43655
|
| posted on 11/9/13 at 07:38 AM |
|
|
Suspension Geometry and Fab
Hi all,
Am in the stage of designing the suspension for my scratch built chassis, and despite having read most of Stainforth's book I always found it
easier to understand by doing it myself.
so here's a Solidworks sketch of my proposed geometry. Width, and chassis mounts are fairly fixed. Wheels i'm undecided on yet, probably
7.5J front, 9J rear.
(pardon the mess, not sure how to hide sketch dimensions and never used blocks yet)
Wishbones will be rose joint chassis side, and spherical bearing on the upright side. (version 1 was maxi joint lower, drag link upper with PU
bushes)
Speaking of uprights, this geometry is based around the Wilwood ProSpindles I have, but really leaning towards the not being ideal. As to get a lower
scrub the hub face has to be at the very back of the hub pretty much. Would i be better off selling these and using a pair of bolt on hub assemblies
(which i already have, off an Octavia)? Can probably have them CNC machined ally or fab'd steel.
What else am i unsure of... ah roll centres, it's a mid engine car am i right to think the rear roll height should be a few inches higher than
the front? It will have IRS too, so just a matter of making the wishbones closer to parallel or slightly down to the upright?
I know i'll be fabricating the rear uprights around bolt-on hub assemblies. Unfortunately all my stuff is based around VW crap unlike the
predominant 80s ford stuff mostly favoured on here 
Look forward to any feedback, as this is really my first venture into car design and fabrication, and it's a hell of a learning curve 
Cheers
-Henk
|
|
|
|
|
hughpinder
|
| posted on 11/9/13 at 09:39 AM |
|
|
Hi,
You could reduce the scrub radius by getting wheels with more offset - ET35/38/40 wheels are pretty common and would reduce the scrub to 21/16/11
mm.
I have also gone with a rear roll height above the front - I can't remember by how much though, but I dont think it was as much as a few inches
(about 40mm I think, I'll repost if it was much different) - will look it up tonight.
Regards
Hugh
|
|
|
ettore bugatti
|
| posted on 11/9/13 at 07:25 PM |
|
|
Lotus Elise has roll centre heights of 30/75mm f/r
This is a quick tool to analyse suspension:
http://www.vsusp.com/
TVR Cerbera (and beyond) and Lotus Elise S2 have nice uprights with bolt-on hubs, But both have KPI of around 10deg and a offset (ET) of around
30-35mm.
So not really different to the WILWOOD spindles, if Im not mistaken.
Im not sure a fabricated upright woudl be that much better (geometry wise) and/ or cheaper.
|
|
|
43655
|
| posted on 11/9/13 at 08:37 PM |
|
|
finding it harder to find tasteful and wide-ish wheels in higher offset
After all i am only 21 so a little picky about wheels, childish as it sounds
Appreciate the note on higher roll centre, doesn't seem to hard to do then
will have a proper go at the suspension link there, it's too late to be converting my geometry setup into inches!
Doesn't really do much that solidworks and excel couldn't do mind.
Will have a look into them
As for the Wilwood spindles, they're a real nice piece of kit and all, but looking at a minimum of 100mm offset from the centreline, i think on
the hubs i made to suit it was an extra 50/60mm ontop of that.
http://www.wilwood.com/Images/Steering/Steering_Drawings-Large/spindle_830-9807-dwg-lg.jpg
In my geometry drawing the offset from the rear bearing was really quite small, so even then i'd have a huge chunk of spindle sticking out the
front of the wheel (see the dimension '99' middle left)
bolt-on hubs have an offset around 50mm from the mounting face, and i'm sure properly designed uprights could be made to give a good scrub.
Is about 10deg of KPI about what i should aim for then? I think the prospindles are 11, so close already
|
|
|
ettore bugatti
|
| posted on 11/9/13 at 09:39 PM |
|
|
Nothing wrong with a bit of dish (and stretch) 
Here is a thread on the Ferrari F355 suspension, food for thought (or confusion):
http://rejsa.nu/forum/viewtopic.php?t=60154
There's a preference tab in Vsusp, where you can go to the proper metrics.
|
|
|
snapper
|
| posted on 12/9/13 at 05:57 AM |
|
|
You can get 3 piece wheels made with any offset and many designs but I would not compramise all the work just to fit a set of non optimal wheels
Makes no sense
I eat to survive
I drink to forget
I breath to pi55 my ex wife off (and now my ex partner)
|
|
|
phelpsa
|
| posted on 13/9/13 at 09:10 AM |
|
|
Don't worry too much about scrub on a lightweight car, it's nice to have the steering feedback. Also remember that your pneumatic scrub if
running reasonably soft, high profile tyres could be a lot less than your mechanical scrub. I wouldn't be concerned with double the scrub radius
you have there, in fact most cars running cortina uprights have in excess of 100mm scrub radius.
I'd be looking to get rid of as much KPI as you can as it causes the wheel to 'gain' positive camber with steer, revert to castor to
gain self centering and steering effort.
With roll centre heights the more important factor is how they move compared to each other rather than their static height. You want to make sure
they're both moving the in same direction a comparable amount through your expected roll angle (3-4 degrees?) or you could end up with some
nasty mid corner balance changes. Try and make sure they don't move left or right too much, if anything moving towards the outside wheel a
little is preferable. Don't about having the front or rear higher or lower (it has some basis on production road cars with tuning ride
frequencies but you probably wont notice it), just make sure they're both reasonably low compared to your CoG (no more than say 20%) and you get
a good amount of camber compensation in roll.
I'd really recommend plugging it into vsusp and at least seeing how the roll centres move.
|
|
|
43655
|
| posted on 17/9/13 at 12:29 PM |
|
|
Aye, I would love to be able to run 17x8j & 10j Azev wheels, but I really don't think it'd be very sensible.
not a fan of stretched tyres though, get out
The Ferrari thread is mindblowing, so much to comprehend, or at least try.
Will get Vsusp filled in soon, this evening if i remember.
I see where you're coming from Snapper, but that doesn't sound like a very locost-y approach either! Don't have that kind of money
to spend I'm afraid.
But I'll probably be looking at Compomotive MO's or Team Dynamics which seem to be fairly popular and pretty sensible shaped. (I
do have a big soft spot for ET0 16x8 XXR wheels though!)
Great info Phelpsa, thanks. Have got scrub down to around 30mm with 5degrees of KPI (would be with custom uprights naturally). Will have a good look
at the behavior of the roll centres when i get the Vsusp analysis sorted out
|
|
|
43655
|
| posted on 28/9/13 at 07:42 AM |
|
|
http://tinyurl.com/q8ngzef
here's as far as i've got with vsusp, front is done off the cad pretty much, rear just been a play about so far.
what is the meaning of roll centre to roll centre location graph?
and what should it look like?
|
|
|
ettore bugatti
|
| posted on 30/9/13 at 06:20 PM |
|
|
Just some personal observations (feel free to disagree):
-Roll centre height looks a bit high to me, I would aim for something between 35-60mm.
-camber change; at 4 degrees roll, your outer wheel has almost 0 degree camber.
Which is quite impressive, I wouldn't be worried if I had 1.5 positive camber in that case.
-roll centre moves quite a lot in x-direction during motion.
-your virtual swing arm length is quite short.
-wouldn't a 17" tyre with the same width and circumfence be a cheaper, more common alternative?
This should be pretty close what the original Lotus Elan had:
http://tinyurl.com/n5e9qnk
|
|
|
43655
|
| posted on 7/10/13 at 05:55 PM |
|
|
just bringing the bottom wishbone mounts down 10mm puts the RC at 42mm.
That's a good thing, so great
Okay, what can i do about that then?
I'd like to run 17s for that reason, but they look a bit big on the car really
|
|
|
43655
|
| posted on 28/11/13 at 05:15 PM |
|
|
any comments on my current CAD model?
|
|
|
rdodger
|
| posted on 28/11/13 at 05:27 PM |
|
|
Only that recently I learned that spherical bearings should be pinned vertically not horizontally.
If pinned horizontally they wear prematurely.
Just be sure that there will be enough suspension travel before they bind. Longer wishbone may be necessary.
|
|
|
43655
|
| posted on 28/11/13 at 05:48 PM |
|
|
I did wonder about that. what makes them wear differently then? Surely you still get the exact same amount of surface area on the bearing?
Yeah that's a bit of a worry on the spherical bearings, the upright will have to be modified a bit. I'd rather use off the shelf sealed
ball joints but not finding much that's suitable. none are really suited to the double-wishbone's inherent awkward shock-on-the-LCA thing
- i.e every bump trying to pull the ball joint apart.
May look into a pull-rod arrangement instead perhaps.
Can't do longer wishbones due to the chassis already being made. Other option is running low offset wheels, which may not be an issue as I
can't decide on what i want anyway!
|
|
|
rdodger
|
| posted on 28/11/13 at 06:02 PM |
|
|
Firstly I am no suspension engineer. I got that from Andy at AB Performance.
I would imagine it's down to surface area. There is a larger part of the bearing used if pinned vertically.
Have a look at the facebook page.
https://www.facebook.com/abperformanceltd?fref=ts
Also Elite Motorsport
https://www.facebook.com/EliteMotorsportEngineering/photos_stream
Anything that looks as good as those two surely has to work!
|
|
|
phelpsa
|
| posted on 29/11/13 at 10:16 AM |
|
|
quote:
Originally posted by rdodger
Firstly I am no suspension engineer. I got that from Andy at AB Performance.
I would imagine it's down to surface area. There is a larger part of the bearing used if pinned vertically.
Have a look at the facebook page.
https://www.facebook.com/abperformanceltd?fref=ts
Also Elite Motorsport
https://www.facebook.com/EliteMotorsportEngineering/photos_stream
Anything that looks as good as those two surely has to work!
The problem isn't whether they are horizontally or vertically pinned, but whether they are in the same vector as the load (ie in line with the
wishbone member). If they are not, then you have two issues. One is that the thread is in shear/bending, and the other that you have an axial load on
the bearing in which your contact area is tiny (the perimeter of the edge of the bearing surface). Neither of these will be an issue if you spec a
large enough rod end, but you can use smaller ones if you use them more efficiently.
|
|
|
43655
|
| posted on 29/11/13 at 01:07 PM |
|
|
well with the rod ends perpendicular to the direction of the wishbone that shouldn't be a problem i hope?
Lowers are M14x1.5 i think, and the uppers are M12x1.5
Would you say what I'm proposing to use is suitable then?
I think i'll modify the uprights, raise the wishbone mounting points on the frame and then bolt down into the upright (as it should be really)
|
|
|
43655
|
| posted on 10/12/13 at 06:54 PM |
|
|
still struggling to design my uprights, not happy having the car suspended by the spherical bearing as that's much worse than having a rod end
'the wrong way around'.
Surprised no one is too bothered by that, seen it doe a fair bit now although with ball joints usually.
So still don't have a proper idea of how the front uprights will look. but got a wishbone design i'm fairly content with.
bought a pair of transit ball joints for £13 which look ideal.
so, thoughts?
ball joint goes in from below and held i with an o-right i believe. little plastic cap on top to keep the sh1t and water out
|
|
|
