Your thoughts on roll oversteer please

[Ian Williams]

John, I believe that Bill Williams made his wide axle by cutting two axles and welding them together increasing the track by about 4", no doubt Stuart can confirm this. I have experimented with a Big seven axle (these are wider than std), with limited success, as Stuart said wheel patter was an issue and I experienced an increase in understeer, I think part of the problem is getting the Ackerman in the optimum position. Alan will hopefully be able to explain this phenomenon better than I, but in layman terms striking an imaginary line each side through the KP and track rod pivots they will converge at a point near the rear axle, altering the for and aft position of this convergence by altering the track rod arms will have an effect on the cars handling characteristics. There are other variables such a slip angle to consider and I doubt If Bill Williams understood much of this, perhaps its a fortunate coincidence that the car now owned by Stuart handles so well.

[Alan]

You can get better turn in by using greater than Ackerman angles on the outside wheel, but you need to work it out carefully because, of course, there is a downside and wheel patter on cornering is one of them. I do wonder about modifying Austin suspension all that much. All the racers here have fairly simple set ups and seem to go quite well. I think at first time spent sorting the engine and gearing would be more productive.

Bob, you are of course right, but we aren’t talking about a normal car. I’m prepared to be proved wrong, but my thinking is that on a seven you effectively have a 3point mounted suspension system. Two at the back and in the middle at the front. The front isn’t very stiff, a softer spring makes it even less so, and the rear becomes dominant. What you have effectively done in softening the front spring is soften an imaginary rear roll bar. Result- understeer., at least it is on mine. Might work differently with more spring camber though.

[Phil Kingdom]

Another point to remember is chassis flex, with a lightweight special body chassis stiffness could be greatly reduced as opposed to the original car. So although one setup may work on one car in may not work as well on another, not that chassis twist is always a bad thing in fact in the world of go-carts some racers have hard and soft chassis to suit the tracks they are going to, a hard setup and tight shocks with a bit of flex in the chassis/body may handle in a similar manner to a softer setup with very little flex. Unless your are building an exact copy of an existing car it is really a case of trial and error.

[DavidL]

Another point to remember is chassis flex, with a lightweight special body chassis stiffness could be greatly reduced as opposed to the original car. So although one setup may work on one car in may not work as well on another, not that chassis twist is always a bad thing in fact in the world of go-carts some racers have hard and soft chassis to suit the tracks they are going to, a hard setup and tight shocks with a bit of flex in the chassis/body may handle in a similar manner to a softer setup with very little flex. Unless your are building an exact copy of an existing car it is really a case of trial and error.

My car would patter going round the long right hander before going down the hill at Cadwell (Chris Curve before Gooseneck) which was quite scary.  My memory says the wheels looked like they were jumping 4ins off the ground in turn, but it was twenty years back.  

The car wasn't raced for long enough to develop it, it had (and still has) the original type damper at the front which was (is...) high up the "to do" list.  What I came to blame was just how floppy the chassis was as the body was (is) very light.  A friend built a heavier car and I can remember the flitch plates (not 100pct sure that's the right word for them, the bits below the bonnet from scuttle to rad shell) added a LOT of stiffness.  With an axle stand under the middle of my front spring I could bounce a wheel like a basket ball, his felt like a car...

I just looked at the http://www.oxfordshiresevens.co.uk/ site where there are body component diagrams.  There's a lot of structure my car lacks.  The chassis is boxed as are the cross members but no extra stiffness in the nose or spring mounts which I remember from the John Miles experiments "recently" make a tremendous difference.  I was always pondering some sort of structure hidden in the rad shell with "sturdy" bracing from the filler cap area back to the corners of the scuttle.

What I'm only now wondering is if I have short chassis steering arms on a long chassis.  From memory the king pin inclination was quite steep (that's how it came out with the drop links I had/have).  It used to be on 15in wheels so perhaps the offset came into the mix as well?   (Its now on 19in wheels).

As least at the back I've got "short" damper arms, beefier springs in the dampers, the leaf springs flat, stiff, and bound.  On the hills it used to be a progressive understeer, I think on the circuits as well unless really REALLY wet, which it mostly was those summers.

[Bob Culver]

It is getting off the immediate topic but Ackerman was done to death a few years ago. It was established that  the theoretical does not suit real cars which operate above walking speed.
The theoretical point about which  all wheels turn lies on the back axle axis extended.  I think this leads to confusion. The line thru king pins axis  (at road level) and steer joints usually converges notably ahead of the rear axle.
 
Maybe with a soft front spring and strong sticky friction s.as. the s.a tends to lift the inside wheel much as does  an anti roll bar, with similar effect. If one  wheel is effectively off the road all load is on the other wheel, whatever the springs.
  
Telescopic sas look awful but lever types reasonable and quite easy to incorporate at the front. Anyone running these? I found on my RP pitching was hugely reduced, although no other obvious effect..
Lever s.as. on the rear transformed handling but difficult to contrive a workmanlike mounting. Some books claim s.as. do not influence roll but I have fitted stronger units to several makes of car and all cornered more flat. On the RP the terror of snap action roll oversteer on metal roads was eliminated. (so few metal roads now few experience!)

[stuartu]

I believe that Bill Williams made his wide axle by cutting two axles and welding them together increasing the track by about 4", no doubt Stuart can confirm this.

This seems to be the case, Ian. The central 9" or so on the wide axle is filled with weld so the exact method is not apparent to the naked eye.
We have moved away a little from roll oversteer but, if I may be indulged, a couple more remarks on wheel patter.
When my father first raced the car with the wide axle it had the front shock absorbers shown in 1937 pictures of the car, basically a standard arrangement with lengthened arms. Despite careful balancing, wheel patter first time out at Silverstone was shocking. This was almost 50 years ago so Dougal wasn't in the paddock but (probably) Vintage Tyre Supplies were there and were most helpful, trying two further pairs of front tyres to no avail until the penny dropped and a brave man ensured the tyres were seating properly by inflating to some terrifying pressure for a short time. Father, ever cautious, quietly edged behind something substantial while that was going on. Anyway it worked; later we fitted separated s/as with the arm linked to the radius arm end as belt and braces.
I suspect I now understand why this was a particular problem with the Aerolite wheels. This should not bother anyone else but the curious who have supported my literary efforts may like to peruse p.110 of the opus to see if they agree.
It interests me to see that this car had separate front s/as linked to the radius arm ends in 1938, possibly to help control the same tendency. I have some photos showing this clearly which are not in the book but you can just see the arrangement in the photo on p.21. I think the separated shockers were transferred to the Jarvis monoposto in early '39, see p. 32.
Apologies for thread drift - my memory suddenly worked!
Regards,
Stuart

[John Bonnett]

Stuart, no apologies necessary; all very good information, thank you. So, continuing the thread drift......

Phil wrote that karts deliberately have flexible chassis but I don't think any car with suspension benefits from chassis flex. It is my thinking that for a suspension to be set up and to operate predictably, it must be mounted to a firm platform that holds everything in position.

I remember that Arthur Mallock used a crude but effective way of checking torsional deflection of a  chassis which involved a bar and a weight. I thought I might check a standard chassis and then experiment with boxing and further stiffening to see what benefits can be obtained. One obvious way to reduce twist is to build a spaceframe for the body but this will add weight and my aim is to build as light as possible. I very much like the way Ian has tackled the rear firewall which must add stiffness but still allows access. I would like the transmission tunnel to be removable so I doubt that will contribute much to stiffness. There's nothing in the way of bracing to prevent chassis lozenge so perhaps a stressed floor might help. JO 66 has an under tray so this might be a possibility. I'm not sure this is practical but I would like to make the underpinning body hoops in aluminium rather than steel tube to produce a light car. Extruded T or L section would work providing that once annealed I can shrink and stretch it to the desired shape. 

John

[AustinWood]

Wheel patter, which I know as axle tramp, is not confined to Austin Sevens. Large and heavy cars are very prone to it. A friend always drove Rover 14s and 16s which suffered from this. I had a Daimler Light Straight 8 which was also prone to it and W O Bentley while working for Rolls Royce testing the new 3-1/2 litre Bentleys had an accident caused by it. On these cars it happens at around 70mph while traveling in a straight line. At a certain frequency the front axle starts to oscillate with the front wheels jumping in the air alternately. It's very scary.
The cause is believed to be lack of rigidity in the front section of the chassis. On these cars there's a big section with no cross bracing because the space is taken by engine and gearbox.
It is aggravated if the shock absorbers are weak and Daimler adopted cross-braced chassis where the bracing extends forward alongside the engine. All these cars were fitted with harmonic front bumpers consisting of a flat spring steel bar across the car with lead weights at the ends. They were tuned to the frequency at which this happened. Akin to the dampers fitted to high-tension cables on pylons.

[Ian Williams]

John, the rebuild if my car will incorporate a full stressed under tray this is to A, add stiffness and B, hopefully reduce drag, you can see pictures of the chassis in its early stages in Cardiff Robs chassis boxing thread. The front firewall/scuttle area of a body can also be made to impart some stiffness to the chassis without necessarily adding significant weight, think in terns of boxes.
Stuart, in my opinion thread drift when relevant to the original subject ( and even occasionally not ) can sometimes provide the real gems of information, keep that memory working please!

[John Bonnett]

Ian that's absolutely brilliant; two jobs done with one part. Boxing of the longitudinal members and a platform for the stressed floor. Colin Chapman would love that. I myself feel encouraged because what you have done is in line with my thoughts so I'm hoping I'm on the right lines. By the way, how did you heat up the rivets? 

Should I try to build the bulkhead/footwell assembly as an aluminium monocoque rather than panelling a frame? And if taking the monocoque route,  glue it together or weld? I can't tell from your photo what approach you used but either way, it must give plenty of stiffness and in conjunction with the stressed floor a huge improvement over the standard set up.

We have difficulty in getting hold of NS3 now so I have little choice but to use 1050A for the formed panels and for lightness 18 gauge. However, for the floor and the bulkhead where only folding rather than forming is required I can use NS4, again in 18 gauge. 

Thank you very much for your considerable help.

John