Announcement

Collapse
No announcement yet.

Sway Bars - How they work and how to choose the right size for your needs

Collapse
This is a sticky topic.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Sway Bars - How they work and how to choose the right size for your needs

    by John Comeskey of SPS and James Walker of scR motorsports ( www.teamscr.com )


    What is Body Roll?

    The chances are high that you have experienced the effects of body roll during the course of your normal driving. It happens during almost every turn when one side of the car lifts, causing the entire vehicle to "lean" toward the outside of the turn.

    The cause of body roll is simple physics: an object in motion tends to stay in motion until acted upon by an outside force. So in practical terms, as you drive ahead in a straight line, you allow 2300-2600 pounds of vehicle, fluids, and passengers to build momentum in a straight line. Suddenly, through input at the steering wheel, you tell everything to change direction. But even though the front tires may change direction, thanks to the mechanical advantages of the steering system, the momentum of the vehicle, fluids, and passengers continues in the original direction. The tires are the only element capable of generating an outside force that can act against this momentum and change its direction.

    At this point, one of two scenarios is most likely to occur. If enough momentum exists in the original direction, and the tires lack enough grip to act against the original forward energy, then the vehicle will slide out of the turn as the tires loose traction. However, if the tires have enough grip at the road surface, then instead of sliding, the vehicle’s traction at the road surface will overwhelm the original forward momentum and act upon the original forces to induce a change of direction. Hence, a cornering maneuver.

    But what happens to that energy? Even though we may have had enough grip to hang-on through the turn, we know that the momentum of the vehicle mass will continue in the original direction. The result is a weight transfer toward the new outside edge of the vehicle – the same direction as the original forward momentum. If enough energy is behind the weight transfer, then this energy will cause the outside suspension (in this case, the spring and strut assembly) to compress while the other side lifts and extends. An engineer-type likes to describe this by saying that one side moves into "jounce" while the other moves into "rebound." The rest of us call it "lean" or "body roll."

    Why is Body Roll a Bad Thing?

    We often hear that preventing body roll is "so important" that we all must rush out and buy "this" product or "that" product in order to prevent it. And many enthusiasts have consequently accepted that body roll is therefore "bad." But what exactly does body roll do to negatively affect vehicle handling?

    For starters, it disrupts the driver. This is probably the effect that most drivers can see and feel during their own driving experiences. And while this is not the most important negative effect of body roll, it is true that the car does not drive itself – no matter how many aftermarket parts you install. So keeping the driver settled, focused, and able to concentrate on the task of driving is a foremost priority for spirited vehicle handling.

    However, the most often misunderstood effect of body roll upon vehicle handling is the effect of body roll upon camber - and the effect of camber changes upon tire traction. Put simply, the larger the contact patch of the tire, the more traction exists against the road surface, holding all else constant. But when the vehicle begins to "lean" or "roll" to one side, the tires are also forced to "lean" or "roll" to one side. This can be described as a "camber change" in which the outside tire experiences increased positive camber (rolls to the outside edge of the tire) and the inside tire experiences increased negative camber (rolls to the inside edge of the tire.) So a tire that originally enjoyed a complete and flat contact patch prior to body roll must operate only on the tire edge during body roll.

    The resulting loss of traction can allow the tires to more easily give away to the forces of weight transfer to the outside edge of the vehicle. When this happens, the vehicle slides sideways – which is generally a "bad" thing.
    90 Mazda 323 - KLZE, fender-flared, right hand drive, 2350 lb fully loaded sex machine. || Pic Thread ||- SOLD
    93 PGT FRANKENPROBE - 10.24 @ 139.9 mph ||545 whp @ 20 psi || Timeslip || Dyno slip|| Build Thread - GONE
    97 GTS - Rear Wheel Drive KLZE, 6-speed, 3.90 Torsen LSD, 2650 lb, daily driver! - Build Thread - GONE
    90 Ghettocet KLiata - forever WIP

  • #2
    How to Prevent Body Roll

    By definition, body roll only occurs when one side of the suspension is compressed (moves into "jounce") while the other extends (moves into "rebound.") Therefore, we can limit body roll by making it harder for the driver side and passenger side suspension to move in opposite directions.

    One fairly obvious method to achieve this is through the use of stiffer springs. After all, a stiffer spring will compress less than a softer spring when subjected to an equal amount of force. And less compression of the suspension on the outside edge will result in less body roll. However, stiffer springs require the use of stronger dampers (struts) and have an immediate and substantial effect on ride quality. So even though handling is improved, they may not be the easiest or most cost-effective way to achieve the objective of reducing body roll.

    For many enthusiasts, the use of anti-sway bars – also known as "anti-roll" bars, "roll" bars or "sway" bars – provides a more cost-effective reduction in body roll with minimal negative impacts upon ride quality.

    How an Anti-Sway Bar Works

    Put simply, an anti-sway bar is a "U-shaped" metal bar that connects to both wheels on opposite sides of the car at the ends and connects to the chassis in the middle. Essentially, the ends of the bar are connected to the wheels while the center of the bar is connected to the body of the car.

    In order for body roll to occur, the suspension on the outside edge of the car must compress while the suspension on the inside edge simultaneously extends. However, since the anti-sway bar is attached to both wheels, such movement is only possible if the metal bar is allowed to twist. (One side of the bar must twist upward while the other twists downward.) So the bar’s "torsional stiffness" – or resistance to twist – determines that bar’s ability to reduce body roll. Less twisting of the bar results in less movement into jounce and rebound by the opposite ends of the suspension – which results in less body roll.

    Factors that Determine Sway Bar Stiffness

    There are two primary factors that determine an anti-sway bar’s torsional stiffness: the diameter of the bar and the length of the bar’s "moment arm" (more commonly known as the amount of leverage that the vehicle is able to apply against the twisting motion of the bar.)

    Diameter is generally the easiest concept to grasp, as it is somewhat intuitive that a larger diameter bar would have greater torsional rigidity. Torsional (or twisting) motion of the bar is actually governed by the equation:

    twist = (2 x torque x length)/(pi x diam^4 x material modulus)


    And since "diameter" is in the denominator, as diameter gets larger, the amount of twist gets smaller. Which, in a nutshell, means that torsional rigidity is a function of the diameter to the fourth power! This is why a very small increase in diameter makes a large increase in torsional rigidity.

    To compare, for example, the rigidity of a Saturn’s stock rear 15.0mm bar to a larger 16.5mm bar, simply use the equation, 16.54/154 which yields 1.46. In other words, a 16.5mm bar is 1.46 times as stiff – or 46% stiffer – than a 15.0mm bar of the same design.

    Add just one more millimeter to the diameter of the bar – for a total of 17.5mm – and the torsional strength skyrockets to 85% stiffer than the stock 15.0mm bar. (17.54/15.04=1.85.)

    However, in addition to the diameter of a bar, there is another very important factor that determines an anti-sway bar’s torsional rigidity. This factor is known as the "length of the moment arm" – or in common terms, the amount of leverage between the vehicle and the bar.

    As with anything, an increased amount of leverage makes it easier to do work. This is governed by the "lever law," force x distance = torque. As "distance" - or the length of the lever - increases, the resulting amount of torque also increases. (This is why it was easier to move your big brother on the teeter-totter when he moved into the middle and you stayed out on the end. You enjoyed increased leverage at the end, while he suffered from reduced leverage in the middle.)

    Because an anti-sway bar is shaped as a "U," the ends of the bar that lead from the center of the bar to the end-link attachment serve as a lever. As the distance from the straight part of the bar to the attachment at the end link becomes longer, the torque applied against the bar increases – making it easier for a given amount of energy to twist the anti-sway bar. As this distance is reduced, torque is reduced - making it more difficult for a given amount of energy to twist the anti-sway bar.
    Last edited by Dan [Sac]; November 11, 2004, 10:28 PM.
    90 Mazda 323 - KLZE, fender-flared, right hand drive, 2350 lb fully loaded sex machine. || Pic Thread ||- SOLD
    93 PGT FRANKENPROBE - 10.24 @ 139.9 mph ||545 whp @ 20 psi || Timeslip || Dyno slip|| Build Thread - GONE
    97 GTS - Rear Wheel Drive KLZE, 6-speed, 3.90 Torsen LSD, 2650 lb, daily driver! - Build Thread - GONE
    90 Ghettocet KLiata - forever WIP

    Comment


    • #3
      this is a good thing that a lot of people on here need to read


      thanks for puttin this up
      /X\ipoc - Chipoc Black 1994 Probe GT - R.I.P.
      1993 honda civic hatch - K20
      1991 honda civic hatch - 12.3:1 C/R balanced for 10,000rpm b16 -Gone-
      1995 Acura integra GSR 4dr - stuff -Gone-

      Comment


      • #4
        It is the lever law that is applied during the design of an adjustable anti-sway bar. By using multiple end link locations, the distance from the point of attachment to the straight part of the bar can be altered. Or in engineers’ terms, the length of "the moment arm" can be increased or reduced in order to make more or less torque against the bar. Using a setting further from the center of the bar increases the length of the moment arm, resulting in more torque against the bar, allowing more twisting motion of the bar, creating more body roll. Using a setting closer to the center of the bar reduces the length of the moment arm, resulting in less torque against the bar, allowing less twisting motion of the bar, creating less body roll.

        The actual impact upon torque can be compared by dividing the center-to-center distances of the end-link attachment points. For example, the center-to-center distance of the stock rear anti-sway bar is 200mm. We can compare this to the 160mm distance of the firmest setting of the 4-way adjustable 17.5mm bar by simply dividing the distances, 160/200 = .8. In other words, a 160mm center-to-center bar produces only 80% of the torque that would be produced by a 200mm center-to-center bar of the same diameter. Or simpler yet, by using the 160mm end-link attachment points, we increase the stiffness of the anti-sway bar by an extra 20%.

        (Note: by now you may have realized that the 17.5mm, 4-way adjustable bar found on the scR ITA racecar in the example above is as much as 105% stiffer than stock. 85% of this is a result of diameter, while the remaining 20% is a result of moment arm length. The dramatic increase in stiffness is why it is recommended for race-use only!)

        So, in summary, less twist = less deflection = less body roll.

        What the Heck is "TLLTD?"

        TLLTD stands for Tire Lateral Load Transfer Distribution. While this term may sound complex, it simply measures the front-to-rear balance of how lateral load is transferred in a cornering maneuver and is commonly used to compare the rate of lateral traction loss between the front and rear tires. You probably understand this already as the concept of "understeer" and "oversteer."

        Put simply, there is only so much force that a tire can handle. When we ask more of the tire than the tire can deliver, it saturates and loses traction. If the front tires saturate before the rear tires, then we call this "understeer" or "push" – which means that the car tends to continue moving in the original direction, even though the wheels are turned. If the rear tires saturate before the front tires, then we call this "oversteer" or "loose" – which means that the rear of the car tends to swing around faster than the front, causing a spin. When neither of these conditions prevail consistently, then we describe the chassis as "balanced."

        We can measure and compare the steady-state understeer and oversteer characteristics of a vehicle by assigning a lateral load transfer percentage of the front relative to the rear. A TLLTD value equal to 50% indicates that the chassis is balanced – or both the front and rear tend to lose traction at roughly the same time. A front TLLTD value greater than 50% indicates that the front tires lose traction more quickly than the rear tires – resulting in understeer. And a front TLLTD value lower than 50% indicates that the rear tires tend to lose traction more quickly than the front - resulting in oversteer.

        (It is important to note that our discussion of TLLTD is only considering steady-state cornering maneuvers, such as a long 270o on-ramp or off-ramp. Moderate-to-aggressive throttle or brake application can upset this "balance" during a transient condition, briefly transitioning a vehicle from understeer to oversteer.)
        90 Mazda 323 - KLZE, fender-flared, right hand drive, 2350 lb fully loaded sex machine. || Pic Thread ||- SOLD
        93 PGT FRANKENPROBE - 10.24 @ 139.9 mph ||545 whp @ 20 psi || Timeslip || Dyno slip|| Build Thread - GONE
        97 GTS - Rear Wheel Drive KLZE, 6-speed, 3.90 Torsen LSD, 2650 lb, daily driver! - Build Thread - GONE
        90 Ghettocet KLiata - forever WIP

        Comment


        • #5
          The Effect of Anti-Sway Bars Upon TLLTD

          You now understand how an anti-sway bar can be used to limit body roll. And you understand that reduced body roll can lead to a reduction in adverse camber changes for better tire traction. But what may not be obvious is the effect of anti-sway bar changes upon TLLTD (understeer and oversteer.)

          In fact, given the above information, one might even assume that a firmer anti-sway bar, which leads to better camber control, would lead to better traction. So if we add a firmer anti-sway bar to the front, traction loss diminishes, so understeer is reduced, right?

          Not quite. Let’s evaluate more closely the meaning of TLLTD – tire lateral load transfer distribution. If said another way, we might describe TLLTD as the relative demand of side-to-side energy control that is placed upon the tires. Because a firmer anti-sway bar allows less deflection, it will transfer side-to-side energy (lateral loads) at a faster rate. As the rate of lateral load transfer increases, additional demands are placed upon the tire. So if we install a firmer anti-sway bar in the front, then we increase the distribution of lateral load transfer toward the front tires. This increases the front TLLTD value, which will result in additional understeer, holding all else constant.

          The same logic also holds true in the rear. A firmer anti-sway bar in the rear will increase the rate of lateral load transfer, placing more demand upon the rear tires, accelerating lateral traction loss, creating more oversteer, holding all else constant.

          "I want a 50% TLLTD on my car, right?"

          For many enthusiasts, it is tempting to jump to the conclusion, since a 50% TLLTD indicates a "balanced" chassis on paper, that this is therefore desirable. And all cars "should obviously come this way from the factory." Unfortunately, this is not the case – and the considerations are not that simple.

          In reality, a car with a 50% TLLTD is literally on the constant brink of oversteer. And there are many factors that can quickly and easily take the car from the brink into a full-scale, out-of-control, spinning-in-circles disaster!

          For starters, consider the effects of weather conditions that might create a wet or icy road surface…or imagine that the driver happens to apply too much brake late into a turn …or consider the effects of varying tire temperatures, tire pressures, or tire wear – all of which will have major impacts upon lateral traction thresholds. And of course, varying weight distribution, as a result of changing fuel tank levels, passengers, or the number of subwoofers in the trunk, will also impact TLLTD.

          With all of these things to consider, the engineers are forced to create a more conservative TLLTD. As a result, they intentionally target higher front TLLTD values so that stock vehicles will be prone to understeer - the assumption being that understeer is safer and more predictable for the average driver. For example, the stock Saturn SC2 is tuned to produce a front TLLTD of approximately 63.4% - a relatively conservative target. (But give Saturn some credit – this is on the aggressive end of the conservative spectrum, especially relative to other front-wheel-drive economy cars!)

          As a general rule, an average street-driving enthusiast is probably willing to accept some compromises – within reason – of a more aggressive TLLTD in exchange for better handling. A suitable target is probably a front TLLTD value of approximately 58% - a value that is considered "aggressive," but suitable for street driving.
          90 Mazda 323 - KLZE, fender-flared, right hand drive, 2350 lb fully loaded sex machine. || Pic Thread ||- SOLD
          93 PGT FRANKENPROBE - 10.24 @ 139.9 mph ||545 whp @ 20 psi || Timeslip || Dyno slip|| Build Thread - GONE
          97 GTS - Rear Wheel Drive KLZE, 6-speed, 3.90 Torsen LSD, 2650 lb, daily driver! - Build Thread - GONE
          90 Ghettocet KLiata - forever WIP

          Comment


          • #6
            So what is the 2nd gen.. 12mm.. compared to a 22 addco.

            Comment


            • #7
              How do I Create the Handling Balance that is Best for Me?

              Obviously, TLLTD and body roll will both be affected by changes to springs and anti-sway bars. So understanding the effects of multiple changes can get confusing. Fortunately, knowledgeable tuners (such as SPS) have access to basic chassis dynamic calculation software that allows the simulation of various spring and sway bar configurations upon the Saturn chassis.

              One of the things that we have learned through experience – and through simulation – is that the installation of a stiffer front sway bar may not be best for many situations. One reason for this is that while the use of a stiffer front bar effectively further reduces body roll, it also increases front TLLTD, inducing more understeer. It also can reduce front traction while accelerating and turning – already a problem thanks to a lack of a limited slip differential. For

              these reasons, at team scR we run a STOCK front anti-sway bar…a fact that seems to contradict the cornering power of our car until you understand the relationships above.

              A few calculations showing the effects of anti-sway bars changes upon a Saturn SC2 with stock springs is shown below. (Roll Gain measures degrees of body roll per g of cornering load. Lower numbers represent less body roll.)

              Anti-Sway Bar -------------------------Roll Gain (deg/g)-------------------------% Front TLLTD
              Stock front, stock rear-----------------------------4.5 --------------------------63.4
              Stock front, 16.5mm rear---------------------------4.2--------------------------59.7
              25mm solid front, 16.5mm rear----------------------3.97-------------------------62.1
              Stock front, 16.5mm rear adj. set on hardest setting-4.06-------------------------58.0


              Summary

              Hopefully this information will allow you take the next steps toward transforming your daily driver in to a canyon-carving g-machine. Small, well planned steps might just make you to the finish line faster than you thought possible…
              Last edited by Dan [Sac]; November 11, 2004, 10:22 PM.
              90 Mazda 323 - KLZE, fender-flared, right hand drive, 2350 lb fully loaded sex machine. || Pic Thread ||- SOLD
              93 PGT FRANKENPROBE - 10.24 @ 139.9 mph ||545 whp @ 20 psi || Timeslip || Dyno slip|| Build Thread - GONE
              97 GTS - Rear Wheel Drive KLZE, 6-speed, 3.90 Torsen LSD, 2650 lb, daily driver! - Build Thread - GONE
              90 Ghettocet KLiata - forever WIP

              Comment


              • #8
                Stock bar = 12mm
                Mazda 626 bar = 16mm
                PT bar = 17.5mm (iirc)
                Mazdaspeed = 19mm (iirc)
                Addco = 22mm



                Edit - the above calculations in red were made for a FWD Saturn road race car - you can easily take the equation in red a few posts up and make the calcs. for the Probe.

                Again, I cannot emphasize enough that there is NO "best" bar out there, but just a bar that will best work for YOU and what your handling goals are.
                Last edited by Dan [Sac]; November 11, 2004, 10:26 PM.
                90 Mazda 323 - KLZE, fender-flared, right hand drive, 2350 lb fully loaded sex machine. || Pic Thread ||- SOLD
                93 PGT FRANKENPROBE - 10.24 @ 139.9 mph ||545 whp @ 20 psi || Timeslip || Dyno slip|| Build Thread - GONE
                97 GTS - Rear Wheel Drive KLZE, 6-speed, 3.90 Torsen LSD, 2650 lb, daily driver! - Build Thread - GONE
                90 Ghettocet KLiata - forever WIP

                Comment


                • #9
                  dan,

                  do you have that in a word document? that would be great for the main site's FAQ.
                  ProbeTalk is BACK! Once again, it's what you use the Internet for!

                  Let's get DANGEROUS!

                  Comment


                  • #10
                    No, but I could make it into one...
                    90 Mazda 323 - KLZE, fender-flared, right hand drive, 2350 lb fully loaded sex machine. || Pic Thread ||- SOLD
                    93 PGT FRANKENPROBE - 10.24 @ 139.9 mph ||545 whp @ 20 psi || Timeslip || Dyno slip|| Build Thread - GONE
                    97 GTS - Rear Wheel Drive KLZE, 6-speed, 3.90 Torsen LSD, 2650 lb, daily driver! - Build Thread - GONE
                    90 Ghettocet KLiata - forever WIP

                    Comment


                    • #11
                      not bad at all. But just a bit of an FYI

                      The calcs listed above for relative torsional stiffness assume comparisons between solid swaybars only. For the swaybars available for the MX6/PGT there are solid and hollow swaybars to consider and require different calculations.

                      Unless someone is planning to have a swaybar made of chromoly or titanium, you can assume the same modulus strength. That is, it will cancel itself out since the materials used by all of the swaybar companies will be the same (cold rolled spring steel).

                      Again, unless someone is planning a torsion bar with clamping adjustable ends (possible...you never know) the moment arm calcs also cancel out since all the remaining commercially available swaybars on the market are the same length and have no adjustability.

                      Oh yeah....The calcs for the relative torsional stifnesses of the swaybars(both solid and hollow)for the 1st gens have been done already. (where have you guys been? )

                      EDIT: Link to 1st gen swaybar calcs. http://www.mx6.com/forums/showthread.php?t=89899

                      Gavin
                      Last edited by gavin; January 8, 2005, 05:52 PM.

                      Comment


                      • #12
                        We know nothing of first gens

                        The 2nd gens have hollow front bars and solid rear, including all aftermarkets. The Saturn stock front bar (for which the calcs were made) are hollow as well, to the best of my knowledge.
                        Last edited by Dan [Sac]; November 12, 2004, 01:23 PM.
                        90 Mazda 323 - KLZE, fender-flared, right hand drive, 2350 lb fully loaded sex machine. || Pic Thread ||- SOLD
                        93 PGT FRANKENPROBE - 10.24 @ 139.9 mph ||545 whp @ 20 psi || Timeslip || Dyno slip|| Build Thread - GONE
                        97 GTS - Rear Wheel Drive KLZE, 6-speed, 3.90 Torsen LSD, 2650 lb, daily driver! - Build Thread - GONE
                        90 Ghettocet KLiata - forever WIP

                        Comment


                        • #13
                          We know nothing of first gens
                          Commie Loser! !st gen are teh fastAr!!111!!!

                          The AWR 19mm rear bar is hollow FWIW. This is curiously one of the few areas where the 1st gen has more options than the 2nd gen. A miracle.

                          Gavin
                          Last edited by gavin; November 12, 2004, 02:36 PM.

                          Comment


                          • #14
                            Actually, truth be told, I liked to autocross my atx PGL over any of my second gens (except for the one with the R tires). The GL was lighter, looser, but a bit low on power.
                            90 Mazda 323 - KLZE, fender-flared, right hand drive, 2350 lb fully loaded sex machine. || Pic Thread ||- SOLD
                            93 PGT FRANKENPROBE - 10.24 @ 139.9 mph ||545 whp @ 20 psi || Timeslip || Dyno slip|| Build Thread - GONE
                            97 GTS - Rear Wheel Drive KLZE, 6-speed, 3.90 Torsen LSD, 2650 lb, daily driver! - Build Thread - GONE
                            90 Ghettocet KLiata - forever WIP

                            Comment


                            • #15
                              great stuff! Anyone have the link to the partn umbers for the 626 bar?
                              MazdaSPEED6

                              Comment

                              Working...
                              X