Anti-Squat and Instant Center 1

[tlmartin84]

Hey guys, just found this forum and I am hoping I can find here what I haven't been able to in my normal "hobby car forums". As an engineer I tend to ask a lot of questions, and they are more geared to "just do this" because it works.

Anyhow I have started installing the rear suspension on my 56 F100 project, and have some antisquat questions. This truck is just for street driving, it is light maybe 3600 lbs, 4200 with passengers. It is light in the rear as well, with a 2:1 weight ratio front to back.

How does the instant centers location along the antisquat line affect the handling?

I would also like suggestions on engineering material for suspension setups, any good reads that explains this?

Here is how I have been modelling it up:

 

[GregLocock]

First books.

The two serious vehicle dynamics books most often used are Gillespie, and Milliken and Milliken, which is probably mode directly useful. Slightly simpler are Carroll Smith's X to win series.

So you've got 3 or 4 roughly parallel longitudinal links at the rear? That's a suspension i am very familiar with. As you lower the front end of the links the tractive forces are reacted into the sprung mass below its cg, and so the nose of the truck pitches up. This is subjectively annoying/amusing but not of any great concern, in moderation. BUT you are also dialling in roll steer as you play about with hp locations, and this can take the lateral handling of the truck from fairly awful, in its original state, to positively dangerous. Roll steer is entirely generated by the recession in jounce curve, that is the tendency of the wheel to move backwards in jounce. That is the wrong direction.

A moments thought reveals this is a direct function of your ic location in sideview, hence you've got a tradeoff.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[tlmartin84]

Yes this is a triangulated 4 - link.

I am more interested in handling characteristics than launching. What can I do to improve on this setup? Pull the IC down, or pull it down and back?

 

[tlmartin84]

Let me rephrase the last comment. I am more concerned with having a well mannered daily driver than a race car. I want something comfortable to "cruise" in.

 

[BrianPetersen]

If it were my vehicle ...

If you are more concerned about cornering and ride quality than launch from a start then the side-view instant-center matters less than other factors such as bump-steer, roll-center height, etc.

Bump-steer is connected to roll-steer. You can have roll understeer, or roll oversteer. Roll understeer (at the rear) works by making the wheel on the outside of the corner toe-in a little. Roll oversteer is vice-versa. A little roll understeer helps the rear of the vehicle follow the front on turn-in. With a beam rear axle, roll understeer requires the wheel (on either side) to pull forward with bump travel. That requires the outer control arm to have the chassis-end pivot lower than the axle-end pivot ... which isn't the usual design. Most diagonally-trianglated-4-link rear suspensions have roll oversteer.

Take a look at the way the leaf springs are angled on a good many newer light-duty trucks, and think about how that influences which way the axle "steers" when the outer wheel is in bump travel and the inner wheel is in rebound. Roll understeer ...

Getting the outer control arms to be angled-up towards the axle instead of the usual angled-down ... is an interesting challenge. But ... I chose the word "outer" intentionally. Maybe you need to make the outer (non-triangulated) control arms the upper arms, and the inner (diagonal) arms the lower arms! It isn't usually done that way ...

The other consequence of doing that would be to lower the front-view roll center. This is also not a terrible thing to do. I hate the ride quality of beam-axle suspensions with high roll centers. There's a peculiar side-kicking motion on every one-wheel bump. Lowering the roll center reduces the amount of that side-kicking action. You may need more roll stiffness ... that's what antiroll bars are for.

Consider the design of the S197 Mustang rear suspension carefully (as compared to the diagonal-4-link Fox suspension that preceded it). Panhard rod instead of diagonal links ... lowers the roll center to roughly axle-centerline height instead of top-of-diff height. Outer arms are close to horizontal at nominal ride height to minimise axle steer. Only needs one upper link. S197 ride and handling is quite a bit better than Fox-body, although the change to the rear suspension wasn't the only thing going on.

 

[GregLocock]

A symmetrical solid rear axle has no bump steer, unless you have a panhard, which is asymmetrical, but can be made to have negligible bump steer.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[tlmartin84]

Well, I picked up Herb Adams book and Engineer to Win by Smith. Still waiting on Milliken to arrive. The Smith book was of no help. I wonder if I need tune to win instead?

Herb Adams suggests a Satchell Link Suspension is the best overall 4 link....mainly because using the converging bars on the bottom of the axle lowers the roll center. I am not sure I understand why though. Can any of you explain this?

 

[BrianPetersen]

That's where I was going, too. Lower roll center - below diff instead of above. Easier to arrange for roll understeer by having the front outer links have the axle-end pivot higher than the chassis-end.

 

[BrianPetersen]

Lower roll center with a beam axle translates to less sideways-kick on bumps. Of course, it may lead to needing an antiroll bar, but that's easier to tune.

 

[tlmartin84]

In my design above, I have the lower bars with fronts higher than the rear. However, these lower bars run parallel to the frame, and the uppers converge.

Does converging the top bars vs the bottom bars affect roll center? If so, how?

I wish I could find a good diagram for a solid axle.

 

[tlmartin84]

Am I looking at this correctly?

Assuming the convergence point is the same in both drawings, and the arms are just flopped the roll center changes just a few inches.

In this one, the lower arms are parallel and the upper arms converge to a point in front of the axle. (this lowers the RC rather than if they converge behind it, CORRECT?)

In this scenario the upper arms are parallel and the lower arms converge to a point in front of the axle. (Satchell).

Am I mapping this correcly?

 

[BrianPetersen]

Most such designs have the convergence at the axle end rather than the chassis end. Both will work. One way keeps the roll center close to constant height above the ground no matter what the chassis does, and the other way keeps it close to the same position in the chassis no matter how the axle bounces around underneath. What's better is beyond my pay grade. One way feeds loads into the chassis near each frame rail. The other way has control arms ending in the middle of nowhere unless you happen to have a good strong crossmember there.

As for the axle steering ... In side view, figure out the X (longitudinal) position of the axle 50mm into bump and 50mm into rebound (body roll). Now draw the axle in top view taking that into account, and keeping in mind which side is doing what when turning left or right. Is the axle "steering" in the same direction as the front (roll understeer) or opposite (roll oversteer)?

 

[cibachrome]

Consider these design considerations. Maybe they don't apply to your truck, but they sure are very important in the Big PU Truck Picture:

1) Trucks usually experience very high load CHANGES in their normal use. Forget "Load Ratings" as a guide. Truck operators will put in the back bed any amount of crap that fits or doesn't eventually vibrate or fall out. Heck, even I can & will put 50 forty lb hay bales in the back of a 1/2 ton pickup, even using the tailgate flopped down to give me a decent overhang. Then take it all out for a local delivery. Customers who come and get it can & will pack even more (like 60).
2) Truck tires are usually aired up to a pressure well above 30 psi (as in 50 to 85, depending on 'rating'. This is done to achieve a legal tire load rating. AND to help get fuel economy driven low rolling resistance. Yet, their peak cornering stiffness occurs well below 35 psi. So, augmenting tire cornering stiffness with roll steer is handy, BECAUSE: the roll UNDERSTEER jacks up the tire's effective cornering stiffness as you add payload. This keeps the steering gain manageable as speed varies and frequency response / lateral acceleration response times fairly low with all payloads. You can add lateral force deflection understeer via plan view spring angles, but it's not load dependent. It's either there or not there. Adding load dependent roll steer is an attempt to supply the rear axle's sideslip gradient with a constant value for as long as possible as you corner more severely. We call this the Rear Cornering Compliance, in units of deg/g. Low is always good.
3) Most truck are equipped with trailer hitches, not only to protect owners from tailgaters, but because they pull trailers nicely. Trailers have hitches set to a prescribed height. The tongue loads on heavy trailers during cornering can and will be very high. You really don't want the lateral load component to increase the sprung mass roll moment because that's how crashes occur. Not accidents, but crashes from jack knifing. For this reason and a few others, rear suspensions of real trucks have rear roll centers well over 450 mm. Some above 600 mm (I have the stats).
4) Also, as an addenda to #3, the front & rear roll center locations amend the sprung masses roll inertia (parallel axis theorem, remember ?). As you add payload, the inertias (obviously) increase. If the roll axis is low, the roll natural frequency will also be lower than if the axis were very high. Besides the unpleasant wallowing roll feel, this low damped roll frequency can and will 'Q' up with yaw velocity's damped frequency and self excitation plus self preservation conditions are known to arise. Yes, you can add a rear roll bar, but then the extra stiffness is unsuitable for your unloaded condition.

So, for running around town on pass car tires at 28 psi, just put it together, but if want to stay in a normal pickup truck's safety zone, don't change the rear suspension far away from a modern truck's layout.

 

[GregLocock]

Point 3 is very interesting, I have never seen that discussed. We get high RCH at the rear as a fallout from using cheap suspensions!

Rear RCH seems slightly bogus to me, in general. We used to have a beam axle (RCH 400) and IRS (RCH 120) on the same car, with much the same front suspension, and on smooth roads I couldn't pick any difference in handling (not that I'm much good at subjective handling assessments). Front RCH on the other hand can be picked by a duffer like me, on centre response was 'crisper' with a high RCH (I don't think crisper is in the ride and handling dictionary).

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[cibachrome]

Chevy, Ford, Nissan, Ram, etc standard pickups are well over 450, Ford and GM 3/4 ton & above right up there at 600 & higher. Don't forget there's a dynamic component to this because the shock absorber damper thingys play into RCH movement.

 

[tlmartin84]

This is a pickup, but it is a 1956 F100. It will never see a load, other than passengers and a suitcase. I want it to handle more like a rally car than an actual truck. Otherwise I would have kept the leaf springs.

 

[GregLocock]

If you want it to handle better you might want to sort the weight distribution out first, at least get it to 60/40. And you'll need softer rates in all the springs.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?