Haynes Roadster – Rear Chassis

Now I have a flat table, progress on the chassis has been swift.

In fact I surprise even myself.

With only 4 hours work, Saturday and 7 hours Sunday the rear of the chassis has totally transformed.

Armed with a bunch of tubes I’d cut and filed during the week, the chassis now proudly boasts 31 new tubes.

I used 25 x 25 (2mm wall) box. This is slightly thicker and heaver than that of the book. A while back now, a friend did some Finite Element Analysis (FEA) to test chassis flex and distortion. Thicker tubes was not the answer; triangulation and more tubes were. Anything thicker than 2mm wall thickness, moved virtually all the stress to the welds. So, a 3mm wall thickness chassis is barely any stronger; if at all; than the 1.5mm, despite being twice the weight. Hence for this build I’ve added several extra tubes and moved many of the tubes, so they are no longer parallel.

By making the tubes non-parallel, huge amounts of strength are added via triangulation. My FEA mate did say by triangulating there was only a small advantage in going from 1.5 to 2mm, but it does give me some kind of dumb piece of mind. He even said that if he had more time he could design a chassis with 1mm that was even stronger, but I guess he had better things to do.

The picture above shows a roll bar that was bought from Fleebay for less than £10. There were no dimensions on the ad and the seller only assumed it was for ‘some kind of kit car’.

Well it was worth a shot!

It turned out that it probably was for a Haynes Roadster style car, but essentially this build has similar dimensions to that of a McSorley 442 and is therefore about 4 inches too narrow for this build.

If nothing else, when I’m standing and staring at the end of the evening, it does get the imagination flowing.

This picture also shows why I haven’t release any of the drawings yet. Why? Every single tube is accurate to the CAD in 3 dimensions all taken from a single reference point to within 1mm. Still not spotted the problem?

I’ll give you a clue, the problem is the Sierra Cosworth diff……

It might only be propped in place but to get it to this point was a very close call.

Have you ever stood back and had one of the ‘Oh Heck!’ moments? One that sends blood rushing to your brain and a results in panic?

I suddenly thought, ‘with all those tubes, how the heck am I going to get the diff into position!’

Basically, by pure luck, as you can see the diff does thread through; at weird angles; with less than 2mm to spare. – But I was worried though!

Don’t look too closely at those welds!

My excuse is I’m not using Argon gas for the MIG I’ve kindly borrowed. I’m using C02. Coupled to the fact it’s been 20 years since I’ve welded anything, it took a while to get all the settings right and back into the flow.

I had to saw through one of my welds (I welded one end of a tube back to front – Doh!), I have no worries about penetration (oh er missus!). It’s just a shame the best welds are the ones you’d never get to see.

A year back I bought a McSorley Roadster chassis for about £50 on Fleebay. As we picked it up to load it into the van, first the floor fell off, then one of the tubes. The seller had welded these parts on with over 30 welds. See here for a picture – Note the missing floor panel. Only 9 tubes of this chassis now remain, the rest is all new.

I don’t think gravity is going to be a problem for the welds on this chassis.

As this build is going to have 4×4 and potentially close to 300Bhp, there is a far bit more triangulation than the standard McSorley 442 chassis.