Board logo

Chassis twist
Nick - 29/1/19 at 06:35 PM

Now got to the point in fully welding up the chassis. I've put extra temporary braces in diagonally and braced the chassis down full width in about 4 places.
Question is even if I weld in diagonally opposed areas constantly I know I will still get distortion. How much twist is considered acceptable?

Mr Whippy - 30/1/19 at 07:38 AM

Depends on what you want to do with the car, track use will need much better than something for the road I've drove some very wonky cars over the years (on the road) and tbh it was never really been much of a problem

You could do what I do with my models, aim for the very best tolerances you can and everything seems to work out just fine. Accept sloppy at the start and things can only get worse.

Nick - 30/1/19 at 08:24 AM

Good comment and good advice.

David Jenkins - 30/1/19 at 09:30 AM

When I was building my chassis, I ended up with a curve along its length that meant if you held one end down the other end was about 5mm off the ground. I mentioned it on this forum (ages ago!) and someone suggested taking a tape measure to a production car and measure distances to the ground at significant points - top of wheel arches each side, etc.

I felt a lot better after doing that!

nick205 - 30/1/19 at 10:14 AM

quote:
Originally posted by David Jenkins
When I was building my chassis, I ended up with a curve along its length that meant if you held one end down the other end was about 5mm off the ground. I mentioned it on this forum (ages ago!) and someone suggested taking a tape measure to a production car and measure distances to the ground at significant points - top of wheel arches each side, etc.

I felt a lot better after doing that!



I haven't welded my own chassis together, but I've measure a few monocoque tin tops in the way you describe. Many of them are not perfect! Quite a realisation as I assumed the component panels to be made accurately and the monocoques themselves are assembled by robots. Again I assumed things done that way must be done more accurately.

As advised above, aim for the best accuracy you can achieve from the start. If you start sloppy it's a downhill route from there!

FatChapChipChop - 30/1/19 at 10:25 AM

Jeremy Philips, when creating Striker Chassis for race use, would often us ratchet and straps to "straighten" a chassis to acceptable tolerances once fully welded up ..

jps - 30/1/19 at 11:31 AM

I've seen posts on here about people doing deflection testing and things to see how much flex there is in their chassis - so i'd assume a locost chassis is not 100% rigid in use anyway. I would think if you are that bothered about precise alignment then plan at this stage to have rose-joints on all suspension arms - that way you can set the alignement of the actual contact with the ground finely, even if the chassis itself is a little out.

[Edited on 30/1/19 by jps]

James - 30/1/19 at 05:33 PM

quote:
Originally posted by jps
I've seen posts on here about people doing deflection testing and things to see how much flex there is in their chassis - so i'd assume a locost chassis is not 100% rigid in use anyway. I would think if you are that bothered about precise alignment then plan at this stage to have rose-joints on all suspension arms - that way you can set the alignement of the actual contact with the ground finely, even if the chassis itself is a little out.

[Edited on 30/1/19 by jps]


From what I remember from the early days when a few people far more knowledgeable than me did Finite Element Analysis of the Locost chassis it's fairly poor.
A clever chappy called Cymtriks did a chassis plan that weighed less than the Ron Champion version and was twice(?) as stiff.

Worth looking up the plans as they were very simple mods and greatly helped some of the weaker areas like the engine bay.

Cheers,
James

Simon - 30/1/19 at 05:47 PM

At the end of the day, so long as all the suspension points are symmetrical it doesn't really matter (within reason) as the chassis is only there to hold the suspension in place and stop the components and occupants sitting on the floor

Charlie_Zetec - 30/1/19 at 06:26 PM

quote:
Originally posted by David Jenkins
When I was building my chassis, I ended up with a curve along its length that meant if you held one end down the other end was about 5mm off the ground. I mentioned it on this forum (ages ago!) and someone suggested taking a tape measure to a production car and measure distances to the ground at significant points - top of wheel arches each side, etc.

I felt a lot better after doing that!


Having rebuilt a Defender 90 from the ground up, I would definitely agree with this ^ statement!

Angel Acevedo - 31/1/19 at 01:54 AM

My chassis is about twisted about 5 mm.
But I welded Sispension Pivots level and square to each other.
I donŽt think IŽll notice when driving.

snapper - 31/1/19 at 06:01 AM

A full cage will stiffen the chassis
I like the idea about ratchet straps
I thought you had to clamp the chassis sections to a flat board as you welded it up?

JimSpencer - 31/1/19 at 08:07 AM

quote:
Originally posted by Mr Whippy
aim for the very best tolerances you can and everything seems to work out just fine. Accept sloppy at the start and things can only get worse.


quote:
Originally posted by jps
Plan at this stage to have rose-joints on all suspension arms - that way you can set the alignment of the actual contact with the ground finely, even if the chassis itself is a little out.


Hi, These two gentlemen have it spot on..

Trying to get a car 'straight' mechanically very rarely works, works ok for tin tops on the road that aren't going to be driven 'enthusiastically' hence why the standard upgrade is to retro-build adjustability in so you can adjust them straight..

Make it all adjustable, (Corner weights, caster, camber, toe etc) build it as true as you can and then set it all up once finished.

ianhurley20 - 31/1/19 at 09:07 AM

When British Leyland were making cars like the Mini etc +/- 1.5" was a pass at the factory. 5mm is nothing

b14wrc - 31/1/19 at 12:55 PM

Mine came out at 5mm difference too, between front and rear over the longest distance, so pretty decent.

Like said above, the adjustment in the suspension can dial out most of the error. I wouldn't be concerned.

jps - 31/1/19 at 01:35 PM

quote:
Originally posted by ianhurley20
When British Leyland were making cars like the Mini etc +/- 1.5" was a pass at the factory. 5mm is nothing




Having said that - the last car I saw driving down the road looking like it was a crab was a Mini....

jossey - 31/1/19 at 08:28 PM

I have a mondeo st220 brand new which I took for suspension change and setup and it was 48mm out from right to rear left from new. My westfield was no worse lol

coozer - 31/1/19 at 09:02 PM

Build the floor first, fully weld it up then manhandle it and straighten it.

Once its straight and flat weld the cage on top then all the other top bits

Nick - 16/2/19 at 08:52 AM

Well dudes, I've welded the whole chassis(apart from underneath), and with trepidation removed all my clamps, cross pieces, ratchet straps and more clamps.
And I have a deflection over the whole chassis from front to back and from left to right on the diagonal of about 2mm.
Happy days, I'm well chuffed with that.

Angel Acevedo - 16/2/19 at 12:32 PM

I reckon that is very good.
I think the chassis will twist a lot more than that under cornering forces.
Congrats.
AA

flak monkey - 21/2/19 at 01:48 PM

Common practice is to weld the whole chassis and then jig the suspension mounts on to it afterwards. That way if there is any twist in the chassis the mounts are still aligned as they should be.

To be honest, 5mm or so of twist is nothing on a Locost chassis and is perfectly acceptable.