The Science of …Front Sway Bars (Roush Fenway and Others)

David Poole has a great summary of what we currently know about ‘Partgate’. I think it’s safe to state from the media reports that a front sway bar went missing from Roush Fenway Racing last year and that its absence was discovered only when a vendor reported to RFR that someone had asked them to make mating parts for the AWOL sway bar. Michael Waltrip has admitted that his team is the one involved, and that the part was returned to RFR.

Jack Roush’s response to the entire situation has been treated with a somewhat comical attitude in many quarters. Jeff Gordon opined that, basically, parts get ‘borrowed’ all the time and this isn’t anything to be that upset about. Roush’s personal issues with Toyota no doubt color the coverage; however, I disagree with commentators like Joe Menzer, who say that Roush ought to lighten up. (I do, however, agree with Dave Moody, who pointed out early on that any accusations made ought to be specific or not made at all.) Some confusion was introduced by Toyota’s Lee White, who apparently didn’t know about the sway bar and thought Roush was talking about an incident in Fontana, where a valve spring from a Roush Fenway engine ended up at Toyota Racing Development’s California research lab.

Jack Roush’s actions are clearly clouded by his personal issues with Toyota and people associated with Toyota; however, to fully understand Roush’s frustration, you have to appreciate what a sway bar does and its increased importance in the new car.

The number of things teams can change has become nanoscopic (a made-up word for almost nothing). You’ve got a company that went from five teams in the Chase two years ago to having one of their cars out of the top 35. Every possible advantage over the competition is important. Gaining a tenth of a second is much more important than it once was. So let’s look at sway bars and what they can do for a car.

A sway bar (also called an anti-sway bar, roll bar, or anti-roll bar) is a long cylindrical piece of metal. The picture above comes from Speedway Engineering.I’ve seen sway bars at the track ranging in diameter from about 3/4″ to maybe 1-1/2″. Some sway bars are solid and some are hollow.

(Side note to Michael Waltrip: A sway bar differs from a jack handle in that the sway bar has splines at both ends. Splines are radial protrusions on the ends of the sway bar.

The arms that hold the sway bar ends have similarly shaped recesses, which means that the round sway bar cannot rotate freely in the arm: When the sway bar rotates, the arm moves (and vice-versa). The end of the sway bar arms not shown in my drawing are mounted to one of the control arms, which means that, when the wheel moves, the sway bar arm moves as well. The sway bar itself runs from the right side, through a tube in the chassis, and connects on the left side.

The primary purpose of a front sway bar is to resist body roll during turns, which shifts weight from one side of the car to the other. When a car turns, weight transfers from the inside tires to the outside tires (link to previous post). The body rolls more when the center of gravity is higher. As I noted in my pre-Martinsville post, you often have a problem keeping enough weight on the left front tire when you’re just coming out of a turn. A tire’s grip is proportional to how hard it is pushed into the track, so lack of weight on the tire means lack of grip and that means lack of speed. A sway bar helps resist the weight transfer in the front of the car, as do the coil springs in the front suspension.

The front sway bar couples the left and the right wheels together. A car can’t really tell whether the inside wheel is going over a bump, or if it is rising because it doesn’t have enough force pushing down on it. The sway bar does this using torsion, which is twisting of the sway bar when a torque (a force that causes rotation) is applied to it. If you put your arm straight out and ask someone grab your wrist and twist gently, your arm is experiencing torsion. A sway bar works the same way, although a thick steel bar has significantly more resistance to twisting than your arm.

If the two wheels move the same way (for example, both wheels run over a bump at the same time), the sway bar doesn’t do anything; however, when only one wheel moves, the sway bar becomes important. Let’s say that the left front wheel moves upward. The wheel moving upward causes the left sway bar arm to move upward, which tries to twist the sway bar. When the left side of the sway bar twists, the right side will twist as well, which causes the right sway bar arm to move in the same direction as the left arm moved, and the right wheel to move in the same direction the left wheel moved.

If the sway bar is very stiff (meaning resistant to twisting), the sway bar will keep the left front wheel from moving as much, which will decrease the weight transfer.
The resistance of the swaybar to twisting is the variable here. A stiffer sway bar gives you more resistance to roll; however, stiffer means larger and thus heavier. A solid sway bar one inch in diameter and 37.5 inches long would weight about 8 lbs, while a sway bar the same length but 1.5 inches in diameter would weight about 18.6 lbs. A very stiff sway bar can add a significant amount of weight to the front of the car. Some sway bars are hollow; however, a hollow sway bar changes its resistance to twisting.

The sway bar has become especially important because the new car has a higher center of gravity. The amount of roll increases as the center of gravity gets higher. The more roll, the harder the car is to turn. If you put a heavy sway bar on the front, you decrease body roll; however, you also add front weight, which is going to make the car looser. You’d have to put more weight in the rear to get balance, but then you have a heavier car and Newton’s second law of motion (F=ma) becomes an issue. As Joe Menzer reported, Roush said:

“One of the challenges this car has is it’s heavy in the front end. It’s hard to achieve the balance you’d like to have and get enough nose weight off the car to let it balance properly from a weight and balance point of view.”

Again, following Menzer’s report, Roush goes on to explain why he’s so agitated about the situation.

“One of the areas that we’ve worked in is the front sway bar. It’s not subject to a NASCAR teardown inspection. It was not a part that would have been mixed up with other Toyota parts–a front anti-roll sway bar. And within the confines–the package that NASCAR gives you, the material, the dimension, all those things–we designed our own part and we did an analysis of it. We optimized the torsional effect of it with minimizing the weight of it through heat treating and material selection and the way the components were machined and the way it was assembled.”

After that, Menzer reported only that Roush ‘continued to drone on’. One person’s drone is another person’s goldmine, especially if both the speaker and the second person have physics degrees. (When my students ask whether you can make money with a degree in physics, I talk about Jack Roush.) I would love to have access to the rest of the “droning” because there were doubtless additional clues about what they were doing with the sway bars.

David Poole reported that the Roush Fenway Racing sway bar was longer than usual sway bars. The fact that MWR supposedly asked a vendor to manufacture arms specifically for this sway bar suggests to me that the sway bar arms have to be shaped differently to accommodate the longer sway bar. If I remember the Millikens’ RCVD correctly (and I’m traveling, so I don’t have it with me), making the bar longer would decrease the effective spring rate; however, you can’t use a scalar formula to calculate the effect of offsetting the arms, especially if there is some built-in asymmetry to the bar and/or the arms.

Most standard swaybars are made from steel. 4140 steel is used commonly in commercial sway bars. 4140 is a chromium-molybdenum steel with very high torsional strength (which means you can twist it a lot before it breaks). Steel is made by heating and then cooling under controlled conditions; however, most steels have additional treatments to optimize certain properties. The treament depends on the application for which the part is intended. For example, torsional strength can be improved by heat treatment. For 4140, heat treatment often consists of heating to about 1550 °F and quenching (cooling quickly) in oil. You can also improve torsional strength by shot peening, which is basically shooting little hard things at the piece you’re treating. The impacts create residual stress (at least at the surface), which increases the torsional strength.

One thing I learned to appreciate from NASCAR is that steel is definitely not a has-been in the world of materials. Although nanomaterials (my area of research) might get more attention, you can’t beat steel for versatility and cost. Post-fabrication treatments are an ancient science–and something of an art if you ask me. We use heat treatments for permanent magnets (the microstructure of the magnets determines their magnetic properties, just as the microstructure of a steel determines its mechanical properties). I remember one in particular that required five distinct temperatures for five different times and in a very specific order. The positive for RFR is that, even though another team might be able to copy exterior dimensions and can probably determine the microstructure of the sway bar, they probably can’t reverse engineer the treatments used to get the material to that state.

There’s also the possibility that the sway bar wasn’t made from steel. You’ll notice that Jack said only “materials selection”, and heat treatments are used on many different types of materials. There are a number of examples in which technologies originally developed for NASA are used by NASCAR (e.g. carbon monoxide filter materials, thermal blankets). Structural materials ought to be one of them: NASA often cites a figure of about $10,000 per pound to put things into space, so it shouldn’t be a surprise that NASA was at least part of the motivation for developing lightweight, but still strong, alloys like titanium alumnides. Titanium aluminum alloys have a much higher strength-to-weight ratio, which means that you get the same strength for less weight. The problem is that titanium alloys are much more difficult to make than steel and are consequently much more expensive. Their cost is why NASCAR doesn’t allow these ‘exotic’ metals to be used in the chassis; however, I’m not aware of any rule that would preclude their use in sway bars. I also don’t know of any reasons why the sway bar would have to be uniform in diameter (or any other physical dimension for that matter).

Many of these changes from the traditional, uniform sway bar would be obvious to even the greenest engineer, so losing an innovative sway bar could be a real blow to a program that thought they had a sure way of getting ahead of the competition. The sway bar could be very important to helping the car turn better, and the ability to turn is key to winning races. Ignoring Jack’s knee-jerk reaction to Toyota, I hope that understanding the role of the sway bar, especially as the number of variables available to teams decreases, helps you appreciate why Roush was (is) so upset about one of his sway bars going missing. If they had come up with something truly revolutionary, having it become public knowledge could be a real setback.

For those of you in the Charlotte area, I’ll be explaining load transfer and sway bars in greater depth during a talk I’ll be giving April 7th at the University of North Carolina–Charlotte. The talk (which is free of charge) will be at 7:00 p.m. in room 281 of the College of Health and Human Services Building on the UNC Charlotte Campus. Parking is available at the new Union Deck.

Finally, everyone at my house (especially Darwin, our beagle/hound dog mix) wishes Elliott Sadler a speedy recovery from his back injury. Been there, done that and know it is no fun. Get better soon!


  1. Finally someone that can redeem Jack Rosh’s anger. Maybe you should go on Raceday or ESPN and explain this. Only then will people lay off Jack.

  2. Thank you for this column and all the others you’ve written. It’s amazing how much I’ve learned from you. I’ve been fascinated for years by the connection between NASA and Nascar…technology at its’ best. I sincerely hope the rest of the media reads this column and quits acting as if this occurence is insignificant. One would think that his thorough explanation would suffice, but apparently his questioners aren’t as savvy as I gave them credit for.

  3. Marc: You make me blush! I guess that’s the logical progression: Assistant Professor, Associate Professor, Professor, Goddess…

  4. Thanks for explaining. Now what is the difference between a sway bar and a Torsion Bar. I had a torsion bar need replacing on a car I owned once (towing accident, bar was used to pull the car, not wise). Same basic idea, just in the rear end?


  5. Thanks for a very informative column which puts at least some of swaybar technology into language that most of us can understand. I hope Joe Menzer reads it. As I told him in a comment on his article; Jack Roush is probably the most knowledgable owher since Junior Johnson and yet the media and, often, NASCAR itself continues to downplay his observations. As you said, it would be nice to have heard the rest of Jack’s “drone”. We all may have learned something.

    Cotton Morgan

  6. Hi PittCaleb:
    A torsion bar is the same thing as a sway bar. Like the torsion bar, a sway bar can be used in the front or the back, although the use in the back is a little different because NASCAR cars use a solid rear axle. Dale Jarrett got into some trouble during his last year with Yates (2006) for a rear sway bar violation. They were using it for something other than anti-roll and Slugger Labbe ended up getting a suspension and fine. Regardless, you are right that using it for towing is definitely not a good idea!

  7. Just got back from two week cruise so am catching up with the goings on. I would bet that a magnet would not stick to Jack’s anti-roll bar. Sorry cannot make the trip to Charlotte for your talk. Keep up your good work, but I do not think that the media really wants to know anything about race cars, because that would not make a good story.

  8. Great job of making a mountain out of a molehill. At the best (in typical engineer speak) you’ve described a way to save 10 or 12LBS off the front end. I don’t care what you do to the steel OR the Alloy, it still JUST transfers load to the other side. On the current car, that means while the car is turning and on the bump stops you have as much as a 1/2 inch of travel. Please explain how magical that bar could be to do anything with that. You may be a terrific engineer, but for all that explaining of how a sway bar works and how a SPECIAL ONE would be missed, you haven’t explained how it could DO ANYTHING SPECIAL IN THE COT. By the way the “Bar” in question is reportedly a 3″ diameter “tube”. Even if you got a Progressive rate out of it it would make little difference with the limits of deflection under which it labors. Jack may be a Brilliant engineer , but this is just plain silly.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.