Tilt Test Outdated. Teams Run on Ragged Edge

[J. Vinella]

Dr. Dale Calkins, our late faculty adviser, wrote an SAE paper in 1994 titled “Lateral Response of a Formula SAE Race Car” dictating that an FSAE car would never go above 1.3G of lateral acceleration. The tilt test is based on this idea while adding a little safety factor, 1.7G. “B6.7.2 Rollover stability will be evaluated on a tilt table using a pass/fail test. The vehicle must not roll when tilted at an angle of sixty degrees (60°) to the horizontal in either direction, corresponding to 1.7 G’s. The tilt test will be conducted with the tallest driver in the normal driving position.“
If a car lifts off the tilt table at comp this is to say the car will lift off the ground at 1.7G. Teams reading this today should laugh at such a value. With improved tires and modeling many teams are achieving much higher values that this.

In 2009 we ran a new suspension package that was the culmination of my work at UW. In the summer of 2010 the team was conducting wings vs. no wings on a new grippy go-kart track. Going into a left hand sweeper on a smooth bit of track the 2009 car rolled. The driver was not harmed. A post mortem (2009 car has not been driven since) was conducted and it was determined that the grip the car was making was too much for the CG/track width and it was only a matter of time and tuning that such an event would take place. At the time of roll, the car was testing the no wing configuration. Even though the wings add to the CG, they do provide a restoring force to keep the car on the ground. The fact that we were lifting with wings shows the grip of the car, but more importantly shows the danger that is now present in FSAE. The 2009 car was retired as it was deemed unsafe.
The images below show some winged teams running their cars with no wings. The final picture is the UW 2009 car at completion. The 2009 Car DNF due to a crack in the inboard CV cup. This crack was caused partly by an inadequate design, but also from repeated inertial slams by the tires lifting off the ground and contacting again.


http://a6.sphotos.ak.fbcdn.net/hphotos-ak-ash4/312092_10150861185565008_724710007_21053101_155326 403_n.jpg

http://a3.sphotos.ak.fbcdn.net/hphotos-ak-ash4/263484_197273356991664_130002227052111_609774_5572 110_n.jpg

http://www.uwashingtonfsae.com/gallery/galleries/Team_20_-_2009/Competitions/California/728.jpg

Bottom line: the tilt test needs to be updated for the times.

[Price]

In 2007, the team that I was involved with took the previous years car to a few SCCA autocross events. One such event was held at at an airport with a very abrasive concrete surface. At that event, while I was driving, the two inside tires picked up on one of the corners. From what i was told, they were off the ground around 8" which may be an exaggeration. That was the only event that we ran on that had a concrete surface, and was also the only time we ever picked up both inside wheels. I always assumed that the difference in grip was the result of the different track surface.

There was some talk when SAE announced that west venue was moving from California to Lincoln, that the surface at the new venue was that same type of grippy concrete that I encountered. Now, I will be the first to tell you that I do not actually know if the Cf of the tire will be higher in Lincoln than it would be elsewhere as a result of the track surface, but it does correlate with the experience I have.

I have been curiously waiting to see if some of the top contenders that are planning on going to Michigan, and Lincoln will have noticeably different setups for either surface. If there is indeed a difference in Cf from one event to another, I would expect that there are some design changed from one event to the next. But, I was also expecting to see a change in the tilt requirement for Lincoln. But that was only because I was under the impression that there would be a difference in maximum performance at that event.

However, like everyone else, I am still learning.

[JT A.]

At SCCA Nationals in Lincoln we didn't see any cars going up on two wheels. Almost every car there had wings, for what thats worth. Maryland, of course, could probably drive upside down and keep all wheels planted.

[Kevin Hayward]

John,

We were getting similar moments in 2004. However I wonder whether we need to ask the question how much lateral g do we want these cars capable of?

I have long been a supporter of a spec tyre for the competition. Something with long life to reduce the cost of competing, this should also allow the competition organisers to decide the level of lateral g capable.

Otherwise there should be some mix of an increased angle on the tilt table and an increase in the minimum track width.

Kev

[Crispy]

Our car is one of the offenders pictured above. We have have never rolled a car, but it is certainly possible for a skilled driver to get it up on two wheels (fortunately, the skilled drivers have always brought the car back down). Without aero, we have seen both wheels come up at around 1.9g, but it has always been on a hot day with a very clean surface in combination with a bump on the track. The rollover design point was originally 2g without aero. Our estimate of when we will lift both inside wheels with aero is about 2.5g.

The tilt test is something that we have had an issue with for several years. Even as the tests are run currently, we don't believe they accurately represent rollover at 1.7g. The reason for this is the lip that is used to keep the car from sliding off the table supports the car and prevents if from rolling.

https://lh6.googleusercontent.com/-5enb1wXGtrc/TovxFZ11EKI/AAAAAAAAAvY/s4MGfK6Pv58/s640/IMG_0846e2.jpg

Different competitions have different size lips for supporting the car when it is tilted, but they all add additional support that will delay rollover. The solution is certainly not to remove the lip, in fact I think it is already too small at some events and cars with softer/rounder sidewalls are somewhat at risk of slipping off. I think the situation would be improved if the "L" bracket that supports the car was larger (5"-6") and hinged at the corner so that the car can rotate about the outside of the tire. It seems like changing the tables would be a practical thing to do even this year.

As for increasing the angle, the g level is pretty sensitive to angle change. 2g is only about 3.5 degrees more than 1.7g. I'm not sure what the angle tolerance is at this point but the setups may need to be improved so that the test is accurate enough.

I certainly agree that something should be changed. As is, building a car capable of rolling over is well within the rules.

Has anyone even ever seen a car not past the tilt test?

[RobbyObby]

Originally posted by Crispy:
Has anyone even ever seen a car not past the tilt test?

Well there was the incident with UC Davis at West last year! Although that was down to operator error, not car design.
Given the average track width of a FS/FSAE car, in order to flip on the tilt table, your CG would have to be over 15 inches off the ground. Most cars tend to have a CG between 10 and 12 inches above the ground (although I have no hard data to prove that), and it seems like it would be pretty difficult to design a car with CG that high. IMO.
I do agree though that 1.7g is starting to become common place. However, very few cars have gotten to the point of sustaining lat g of that high. Most instances that have been brought up are usually a combination of track surface, bumps, etc. that caused a peak G higher than expected at that instant.

[NickFavazzo]

I think it is part of your vehicle design, if you don't design the track width to suit the expected G's then you have not designed it well.

We recently had our 2008 vehicle do 2.15 on Barbagallo (race track), no lifting of wheels. We take into account when designing the car that we don't only use it for SAE, we also take it into account that the cars will be hammered by learner drivers, put through their paces in other events and such. As a result we don't only think about SAE rules and safety at SAE speeds, safety at all speeds the car is driven is of paramount importance.

I think the tilt test is outdated but regardless, the tilt test should not be the point where teams find out their design is wrong, you should know long in advance of competition. Teams that go above the tilt test value know that. teams that don't venture that high may not. I think it is still important for the newer teams who are still learning, if their car only ever makes 1.8 max then why test at 2, they will fail even though their car will never make 2.

[StevenWebb]

I feel I should clarify why we were tipping over in that top photo. The car was running 15mm higher than it was designed for, and in that configuration it would not have passed the tilt test.

I personally don't think the tilt test is outdated. Keep in mind that it's the Coefficient of grip is the defining factor to tipping, not the G's. Even the top teams won't see grip coefficients above 1.6 at competition (we even swapped data with other teams in 2009). There would need to be a ridiculous level of grip on the surface to get up to 1.7. Any car that passes tilt should not lift both wheels at the competition unless there is something (dampers, jacking force, etc) lifting the cars cg up.


A concern for winged teams is that if any side gusts hit the rear wing end plates. Our 2005 car car would have passed any reasonable tilt test, but add a strong side gust and the result:

http://img403.imageshack.us/img403/9597/2wheels.jpg

[Mike Cook]

This is an interesting post. Thanks for posting the pictures up Jack. If I recall correctly, the tilt table rules were more relaxed up until 2004 when a certain team, wink wink, flipped their car at competition. I think it was 55 degrees or 57 degrees back then. Anyways, given certain banking, coefficient of friction, wind conditions, and bumps, it is possible to make most cars pick up two wheels on course. I haven't been that concerned and I have seen our car on two wheels several times.

However, add concrete and a hot June day and Lincoln probably has the recipe for the highest g values witnessed in FSAE to date. Obviously it is too late to modify the rules for this year, but it would be very interesting if during tech inspection, all cars were tilted to roll over and the angles were recorded. This data could be used in the future to see if a rule revision is required and if so, how much do we really need to change the tilt angle by.

I'd really like to keep away from mandating a minimum track width. Teams with lower CG's should reap the benefits of a narrower track.

[J. Vinella]

Originally posted by Mike Cook:
I'd really like to keep away from mandating a minimum track width. Teams with lower CG's should reap the benefits of a narrower track.

Mike,
They already reap plenty of benefits, but I agree from an engineering design stand point.
I like your idea of taking data for this year, kind of like the templates.

Kev,
Oh I know. Our '04 car had very narrow track width and we lifted wheels. Add 6 years, better tires and an evolution in tire testing and tuning.

I believe the tilt test should be for at least 2.5G When I look at the grip UW is making and the CG of the car with wings, it would be hard to pass tilt with our current track width and CG. It would have to go way up to be at all safe. Some might bring up the point such a test takes the wings at speed out of the equations, but you should be pulling high Gs everywhere, not just at speed. Good ol' fashion mechanical grip.



StevenWebb,
I personally don't think the tilt test is outdated. Keep in mind that it's the Coefficient of grip is the defining factor to tipping, not the G's. Even the top teams won't see grip coefficients above 1.6 at competition (we even swapped data with other teams in 2009). There would need to be a ridiculous level of grip on the surface to get up to 1.7. Any car that passes tilt should not lift both wheels at the competition unless there is something (dampers, jacking force, etc) lifting the cars cg up.


Steven you have an awesome car. I however don't agree with much of what you have posted. If I am pulling 1.7G, what is my the average coefficient of grip? Hint: F=MA...F=Mu*M...A=???

You didn't share with the right teams. More math...if team X pulls X in skidpad, what was team X's G? And that is skidpad, not transients, where both inside and outside tires are making massive grip together.

[AdamM]

"If I am pulling 1.7G, what is my the average coefficient of grip? Hint: F=MA...F=Mu*M...A=???"

There's something missing here... Fy=Mu*Fz=Mu*(mg+Faero). If you add downforce, your normal force and cornering force increase in proportion - that's assuming a constant friction coefficient. Whether or not wings make the situation worse depends on whether the increased CG from the wings is compensated for by the decreasing coefficient of a real, load sensitive tyre.

[Kevin Hayward]

Mike,

A team with a low centre of gravity already has benefits regardless of whether they decide to take it as less weight transfer or being able to take a tighter line by using a narrower line. There was a time when no teams were near to the minimum wheelbase rule, now there are a lot of teams nearly on the limit. Maybe a starting track limit could be 1m, which shouldn't affect very many teams at all, but at least sets a hard line.

Personally I think the better solution is to start using a spec tyre. Before everyone cries foul it is helpful to think that there are a lot of rules in the rulebook to control speeds with the idea of protecting students some of the main ones are:

- Restrictor size
- Engine size
- Aerodynamic limits

This is not outside the level of control that should be offered to the rule makers in order to protect our students.

We are driving a very fine line if teams end up designing to the limit of roll-over in order to increase vehicle performance. Some teams will be able to do so with ease, but many will just get it wrong. Anyone who has spent a fair bit of time in the FSAE paddocks would appreciate that there are a lot of mistakes made when building cars to such a tight deadline with limited starting knowledge. The last thing we would all want to see is an increase in student injuries.

This is a fairly unique situation. I do not know of any other form of 4 wheeled tarmac motorsport where there is a significant advantage in narrower tracks, usually teams have to be limited by a maximum. This has largely been created as a side-effect of trying to limit top speed by creating twisty tracks and forcing teams to focus on mechanical grip.

Kev

[StevenWebb]

Originally posted by J. Vinella:


Steven you have an awesome car. I however don't agree with much of what you have posted. If I am pulling 1.7G, what is my the average coefficient of grip? Hint: F=MA...F=Mu*M...A=???

You didn't share with the right teams. More math...if team X pulls X in skidpad, what was team X's G? And that is skidpad, not transients, where both inside and outside tires are making massive grip together.

as adamM pointed out, downforce is gonna change it quite a bit. If a 200kg car is pulling 1.7g with 35kg of downforce, your Mu is only 1.5.
You would also need to be over 1.9 g to tip a car that barely passes the tilt table (unless something else is helping).

I don't think a minimum track would help at all, Teams would just focus less about their CG heights, but be just as close to tipping. The benefit of lowering cg simply isnt that high with most SAE tires unless you can make your car narrower.

A spec tire will NEVER work in fsae. If all teams had to get the same tires to work, the cars would end up very similar to each other which is not what FSAE is about.

[MalcolmG]

I'm not convinced that spec tyre will make everybody's cars the same. Pick a common tyre (e.g. the 13" Goodyears or Hoosiers), and look at the range of different designs using that tyre. Every time a new rule is brought in, there's the argument that it's going to force everyone to go down a certain route, but I haven't witnessed that yet - fuel economy rules haven't sent everyone to singles, more liberal aero rules haven't put wings on every car, the templates didn't make everyone's cars too big and heavy to be successful with singles - I'm sure a spec tyre would also have no such effect.

Also FWIW we've had our non-winged cars up on 2 wheels on a sticky go-kart track, which requires a mu greater than 2 - the surface at Werribee would probably never let you get that high but I wouldn't be surprised if there were other competitions where it could be done

[Crispy]

Originally posted by StevenWebb:
as adamM pointed out, downforce is gonna change it quite a bit. If a 200kg car is pulling 1.7g with 35kg of downforce, your Mu is only 1.5.

I think Jack's point is that you don't need aero to get to 1.7g and I agree with him, and we have a lot of data that agrees with him. The coefficient of friction really can be that high (or higher).


Originally posted by StevenWebb:
You would also need to be over 1.9 g to tip a car that barely passes the tilt table (unless something else is helping).

As I mentioned earlier, I don't believe the current tilt test even test to 1.7g due to the side support used on the loaded tire. The support pictured in my post above above effectively lowers your CG by about 4 inches.

[spierson]

Originally posted by Crispy:
[QUOTE]Originally posted by StevenWebb:
as adamM pointed out, downforce is gonna change it quite a bit. If a 200kg car is pulling 1.7g with 35kg of downforce, your Mu is only 1.5.


I think Jack's point is that you don't need aero to get to 1.7g and I agree with him, and we have a lot of data that agrees with him. The coefficient of friction really can be that high (or higher).

Yep, we've been seeing ~2g on our last 2 cars (1.9 in 2010, 2.1 in 2011), no aero. Part of our design constraint is that when we go to cali or in this case Lincoln, the car is going to hit higher g's than we see in testing due to the poor surface quality lots we have. Our team looks for a balance between g-force and track width so that we're still fitting thru the cones yet not putting our driver/car at risk. Our 2010 car was too narrow for its cg height. The person that did the suspension design decided to neglect some things and based his data off old cars hitting 1.7g max. Go figure, we were on 2 wheels in Cali in autocross. We were consistently hitting 1.9 in that car in testing on crappy lots. The heat and surface at Cali was the kicker. I actually had to purposely drive slow during endurance to avoid a DQ/flipping. It sucked.

It is up to us engineers to figure out when these cars will lift wheels/flip over. If you know that your cg is going to be low enough to support a small track width, then good for you, you're doing something right and deserve to dominate.

[Kirk Feldkamp]

Originally posted by spierson:
I actually had to purposely drive slow during endurance to avoid a DQ/flipping.

No way! I use that excuse when I can't keep up too! It works at all levels of racing! http://fsae.com/groupee_common/emoticons/icon_wink.gif

-Kirk

[spierson]

Yea. We always let people win. In fact, it's pretty much why I joined. It makes the other teams feel good. We get a lot of karma points here, especially in our static events.

[J. Vinella]

I purely want to raise this issue. I do not mean to offend anybody or any car. I knew this would not be popular, but I have been thinking about this for a few years and would hate to hear about somebody being killed or seriously injured in FSAE due to this issue.

Kev,

I think your spec tire idea is a great solution. I see some major drawbacks that would need to be over come.

Some teams will have issues getting the tires or cannot afford them. I heard of a team once getting a tractor tire company to make their tires because of money and import issues.

Hoosier cares about Formula Ford and only lucky makes a badass tire for SAE. This might not be the case any more, but it was when I worked with them. If Goodyear made the tire they would make more money, but have little incentive or business case to improve it. This goes for any company that got the contract. Along the export lines, cost would benefit one region (it already does, but this is the nature of any product).

With no tire change or improvement the TTC would be less interesting every year and would resort to doing silly things like wet testing.

While FSAE is a series, how would SAE (International) make this choice? Not necessarily a draw back but a huge issue.

Design challenge wise it would be an interesting experiment. You would truly see the teams that got it right, without the need for budget to try different tires every year.

Chris,

Your point about the current testing technique is noted and I think is why this test has passed so many cars in the previous years. At the very least this need to be addressed this year. Now let us see how many cars pass.

[Pete Marsh]

The vehicles suspension will move into droop and raise the CoM during the tilt test, due to the jacking forces (although less than normal) and also misdirection of the 1G normal force reducing that.

So the hinged support might tip early compared to what the car will do on track. Perhaps that change alone would achieve the increased stability you are calling for, although it would penalise soft cars more, and probably need a gentle introduction so teams can still run if they fail the new way but pass the old test. Pretty hard thing to fix once your at comp, and it's not fair to change fundamental stuff like track width mid build.

I must say I'm surprised how close to the stability limit teams are intentionally running. As the outside tyre goes into positive camber, and it's performance reduces, a stable system should be possible? Or can you achieve these high CoF at 3 or 4 deg positive camber?

Pete

[MalcolmG]

Originally posted by Pete Marsh:
I must say I'm surprised how close to the stability limit teams are intentionally running. As the outside tyre goes into positive camber, and it's performance reduces, a stable system should be possible? Or can you achieve these high CoF at 3 or 4 deg positive camber?

Pete
Remember that as soon as you lift a wheel off the ground, your CG height increases, the question is whether this increases at a rate faster or slower than the grip dropping off due to the positive camber. Anyone want to do the math? http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Kevin Hayward]

In 2001 most top teams were running tracks of around 1250mm, the commonly used tyres could see peaks of 1.5g (feel free to correct that). Fast forward to 2011 and teams are reporting gs of 1.9g or higher and tracks have reduced to around 1100mm. That means that a 2011 vehicle is about 40% more likely to rollover, for the same COG. To remedy this by COG alone would mean a reduction in COG height of 90mm. As good as these cars have become they are not 90mm lower. Anyway we look at it the cars are much closer to rollover and have been marching towards that point by improvements in weight, tyre grip, and narrower tracks.

Carroll Smith predicted a potential problem with rollover due to bad corner weighting and too much caster jacking. Teams generally run less caster now than ten years ago and arguably more cars a better setup. However the rollover design point should take into account setup and caster jacking. Once one wheel is lifted the COG rises at a reasonable rate increasing rollover potential. Add that to the fact that these tyres generate maximum grip at quite low pressures and you see a reduction in track due to lateral compliance of the tyres, put in some overturning moment in the tyres and your simple calculations for rollover resistance do not cut it. There is enough going on that it is very easy for teams to get this wrong. Even the top teams who might go close to the line may have a tyre deflate during the event. I am very sure that we do not want students designing even remotely close to the rollover point of the vehicles.

It is not accurate to claim that the good teams will get it right and will design far enough away from rollover and they should be able to take advantage of their brilliance. If you look at the photos they are all highly ranked teams. There is evidence to support that the better a team is doing the more likely they are to rollover.

I have been vocal of my criticism of changing the rules in ways that limit the variation we see in FSAE vehicles, however I do not see that limiting the tyre grip (either by spec or a set of approved tyres) will do anything to stop innovation. Just about every form of motorsport has spec tyres for simple safety reasons.

In the very early days of the Australian competition (2000, 2001) we effectively had spec tyres. You could run want you wanted but tyres from Goodyear (R095s) were supplied free of charge. Most teams took that option and none complained. It would be easy enough to include a set of tyres in your competition entry fee, to be mounted at competition. That way the organisers can have a closer control on tyre performance, and order in bulk to save money. The tyre, like formula ford could be designed to run in both wet and dry conditions. This has the added advantage of reducing the costs for teams, including the reduction in the number of wheels required per car.

Kev

[NickFavazzo]

I think most teams design and set-up the car for grip levels at comp though, when it comes to running on grippy surfaces they make more G's than they are set-up for and come closer to rolling, the photos on page 1 are all on grippy tracks (with the exception being the side gusts on the wings). Teams know the grip levels of Aus and should set up for that. When you go somewhere with a better surface your car is no longer set-up for that surface.

I think having a choice of tyre is good, there are plenty of choices, engines, chassis why not tyres? Maybe having 2 spec tyres to introduce the control but still letting teams have a choice, Hoosiers and Goodyears are completely different tyres and suit teams differently. Getting a set at comp will ensure all teams get a new set for the competition, not just the teams who budget for a fresh set.

I also think that having a slicks or a wet/dry tyre only would be interesting, especially how teams respond to the changing surface conditions, although if throughout the day the track drys teams will be disadvantaged, running both a wet and dry balances that out to some extent. there would be limits on how wet the track becomes before driving has to stop.

[Derf]

Rather than having a spec tyre define a spec compound that is distributed to and utilized by all manufacturers. I don't want to lose any of the great manufacturers that continue to support the competition. This leaves open a lot of options and variety in carcass setup. Size is left open as well for teams to deal with how they want it to behave thermally. TTC would still have an important use as well in this format.

As stated before, the variety of options available is one of the most important characteristics to this competition. I believe this would establish a safety limit while not being overly restrictive.

Is such an approach viable from a tire manufacturer's perspective? I can't say.

[RANeff]

I am by no means our Tire/Suspension guy. I do feel confident enough in my understanding, however, to state that I feel a spec tire would be detrimental to FSAE as a whole. Teams *usually* pick a tire, then start designing the car. They should know the levels of grip they can generate, and design for that expectation. FSAE is all about design freedom, and tire choice is a large part of that.

This isnt your local kart track, teams are going to try and make their cars as fast and agile as possible. What that means for SAE and future rules changes... Who knows.

[PatClarke]

Kevin, you are making a lot of sense.

One option would be to specify any homologated tyre for Formula 3, including wets.
Sure there are a few diferences beween the spec tyres from Kumho, Bridgestone, Avon, Dunlop etc, but fundamentaly, they have lower grip than special hillclimb or autocross tyres because, being of a 'spec' nature, there is no need for the makers to push the limit.
This makes the tyres safer and easier to understand.
The only downside I see is they are only available in 13" diameters.

This is not my suggestion, just another option to throw in the pot.

Pat

[Mbirt]

This thread is an awesome example of the potential of this forum and easily makes up for 13 threads asking "What is optimum formula for to design chassis?"

[kcapitano]

There is a lot of talk about spec tyres here, but I think there is a better solution. Why not impose a minimum track width? Yes team with a lower CG will lose some benfits, but there a more benefts to a lower CG than just track width. From an organizatioanl stand point this would be an easy rule to implement and It wouldn't limit the teams design freedom either.

Besides, I think this argument is a bit of a moot point anyway. There are already strict rules on rollover structure, as well as a mandatory roll hoop. If a car were to roll in competition the driver would be unscathed, the car a mess, and the team would never make that mistake again. As an engineering competition we are all constantly searching for the limits, sometimes in our search we push past them. Some of you may not agree, but IMO the rules should be a tool to prevent injury in case a limit is exceeded, not to prevent us from finding those limits.

[Xeilos]

I do not agree with the spec tyres for very much to same reasons as have been stated above. There have been massive advances in FSAE-spec tyres and it is fueling innovation and development not only in our racing series but I am sure that knowledge has some effect to the tyre manufacturers themselves and carries to other classes of racing. Jersey Tom, can you provide some insight here for Goodyear?

Now I equate the entire track width/rollover argument to building anything mechanical/structural on a racecar:

If it does not break, then it was too heavy. If it does break, then you built it too light.

Similarly for track width:

If it does roll over, then it was built too narrow. If it does not roll over then you built it too wide

This again feeds into the entire purpose and intent of the rules themselves. Are they there to restrict innovation in a race class where there is no shortage of it. But nor is there too much; reliability is far more important in the grand scheme as it is difficult enough just to finish endurance (observance).

Or are the rules there to foster innovation while keeping everyone safe so that they can try again?

One other avenue, would be to go to the manufacturers and request that they decrease the performance of the tyres in some regards (ie. similar to the new Pirelli's in F1 at the request of Eccelstone/FIA) to make a narrower track width less appealing?

[Buckingham]

If a car were to roll in competition the driver would be unscathed

Unfortunately, that has never been universally true in any racing series ever.

We as humans should never stop respecting the danger present in our environment (whether it's racing, the workplace, or anywhere else). As soon as we do, we open the door for people to get injured or killed.

[Adambomb]

Just my $0.02: Making a spec-anything tends to spoil the nature of the competition a bit. One of the things about FSAE that is a blessing and a curse is the rules flexibility. You have the flexibility to do some amazing things, and also the flexibility to completely over-think things to the point that you don't get the car done or aren't prepared enough, which is an outstanding life lesson for a lot of young engineers IMO (even though it can result in, say, possibly not getting the car done, which really sucks). It really emphasizes the management side of things.

So, given our tendencies as young engineers to religiously over-think things, instead of inventing a new standard that will undoubtedly add some level of compromise to the competition, what about just simply revising the tilt-table test? What about either making it a larger angle, or something else that is representative of a car's tendency to roll?

To further discussion, I'll add a small jewel of knowledge from our old vehicle dynamics professor at ISU and a very wise man who helped pioneer computational vehicle dynamics research in the '70s, Dr. James Bernard. He has spoken to both our team and our class about the hot topic in the auto industry of rollover (he has also spent some time with the NHTSA, and has been called in as an expert witness for several large court cases), and the industry-accepted solution is to look at a simple factor:

t/(2h)

where t=track width and h=cg height. That is the acceleration, in g's, for steady state rollover. Note that actual rollover will occur before this, as mathematically steady-state cornering is not representative of real-world conditions, as has already been mentioned here by others. That simple factor has been proven to be the most reliable factor in determining rollover frequency (this is also described in more detail, with plots, in Thomas Gillespie's book Fundamentals of Vehicle Dynamics).

OK, with that bit of math out of the way, how would we measure that? Track width is easy, and has even been mentioned as a proposed solution. CG height is much trickier, especially when you start getting into the realm of systems where the driver makes up 30% of the car's total mass. The only way I can think of at the moment to measure this effect, quickly and accurately, is with, you guessed it, a tilt table test. Now if someone can think of something better than that, IMO they have earned the right to have the new standard named in their honor. http://fsae.com/groupee_common/emoticons/icon_biggrin.gif (although as a fair warning, use of that naming convention at ISU, which was copied shamelessly from realtors, has resulted in parts gaining officials names like Jon Bloq, Seahorse, and Sprogina)

So in lieu of a better solution, my vote would be to simply update the current tilt table test to something more representative of current conditions, perhaps 2g? Something to compensate for aero side-loading on wings? And in addition, there has been some discussion of the lip the tilt table that contacts the tires being too large, necessitating a hinged mechanism or something? Also worth noting, it is a fairly dangerous test, and tech officials are not immune to error, so if the solution also included something to, say, more reliably prevent a car from rolling off the tilt table like in that video I saw a while back, that would be nice too :/

Also worth noting...A quick look at sensitivities will also tell you that changes in cg height are twice as sensitive as changes in track width. On top of that, the tilt table test calls for having the tallest competition driver in the car, but I get the impression a lot of the cars 2-wheeling it in the photos were NOT doing that at competition. Cyclone Racing, like many other teams, will use their more spritely drivers (which come in at about 6' tall, 210 lbs in these corn-fed parts) at competition, but at SCCA events will have some rather tall folks drive who come in at around 6'5", 300 lbs. Something to consider...

[Zac]

IF the competition moved to a spec tire, I think the way to work it would be to have an all new tire every year, with a corresponding data package (F&M, footprint, and springrate information) released to teams in August of the year before the competition. Tire changes could even be as drastic as moving between bias and radial designs. This way teams are forced to make major revisions to their designs every year and the competition doesn't become static.

That said, I don't think moving to a spec tire really solves the problem of potential roll-over. If you give any vehicle a low enough ratio of track width to CG height then you stand a chance to flip it no matter what you set the grip level at. I know I've ended up wrong side up before while hooning a golf cart...on old and worn out tires...on wet pavement.

Come to think of it, running autocross and endurance on wet surfaces might be the ticket. Reducing the grip level this way will slow cars down (improving safety across the board), reduce the risk of roll over, increase the impact of having a good setup, AND still allow teams the choice of running whatever trackwidth or tire they want (hint: the tread pattern isn't molded into the tire).

But of all the other suggestions listed so far, I like the idea of just increasing the severity of the tilt test.

[J. Vinella]

Derf,

Call up tire companies and ask them how close they get compounds batch to batch. Methods are different from company to company. The compound is a black art. Give me a Goodyear construction with a Hoosier compound and I'll show you how to easily one can roll a car.

kcapitano,

This is the easy way out. Race track design would have to be taken into account and the cone width dramatically increased. Who wants to run 54"+ track width?

We were lucky that our driver was unhurt and we were testing on a wide go-kart track with professionals and emergency vehicles very close by. Add curbs, light poles and the like...the structural rules can only go so far.

Xelios,

Built too narrow? You are missing the point. Teams can build a car right now within the 1.7G rule and flip over. They are not going to widen the track width because their sims tell them that the track they have is already too wide. Teams need to design a safe car and this rule is not helping, but hurting.

If I am a new team and design my car to the 1.7G rule thinking this is some extreme, only to find out 6 months later that this is more like a FOS under 1, what does that say?

[J. Vinella]

Originally posted by Zac:
Come to think of it, running autocross and endurance on wet surfaces might be the ticket. Reducing the grip level this way will slow cars down (improving safety across the board), reduce the risk of roll over, increase the impact of having a good setup, AND still allow teams the choice of running whatever trackwidth or tire they want (hint: the tread pattern isn't molded into the tire).



UW would like it. It rains "9 months out of the year" here. Teams everywhere else would have to heir a water truck to do their testing.

[Derf]

J. Vinella,

Do you think a spec compound would be as soft as the current crop of compounds? Anything of this sort would ideally contain no proprietary 'black art' as it would be designed to limit performance. The manufacturers would have to agree on a simplistic compound that they could all produce. A lot less fun, pretty detrimental to the competition overall and something I brought up solely as an alternative to the draconian approach of an all out spec tyre to limit performance. As you would probably agree with, both are an improbable and unsatisfactory solution in the end.

The tyre conversation was a bit off topic, and I'm glad Adambomb brought back good points on topic. How do we change the tilt test to assure accurate measurement of the stability of the vehicle? Track is easily defined. The capability of the tilt test to accurately assess the ratio of track over cg height is hindered by the design of the test as noted previously. Increases in the angle of the test will provide a higher testing limit for enforcement, but the ambiguity of the results remains the same, if not increasing, as the testing method has not been changed.

A solution for measuring the cg height in a quick, more accurate manner with little relative increase in labor could have interesting effects beyond the safety issue here. Since total mass is already being recorded at competition, one step is done. Following the weigh in, a simple rigid pendulum setup utilizing knife edge pivots and a precise angle measuring device could quickly and accurately estimate cg height without subjecting the vehicle to excessive inclines. The vehicle would be loaded, angle measured, known mass added at known location, angle measured, mass removed, angle remeasured and finally the vehicle would be removed. Other equipment needed to hasten the process would be ramps and stabilizing jacks to steady the platform during loading and unloading. Since mobility will be needed between competitions a modular design would be needed that could be broke down and reassembled. The design judges might be interested in this data as well. In the end this is probably more complexity than what is really needed.

In the short term improvements to the test stand can bring some relief. A solid stop is still needed to keep the cars from sliding off. The portion touching the sidewall should be as minimal as possible. A taller section can still be present, but implemented in an offset fashion to 'catch' the tyre's top edge after liftoff has occurred. If memory serves me, the rig has a smooth platform that could be enhanced with either grip tape or replacement by the something similar to the aluminum extrusion seen in small equipment loading ramps. All effort should be made to minimize the load carried by the side rail obviously. Such changes simply improve the test's compliance with the written rule and do not change the requirement for this year. If any change is applied a proper announcement should be made as well.

A rigid pendulum can be seen at work, implemented by SEA Limited. Their brochure for that division of their business features an older Ohio State car on said rig. They hosted an open house for our local SAE section a couple of years ago and some of their expertise is rather applicable to this discussion. I would like to know their thoughts on the subject of what the maximum limit should be, if thoughts of increasing it are entertained. If an inquiry was made they could be an expert in helping to resolve this issue. I'm sure there are other organizations and businesses that have expertise in this field; what other experts, from your perspective, would be useful in determining this limit?

[PatClarke]

Come to think of it, running autocross and endurance on wet surfaces


Aaah, a suggestion made by the late great Carroll Smith many years ago!

Pat

[NickFavazzo]

What about introducing a penalty for side area and centroid height? All teams will be subject to it and it will put aero against non-aero, even wings against floors. This can be used to work out the side gust effect on rollover with roll centre etc.

And what about not having tilt at all? Why not make teams test the rollover of their car and provide it in the design specs? Judges in design will penalise bad teams and the tech inspectors will do some calcs before the event to make sure cars have reasonable rollover stats, then if the track conditions change the "baseline" stats can be changed accordingly, ie if a team did not meet the "dry" baseline and it rains they have the chance to meet the "wet" baseline.

or adjust their track..

[Kevin Hayward]

Adambomb,

The equation you used is incomplete as it absolutely needs to include tyre friction. Also note that COG height is not twice as sensitive as the track width.

I think Pat's idea of a homologated tyre like F3 would be a good one. The tyre testing consortium could be used to approve a tyre built to certain limits. Goodyear, Hoosier and Avon would all have compounds they already use that would suit.

That way there would still be differences in compounds and construction to fuel all of this tyre based innovation that so many people says currently exists.

Kev

[Kevin Hayward]

I would also add that raising the tilt test angle significantly would likely be very detrimental to the competition. There are hundreds of teams that would operate well away from the rollover point of their vehicles. These are teams with a lot more weight, running worse tyres and getting a lot less of the potential out of their cars.

Many of these teams would not pass a tilt test raised to account for 2g (or 2.2g if the rules were to build in some future proofing). I don't see hoe this would benefit those students, or competitions reduced from 100 potential entries to 20.

Kev

[Zac]

Do you think a spec compound would be as soft as the current crop of compounds? Anything of this sort would ideally contain no proprietary 'black art' as it would be designed to limit performance. The manufacturers would have to agree on a simplistic compound that they could all produce. A lot less fun, pretty detrimental to the competition overall and something I brought up solely as an alternative to the draconian approach of an all out spec tyre to limit performance. As you would probably agree with, both are an improbable and unsatisfactory solution in the end.

There's more to grip level of the tread than just the compound formulation itself. All of the other processes needed to take the green compound, extrude it into the desired shape, build it into a green tire, and cure that tire in a mold will influence the final physical properties of the cured tread.

Build to build variation generally isn't a big deal with production tires. But it can be tricky to get the same tire out of two different tire plants, and that's two plants run by the same company. I can only imagine the nightmare in trying to get two companies that don't particularly like each other to build tires with the "same" compound.

[Jersey Tom]

I'll chime in. To someone's earlier inquiry, I no longer work for Goodyear. Just use their products.

Interesting discussion, but I see no reason to avoid the simplest and easiest solution - gradually increase the 'severity' of the tilt test. To Kevin's concerns about reducing the potential of the sub-limit cars... raising the requirement is going to affect all vehicles. If your current car is borderline for passing a raised spec... build a wider track width. Simple. If it makes the whole field a tenth or two slower on average on a cone course... so what? Might even make skidpad a bit faster.

To the idea of spec tires... wouldn't be that bad. Really don't think it would limit creativity at all. On the other hand, it's really not going to save any appreciable cost. Won't change the unit cost of the tires from the suppliers. Don't think a lot of teams do much for back-to-back tire testing. Those who do... if I had the money to do back to back testing, I'd just use that funding elsewhere (like on another set of tires for driver training, testing and tuning).

Talk of control compounds and stuff like that... super impractical IMO. I'm sure there's IP issues in sharing compound booksheets... companies might have to go get materials they don't often use... all for a very very small market. FSAE teams are extremely fortunate to have the level of support they do from tire suppliers. I wouldn't try stretching that.

Doing something to knock grip levels down at competition seems un-necessarily difficult and impractical as well... having to consistently water down an event area that isn't set up for it.

Just one man's opinion here... but if there's serious roll-over concern, just raise the tilt test angle a bit. Seems like a no-brainer to me.

[spierson]

Maybe there should be a blurb in the rules that tells teams to check their track width against their cg height. It's not Hoosier's fault, it's not SAE's fault, and I'm not going to rally for a rule to slow the cars down, we aren't going THAT fast.

We as engineers need to take responsibility for our designs, it's not SAE's job to cover our asses on everything. This scenario happens in the real world, and there are some things that you do need to protect your user from. If you aren't checking weight transfers before you go out driving, you're putting your driver and car at risk.

Right now, we all know for a fact that we will be seeing higher g's at Lincoln. Design your car to not flip at these higher g's, not at 1.7g. One reason we pass the 1.7g rule so easily is because we know for a fact we'll be hitting more than 1.7g on the track.

My biggest issue with the minimum 1.7g/minimum track width rules are that people like to design to minimums even if they shouln't. Just because you can doesn't mean you should.

[Jersey Tom]

Right now, we all know for a fact that we will be seeing higher g's at Lincoln.

I'd be careful with that mentality.

[J. Vinella]

To organize the thoughts here is what I see can be changed and ideas and repercussions:

Tilt test spec:
The more I think about it, should probably be somewhere around 2.8-3.0G

Min track width:
This would have repercussions on track cone width and for some teams facilities issues. I know for UW we are ok getting our car in and out of the building, but other teams have to design the car differently to maneuver. Wanna sell all our trailers, take tires off, or tilt the cars at an angle for transport?

Tires:
Min pressure.
Min width.
Start carving straight grooves in.
Always run on wet tires with a harder compound.
Spec tire.

[Pete Marsh]

I don't think drastic changes are needed here, and I don't really see how any reduction in grip prevents teams building cars too close to the roll over limit. You can simply make it narrower, as you would with lower LLT from lower grip, and more importance on fitting through the tight stuff.

I can see room for improvement in the tilt test, by hinging the side support beam so the point of rotation is properly controlled.

Why not simply make it clear to teams in the rules that they are expected to consider the roll over stability of their cars and build them accordingly. At the moment there is nothing to guide a team who's risk assessment of a 2 wheeling car is different from those who are calling for change here today. As the rules stand if you think you can get your inside wheels to clear the cones you can cut the corners! How about an addition to the vehicle integrity rule D1.1.2

"......dangerous handling, eg lifting two wheels simultaneously...........valid reason for exclusion"

I think, if enforced, this would provide sufficient motivation for this issue to go away. No one wants to be excluded, or have to slow down to avoid it.

Pete

[Adambomb]

Originally posted by Kevin Hayward:
Adambomb,

The equation you used is incomplete as it absolutely needs to include tyre friction. Also note that COG height is not twice as sensitive as the track width.
Kev

No, tire friction is very intentionally not included. This isn't any sort of "simulation" equation, it's an incredibly basic statics "slip or tip" equation. Slip is tire friction, tip is t/(2h). You have one or the other. If you have slip, no problem (assuming you don't slip right into something). Tip is the AMOUNT of tire friction (or other sources of lateral acceleration that may include things other than tire friction) that cause things to go very badly.

Again, this isn't even my equation, this is the factor spouted off in books about FUNDAMENTALS of vehicle dynamics, and also chosen by the NHTSA. And if you have "t" by its lonesome in the numerator, and 2x "h" in the denominator, how is "h" not twice as sensitive?

[Kevin Hayward]

Adambomb,

You could rewrite it as:

rollover resistance = k * t/h, where k is a constant. Rollover resistance is proportional to t/h.

If h is 20mm then reducing it by 10mm will double the rollover resistance. however if h is 100mm then reducing it by 10mm will only increase rollover resistance by 10%. In the real case of the car any small change to h is worth more than a small change in t, but not as a ratio of 2:1.

As a % change both the track and COG height have a similar effect. Double the track and you double the rollover resistance, halving the COG height will have the same effect.

You essentially have in your equation the % of lateral load transfer, which is complete if you add the achievable Ay which is primarily defined by the friction of the tyres.

To intentionally not include the tyre friction in this case is silly, even if the NHTSA doesn't. We are talking about a situation where rollover is being caused by the reaction to lateral tyre forces.

[Edward M. Kasprzak]

The equation in question is AY = t/2h, where AY is the lateral acceleration where the inside wheels carry no load.

You can derive this from the formulas on page 678 of RCVD. The "2" comes from having to transfer half the vehicle weight from the inside wheels to the outside wheels.

It does not indicate that the solution is twice as sensitive to CG height as to track width. It also has nothing to do with tire/road friction coefficient.

[coleasterling]

Here's a question...What percentage of teams are consistently lifting both inner wheels at official FSAE competitions? I would bet it is well under 10. The simple answer to this "problem" is that your team designs a car that is stable at the levels of grip you are expecting. Most teams aren't anywhere close to the examples given here.

[Adambomb]

Kevin...Dr. Kasperzak has explained this better than I did. This isn't some random "rollover resistance factor" we're pulling out of thin air, there is no need for some arbitrary "k" constant, and the 2 in the denominator does in fact serve a purpose. t/2h is the value of Ay where the inside tires come off the ground. To re-derive this, check out RCVD as Dr. Kasperzak has mentioned and "plug and chug," or simply just do a FBD on a car in quasi-static cornering at the point where the inside tire force goes to zero, then solve for AY. THAT is where t/(2h) comes from. Again, there is an entire chapter on rollover in Thomas Gillespie's book "Fundamentals of Vehicle Dynamics." FWIW, Gillespie's book is focused a little more on the passenger car side of things, where, of little surprise, rollover is a pretty hot topic.

Tire friction contributes to AY. But AY can also be caused by, say, side-loading on wings, other odd disturbances in aero, etc. Although tire friction is, as you said, at a first-level guess, a good predictor of maximum AY, there may be other factors involved.

~~~~~~~~~~

I suppose there are two distinct sides of the discussion:
1) Limit performance by keeping AY small through tire friction so that AY does not exceed t/(2h)
2) Increase rollover resistance by keeping t/(2h) larger than AY.

Which sounds more enticing? Which sounds more reliable? Which sounds easier to implement, measure, and enforce?

~~~~~~~~~~~~~



Originally posted by coleasterling:
Here's a question...What percentage of teams are consistently lifting both inner wheels at official FSAE competitions? I would bet it is well under 10. The simple answer to this "problem" is that your team designs a car that is stable at the levels of grip you are expecting. Most teams aren't anywhere close to the examples given here.

I've never seen it happen myself, so I'd say it's pretty rare. Now I pose a second question: How many rollovers at competition can be considered acceptable? What would happen if an FSAE car rolled at a location other than competition, such as in a university parking lot or at an SCCA event?

[MalcolmG]

I'm really not buying this "teams should design the cars so they don't lift both wheels" stuff. It's not that I don't agree with it, but by the same argument you could say teams should design cars that have sufficient braking capacity, yet there is still a defined rule and a test to ensure they do so.

At the moment it looks like the solutions presented are:
1 - Ignore it
2 - Increase the angle for tilt test
3 - Include a rule that allows for disqualification if a team is lifting wheels
4 - Specify a minimum track width
5 - Run events on wet surfaces
6 - Have a spec tyre (or homologated F3 tyres)

Assuming there's actually a desire to address this issue, I think #1 can be ruled out.

As Kevin has pointed out, #2 could unfairly penalise teams who are simply not at the level of exceeding the existing 1.7G limit, potentially forcing them to significantly greater track widths than can be competitive in FSAE, while also placing greater importance on achieving a low CoG

#3 could work, but I would be concerned about how it may be enforced. You end up with the unenviable situation where you essentially have a rule which isn't checked for compliance until it's too late to do anything about it. Imagine the uproar if a front-running team got black flagged and told to slow down halfway through endurance because the surface was allowing higher lateral Gs than they had planned for/tested in? It also doesn't address the possibility that a car could conceivably roll before being stopped

#4 seems somewhat viable, but it certainly runs the risk of constraining design options. If you were going to specify a track width that would be a safe minimum for all cars, you'd be looking at probably 1200mm or more. I know at least one team whose entire design philosophy would have to be thrown out the door if this happened (which makes it 100% of teams whose design philosophies I'm intimately familiar with http://fsae.com/groupee_common/emoticons/icon_wink.gif ).

#5 is an interesting idea, it certainly addresses the issue well and creates some new design challenges. The logistics of implementing it possibly make it unfeasible though? Imagine the water consumption if teams had to do all their testing in wet conditions!

So we come back to tyres. The more I think about a spec tyre, the more I'm liking it.
- Supplier could be put up for tender every 3 or 4 years.
- Various attributes could be specified by SAE, such as maximum coefficient of friction on the Calspan belt
- Has the potential to reduce cost if the tyre manufacturer could use a lower cost compound due to lower grip requirements
- Can also lower cost by improving wear characteristics of the tyres so they last longer
- SAE could specify that manufacturers provide both a 10" and a 13" tyre option, probably with the same compound and possibly with a couple of options for width
- Perhaps you could also add the option of using a street-legal tyre, so teams with very small budgets could still use cheap tyres

I think the last point could be quite useful, as it would at least provide an option so teams could consider which size would best suit their needs, and allow some more design freedom. At present I hardly think there's so much variety in tyre options and any real association between a particular tyre and a particular vehicle concept that if teams were restricted to a single option that suddenly the cars would all become cookie cutters.

I guess the big question is how feasible it would be to get a tyre manufacture to commit to production of a spec tyre - the guaranteed sales would undoubtedly be a nice carrot but would they be able to churn out enough to meet the demand? Would the same compound offered in two diameters and two different widths be a realistic request, and would the performance be similar enough that one wasn't clearly better than the others?

[Kevin Hayward]

Adambomb,

The only reason I pulled the 2 out in a constant was to show that the COG is not twice as sensitive as track. We have certainly been talking about the same condition, which is 100% lateral load transfer due to cornering forces.

Unfortunately (as mentioned previously) it is too inaccurate to apply statically. With a car track of 1100mm and a COG height of 250mm 100% Lateral load transfer should occur at 2.2g. I have seen a car with these dimensions up on two wheels well before 2.2g. This is due to a dynamic increase in COG height, tyre deformation etc. The error is too high to use the simple equation as anything but a very rough approximation.

There has been quite a lot of talk on these forums about the differences between the friction of tyres found on the test rig vs. the friction found at the track. They are quite noticeably different, additionally the tyre companies have new compounds released regularly that teams use before testing. With these facts it it not possible for teams to determine an accurate Ay before on car testing of the tyres. Even when that testing is done it will not be done on all representative surfaces. It is very possible to unintentionally design a car that will lift two inside tyres.

Kev

[Jersey Tom]

Originally posted by MalcolmG:
I guess the big question is how feasible it would be to get a tyre manufacture to commit to production of a spec tyre - the guaranteed sales would undoubtedly be a nice carrot but would they be able to churn out enough to meet the demand?

Again not really my place to say anything from the position of a tire manufacturer, and my opinions are in no way associated with my previous or current employers.

Consider this however: Total tire consumption for the Formula SAE events (literally those sanctioned by SAE within the US) is measured in what.. some thousands of units in a year? Order of magnitude approximation.

A pro racing series - I'll use NASCAR as an example - goes through several (3-10?) thousand tires a weekend, and there's a race every weekend from February through November.

A single "normal" tire factory is capable of producing quantities measured in the tens of thousands per day.

Point is, the carrot you propose to dangle is a very small one. This goes back to my earlier opinion that the level of support FSAE gets from tire companies is incredible. They're not there to turn a profit. So the proposals of spec tires, requiring multiple widths and sizes... for the size of market here those are big requests IMO.

On the assumption that there is an actual need here and there is a high potential for rollovers at future events... I still don't see why raising the tilt test angle isn't the most logical choice. Nor do I understand how this would punish some teams but not others. IMO it affects everyone equally.

[Pete Marsh]

What is it about these teams that "can't achieve 1.7G" and would be disadvantaged by an improved tilt test?
This is all due to better tyres in recent times, wont they fit on these team's cars? If teams are so inept they need new rules to stop them hurting themselves when trying out a new tyre, or any other feature that might increase Ay, how can you be so sure "these less than 1.7 teams" won't try the new tyre, or damper etc, one day and hurt themselves. IMO such a change is MORE relevant to a poor performing team that has the potential for a massive improvement that was not foreseen in the basic layout of the design.

I appreciate the rules have a role to play in participant safety, but I live in a "nanny state", and I can tell you, despite plenty of rules, stupid still hurts.

Many well intentioned changes often have unforeseen effects that create a new problem, sometimes worse than the original problem. I can see that happening here. In fact this very issue is a result of smaller cars with better fuel economy not wanting to slow down for the tighter sections of the track. IMO the best strategy is to try to remove the advantage/reward for the behaviour you don't like.

In this case, cars designed and built at or near their roll over stability limit.
This is done because a narrow car takes a faster line in tight chicanes and slaloms.
That is good because you don't need to burn fuel to speed up again. (and it's faster)
They were put there to slow down the cars, especially at venues that can't do long curves.

Items 5 and 6 on Malcolm's list don't address the motivation individually, and 4 does not ensure stability,although 4 and 6 together might.

Pete

[Luke Phersson]

Here's a bit of a diagram I just knocked up to help people understand the calculation:
http://img830.imageshack.us/img830/471/rolloverg.jpg

Edit: ^Wheelbase = Track..

T you could take as track width, although it may be more or less depending on the stiffness of the sidewalls. As you can see I've incorporated downforce(DF) into the calculations, which means Mu =/= G's. Downforce is irrelevant for rollover calculations as the downwards restoring force is negated by the greater potential lateral force generation, however you could argue it reduces the potential Mu due to load sensitivity.
So say we have an average track of 1200mm and a CG of 320mm to be conservative, that requires the tyres to give a Mu of 1.875 to get 100% LLT (rollover initiated). Having driven the Monash car on the same day as when the pic on the first page was taken, it wasn't particularly scary/difficult to get the car back on 4 wheels again, just a small steering input with plenty of time to react - as the LLT was fairly gradual (tyres don't enjoy binary load increases).


Another interesting thing, typically you will see an aerodynamic restoring moment due to the chassis slip angle relative to the velocity vector - probably negligible on non-aero cars, but with big endplates we've got data showing we're getting 50Nm and 110Nm restoring moment at 5 deg and 10 deg yaw at 60km/h.

This model is pretty simplistic as it's assuming your CG height doesn't change as you roll (I guess you could say CG height @ rollover), or ride up/down from track irregularities and compliance. I know we've seen (as Steve pointed out earlier) increases in ride height of 15mm+ with incorrect damping. We've also encountered average ride height increasing with roll angle from high roll centers and jacking forces.
You also have to consider what happens when the inside wheels hit a cone and you get a force impulse, or the outside tyres hit something protruding from the ground. I think these factors are more dangerous than just 'lots of grip' as they're sudden impulses with little reaction time.

I'm still a little dubious of people achieving Mu's of ~2 (Or G's with no Aero), although I acknowledge it could be possible on an EXTREMELY well rubbered surface. Is this a sustained G trace (0.5s +) or a spike? I agree you can see spikes over 2G. Also, if your G-sensor isn't located on your CG, do you correct for yaw accelerations, as this will read as extra lateral acceleration? I'd expect people to be getting faster skid-pad times from these highly refined geometries.. I think Monash may have the fastest skidpad time of 4.8s (http://www.natsoft.com.au/cgi-bin/results.cgi?12/12/2010.FSAE.P2) at a competition (bar Cornell '88 sucker car), and that's only 1.6G with significant downforce. Perhaps other cars work a lot better with hot tyres.

All of that said, I do agree that cars are approaching the roll-over limit and something needs to be done, I think increasing the tilt test to simulate a Mu (Not G!) of 2 would be sufficient - an increase of 0.3 lateral coefficient is huge.
In regards to the spec tyre, and along the lines of what Pete was talking about, you could have a tyre that is extremely camber sensitive so you essentially create a stable system such that when you have (or approach) 100% LLT the outside tyre coefficient will drop off due to excessive camber, bringing the inside tyres back down.

Luke.

[Adambomb]

Kevin, I see. I know what you're talking about with it not being a "gospel" number on its own, in fact if you go back to my first long-winded post on the subject that is something I bring up. Transients will always push that number lower. However, even if you can't count on a vehicle being stable in roll all the way up to t/2h, it is still, statistically speaking, a good predictor, even if you know that it is only a theoretical maximum. I'm taking a guess that's where the old standard came from: as someone else posted, back in the olden days a professor wrote a paper that FSAE cars would never go over 1.3g, so the tilt table test was set at "something safe," like 1.7g. You could try to scale it by a factor I suppose, but I would not be so quick to just toss the 2 out of the denominator...

OK, back to sensitivities: My first guess, just looking at the 2 in the denominator, was clearly incomplete. So instead, I decided to whip up some plots in Excel, one showing AY as a function of h for a track width of 48 in. and h from 8.5-13 in., and one that shows AY as a function of t for an h of 10.5 with t going from 42-52 in.

Given that h is already much smaller than t, when you look at a range of reasonable values of h and t for FSAE, dAY/dh is actually more like 5 times as large as dAY/dt. I would interpret that as saying that h is much more worrisome than t, so IMO minimum track width won't be as reliable a predictor, especially when taking into account the variations in h that can occur by switching drivers. Just looking at track width just won't be as effective, but that is no problem, as the current tilt table test already looks at both t and h in their most representative form, and in a cheap, fast, simple, repeatable manner.

My resistance to the "limit the performance with a spec tire" strategy stems from the fact that I fear that mandating a single tire would add a good deal of politics to the competition, because let's face it, it would be quite a powerful decision. And the minarchist in me would rather not see the competition become susceptible to the sort of bias and corruption that is associated with those kinds of politics. If you specify a maximum coefficient of friction on the Calspan belt, now you've effectively disqualified any tire that hasn't seen TTC testing. That doesn't sound healthy for the spirit of the competition at all IMO. And just like Jersey Tom said, the profit tire companies see from FSAE involvement doesn't come from tire sales, it comes from buying the respect and good will of future racing professionals. I'm sure Goodyear's program is a tough sell to the bean counters at the top. Add certification to the mix and you just might lose them!

Also, aside from the politics involved, you have also added cost in that not only will tire suppliers have to perform another level of certification to show that their tire meets SAE standards, but you will also have administrative overhead at the SAE level as they not only develop a new standard, but also routinely check up on it, do a bit of testing on their own to determine that it is representative and is being upheld, etc.

Finally, if you do make a spec tire, what is Joe Racer going to do? Well the first thing I would do is prep the hell out of it to make it stickier. If they gave me a fresh set at competition I would have to be somewhat sneaky about it. Then as soon as comp. is done I'd throw a set of R25Bs on it. And then, as I've asked before, what happens if we roll a car in the university lot (not likely in ours I know, our lots suck...but still a possibility), or again, what happens if the car rolls at an autocross on a nice sticky airstrip with much higher grip levels than seen at MI? Have to think beyond competition...

Or I suppose you could make a spec 175/70R13 street tire. While you're at it I suppose you could also add a spec 800cc Suzuki engine. That way our cars would be eligible for SAE Supra India as well!

Or you could just slightly modify the tilt table test.

[Kevin Hayward]

Adambomb,

I never suggested that h was less sensitive than t in this case, merely that it wasn't twice as sensitive, nor was I saying the 2 should go. I am well aware that a linear change in COG height is worth more than a linear change in track for the given ranges that we work with. That being said I think we both have a handle on the lateral load transfer calculation, and anyone else who has been watching this thread.

Also as suggested by Pat the spec does not have to be a single tyre it could be any one of a set of homologated tyres. There would be no need at all to create a tyre that already doesn't exist. A vast majority of teams run either a Goodyear, Hoosier, or Avon. All of these companies already produce tyres with lower available grip than those commonly used at FSAE.

With data already available from the TTC, tyre companies will not need to do any other level of certification and organisers just have to check the markings on the tyres at comp. The homologated tyres list could be maintained almost indefinitely without modification. Tyre treatments are already illegal, and teams may be breaking that rule now for all we know. It will however be easy to spot if teams on tyres of a known performance level are suddenly pulling an extra g on the skidpad. There are ways to implement these sorts of rules sensibly as has been done in many racing series.

As for thinking beyond competition we see all sorts of stupid things being done by teams. restrictors removed, safety gear not being worn, ramming cars into tractors to test impact attenuators. I do not expect SAE to be able to enforce safety outside of the competitions, I do however expect that the competition itself to be safe. You can bet that if in ten years power had increased by an average of 40% that we would have seen another restrictor size decrease.

As mentioned in my very first post in this topic I think that without a spec tyre there should an increase to the tilt test angle and possibly a mandated minimum track.

Pete is spot on about the unintended effects of rules. The layout of tracks is meant to lower speeds of the cars to improve safety. Unfortunately it creates situations that promote the onset of rollover, fast changes in direction and repeated cornering increasing the effects of jacking. However in this case I am struggling to figure out how the unforseen problems of returning grip levels to what they once were will be worse than potential rollovers. I would be very interested to read about the potential issues though.

I would still like to pose the questions from the first page:

From a safety perspective should there a limit on cornering capacity of these cars? (in the same vein that we limit power, downforce, and average speed)

If so what should the limit be? 1.7, 2.0, 3.0, 5.0g?

Kev

[NickFavazzo]

Originally posted by coleasterling:
Here's a question...What percentage of teams are consistently lifting both inner wheels at official FSAE competitions? I would bet it is well under 10. The simple answer to this "problem" is that your team designs a car that is stable at the levels of grip you are expecting. Most teams aren't anywhere close to the examples given here.

I think Coleasterling's thought are spot on, I don't know of any rollovers happening at comp but what about a whole corner breaking off? Brakes failing? Wishbones breaking?

These can be bad engineering design, or from effects outside our control, for example, running aluminium springs, bad design, or a massive side gust, more than would be reasonble expected, rolling a vehicle over.

What about running the static events after dynamics? Big change up but then the Judges can see cars that bicycle and speak to the teams about that, the penalty will be in the design, where the error is made.

#5 would be difficult to make viable as it is water. Runs away, evaporates etc.

#6 will be hard to implement, and I think the support we get from tyre manufacturers is great not only for the competition but also for us students.

#4 will help prevent rollover, but then Aus becomes a very slow track, in order to get the average speed goals there will be fewer cones Aus is already a tight track.

#1 is not really fixing a problem, although It has not been a problem yet...

#2 increasing the angle of tilt teams will not effect the target group, those that were near the limit will redesign their next vehicle to be closer to another limit, is this competition not about pushing the limits?

#3 I think this rule could be harsh (how do you see if wheels are up by say 5mm?) It would surely make teams sure their car does not roll though

[sbrenaman]

Set a "t/2h" minimum of X in the rules. Ballast the cars to meet that minimum at comp.

Who knows, maybe the ticket will be a 200# car w/a narrow track and an absurd amount of ballast and a 3" CG height :P

[J. Vinella]

Originally posted by Jersey Tom:
Point is, the carrot you propose to dangle is a very small one.

I agree with most of your point, but if Goodyear can justify making specific FSAE tires I bet having the opportunity to have a monopoly on a series, for tires they already make, is not that hard to sell.


Originally posted by Luke Phersson:
Downforce is irrelevant for rollover calculations as the downwards restoring force is negated by the greater potential lateral force generation...

I'm still a little dubious of people achieving Mu's of ~2 (Or G's with no Aero)

I think Monash may have the fastest skidpad time of 4.8s at a competition (bar Cornell '88 sucker car), and that's only 1.6G with significant downforce. Perhaps other cars work a lot better with hot tyres.


Not correct. Look at a plot of Mu vs normal force for a tire.

See Page #1, Picture #1. It is your car, what was the G level there? Do you correct for roll at that angle?

1.6G with no transients, cold tires, and a small radius....and you are not making well over 2G...really?

Kev,

If we would have rolled in '09 I think things would be very different now. Me and the other endurance driver came very close. I don't think the comp is safe.

[Pete Marsh]

@ Jack


If we would have rolled in '09 I think things would be very different now. Me and the other endurance driver came very close. I don't think the comp is safe.



So fess up, what was the process that led to the dangerous car at comp?

Design oversight? - Did you know before enduro? Why not add ballast until stable. Maybe remove the wings?

Intentionally on the limit? - Did you know before comp? Would you have risked it if it meant exclusion?

Why keep pushing? - How did the second driver not see the danger and slow down? Did the desire for maximum performance outweigh the risk, even if scary?

What is your comment on addressing motivation rather than trying to control solutions? How would the proposed ideas have played out in your '09 team?

Pete

[J. Vinella]

Pete,

The process was years of understanding of tire data. I focused on race engineering and tuning and synthesized many smart people's ideas into our car. I don't mean to sound uppity or elitist, if I am coming across like that it is my poor choice in words. People who know me on this forum, know that I am very humble.

We were intentionally on the limit. I would not have risked it if it meant exclusion. If there were a rule saying no lifting of inside tires, like no dragging body panels I probably would have made the track a much wider, or tried to lower the CG more. By there were very good design choices made for the CG and track.

Being an engineer I looked at all our data and knew we were within the 1.7G rule and could not go wider on the track, as I was running 2" more than I wanted to anyway.

We caught glimpses of lifting before comp at the same kart rack we rolled. The remaining weeks before comp we tested on our ice rink of a football parking lot. Tuning at these low levels of grip helped and when we got back on sticky tarmac the grip came way back. Take into account hotter track surfaces and a few more variables and we started lifting.

Why keep pushing? Our mentality has always been to go for it at comp. Looking back I am happy we did, I would rather know that we tried then regret lifting and wondering what if. Some might say young and foolish, but I think we were within reason for the situation.

Like I said before, the car has not been run again. It probably could but would need to be de-tuned.

I think telling the team we need to back way off because we made a good car within the rules but dangerous would not be a good motivator. Designing to 1.7G is crazy these days, but for an engineer designing to 2.5G when the rule is 1.7G is crazy too.

The team would have been fine with many of the above solutions to this problem.

I think a spec tire is a good idea, but logistically very hard to execute. Keeping everybody safe is better than compromising so the back of the pack can go faster.

I started this tread to prompt a good discussion and so far so good...

[theTTshark]

Originally posted by J. Vinella:
If we would have rolled in '09 I think things would be very different now. Me and the other endurance driver came very close. I don't think the comp is safe.

The competition is not unsafe. Your design was. You don't need to lift the inside wheels to be fast. We placed 1st at Michigan, 2nd at California, two 1st and a 2nd at UTA, and 2nd at Solo Nationals (all autocross placings of course) and we have never lifted both inside wheels. Inside fronts? Yes, but never unloaded the inside. So does that mean that cars have to be designed on the "ragged edge"? I highly doubt it.

Change the tilt test. No need to implement spec tires or anything, just increase the safety standards. If you pass then you pass, if you don't then you can't race. Pretty simple.

[woodsy96]

Interesting discussion.

It seems to me that an increase in the tilt test is the easiest way to go. It might be a bit of a patch fix but it will reduce major concerns. In two years time review the rule and if it is still an issue, raise it again.

As for less teams on the track : If teams fail tilt test at comp - bolt lead to the bottom of the car. You immediately get a penalty from increased mass (slower acceleration, turning, braking, more fuel etc), and you still get to compete.

[J. Vinella]

Originally posted by theTTshark:
Your design was (unsafe).

I totally agree and I see others approaching the same design solutions.

[Luke Phersson]

Originally posted by J. Vinella:
Not correct. Look at a plot of Mu vs normal force for a tire.


Originally posted by Luke Phersson:
Downforce is irrelevant for rollover calculations as the downwards restoring force is negated by the greater potential lateral force generation, however you could argue it reduces the potential Mu due to load sensitivity.



Originally posted by J. Vinella:
See Page #1, Picture #1. It is your car, what was the G level there? Do you correct for roll at that angle?

The G's spiked to about 1.9, ignoring roll angle.
Typically we don't bother accounting for roll angle as we run a pretty high roll gradient - 0.86% error at 2G cornering.
As for not accounting for yaw acceleration, here is a random slalom from a testing day... I've seen much higher yaw acceleration than this, too.
http://img836.imageshack.us/img836/5921/angularaccel.jpg
Blue is the uncorrected Lat G trace from a YRS3.
Yellow is the uncorrected Lat G trace from another G-sensor nearly on the vehicle CG.
Red is the corrected G-trace from the YRS3.
More than 0.86% error.


Originally posted by J. Vinella:
1.6G with no transients, cold tires, and a small radius....and you are not making well over 2G...really?


We can pull well over 2G. We can not get a friction coefficient over 2. Big difference.

I think you might be missing my point. If you take Mu as the driver for rollover, you remove the aero variable. You can't say the car will have 100% LLT at xx G's - you CAN say the car will have 100% LLT at xx G's and yy m/s. You CAN say the vehicle will roll if you achieve a Mu of xx.

My previous post is deriving exactly what Dr. Kasprzak and Adambomb are talking about, except with provisions for downforce. The controlling factor is what coefficient you can get out've your tyres - not a lateral acceleration. Without aero, they are essentially synonymous.

Example:
M=260kg; CG=0.31m; V=31.4m/s; T=1.075m; Mu=1.5; CLA=5.2;
You have enough grip for a lateral acceleration of 3.34 G. 93% LLT - no rollover.
M=260kg; CG=0.31m; V=10m/s; T=1.075m; Mu=1.8; CLA=5.2;
You have enough grip for a lateral acceleration of 2.02 G. 102% LLT - rollover.

Identical weight, identical track width, identical CG height.
The scenario with a higher lateral acceleration (3.34G) is not getting 100% LLT, where the scenario with only 2.02G lateral accel is.
So what's different? The friction coefficient and velocity.

I'm by no means an expert and others here have a lot more experience than I - look at the diagram I posted on the previous page from first principles. It's relatively simple math and I may have made a mistake, if so please correct me.

Luke.

[J. Vinella]

I see your point Luke.

What type of accel are you using? What are you sampling at? The G chart looks very smooth. Did you apply a filter? Yaw looks very rough.

[Luke Phersson]

The lateral accel is in G's, with the yellow curve from a 3-axis Crossbow accelerometer, which is soft mounted with rubber and unfiltered in i2 (I think the crossbow might have some internal filters?). The other lines are from a Bosch YRS3 sensor (giving lat, long, yaw rate and yaw accel) which is hard mounted and gets a lot of vibration (will add rubber mounts in future). The YRS3 lat G has a 5 sample filter on it to get rid of some of the vibration noise. I think G's are logging at 50Hz.

Yaw accel is in deg/s/s, and logging at 100Hz. Yeah, the yaw accel looks a bit rough, unsure if it's because it's a pure digital signal or just logging too slow - we have to balance the logging rates a bit as we run out've memory easily (lots of data aq. at the moment).

Edit: Turns out it was just a funny up-sampling setting on the data display for the yaw accel.

Luke.

[Adambomb]

Originally posted by Kevin Hayward:

If so what should the limit be? 1.7, 2.0, 3.0, 5.0g?

Kev

I had to think about this a good long time...although now that I have a solution, I feel embarrassed it took so long to come to.

Basically, how fast can you go? Personally, if you can keep the shiny side up, my thoughts are the limits on cornering should be based on physiological factors, like for example when your retinas start detaching from your eyeballs. http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

[Tom Wettenhall]

First up, watering the track will never EVER fly in Australia, we tend to take a dim view of dumping megalitres of water on to things which aren't going to grow. Tarmac doesn't grow no matter how much you water it. Can't speak for elsewhere. Besides, you leave yourself open for someone to go running on a dry day and have a spectacular accident.

I have to disagree with spec tyres. The examples given of rules to slow the cars down don't rob teams of avenues of design. Spec tyres do, and in a big way. I think that, as an engineering design competition, FSAE needs to keep as many options open as possible because it promotes teams actually having a close look at what's available and how different options are going to affect performance. Engineering in other words.

The way I see it, what we're trying to assess is whether a given car is going to slide or roll. Increasing the tilt angle will make it safe, but doesn't take in to account cars slipping before they roll. So: Rope the car to the top of the tilt table, take the lip off and see if it tips or drops. Destructive, dangerous and terrifies the daylights out of the driver all in one!

[flavorPacket]

Originally posted by Tom Wettenhall:
So: Rope the car to the top of the tilt table, take the lip off and see if it tips or drops.

And how do you plan to heat up the tires on the rig?

[Adambomb]

Originally posted by flavorPacket:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Tom Wettenhall:
So: Rope the car to the top of the tilt table, take the lip off and see if it tips or drops.

And how do you plan to heat up the tires on the rig? </div></BLOCKQUOTE>


...and why would it matter if it slipped? Would you then fail tilt test for not maintaining mechanical grip to the tilt table? (not to mention cold tires as flavorPacket mentioned)

[Tom Wettenhall]

Give the inspectors tyre warmers and concrete the base of the table? If it slipped you'd pass, as it'd be deemed incapable of rolling. I wasn't actually being serious... For instance, a car which passed this test would still flip happily over a bump.

[J. Vinella]

Originally posted by Luke Phersson:
The lateral accel is in G's, with the yellow curve from a 3-axis Crossbow accelerometer, which is soft mounted with rubber and unfiltered in i2 (I think the crossbow might have some internal filters?). The other lines are from a Bosch YRS3 sensor (giving lat, long, yaw rate and yaw accel) which is hard mounted and gets a lot of vibration (will add rubber mounts in future). The YRS3 lat G has a 5 sample filter on it to get rid of some of the vibration noise. I think G's are logging at 50Hz.

Yaw accel is in deg/s/s, and logging at 100Hz. Yeah, the yaw accel looks a bit rough, unsure if it's because it's a pure digital signal or just logging too slow - we have to balance the logging rates a bit as we run out've memory easily (lots of data aq. at the moment).

Luke.

Those sample rates are pretty low. For questionably the most important piece of instrumentation I would think you would want to have a dedicated analog channel and sample at like 300-400 Hz. You can get much more out of your CG accel if you sample it higher even, tune dampers and the like. Do you have pots on the dampers? Is that how you are tuning your dampers? What are you sampling those at. The yaw looks nothing like 100Hz more like 1Hz.

Don't let engine dictate your sample rates. I bet they are measuring water temp and other nonsense at very high rates. Your set up has promise but I have many questions. The data is only as good as what you can see. You are hiding some real gems with those sample rates.

Back to the topic at hand.

[mech5496]

Well, definitely an interesting discussion here... Increasing the tilt angle by a few degrees to simulate 2G's won't hurt and it's pretty easy change to implement. However I think that discussing the matter of rolling over has gone too far. After all, that's why we have roll hoops for, right? If a team is afraid of rolling over, they can do so by numerous ways (increasing track, lowering their CG, driving slower). Introducing spec tires or a minimum track, or even a rule regarding DQ for lifting wheels to "make comp safer" is simply not the way to do so! Increasing the minimum requirements for the roll hoops and crash zones may do so...even though incidents like the Lions crash back in 2008 proves they are adequate! After all if you want to prevent any risk, just stop all dynamic events and keep only the statics. That could bring the cost and environmental impact down too.....

[woodsy96]

However I think that discussing the matter of rolling over has gone too far. After all, that's why we have roll hoops for, right?

Thats a fantastic approach. I'll remind INDYCAR that next time safety issues are raised that they should stop talking and get on with racing. After all, they have roll hoops for when they crash, right?

[/sarcasm]

RIP Dan Wheldon.

Also, the comment "If a team is afraid of rolling over" is ridculous too. Teams that push to the limit may occassionally cross it, and some newer team may not understand/ conisder that a possibility or its implications - not just on the driver/spectators, but on the competition and project as a whole.

[mech5496]

...and it is exactly that Dan Wheldon incident that got me thinking! That roll hoop should not be able to collapse in such a crash. How many times have you seen an F1 hoop collapsing in the recent history for instance?! The last one I remember was Pedro Diniz crash back in 99, which caused some rule changes... I didn't say to ignore safety, I just think that we should concentrate on making cars safer and not slower....

[Tom Wettenhall]

If you try and make the cars safer, most of the teams will agree with you. If you try and make the cars slower, teams will fight every step to find a way of going even faster.

For instance, cutting grooves in F1 tyres didn't make them turn the power down, it forced aero and suspension development. Water always finds its level, so any attempt to reduce speeds at comp will only be temporary.

I'm liking the minimum track idea more and more, it doesn't hurt the slower teams too badly, and the fast teams who are likely to roll will benefit. It'll also drop the spring rates as cars will take wider lines through slaloms, reducing the tendency to take bumps very badly, i.e. flip, so the benefit is twofold.

[Kevin Hayward]

We should be careful about saying that reducing the tyre friction will make the cars slower. SAE already monitors average speed during the competition and modifies future tracks to keep the average speeds outlined by the rules. As the cars get more grip they slow them down by making shorter straights and putting in more corners and slaloms. This makes the problem worse as teams compensate by making the cars narrower.

Higher friction -> more turns -> narrower cars -> speed back to what is was -> cycle restarts

By slowing the cars in the corners you can increase the length of straights put less turns in (or wider turns) to meet the average speed shown in the rules. With the changes to the tracks the cars are also less likely to rollover.

mech5496,

You cannot prove by example, the Lions crash only shows that the structure was okay for that particular case.


I would love to hear the thoughts of someone involved with organising FSAE. Having spoken to a number of them over the years it is pretty clear to me that they consider the student safety fundamental to the running of the competition, and so far have managed to maintain both the interest and safety of the competition.

I agree with Tom that teams will find a way to make the cars go faster, however motorsport history has shown that they will often do so at the cost of safety, even by introdcuing new problems or pushing up against known safety boundaries. Overall safety of a competition is largely dictated by the organisers and the rules rather than the competitors.

With the recent Indycar accident I am sure we expect there to be rule changes to improve safety rather than relying on the good graces of the teams to slow there cars or willingly add extra weight or cost to improve safety.

Kev