Board logo

questions on design...
A1 - 11/12/08 at 09:52 AM

got a couple of questions about designing a car...cant decide where to put them?

basically what are the best steps to take after pencil diagrams?

Mr Whippy - 11/12/08 at 10:00 AM

if you have a chassis design drawn out I'd recommend you make a model out of balsa wood and make a replica about a foot long to see if you have the bracing tubes correct, assuming it a space frame or similar. Any weakness in the stiffness of it will be very apparent in your hands and best of all you can quickly correct any faults and test them for pennies. Something that is difficult to do even on a computer. You'll also pick up the correct order in which to assemble the real thing

balidey - 11/12/08 at 10:03 AM

You could post up your designs on here and then everyone can tell you what you've done wrong

mcerd1 - 11/12/08 at 10:08 AM

where abouts are you in scotland ?
(or does your user name give it away )


quote:
Originally posted by Mr Whippy
if you have a chassis design drawn out I'd recommend you make a model out of balsa wood

or a 3D CAD model for those of us who can

[Edited on 11/12/08 by mcerd1]

dubstar_04 - 11/12/08 at 10:12 AM

I would go for a cad model and then some CAE analysis work.

iank - 11/12/08 at 10:16 AM

quote:
Originally posted by mcerd1
where abouts are you in scotland ?
(or does your user name give it away )


quote:
Originally posted by Mr Whippy
if you have a chassis design drawn out I'd recommend you make a model out of balsa wood

or a 3D CAD model for those of us who can

[Edited on 11/12/08 by mcerd1]


Computer model takes 100 times longer and doesn't show you how stiff/floppy it is - it's truly amazing to find how changing one or two tubes can make such an enormous difference.

Computer model is the next step especially if you want to make more than one. If it's just for you, copy of the balsa model in steel and drive it

iank - 11/12/08 at 10:20 AM

quote:
Originally posted by dubstar_04
I would go for a cad model and then some CAE analysis work.


you need a lot of experience to get useful information out of CAE. Just putting in a model and pressing a button can give massively wrong answers. Best way to understand spaceframe stiffness is to twist a physical model in your hands. Once you've done that CAE can be made to give a 5 decimal point answer that isn't actually very useful for anything on a homebuilt car.

Problem with computers (and I work with them for a living so am quite biased towards them ) is it takes 100 times longer to do by the time you've futzed around getting everything just right, by which time you've lost interest and momentum.

Mr Whippy - 11/12/08 at 10:23 AM

a real model in your hand tells you far more and quicker than a computer model on a screen

mcerd1 - 11/12/08 at 10:32 AM

quote:
Originally posted by iank
Computer model takes 100 times longer and doesn't show you how stiff/floppy it is - it's truly amazing to find how changing one or two tubes can make such an enormous difference.

Computer model is the next step especially if you want to make more than one. If it's just for you, copy of the balsa model in steel and drive it

that depends on what CAD you've got and how good you are with it

CAD (as in computer aided drafting - eg autocad) is slow and won't tell you much
but a CADD package (computer aided design & drafting - something with static, dynamic and/or FEM analysis) could tell you allot more than the wooden model and a good one will let you model it quickly

if you can't get your hands on something like that (and the experience at using it) then a balsa model is probably much quicker and easier




when I get the time, I'm going to have a go at modeling up some versions of the book chassis using Tekla / Staad


[Edited on 11/12/08 by mcerd1]

richardlee237 - 11/12/08 at 10:37 AM

I 2nd Mr Whippy.

Anyone who has built model aircraft wings will understand the benefit of a model when determining structural stiffness.

I am quite proficient in Autocad but could build a locost model chassis faster then I could draw it and the drawing tells you nothing about torsional or longitudinal stiffness.

Richard

A1 - 11/12/08 at 11:08 AM

im in edinburgh...nah, thats my real name believe it or not!!
cheers, ill try doing a balsa model sometime soon!

mcerd1 - 11/12/08 at 11:23 AM

so your just up the road from me then

feel free to give me a shout if you want a hand with some cad stuff (once you've got a bit further with your design that is)

[Edited on 11/12/08 by mcerd1]

A1 - 11/12/08 at 11:29 AM

thanks very much! wherabouts are you?

mcerd1 - 11/12/08 at 11:33 AM

Haddington

there are a few of us in and near Edinburgh, we should probably organize some sort of meet again
(I was thinking about karting next year sometime ? )

A1 - 11/12/08 at 11:45 AM

sounds good! i was thinking of trying to organize a trackday at knockhill or something...

mcerd1 - 11/12/08 at 11:52 AM

sounds good too, but I've only got a boring car at the moment

(yes I know before anyone says it, get my finger out and build the thing.........)

smart51 - 11/12/08 at 12:17 PM

Are your pencil drawings sketches or technical drawings?

I'd take some measurements of:

1) me
2) the engine and transmission
3) the wheels, suspension and steering

then draw it out to scale to make sure it all fits.

A balsa model of the chassis is a good idea. I then made a seat and pedals out of plywood and foam to check that there was enough room. Still, I found that there wasn't space for the master cylinder below the bonnet. The more checking you do on paper, the less you have to redesign during the build.

iank - 11/12/08 at 12:47 PM

quote:
Originally posted by mcerd1
...
CAD (as in computer aided drafting - eg autocad) is slow and won't tell you much
but a CADD package (computer aided design & drafting - something with static, dynamic and/or FEM analysis) could tell you allot more than the wooden model and a good one will let you model it quickly
...


My points are
1. It takes an evening and about £5 of balsa and glue to get to a chassis design you've proved is stiff. You just can't do that with any cad package. Certainly worth doing just to have something sat on your desk for people to ask about.

2. Every 3d cad/modelling package I've used takes days/weeks to learn to use at any speed, and you still won't produce a full chassis in a couple of hours. I certainly wasn't thinking anyone would use autocad for the job.

3. Just because you've got a software package that can calculate stiffness it doesn't give you the experience to set it up and to be able to tell what the numbers it's giving you actually mean. All the professional engineers on here who've commented on them say it's not for amateurs - not saying that you couldn't use one successfully as I don't know how experienced you are, but someone asking what to use going from a pencil drawing has little chance of knowing what the figures mean and how to make them better.

Finally - and I know most people here don't care in the slightest - to get a £5000-20000 software package for most of us actually means getting a cracked version from the net for free which means you're breaking the law. At some point someone's going to get caught and paupered by a legal firm for that.

Mal - 11/12/08 at 01:12 PM

For a balsa model, make it at 1/8th scale.
This is convenient because 25mm square rhs can be modelled in 3mm square balsa.
I found it good for accessing torsional stiffness and for working out how joints will be made.

mcerd1 - 11/12/08 at 01:53 PM

quote:
Originally posted by iank
My points are.......


points taken, and basicaly I agree with you

I guess I just favour CADD as I'm crap at woodwork




I recon I could have a go at stiffnes and strength analysis using the structural design packages Ive got at work (but these arn't ideal as they don't really do dynamic loading )

so just to be clear to A1, when I'm offering help with CAD its really on the drafting / detailing side - with the aim being to produce part and assembly drawings like the ones in chris's 'new' book

A1 - 12/12/08 at 12:37 AM

my drawings are technical, only on a4 paper though but done to scale, one from front, back, side and top. not got any decent diags of the chassis yet, but ive done a few rough sketches...
dont have any measurements for engine/tranny etc, but working on it...