83 replies to this topic

[NeilR]

Steel floor would be welded to the chassis and the body bonded to the chassis with polyurethane adhesive...much kinder on old glass than bolts

[Catalina Park]

For club registration the only part of the chassis that you really need to keep is the chassis number, it is pretty easy as long as you are a member of the right club. (you may need to start the right club, that's what I did!)

[NeilR]

That's what I'm given to understand. Curiously the Triumph chassis does not appear to have a number, the body does though.

[Kelpiecross]

You have mentioned  "old glass"  several times  - does fibreglass actually get old?   I wouldn't have thought it did to any real extent.   I have seen reports of boats made of FG from around 1950 that still haven't shown any signs of problems.     

[NeilR]

Resin in the 1960's was not as understood nor as well formulated as the current product and there is always the issue of how well the maker understood mixtures and processes. But part of the issue for all old glass is that polystyrene gains some of it's stiffness from the shrinkage that occurs during the cure - however things like ester loss and micro-cracking (resin fatigue) tend to weaken it, as well as degradation due to UV light. It gets less stiff and more prone for tears/fracture from a bump.

[carlt]

That's what I'm given to understand. Curiously the Triumph chassis does not appear to have a number, the body does though.

looks like you need an old scrapper for the body then ;-]

[NeilR]

I think so too. Any sort of rust bucket may do. I'll grandfather axe the chassis tot he same dimensions but with better metal and joints and use the id plates.

[NRoshier]

Holy thread revival batman!

Was it really that long ago I posted this?

Well time HAS moved on and my oldest son William has now started university! The car is officially 'his' now, though he has a 'proper' car for everyday use in a Fiesta XR4/ST150. He has stripped the car down to components and de-rusted, cleaned and painted. Meanwhile body repairs have taken place and we are learning more about the car. It is definitely a one-off special. The windscreen is steel and the outer skin has been unpicked from the donor and 'bonded' to the fibreglass body using the well utilised 'body filler and hope' method! I am in process of rectifying some of the construction and filling in the missing bits as well as repairing damage. In a curious twist of rules there is a pathway to improvement on the chassis front: local rules for 'historic' cars allow for 'period modifications' which can be reasonably proven for the period. So the plan is to make a backbone chassis out of the original as per the image below. This is copying the layout of the Spyder branded chassis made for Lotus Elan cars in the UK. As you can see we have made a model to test aspects of it - in fact we've made five different models. We need to improve the rear damper load pathway to the tunnel and a diagonal from the bottom corners upwards to the rear corss member seems to do the trick. Again this is also in the Spyder chassis. The roll hoop is there because it's a Triumph Herald with a swing axle rear suspension! It also gives a secure mounting point for seat belts/harnesses. The front/rear bulkheads are there for anti-wheel intrusion in case of a crash. The outrigger/lateral sections of the chassis that are used for body supports normally are going to be replaced with 75mm x 50mm x 1.6mm RHS steel and a 1.2mm steel steel floor added. There is not much in the way of any torsional advantage to this, but the old structure is C section steel and corroded. We cannot cross brace this area as the floor sections in the passenger seating areas need to be lowered to allow for added headroom. The exhaust will run along the LHS sill and not under the car. Hope this explains it. Finding old triumph parts is difficult - we need a late model differential and driveshafts. We also need to crack test a lot of things and shot preen other bits. The rear suspension is now a 'swing spring' - the best 'period' solution using OEM parts and disc brakes have been added to the front again OEM. The brake system will be updated to a split system, but a collapsible steering column is not possible, but desired.

 

https://i.ibb.co/k48...der-chassis.jpg
https://i.ibb.co/XVQ...epair-small.jpg
https://i.ibb.co/80q...0712-111953.jpg
https://i.ibb.co/VCv...ore-sanding.jpg
https://i.ibb.co/jHs...epair-small.jpg
https://i.ibb.co/6JP...-lake-small.jpg
https://i.ibb.co/C04...0627-152205.jpg

Edited by NRoshier, 28 July 2020 - 06:24.

[Greg Locock]

Your links are busted

[NRoshier]

Thanks Greg, I think I've fixed them.

Edited by NRoshier, 28 July 2020 - 06:31.

[Greg Locock]

Yup, all good. The Lotus backbone chassis wasn't especially stiff. For the intercooled Esprit I put a pair of  braces between the front of the chassis and the body. This let the body provide some torsional stiffness, enough that they could dial in a lot more understeer. Having said that anything is better than the Herald/Vitesse/Spitfire chassis, and I think boxing the spine in and your back end cage is a good move. I wonder why the Elan had such long rear springs?

[NRoshier]

It's a good question, the spring/damper mounts do hang out in the air.

The Spyder chassis is reported at being 150% as stiff as the original chassis, but who knows how or if this was actually tested and by who?! But supposedly 4000lb/ft per degree ... the balsa model is very significantly stiffer than the original. In fact it is startlingly stiffer and it's interesting how much missing a joint has an effect. I was concerned to address some of the weaknesses: The Herald has a raised stress area where the body/bulkhead ends and the chassis continues forward. This is just where the lower suspension arm bolts through - and this early chassis often cracks there. The rear diff mounts also often fail on this model of chassis. The Mk2 chassis had deeper main rails and internal sleeving at the front as well as a better diff mount/damper mount at the rear. Unfortunately I don't have one and they're very rare in Australia. If we can match the Elan for stiffness that is a huge improvement.

The fibreglass body has very little stiffness at the A-pillar so I cannot imagine any way to use it for torsional stiffness. Bolwells are gluing in the windscreens to try to improve this area, but the fibreglass A pillar is still a weak point. However I'm trying to improve what we have - this is never going to be something to chase

[Greg Locock]

"4000lb/ft per degree" That's enough. A full size sedan sees improvements up to about 10000 and it is a bit of a fetish number for some manufacturers. But the requirement drops off proportional to mass so for your lightweight I think that would be a great target.

 

Torsion is a weird case to design for, even working out the torsional stiffness of some sections is hard, never mind assemblies.

[NRoshier]

It will certainly be better than the original chassis by a long margin. I also wondered about cross linking the sills, from the outer bulkhead corner to the middle, then attaching prior to the diagonal from the ROPS tube on the other side. This way we can still lower the seat base, but it would have to be a modest tube. I have a spare model base so I'll give it a go without the tunnel added. There would have to be some connection between the two centre tubes, but if I join with sheet top/bottom then that should be enough. I'm following the British Steel design guide for the joint design, otherwise I feel I could lose a lot of potential performance through bad joint design.

I wonder if we'll really notice the difference with a stiff chassis given the rear suspension performance?

[gruntguru]

The rear suspension will be OK as long as you don't let it move.

[NRoshier]

LOL. The swing spring is a curious setup that's for sure.

[carlt]

All those models of Triumphs had collapsable steering columns. An internal column could slide inside the external tube column, with an adjustable pinch bolt.

https://www.google.c...=ALXFDT8MQ1a44M

 

If keeping the transverse rear spring, I would suggest using later Spitfire one (MK4?) as it is flatter (technically less spring camber) so inherently less spring jacking effect. Or purchase a proprietary handling kit from one of the UK Triumph racing specialist. The Herald is maybe the worst rear handling design ever !  A real contrast with the front which is fantastic.

[gruntguru]

Ouch - sounds like a recipe for oversteer - and probably oversteer of the suddenly variable intensity kind!

[Fat Boy]

Ouch - sounds like a recipe for oversteer - and probably oversteer of the suddenly variable intensity kind!

All the cool kids are doing it.

[carlt]

It was certainly an interesting 1st car to own at 17. You either learned great car control or crashed - nothing in between. Fortunately for me the learning process didn't involve hitting anything of importance. 

[mariner]

In terms of trying to get more torsional stifness, boxing in the central backbone should help things, you would have to feed the propshaft in but it turns two flimsy sections into a single much bigger one. Closing of the bottom of the Vette propshaft tunnel with a panel stitched on with bolts helped a lot t I understand 

[NRoshier]

We had plans to make progress, but the CV19 lockdown we are in has scuppered those.

So our planning has changed to what to show the the engineer for approval. The front/rear bulkheads are not intended to make the floor contribute - they are really there to be wheel-intrusion limiters.

We have Mk4 spitfire rear swing spring already in the garage.

 

 

 

https://i.ibb.co/80q...0712-111953.jpg

Edited by NRoshier, 23 August 2020 - 10:44.

[NRoshier]

Coming back to threads like this is worrying because I can see how so much time has passed by!

But I'd also forgotten how hard it is to post images!

OK, original chassis has this sort of torsion 'performance': https://lh3.googleus...6pjQ_s2Dg=w2400

 

It's not pretty, but that's the turd we have to polish. Our VASS engineer has requested a linear torsional performance - which is a challenge.

We have moved from the 1/10th scale model to mocking up full size to check fit with all of the components that will be used as I don't want something impossible to fit things into or service.

The mockup looks like this:

https://lh3.googleus...RzTFIgckw=w2400

 

The body has improved and now looks like this:

https://lh3.googleus...THIG9Pquw=w2400

https://photos.app.g...jXuFxdTa9VT1gMA

 

Chasing 3000nm/deg with a linear torsional performance - yes, I'm probably overbuilding it, for the simple reason I'm making it up as I go!

It has been suggested to use a steel steel tunnel - but how would you feed loads into it and stop it from buckling?

Edited by NRoshier, 08 March 2021 - 12:30.

[Greg Locock]

So the way my collaborator measured chassis stiffness was to rawlbolt vertical posts into the slab at 3 of  the spring towers, and then used bags of concrete on the 4th. He measured deflection at several points along each chassis rail. Obviously this is single wheel loading which is hard to refer back to a torsion stiffness, so I repeated the test in FEA and then could extract the 'pure' numbers. To get good correlation I had to model the posts as well as the chassis, which is quite common in my experience. OTOH there is an FSAE paper on measuring torsional stiffness directly. It's a bit of a science project.

 

You feed torsionals into a thin wall round tube via ring bulkheads.

 

Body is looking great.

[NRoshier]

Thanks, body is pretty - and not badly constructed for a one-off - but some of the past restoration leaves a bit to be desired - like 2" nails added to bodyfliier in seams. Still have not been able to track down the origin.

The masters thesis that provided the torsion pic used a similar method and measured stations along the chassis. Physical measurement was used to correlate FEA model.

I think the sheet tunnel will be easier to get wrong ... so may stick with the tubes.

[Lee Nicolle]

Pretty much anywhere a live rear is raced, they put a certain amount of camber and toe in the housing. It generally doesn't do the axles any favors, but it makes the stopwatch go in the right direction. You gotta do what you gotta do.

Bending diff tubes can be an advantage. Toeing in one car I raced stopped it eating the rear tyres. Though that was more a tyre issue.

Many GpC cars here had 'bent' diffs to get some camber. Some Nascar uses ball end axles on the outer end which makes life easier.

Not something to do though without full floating axles.

[Lee Nicolle]

All those models of Triumphs had collapsable steering columns. An internal column could slide inside the external tube column, with an adjustable pinch bolt.

https://www.google.c...=ALXFDT8MQ1a44M

 

If keeping the transverse rear spring, I would suggest using later Spitfire one (MK4?) as it is flatter (technically less spring camber) so inherently less spring jacking effect. Or purchase a proprietary handling kit from one of the UK Triumph racing specialist. The Herald is maybe the worst rear handling design ever !  A real contrast with the front which is fantastic.

Herald is bloody terrible. The front stubs seem to be used often for older Clubmans and they crack.

GT6s have to be run very low or they get camber tuck as well. All of those Triumphs suffered that.

[Greg Locock]

Falcon nearly got into production with toe and camber on the live axle. Everest does.

 

Yeah, I'd be very wary of using Herald/spitfire suspension parts with modern tires.

[mariner]

here is nothing intrinsically wrong with a Herald/vitesse front suspension and modern ROAD tyres as long as the weight is low. My Marcos used a Vitesse front end and Yokohama A008 tyres which gave amazing grip as the soft compound suited the low weight which was 825kg.

 

There are two problems though

 

- Most  cars aren't only 825kg.

 

- The modern replacement stub axles are of very variable quality. What might be OK for a Herald restoration only driven to shows isn't any good for Historic racing etc.

[carlt]

there are after market 'big bearing' stub axles available 

 

friend of mine runs cars in various classic motorsports for customers (Scimitars/Marcos/TVR/Lotus) using these components with ultra sticky slicks. 

 

https://iaindanielsc...t.wordpress.com

[Greg Locock]

Back in the day car suspensions were designed for 1g 2g 3g. This is fine for circuit racing on narrow tires even now, but is under strength for roads with potholes and kerbs. 

[carlt]

Back in the day car suspensions were designed for 1g 2g 3g. This is fine for circuit racing on narrow tires even now, but is under strength for roads with potholes and kerbs. 

 

are you suggesting that 'back in the day' they didn't have potholes and kerbs, or indeed a larger proportion of gravel surfaces ?

[Greg Locock]

No. They accepted suspension failures that are not acceptable now.

[NRoshier]

I'm acutely aware of the potential for axle breakages and things such as the wheel studs, - they're 3/8" in size, and of course they're 60 years old now.  So they will be replaced with 12mm versions, which means new wheels. Tyres will be of a modest size.