction, use or fabrication of this or any other similar design. That said I hope there is some inspiration here for others facing the challenge of making or finding uprights with suitable geometry for their cars.

The story so far, I’m building an exo-skeleton type car, inspired by the likes of the Warner R4 and the Arial Atom. Like those and others before I’ve needed to make an early decision on what uprights to use because this dictates suspension geometry, this dictates mounting points, which dictates where chassis tubes need to go, basically I’m really designing the car front the 4 corners and working my way inwards. Decisions so far 15” by 7” wheels, wearing 50 profile tyres, with a dynamic rolling diameter of about 560mm.

I started off assuming I’d use something similar to a Haynes roadster of Locost build I.e. Cortina, Sierra or MX5 parts. The first two were discounted because of perceived short comings in their geometry, the MX5 looked favourable for a while and I even got myself a set to start with. However the more I tried to finalise the suspension design the more I started to resent the compromise these were imposing on me. Particularly I didn’t like the rack position or the fact the lower wishbone needed to be mounted higher that I would like. Nothing insurmountable but I just figured finding something else or making my own might be a better option, so I started researching the making option.

My initial idea was to fabricate something based around a tubular design, a little like the MK produced ones, I figured I could use a simple rear hub from a small fwd car, possibly even using the same stub axle, turned down to pass through the main tube, see pic below.

Whist I really like the simplicity of this design it does require several turned parts and I don’t have a lathe but that might change soon. Also care has to be taken welding or brazing the stub axle in otherwise you effect the hardening. If no suitable axle could be found an axle would need machining again not rocket science but ideally it should be made from something better than regular mild steel bar. Critics would also point-out that it also has the rose-joint bolts in single-shear however many have successfully made this design work. Anyway whilst researching hubs bearing etc. I came across this:

a rear stub axle from a Nissan Micra K12 (about 2003 onwards) There are several things I really like about this part; its short, compact, takes a equally compact bearing and most invitingly has symmetrical 4 by M10 bolt mountings on a flat square base crying out to be bolted to either a CNC machined alloy or folded steel type upright.

Whilst I, like many, really like trick CNC stuff and I have got access to CAD to pen a design I don’t have easy access to the machining and wanted something I could play with and tweak without wasting block after block of Alloy. So that eliminated one option, I could have chosen to fabricate something like the Haynes rears with several cut pieces of plate but I prefer the look of some of the neatly folded parts I’ve seen. However again the reality check was that I didn’t have direct access to trick accurate CNC folding equipment so experimentation would again be about paying for repeated parts to be made. Another option I’d seen was based on stock box-section sizes see pic below

There are several aspects about this design I like and whilst the person behind it cnc milled it all, it equally lends itself to home workshop methods of cutting and shaping. The conclusion I came to was that although I didn’t entirely like the finished design I though the construction technique could be developed further to give more the look of a folded part and if I could get a suitable sized box-section would be the perfect companion to the Micra stub axle.

Steel Selection. The micra stub-axle has a square base 68mm across and the PCD of the threaded holes is 65mm meaning it would bolt perfectly to a piece of 70mm wide box section, rather than mimic the other part I’d seen and go for rectangular section I decided square would be better because it would allow me more flexibility on KPI. I knew 70 by 70mm was widely available with a 3mm wall but hoped to find something a little more delicate, after much searching I sourced some 2mm ERW, If I wanted to go lighter I could probably find 1.5mm but I figured 2mm would be a good trial size to test things out.

Basis of the Design: The fundamental objectives of the geometry are minimal scrub, KPI under 8 degrees, caster of 6 degrees, Ackerman steering angles and a distance between top and bottom rose-joint centres of 250mm, time for some metal origami…. Not quite I’m still waiting on the elusive 70 x70 x 2mm box which should arrive next week, so for now I’ll share the cad pictures of what I intend doing. Like I say I fully expect to tweak things further as I go.

First cut triangles to allow the box section to fold


Weld and repeat in the other plane


Then use standard size section to add rose joint brackets and steering arm 50 by 25mm and 50 by 30mm pictured


Anyway more to follow once i've added a hub and brake rotor etc

This is really interesting stuff, thanks for sharing it and in such detail. Going to have another read through now

quote:

---

Originally posted by orton1966
Decisions so far 15” by 7” wheels, wearing 50 profile tyres, with a dynamic rolling diameter of about 560mm.

---

by 'rolling' i'd think he means rolling circumference?

interesting stuff indeed

quote:

---

Originally posted by blakep82
by 'rolling' i'd think he means rolling circumference?

---

Thanks for the post Orton, will follow with interest

Cheers

Fred W B

I'm really interested in the folded steel / micra options, seems like a nice solution. I'm used to looking at billet parts but a bit of folded steel seems easier and cheaper to say the least. the ones we work with are ~£100 a piece materials only. What's the cost of these approximately?

Have you done any FEA on the rigidity of it?

[Edited on 14/2/11 by coyoteboy]

quote:

---

Originally posted by interestedparty

quote:

---

Originally posted by blakep82
by 'rolling' i'd think he means rolling circumference?

---

He did say rolling, although he says diameter whereas the calculators normally give circumference. Easy enough to divide by pi though. Which is what I did with this calculator which gives rolling circumference

http://www.carbibles.com/tyre_bible_pg4.html

---

quote:

---

Originally posted by coyoteboy
I'm really interested in the folded steel / micra options, seems like a nice solution. I'm used to looking at billet parts but a bit of folded steel seems easier and cheaper to say the least. the ones we work with are ~£100 a piece materials only. What's the cost of these approximately?

Have you done any FEA on the rigidity of it?

[Edited on 14/2/11 by coyoteboy]

---

Orton,
That's excellent information.
Are you aware of these, which take your Mirca stub axle a stage further by including a hub and bearings.


VAUXHALL ASTRA MK4 98-04 FRONT WHEEL BEARING 4 HOLE HUB


If you do a search on Ebay you will find pictures of them. They would need a bolt ( a simplification) down the centre to do the job of the drive-shaft in holding it all together axially.

Mal

quote:

---

Not yet but once I have a finished model I'm happy to let someone have the cad work to give it a go. Although saying that I'm not really concerned about the rigidity, it's a 70mm cube with a pyramid on each end, it'll be stiff!

---

quote:

---

Originally posted by Mal
Are you aware of these, which take your Mirca stub axle a stage further by including a hub and bearings.


VAUXHALL ASTRA MK4 98-04 FRONT WHEEL BEARING 4 HOLE HUB

---

Interesting experiment.
I'm wondering about the structural integrity of the gizmo though.
It seems likely that stub axle will bend the simple sheet mount plate.. pretty well on the first bump, or worse, corner.
A wee fraction of a mm deformation will result in significant (frightening ?) shifts in the susp geometry.
IF mine i would through bolt the thing and use welded in bushes between the two upright mounting faces.
But then I would prefer to use a shaft type axle, through bolted with a tube bush welded in between the two upright faces.
I'm thinking that stub axle could be better used in a Utility trailer.

quote:

---

Originally posted by Bare
Interesting experiment.
I'm wondering about the structural integrity of the gizmo though.
It seems likely that stub axle will bend the simple sheet mount plate.. pretty well on the first bump, or worse, corner.
A wee fraction of a mm deformation will result in significant (frightening ?) shifts in the susp geometry.
IF mine i would through bolt the thing and use welded in bushes between the two upright mounting faces.
But then I would prefer to use a shaft type axle, through bolted with a tube bush welded in between the two upright faces.
I'm thinking that stub axle could be better used in a Utility trailer.

---

Sorry for the late post I've been rebuilding gearboxes Again!

This photo shows my first attempt

high kpi side
You can see that I've use 50mm box section and bolted a stub axle through the upright. The thinking here is to strengthen the upright a bit like bamboo. Also this axle is much lighter than the bolt on plate pattern.

high kpi assembly
here you can see the holes and stub axle. I used a hole saw for the larger hole with a bench drill braced to make sure it was exactly square. The stub axles are standard cortina bolt in items used by grasstrack racers with a longer bolt and a spacer tube so that it bolts up hard to both faces of the box section.

low kpi rear
This version is when I decided to go for a low kpi angle and wanted to see the clearances. You can see the bosses welded on for the calipers on both sides. The other side will have the steering arm bolted on. Balljoint positions were not drilled as I had not decided on them at that time.
Currently as I am going for a high castor angle the trail would be too high so the next version will incorporate axle lead, a bit like in a bicycle to reduce the trail to around 15mm.

upright front
This is what it looks like from the front and yes the discs are 280mm diameter and they do fit inside 13" rims and yes it was bl00dy difficult!

Before anyone asks no I have not done a FEA just manual calculations (yup shock horror using a calculator, pen and paper!) principally 'cos my autocad does not support FEA and I can't find a more up to date version
Cheers!

Hi David

I'm always hesitant to make posts like this, but when I have a genuine concern I feel I must.

Are you very sure you have good penetration in the weld at the mid point of the upright? To me the weld profile looks to sit up too much, and looks like you could have used more heat input. Did you have a bevel prep on the butting edges?

In such a critical component this weld has to be perfect.

Cheers

Fred W B



[Edited on 28/2/11 by Fred W B]

No probs Fred I appreciate your concern. I think the penetration was OK when I did it I'll check the reverse side again tomorrow to double check.
Cheers!

Hi chaps,

I decided to test the upright until it failed. See the separate post today if you want to find out how many g it took! - I'd put a link if I knew how>

Cheers

I know I haven't posted much since my initial cad pictures but I now have steal and the other parts so Intend making a start and will be posting some pic' next week. in the meantime I’ve been playing with geometry on cad and I am confident I can get KPi as low as 6 degrees (possibly less) and still have zero scrub from a regular 37mm offset wheel. For comparison the MX5 needs 11 degrees to get the same centre point steering

Are you sure you want zero scrub?

I'm looking to copy the Elise numbers as I don't think anyone will argue that they don't handle/work well.

Previous thread on the subject here.

Good link thanks. But there were loads of arguments both ways there it was far from black and white.

I didn't read zero scrub being a criteria from it in fact I thought it was the opposite but arguable. Quite a few good cars were discussed there and they were all quite different presumably for different reasons

Cheers

quote:

---

Originally posted by v8kid
Are you sure you want zero scrub?

---

That is a truly excellent idea!
So what are the variables we want to change? For me it would be scrub, trail,kpi, caster and camber plus of course the vertical height of the balljoint positions wrt the spindle.
I found it incredably difficult to package my requirements inside a wheel and ended uf getting split rim wheels so I could bury the stuff deep enough to get a low kpi.
I suspect it is this practical aspect of packaging which determines the geometry in production cars. As locosters we can do things manufacturers cannot even that superp car the elise.
For example my split rim wheels were refurbished and assembled from bits over 3 eBay purchaces and the tyres cost 99p!
Manufacturers would be prevented from doing this due to cost and ruggedness and I feel we should leverage advantages like these.

Don't get me wrong I have no illusions I can get anywhere near the superb elise in any respect but I can have some fun and learn something at the same time.

Cheers! and congratulations on the design work