B) most sports cars can corner at higher G loads when accelerating rather than at constant velocity around a skid path. That is why the same cars that get 0.95 to 1.0 Gs in the skid pan are often measured at 1.15 to 1.22 Gs in operation on actual race tracks.

Not true, for max lateral G's you want all the reaction force of your tires for the turn. Any amount of traction you are using to accelerate or deaccelerate takes away from your cornering ability. Racers acclerate through the turns to help get the best lap times, but if you want to max lateral G's best way is to use all traction for turn...

Something to try...

I haven't tried this but I'd like too:

Find a big flat area, drive in a circle at a constant speed, if above calc are correct, drive in a 200 foot arc at 55

In theory, you should be to maintain 1 g while turning, now see if you can get something to stay on your door....or course it is still being pulled down with its weight, but it also has it's weight pushing against your door, if its static friction coeff is high enough, it should stay on your door.

 

Exactly.

One other point. More precisely, 1 g is 56.5 mph at 11.1 seconds. At 11.4 seconds is 55 mph which is .95 g. So you can see that a very small difference in mph makes a large difference in g's because it changes with the square of the velocity. So trying to generate 1 g by looking at your speedeometer would be very difficult to say the least. A stopwatch is much more accurate.

 

Hey TrayC6 give some insight here!!!!

I know you state that you have pulled 1.3 so far on your new C6....give some info on where you were, banked ramp, curve, flat surfac etc.

Local boy

 

Alright C6NITPICKER, you really are "Nit picking" here. The lateral G test is not just selling the tire, it is reflective upon the the entire set-up of the car not just the tire. This would include, ride height, caster, camber, toe-in, toe-out, sway bars, springs, CG, Roll Center, weight balance, and tire selection.

 

Ahh, yes but I'd figure most magazines use something at least as complex as a gtech competition pro, several use data loggers with differential GPS getting velocity infomation with accelerometer backup which has 100hz or more. You can now buy for $800 a data logger with GPS and accelerometers that will give you very accurate 100Hz information for racing. I have the gtech competition pro and will check some things out on my trip back from the museum. I also hotlap several california tracks and will get you some information on various tracks. The streets of willow springs has a nice skidpad. The problem is that it's reasonably difficult to do skidpad runs without practice. Most people just use the skid pad to feel out the cars under/oversteer behaviour. Even a terrible run will yeild me max lateral G. And I can get max G in 2 directions for the entire track day. I'll be at streets in June with the new vette and will give a summary. By then I will probably spurge on a full GPS datalogger for track mapping and full information. There are two I like that are under $1,000 which to me is well worth it for motec like information about any run you wish to make.

-mikey

 

Cool. Let me know what you decide on and a link to where I can check it out.

Things like the Gtech are too subject to instantaneous measurements from a bump or any jerky movement and will give you a max that's far greater than reality. So I'm curious what else is out there.

 

Actually the current gtech is very good and not subject to problems like you describe.

You control the filtering of spikes. It also downloads very nicely in to Excel and gives you 100Hz data to play with. It has 3 accelerometers aligned in 3 axes’ to make it very simple to put in the car. There's no need to line up or change it. You simply mount it at any reasonable angle on a flat surface and it uses gravity to determine the orientation of the units 3 accelerometers. Then it creates a new ideal coordinate system with normalized virtual accelerometers that are combinations of each real one. Doing this it is very simple to set up but able to capture perfect accelerations down the axis's you care about.

These are very accurate accelerometers and very fast. It's easily good enough for any skid pad testing you care to do. The main problem is that you can't account for extreme body roll or pitch. For dyno/speed runs you enter in the pitch amount that the car will make during acceleration. This is then compensated for. With roll there is no built in compensation if I remember correctly. Even good GPS systems with accelerometers don’t deal with roll or pitch well. Same with hills, they can't deal with hills very well. But they are still useful.

You would need differential GPS to deal with this really accurately. With 3 gps readings at very high sample rates in extreme locations on the car and say 3 locations on the track you'd be able to account for everything EXTREMELY accurately. This currently is way beyond the $1000 mark. But the difference in practical use is very tiny. Your drift from reality will be small on the $1000 setup.

What does this mean for the Gtech? It's great for dynos, great for timing runs on flat surfaces. It's also great for tracking G load and gear changes on a road circuit or any canyon run. The only thing it cannot do is provide accurate track mapping and perfectly accurate G loads on non-flat tracks. However since you are using the device for a single car at a single track it's still very useful for you to compare to yourself. The hill will always be in the same place. What it does to you won't vary from lap to lap. So while the g measurement may not be right for comparison to a full gps setup, it's great for comparison lap to lap. Did I do better in that corner this time vs. lap 5? Also many people using Gtechs at the same track can compare numbers because they are all dealing with the same hills. As long as they accurately put in compensation numbers for their cars pitch, and maybe roll have to check that. I'll give you guys some links later tonight.

Even with a full GPS setup I very highly recommend the Gtech competition pro full up model. It's cheap and it is fantastic. Don't worry about reality worry about comparing to yourself. It's just like a dyno, how well they compensate for all the variables for comparison purposes is somewhat up in the air. Usually they state what machine did the Dyno so you can compare to that guy.

Gtech is like that, it should only ideally be compared to itself on a particular vehicle and only to other gtechs barring that, usefulness to other dynos and real world numbers??? That's to be taken with a grain of salt.

For instance if the user inputs weight wrong or calibrates the RPM badly the numbers will be way off reality. But if they don't care about reality and input the same numbers for the same situation each time that are ballpark, then they will have a great history of that vehicles performance against itself.

-mikey

 

The C6 HUD display uses a accelerometer. There is also one in the fore/aft direction, but I don't think there is one in the up/down direction.

Any banking would reduce the lateral G-force because it turns the outward force downward. In fact, it is easy to come up with an equation that will tell you the angle of bank that will eliminate any lateral force, meaning you would go through the turn with your steering wheel straight and not feel any lateral force. Planes bank when they turn, but they usually match the bank and speed pretty well, that's why if you look at a glass of water it stays relatively close to level when the plane is banking...

 

The problem is that most logging systems don't have an accurate way to measure the pitch and roll of the vehicle. A differential GPS system uses multiple positions to campare with eachother. This makes each measure EXTREMELY accurate and also because they are calibrated known positions at rest on the vehicle allows you to know EXACT attitude. You get pitch and roll too. With that and the accelerometers you can calculate everything your interested in.

-mikey

 

Guys/Gals I initiated this thread as the devils advocate. I wanted to start a debate that begin between me and a friend, I posed the that you can exceed 1G, he is the one that said you could not (he is the guy that tried it at school). In any case I found some of the post educational and I appreciated the debate.

 

I see more than 1 G in the HUD routinely. Now if I could just accelerate after the turn, I might actually have some fun.

 

I should be packing for my trip tomorrow but why do now what I can put off for 10 minutes.


NCM on TUESDAY!!!

I have the gtechpro competition and love it!
The gtechproRR replaced it and is basically just colored differantly. The usage is excactly the same.
gtech pro

advanced racing data logger

Several magazines use products by this company. This product is thier new affordable unit and it is very capable. It is my second choice in a data logger right now.
race pak G2x

Nology data logger

Road tune data logger

This company is top notch and makes by far the best stuff I've seen. This is out of my league.
Racelogic

This company makes my favorite data loggers. I've seen these in action, they are getting really good. They work great!
race-technology
This is my first choice in a data logger and my birthdays in July when it comes out, so I may just have to get one!!!
DL-2

Hope you find these link useful, I'll repost these outside in case anyone might find them useful.

-mikey

 

I got my 1.3 G on a flat surface.. took a sharp turn around 40-45mph around 90 degrees... just getting ready to lose the rear end (tires barking).. ... felt good... too bad you couldnt maintain that for a while.. was a quick rush... .. Its not too hard to do actually.. you just have to have the right ratio of speed and curve...

 

With a 50%/50% split of front and rear weights and bigger rear tires than front tires:: when the car is in pure cornering the rear tire contact patch is underloaded relative to the front contact patch.

By adding enough throttle so that some weight transfers to the rear, both ends of the car can be setup such that the contact patches are equally loaded and, thus, slightly more latteral acceleration can be achieved while accelerating than while cornering at constant velocity.

Every once in a while the C&D or R&T magazines take a bunch of cars to the road race track and measure various parameters; two of which happen to be skidpath and dynamic cornering. You will find that several of the best sports cars have higher dynamic cornering numbers than they have skidpan numbers (both measured in Gs).

 

My response was based on a vehicle's dynamics class I took while obtaining my engineering degree. This makes it based on the models and physics we used for simulation, which are very good but have missed some differences between the theoretical modeled world and real world. But the basic principle is this:

1. Your tires are normally not sliding relative to the road because they are rolling. As such each wheel can exert a force equal to its static coeff of friction times it's normal force. (In the models for that class we modeled the tires as a combination of springs and dampeners with a more complex friction coeff, but you will get pretty good results with the above equation)

2. The max force from each tire patch can then be determined, which can be broken down in a component in the lateral direction we'll call Y and one in the fore/aft direction we'll call X. The total force for that tire would then be F=SQRT(X^2+Y^2) and must be less than or equal to the max force for each tire patch under it's current conditions.

3. From the above equation, it is easy to see that to get all the force from the tire patch in the lateral direction, you want the fore/aft force (X) to be zero, meaning no acceleration or braking.

4. You bring up an excellent point, by accelerating/deacceleration you can change the max force of the front/rear tires by shifting the weight (changing the normal force); however, any use of force in the Y direction will decrease force that could be used in the X direction. As a result, you'd want to shift the weight with a moment of acceleration/deacceleration to optimize each tires max force, then turn the vehicle with no acceleration/deacceleration so all the force is in the lateral direction. You may not be able to turn for very long before the weight transfers back, but that point should be where you can achieve max lateral force. Or, you could install weights in ideal places on your vehicle to get the ideal weight distribution for cornering.

5. Finally, if you think of the front and rear tires of the beloved C6, the rear tires are larger and with a 50/50 distribution we'd expect the max force of them to be larger than the fronts. As a result, the rear wheels can handle the additional force of acceleration, but to get a max lateral acceleration you'd be better off increasing the normal force on the front tires to bring their max force up.

Examples:
If we go with a simple model ,a typically coefficient of static fiction for rubber to asphalt is .75, however, this would limit the lateral Gs to a max of .75, so, since we know a Corvette can do 1 G, we'll use that as a coefficient.

The C6 weighs 1442 kg and with 50/50 weight dist., that is 721kg for front and rear. Now, based on this simple model the max force for each tire is the same, .5*721kg*9.81m/sec^2*1=3537 N. In reality the rear wheels can handle more because they are wider. For many materials friction coeff does not depend on surface area, this is because if you decrease the surface area you increase the pressure, but because tires are more complex it falls short. Let's do a quick check to see how close this is, if each tire could provide a max force of 3537 N that would mean during straight acceleration the max force is 7074 N which correlates to a max vehicle acceleration of 4.905 m/sec^s and would result in a max speed of 44 MPH in 4 seconds, which is close to the 60 MPH we know is possible. To get 60 MPH, we would need a force of 9725 N which would correlate to a coefficient of 1.37 or a weight transfer of 31/69% with a coeff of 1. That weight transfer doesn't seem too bad, but in reality the result is inbetween, we'll use 1.25 for rear and 1 for front, this should be a good approximation.

So, starting off the front of the car can provide a max force of 7074N and the rear can provide 8841N. So, with this setup the max lateral acceleration will be when all the 7074 of the front is used in the Y direction along with 7074 of the rear and will result in a lateral acceleration of 9.81 m/sec^2 or 1 G. The rear could still provide another 5305 N in the X direction (acceleration) before breaking loose; however, note that any weight transfer to the rear will decrease the force the front can provide and cause the front tires to slide out. What happens most often is we use up all the rear tire force and the rear breaks loose first.

To maximize lateral force we want to increase the max force of the front tires by transfering weight forward, right now the model equation for the front is Max Force=Coeff*Mass*Gravity, or Max Force=1*front mass*9.81m/sec^2 and for the rear Max Force=1.25*rear mass*9.81m/sec^2. Solving this equation for the max lateral accleration results in 801kg forward and 640 kg rearward, which is 56/44% weight distribution and would result in front and rear lateral force of 7859 N and a lateral acceleration of 10.9 m/sec^2 or 1.11 Gs.

Basically since at the current 50/50 distribution the rear wheels can provide more force, to optimize lateral accleration weight transfer to the front tires would allow the vehicle to achieve a higher lateral G Force. There are many other things in reality that come into play, including the side to side weight transfer, understeer/oversteer, and many other things....

Hope this wasn't too boring...

 

And just to complicate things further, keep in mind that under static load conditions, that same car with 50/50 wt. distribution with the huge tires on the back has the exact same size tire contact patch front and rear.

 

That is correct, I didn't include that because I didn't want to complicate things further, but to determine the rough size of the tire patch simple use the tire inflation pressue and the weight of the vehicle. Corvettes weight about 3172 lbs with 30 psi per tire results in about 26.4 square inches per tire (actually less due to tire stiffness). The front patch and rear patch will have a different shape due to the width difference and that along with the spring/damp. model explain why wider rear tires help with acceleration and also why they have less traction on snow/ice but more traction on bare concrete, which also contradicts simple theory.

 

My data is from studying Milliken and Milliken "Vehicular dynamics" and by reading the R&T and C&D test reports. I don't keep the magazines so I can't give you a date to look up the useful data. However, at the time of the C5 introduction (e.g. 6 years ago) C&D (I think) took the top 6 sports cars {Ferrari F355 Spider, C5, 911C2, Viper GTS, M3, Lotus Esprit V8,...} and ran then through a battery of tests. On the skid pan several of the cars were hovering in the 1.0 G range {0.95 to 1.05}. However, with Mario Andretti at the wheel, and a data logger in the car, several of these same cars would generate 1.22 Gs latterally in the 2D G diagram for short periods of time. The younger Milliken was supervising the data acquisition system during this test day.

I don't have an issue with the math or data you presented, except to say that as one adds load to a tire its coefficient of friction changes (more load == less CoF) which is detailed in Milliken and Milliken. I suspect that this change in CoF and the different widths of tires F/R results in higher latteral accelerations due to more even loading of the contact patches under acceleration (and thereby higher total cornering force) than at constant turning velocity (skid pan).

There might be some other issues in play here also:: the turn measured by the DA system did not have a comparable radii of curvature with the skid pan and therefore the slip angles and steering angles would be different; the velocity on the skid pan was lower then the velocity through the actual race track turns, finally, the skid pan and race track might not have the same CoFs even though they came from the same batches of asphault (track used a lot, skid pan occasionally; or vice versa).

I might also note (from VanValkenburg "Race Car Engineering"):: when a car is setup on a skid pan for maximum latteral acceleration, the driver is often surprised that he cannot accelerate out of turns (oversteer) since all traction is being devoted to latteral grip and none is left for acceleration--in agreement with your equations and argument.

Mitch

 

Another thing is transition G-force, I'm able to get the G-meter up pretty high with a sudden jerk of the wheel...A lot of things change when you move from a model to a real world, but I find the whole thing interesting...