Anti-Squat Physics

[63ImportBeater]

Could someone please explain the physics of ANTI-SQUAT? I am a Mechanical Engineering major so I can appreciate a lot of very technical information, so throw it at me. I am interested in designing my own 3 or 4 link for my 63 nova and need to know the physics of 'Anti-Squat' so I can decide how to best use it in my application.

[MrQuick]

anti squat is a term used with rear wheel drive cars and is the amount of compression that is applied to the drive wheels under acceleration.a suspension with 100% anti squat will exibit no compression of the rear suspension during acceleration. the chassis and suspension links take 100% of the load.
A car with 50% anti squat will plant the tires harder than a car with 100% anti squat.wait... a car with 100% anti squat will squat under acceleration so the rear of the car will drop. I believe this squat is action that goes into the suspension movement instead of tire traction. I think I just confused myself? were's my books? :rtfm:
(EDIT): more info not from memory...
Heres a section from Herb Adams. Memory is bad so books are good...

Anti-squat is not difficult to calculate, however certain values on the vehicle must be known, such as the center of gravity between the front and rear of the car, and the height of the center of gravity . With this information, a scaled drawing of the tires and this information can be made. Once a drawing is made, then the rods of the 4 bar system can be plotted onto the drawing. According to Adams, “The instant center is that point around which the linkage can be assumed to react” . To determine the instant center, the paths of the 4 bar rods are extended forward to find their intersection point. This is the point around which the rear axle would rotate if the 4 bar connected to it.


The anti-squat value is based on where the instant center falls. Using the same drawing as before, a line is drawn from where the rear tire touches the ground; to a point at which a vertical line is drawn through the front axle and the center of gravity meet. If the instant center is on this line, then there is 100% anti-squat in the rear suspension. Remember from earlier, more or less than 100% and the vehicle will rise or squat, respectively. This amount of rise or squat is relative to the instant center’s height in relation to the “100% line”. hope this helps and hope someone eles jumps in here.

[Norm Peterson]

Figures 17.13, 17.14, and 17.15 in Race Car Vehicle Dynamics are sketches showing all of the "antis". The percentage can be graphically determining pretty much as Vince has already described. But if you're more comfortable with just numbers, "anti" percentages represent the relation between longitudinal load transfer and vertical loading through the suspension linkage (rather than through the springs/shocks).

Longitudinal load transfer is a function of h (the vehicle CG height) and L (wheelbase). Graphically, that's the intersection of the "vertical line is drawn through the front axle and a horizontal line drawn through the center of gravity". Or, h/L, and this represents a vertical:longitudinal force ratio.

Load carried through the suspension linkage is based on the SVIC* location. Knowing the vertical and longitudinal coordinates of the link pivots means that equations for the lines between them can be written, and their virtual intersection (the SVIC) determined. The suspension (in side view) can now be considered as a virtual swing arm pivoted at the SVIC. Since there is (obviously) zero moment about the SVIC, the relation between the vertical and longitudinal force components depends on the vertical and longitudinal offsets between the IC and the point of application of the basic fore/aft force**. Using the xyz = longitudinal/lateral/vertical convention with the origin at the contact patch directly under the axle (for a stick axle), the vertical:longitudinal force ratio for the suspension linkage becomes [z(SVIC)]/([x(SVIC)]. RCVD discusses this in terms of the tangent of angle theta, for reasons related to the ** note.

Anyway, since all of the longitudinal acceleration load is resisted by the longitudinal linkage load, the anti-squat percentage ultimately boils down to being the ratio of vertical force through the linkage to the longitudinal load transfer. Or the ratio of the tangents of the two angles.

* Side View Instant Center

** this is at ground level for a stick axle, but at axle height for an independently sprung drive "axle". Specifically, it's the point where the traction force only is applied (zero moment).


Using it best in your application kind of depends just what that application is. Understand that it's entirely possible to have too much A/S, and that one downside is the possibility of brake hop occurring, or occurring too early. If you're only going to drag race and drive no harder than moderately on the street, that probably isn't going to be an issue, and 100% or slightly more A/S may be the ticket. But for any driving that includes hard braking it isn't. Compromises to tame brake hop are to either lower the A/S% and ignore or drive around a somewhat poorer launch or reduce the rear braking proportion (which increases the front braking effort as well as stopping distances relative to what they'd be with ideal brake bias and no hop). A/S% figures closer to 50% are generally more road-race friendly, and I've heard of lower values being used in a very specific non-streetable application for a well thought-out reason (and with more than a little success).

There's more, actually, as the longitudinal location of the SVIC has its own influence even with the A/S% held constant. Where do you want (need) to apply that vertical force?

Don't get so carried away with A/S that other aspects such as axle rollsteer get compromised too much for your application. Playing games with the inclination of the lowers will affect that, or force you to raise the rear RC (which throws you into another bunch of iterations).

Suspension engineering is just that - finding the right balance for a given situation, which involves some compromises. It's why you won't find any "fits all" solutions but will see lots of discussion.

Late edit for clarity. Thanks, wendell.

Norm

[wendell]

It's a lot easier to explain when you can use pencils for links and beer cans for tires but here goes...

Newton said, that drinking enought mercury will cure anything. He also said that for every action there will be an equal and opposite reaction. We should focus on the later.

When the average car takes off it transfers weight rearward (because objests at rest try to stay at rest). The rear weight transfer causes the back of the car to squat DOWN. The equal and oposite reaction to that is tire force pointing UP. That's bad.

A properly designed suspension can reverse these properties. As weight is transfered back, tire force is pointed DOWN, loading the tires with the newly transfered weight. 100% anti squat is neutral. The weight transfered to the back of the car is the same as the force on the tires. Visually the back of the car stays level. At a value greater than 100% the force on the tires is greater than the weight transfer and the back of the car can be seen lifting up. At less than 100% the opposite is seen.

Anti squat is calculated by extending a line from the rear tire contact patch forward through the rear IC untill it intersects the vertical plane of the front axle. The ratio of the height of this intesection divided by the height of the COG X 100% is the % anti squat. Example: COG is 17 inches (a fairly well developed 1st gen) and the line from the contact patch intersects the front axle line at 15 inches, you have 88% anti squat.

That's the best I can do for you. I'm going to put on my helmet and mouth guard and get ready. The best thing you can do is get some books. The herb adams book is good, so is Paved Track Stock Car Technology.

Or just read the above post!

[parsonsj]

Anti-squat is a term that indicates how the vehicle reacts under straight-ahead acceleration. The value (say 50%) is the amount of the force that isn't translated into body squat.

It's kind of a backward term: you could just as easily call it "body lower" or some such and turn the value around.

The calculation itself is based on the lever length of the control arms divided by the lever length of the force generated by the tires.

One other thing to consider is that under vehicle braking, the forces reverse. In this case too much anti-squat is harmful: it will cause wheel hop. That's why 50% anti-squat is a good starting place for road racing and street cars: it's a good compromise between forward acceleration and braking (from a physics point of view: positive and negative acceleration). Finally, since most of the vehicle's braking is done by the front brakes, many racers find some success by increasing the anti-squat beyond 50% to give their cars more forward bite and giving up some rear braking effectiveness.

jp

[nutley]

Hey guys i'm a 4x4 guy, and my cousin that I live with wants me to delvelop and suspension for his 68 camero. I thought I would share with you guys what we have developed in the 4x4 world for link calculations for suspention geometry. What are your recomendations for anit squat values and roll steer? 50% anti squat seams to be a recuring number in this forum for what little I have read. Should be an easy value to hit with such a low car

here is the link to the forum to the link calculator that has been developed.

http://www.pirate4x4.com/forum/showt...ink+calculator

and here is the newest version of the calculator itself.

http://mysite.verizon.net/triaged/fi...arLinkV3.0.zip

Here is a good explaination of anti-squat and diagrams

"When a vehicle accelerates, the weight shifts rearward. The higher the weight is above the ground, the greater the shift off the front and onto the rear. This extra weight will compress the rear suspension, make it SQUAT. But there is also a torque on the rear axle, equal and opposite of the tire torque. The pinion will try to rotate upward. The links resist this torque by putting forces into the frame. These forces may try to lift upward. The upward force will cancel some of the squat from the weight shift, ANTI-SQUAT. The easiest suspension to understand is a ladder bar/radius arm setup. if the axle trys to rotate backwards, the end of the arm will push upwards. The shorter the arm, the more upward force it will create. The greater the force, the more anti-squat it has. For understanding anti-squat, a 4 link can be visualized as a ladder bar, wherever the links would cross is the theoretical length of the arm. This picture is a ladder arm and 4 link with the same anti-squat geometry



For more antisquat, the 4 link should have more angle, making the intersection shorter, for less anti-squat the links should have less angle. The height of the intersection is also important, the higher it is the more antisquat it has. Another pic, the red setup has more than 100% anti-squat, the blue has less. The drawing has a horizontal line through the CG, and a vertical line through the front wheel. Then a line is drawn from the rear tire contact patch through the 4 link intersection, till it crosses the front axle. In the drawing the CG happens to be 38" the red suspension crosses the front axle at 48" 48/38 = 1.26. Thats 126% anti-squat."



It should also be noted that the COG measurment is of the unsprung weight

[BRIAN]

Nut, What about a parallel 4 bar set up where the bars do not crate an intersect point?

[nutley]

That is a good question. It largely depends on the angles of the links not just the intersection. If the links are parallel as well as parallel with the ground, the antisquat at that instant would be 0. This is bad because when you get on it the suspension will compress and this changes your values to a negative anti-squat number, continuing the squat to progress. However if the links are parallel and angled up to the front of the vehicle you end up with a positive anti-squat number. Remember that length of the links will effect their angle toward each other as the suspension cycles. Also an intersection behind the rig is negative antisquat, mean that the axle will actually compress the suspension when power is applied.

This may not apply to short travel vehicles like slammed race cars but hang on for a wild ride into choosing a suspension that works consistently for suspension travel. To determine link length and location, plot all the critical values: WB w/ proper tire size, frame location, unsprung COG height and then the anti squat projection line you are shooting for, and link placement at the axle. From there you can plot the links so they intersect with this line. Then plot your axle end mounting points at a different Z (up or down) as the suspension cycles, then hit your antisquat line again at a different location farther up on the vehicle. Where your upper and lower links at these different projections intersect gives you your location and link length to achieve constant anti-squat for that suspension cycle. I'll try and find the link where a genius derived this with pix.

check out the link calculator, it has way more info than you could want, just play with the numbers and you'll see how seperation and angle effects squat

I'm not an expert but trying to study and get a good grasp on suspension design, and try and help my cuz out.

[nutley]

found it!

http://www.pirate4x4.com/forum/showt...ight=AntiSquat

posts by strange rover, towards the bottem. Cool stuff

[Norm Peterson]

When the links are truly parallel, their "intersection" is assumed to be at infinity. That makes the anti-squat construction line parallel to the links, as it now has to intersect them "at infinity". If you are somewhat uncomfortable with the use of infinite distances, or if your software/spreadsheet pukes at the concept, you could make the links converge at some huge distance (100 yds? 1 mile?). The error in A/S% would be so small as to have no meaning, and the consequences of realistic fabrication tolerances would probably have a greater effect.

On edit, I'm getting a

There seems to have been a problem with the database.
Please try again by clicking the Refresh button in your web browser.

error with the pirate4x4 link. Is it simply a case of the site being down for maintenance or is membership required to access the message board?

Norm

[nutley]

Try this

http://www.pirate4x4.com/forum/showthread.php?t=208915

3rd page

I noticed there was an antisquat in the url, maybe adds a seach fuction which you would have to be registered for.

[Mean 69]

Some outstanding explanations, nice work guys. I'd add one thing, which we feel is very important, and is a big influence in how we design our stuff.

Anti-squat is derived, as stated using several parameters, one of which is the "instant center." This is important, dynamically, because what the A/S value is at one instant, for instance static ride height, it will change as the suspension bumps, droops, and rolls. The IC can move up, back, forward, down, you name it, it depends upon how the links are oriented, and is pretty sensitive to link lengths. The general case is that the longer the links, the less the IC will migrate dynamically, and the resulting A/S migration will be "quiet" relative to a similar system with short links.

Once you start modelling this stuff out, you will learn pretty quickly that there are several things in tension with a (coupled) link suspension. For instance, the enemies of high A/S are axle (roll) steer, and the potential for brake hop. Suspension design is kind of like poking a balloon, you nudge one area, and another area pops out, you have to pick your battles for you individual application.

Mark

[BRIAN]

Great info. I have set up several 4 bar parallel bar set ups and every manufacturer has said to set them up parallel to the ground at ride height. The last set up was a S&W kit that had short (approx 17") bars. I think I remember being told that to leave the bars parallel or for a harder launch to raise them both approx 1-2" each the same measurement. They had 4 equally spaced holes on the front brackets.

I kind of understand the above but will take a little to sink in.

I have to say I always stick with a system and because of all the complex design elements that go into a suspension design you have to be scared with all the cars being constructed where guys are just making parts with no thought of how they actually effect the vehicles handling characteristics. Even worse they think they have improved upon them.

[parsonsj]

Brian,

The reason parallel 4 bars need to be level at ride height is to minimize roll steer. As the body rolls, one side of the parallel bars goes up, and the other side goes down. If they start level, then the amount of "shortening" that occurs as the bars go up/down will be same on both sides, keeping the rear housing perpendicular to the vehicle center line.

jp

[andrewb70]

Well this info is fine and dandy for solid axle cars, now what about IRS?

I am specifically interested in my RX7. I would like to optimize the rear suspension for drag racing. What info is needed to determine what I can chenge in order to increase the rear bite?

Right now when the car is launched hard with sticky tires, the rear of the car squats down very hard. From the info posted above this would indicate that the tires are being pulled up, since for all actions there is a reaction.

Attached is a picture of the junk that I have to work with.

Video of the car launching

Andrew

[MrQuick]

Ok start, not too bad.
Beef up the pumpkin mounting abit. Maybe fabricated rear suspension cradle that uses longer link style control arms. Is that a spare or did you start already?

[Norm Peterson]

Well this info is fine and dandy for solid axle cars, now what about IRS?

I am specifically interested in my RX7. I would like to optimize the rear suspension for drag racing. What info is needed to determine what I can chenge in order to increase the rear bite?

Right now when the car is launched hard with sticky tires, the rear of the car squats down very hard. From the info posted above this would indicate that the tires are being pulled up, since for all actions there is a reaction.

Attached is a picture of the junk that I have to work with.

Video of the car launching

Andrew

The geometric construction is only slightly different from what's been posted above, but it generally represents a huge difference in launch behavior (as you already know). The anti-squat line for IRS passes through the rear axle rather than the rear contact patch.

See the ** note in the November 28 post in this thread. Redraw the constructions to suit and you'll probably find the A/S line is at a fairly shallow slope. Perhaps somebody has C5/C6 rear suspension data; I suspect that it runs a bit more A/S than the RX7s did.

Other than doing things to make a big change in SVIC location - maybe some rear shocks with lots of bump damping will help?

Norm

[andrewb70]

The geometric construction is only slightly different from what's been posted above, but it generally represents a huge difference in launch behavior (as you already know). The anti-squat line for IRS passes through the rear axle rather than the rear contact patch.

See the ** note in the November 28 post in this thread. Redraw the constructions to suit and you'll probably find the A/S line is at a fairly shallow slope. Perhaps somebody has C5/C6 rear suspension data; I suspect that it runs a bit more A/S than the RX7s did.

Other than doing things to make a big change in SVIC location - maybe some rear shocks with lots of bump damping will help?

Norm

The launch in the video had a 1.685 60' time. I should list the things that I have already done.

Tokico adjustable shocks. The rebound and compression are always 50/50, but I can stiffen or soften them. The stiffest setting works the best.

The stock non-adjustable links that control the camber have been replaced with adjustable ones. Most of the negative camber has been dialed out.

There is a pinion snubber installed to control the upward rotation of the rear housing. The stock rubber front diff mount is notorious for breaking.

Ok start, not too bad.
Beef up the pumpkin mounting abit. Maybe fabricated rear suspension cradle that uses longer link style control arms. Is that a spare or did you start already?

So how does one change the SVIC location? How do I calculate it given my current configuration?

I am not at all opposed to fabricating various control arms that will transfer the load of the rotating diff assembly to a different point on the chassis.

Thanks,
Andrew

[Norm Peterson]

Heavy bump damping keeps the chassis from dropping as quickly, and sort of mitigates the lower value of the A/S associated with most IRS arrangements. More specifically, I'd think heavy low speed bump damping. High speed bump damping would be probably best left relatively soft, so that the occasional bump in the track isn't quite so successful at unloading the tire. You want the relative [vertical] positions of the chassis and axle to remain relatively constant under body inertia motions (launch) but have minimum force response from wheel displacement effects.

In the case of IRS, increasing A/S typically involves relocating one or more of the chassis side pickups and perhaps an upright pickup (per side, natch). It's a little more difficult to visualize this than moving the SVIC for a stick axle with a linkage - with IRS you have to think in terms of the planes containing the control arms and their intersection within the vertical plane containing the rear wheel. More A/S means either the LCA plane needs to slope upward more (think side view), the UCA needs to slope downward (ditto), or a little of each, all the while trying not to disturb or having to correct other aspects such as toe steer. Figures 17.15, 17.21 and 17.34 in RCVD give a pretty good picture of this, BTW.

Norm

[andrewb70]

Heavy bump damping keeps the chassis from dropping as quickly, and sort of mitigates the lower value of the A/S associated with most IRS arrangements. More specifically, I'd think heavy low speed bump damping. High speed bump damping would be probably best left relatively soft, so that the occasional bump in the track isn't quite so successful at unloading the tire. You want the relative [vertical] positions of the chassis and axle to remain relatively constant under body inertia motions (launch) but have minimum force response from wheel displacement effects.

In the case of IRS, increasing A/S typically involves relocating one or more of the chassis side pickups and perhaps an upright pickup (per side, natch). It's a little more difficult to visualize this than moving the SVIC for a stick axle with a linkage - with IRS you have to think in terms of the planes containing the control arms and their intersection within the vertical plane containing the rear wheel. More A/S means either the LCA plane needs to slope upward more (think side view), the UCA needs to slope downward (ditto), or a little of each, all the while trying not to disturb or having to correct other aspects such as toe steer. Figures 17.15, 17.21 and 17.34 in RCVD give a pretty good picture of this, BTW.

Norm

Norm,

Thanks for the help, but I am having a little trouble visualising. The RX7 suspension is a pretty simple swing arm design. There are no upper or lower control arms. Please see the picture above.

At one point I was thinking about fabricating a torque arm that would attach to the center housing and extend forward to the rear crossmember. Is that something that might help?

Andrew