Nate Tempest

The beginning of that article is correct, but this part is flatly wrong (and pretty surprising coming from a spring company!):

A higher rate spring on the inside does not push up harder; in fact, it resists roll just like a higher rate spring on the outside does. The math is pretty simple: imagine you have a 700lb/in spring, and you go around a corner that reduces the load on that spring by 350lb. The spring will be 0.5" less compressed than it was at rest, so that side of the body will be able to rise by that amount (divided by motion ratio). Now say instead you have a 350lb/in spring on the inside, instead of 700. With the same corner and same amount of weight transfer, that spring extends by 1" instead of 1/2", and so contributes twice as much to body roll.

Dual rate setups can be used to provide a more compliant ride while minimizing roll, but it's because of the increase in spring rate on the outside that roll stiffness is improved (compared to just running softer springs), not because of the lower rate on the inside. Running stiffer springs all around without the tenders will increase roll stiffness more than dual rate (at the expense of less compliance over uneven or rough surfaces).

As for the compression of tender springs, imagine you have a 700lb/in spring sitting on a table, with a 150lb/in tender on top of it, and a 150lb weight on top of that. Both springs will be loaded to 150lb. The weight pushes on the top spring, and the top spring pushes down on the bottom spring the same amount. The bottom spring pushes on the table. And of course, the table pushes back on the bottom spring with an equal and opposite force and the bottom spring pushes back on the top spring. Newton's third law!

Nate Tempest

Thinking about this some more, there is a third way tenders could be used, which might be more beneficial, but also would be tricky to set up. As I mentioned before, two ways are

1. Tender is extended at static ride height, compresses to solid under cornering load. This gives compliant ride, and adds stiffness on the outside in a hard corner (compared to just using soft springs). Good for road cars, not great on track.

2. Tender is solid all the time, unless a wheel hits a dip, ie by cutting a corner—in these cases is extends, providing a bit of potential benefit with no real downside.

However, I was just thinking, there's another way to do a dual rate setup, which would be to have the tender be solid under static load, but set such that it just starts to extend during a max-G corner. So for example if and inside spring sees a minimum load of ~425lb when cornering hard, something like a 150lb/in spring with 3" of travel. At max G, that tender spring would extend by about 0.17", with the wheel drooping an additional ~0.3" (0.17 / motion ratio) as a result. So you would get that tiny bit of extra body roll. However, any dips OR small bumps—ie from riding a curb—would see the much lower spring rate, and so would be absorbed better.

So that could legitimately be useful. The downside is, it would be tricky to do unless your max cornering acceleration is very consistent, and you know exactly what the resulting weights on the unloaded springs are—not just estimates like I calculated in this thread. The reason is, if you get the spring slightly too soft (or short) for the amount of weight transfer, and it will be solid in the corner, as in scenario 2. Get it slightly too stiff (or long) and instead of adding a tiny amount of body roll, you'll add a fair bit. For example, if it turns out our inside spring actually sees 375lb in a corner rather than 425, now the wheel is drooping almost an extra inch, instead of just 0.3". So you could be in a situation where any time you adjust your sway bars, you'd ideally want to swap out your tender springs.

So, for a purpose built race car, where swapping out tender springs is no big deal, I could see that being a great way to go. For an autocross/hpde car though, scenario 2 is probably still the most reasonable. Still interesting though. If my car were set up for TT rather than autocross I'd consider it.

s2000Junky

All this is moot if your sway bar weight exceeds the ability for a given helper spring weight to do any work. From experience this is the case most of the time. The inside wheel just wont droop enough, but rather just pick up the inside wheel due to sway bar tying both wheels together. 150lb you mentioned was your rear sway and that is light, so your going to have more independent articulation then one of the typical oem rear bars, but how much really is the question on having a helper of any weight offer anything of added droop over none at all. Do a quick and dirty static test using the cars own weight and jack up the whole side of the car and see how much the side in the air droops, if the spring even unseats itself in current config without the helper, that will give you a good indication of what your max potential is in your most severe dynamic scenario and if your going to achieve anything other then as I mentioned before (keeping the spring seated if you unweight both wheels at the same time over a dip/bump) All this math is usually pointless, because its rare the basis of the formulations the person is calculating are even correct, or accounting for all real world variables, and even if correct then they don't always align as expected. At some point(usually sooner rather then later) just have to get out there on the road or garage and experiment. Thats how I operate, trial and error, experience and then I know and im usually better at anticipating without math what can be expected. It helps to understand what components do and how they influence one another. A educated guess is usually all thats needed to get you in the ballpark and can go from there.

BuggyofMildDiscomfort

This is generally how rally cars, safari cars and things like F1 heave springs are setup. It's how my safari buggy is setup as well.

I will say that going back a bit in the thread to the acceleration vs cornering droop - you can't really assume that you can have a max of 1.2g cornering force and 0.3g acceleration, so if you accelerate out of a corner you'll only see 0.9g cornering and hence the forces work out the same - almost all tyres generate more grip longitudinally than they do laterally. Tyre traction circles are generally elliptical, even for full on race slicks, because of sidewall distortion, wrinkling, etc, etc - you can see that on a Top Fuel drag tyre for example, which is an extreme example, those tyres can often generate 3 times the longitudinal traction as they can lateral grip. That also leads into other order effects such as longitudinal slip (wheelspin) giving much more gradual drop in friction than sideways slip that is often a much more abrupt loss of grip and sends you towards some friendly armco.
So you might find that at 0.3g of acceleration out of a corner you can still maintain 1g or more cornering load. That's why if you see a race car lifting an inside front wheel it's almost always just after the apex of a corner right as they get on the throttle. (Last decades F1 cars did a lot of it iif you want some easy examples, before the setups were limited away from it by changes to the steering regs/pickup points).
If we want even more fun tyre info - some tyres, particularly things with stiff tread and shoulders, like slicks and semi-track tyres, and especially rally tarmac tyres with their larger sidewalls - can actually generate a bit *more* lateral load when they have a little bit of acceleration or braking load on them compared to neutral. Tyre's are weird things.

That should also go some way to explaining why everyone is happy with setups that you would otherwise think have massive understeer - they are actually just balancing the front against the rear when you've got traction in play as well on a corner exit.