PDA

View Full Version : G-force reading


marc
2007.08.14, 09:34 AM
You know what would be cool? You know how Mythbusters use that G-force reading device? It would be interesting to get a tiny one mounted on the Mini-Z's, and see which ones generate the most G's on excelleration and turns. Which type do you think would generate the most? AWD? MR01, MR15, or MR02?

idreamz
2007.08.14, 02:30 PM
I think the MR02 (which has the lowerst CG) will have the theoretically highest G rating, at about 2.6G. Then it'll flip. You can estimate that by tipping the car until it flip, note the angle at which it does and take tangency on that value.

An AWD does around 2 G before it flips over, and mine has no problem doing that on track.

I don't know how one of these G-force gauging apparatus looks but what about a homemade variant? say a pendulum that somehow makes a mark on a paper slip? you only get the maximum reading but that's ok?

mini-z
2007.08.16, 11:16 AM
Or a hacked Wii remote? ;) :D

marc
2007.08.16, 01:54 PM
How would you use that on a Mini-Z? And, how would you read it?

benmlee
2007.08.16, 02:20 PM
I think the MR02 (which has the lowerst CG) will have the theoretically highest G rating, at about 2.6G. Then it'll flip. You can estimate that by tipping the car until it flip, note the angle at which it does and take tangency on that value.

An AWD does around 2 G before it flips over, and mine has no problem doing that on track.

I don't know how one of these G-force gauging apparatus looks but what about a homemade variant? say a pendulum that somehow makes a mark on a paper slip? you only get the maximum reading but that's ok?

A pendulum would swing too eratic to work unless you are a watch maker that can put tiny dampers on it. I tred that on a different project in college. Driving the car in a circle is probably the best method to measure the g force. Putting the car on a sloped surface might not give the the right answer. With the car tilted, the weight of the car is no longer spread evenly on the left and right wheels. Also wheels are not rolling, so there is no account for tire deflection.

idreamz
2007.08.17, 02:34 PM
A pendulum would swing too eratic to work unless you are a watch maker that can put tiny dampers on it. I tred that on a different project in college.(...)
That's serious good points. I haven't thought that the pendulum might start oscilliating creating greater false readings than actual Gs. I just thought that the device has to be small and light as not to alter the cars handling with it's weight.

(...)Putting the car on a sloped surface might not give the the right answer. With the car tilted, the weight of the car is no longer spread evenly on the left and right wheels. Also wheels are not rolling, so there is no account for tire deflection.
Agree that the stationary wheels is a weakness with my idea. But I don't agree that the weight is spread evenly on the left and right wheels during cornering (as I was trying to simulate with the tilting). I find it more natural to think that the weight shifts towards the outer wheels when cornering (compare this to what you feel when you sit in a car that's cornering). So when the car turns left, the weight will shift to the right wheel.

As long as the car's weight stays between the left and the right wheels, the car stays upright. But if the car corners so hard that the weight shift gets outside of the wheelpair, the car will flip. In real cars this won't happen, as the tire grip is more limited meaning the car will start to slide before that. But Mini Z are different, see this is what happens to my AWD on track. It flips during hard cornering, meaning it's maximum G readings dictated by it's high CG rather than tire grip.

So to get back to the subject, my model is based on the very instant when the car is about to flip. At that instant it has all it's weight shifted exactly on the outer wheels, balancing on them with the inner wheels taking no load at all. The same situation occurs if you tilt the car manually by lifting the "inner" side, so at a certain angle you'll apply no force to lift the "inner" side since it's starting to tip over itself, balancing all it's weight on it's "outer" wheels. That angle is then corresponding to the "angle" the car has when it's cornering hard and about to flip. With this angle you can calculate the ratio between the downward force component and the lateral force component. Since the downward force is 1G, multiplying with that ratio will give you the lateral G rating when the car starts to flip in a corner.

As an end note, this model only works of the car flips before it skids, for the opposite I think your circluling idea is way better. And my ideas are based upon my high school physics lessons, so they propably miss out a lot of factors.

(...) Driving the car in a circle is probably the best method to measure the g force. (...)
Though even here I have an objection. On track cornering, the cars will not drive in perfect circles, meaning they might pull higher Gs or lower. I for one can't get my AWd to flip when driving in circles... well ok that might have to do with the surface though.

marc
2007.08.18, 10:06 AM
I bet the F1's would generate the most forces on turning. They seem to be quicker, and run flatter, meaning they don't have the tendincy to flip.

idreamz
2007.08.19, 10:42 AM
Hmm, yes seems we need a G-force meter anyway to see if these theories are good enough.

HaCo
2007.08.19, 02:13 PM
Maybe with the right sensors this thing could do the job...
http://www.eagletreesystems.com/MicroPower/micro.htm

benmlee
2007.08.19, 03:25 PM
Agree that the stationary wheels is a weakness with my idea. But I don't agree that the weight is spread evenly on the left and right wheels during cornering (as I was trying to simulate with the tilting). I find it more natural to think that the weight shifts towards the outer wheels when cornering (compare this to what you feel when you sit in a car that's cornering). So when the car turns left, the weight will shift to the right wheel.




Got to remember the weight of the car is always downward and is spread between left and right sides. When the car is cornering, an extra centrifugal force comes into play. That acts in the horizontal direction on the CG of the car. You have to account for two forces. By tilting the car, you still only have one force, that is the weight of the car. You are just changing the direction of the force, but you are not adding the centrifugal force to it. The tilting experiment can help you find the center of gravity and maybe the coefficient of friction of the tires.
In any case, the 2.6 g sounds a little high. When I did a rough test of the tires, they slip at a little over 1.2 g.
See:
http://mini-z.home.att.net/tires.htm

mini-z
2007.08.19, 05:26 PM
Another idea: I've noticed that in JGTC, instead of the digital g-force meter that we see in F1, they seem to have a sort of cylinder with several red balls in it - I don't really understand how it measures g-force vs. just its direction, but, it looks like a fairly simple method if you are able to find more information on how it works.

HammerZ
2007.08.19, 06:51 PM
All I can think of is search the model rocket sites for a thrust meter. And try to adapt that.

marc
2007.08.19, 08:52 PM
Interesting idea, if it's size is suitable to Z's.

HammerZ
2007.08.19, 10:33 PM
I searched around, the best I found was a book on model rocket projects. I was looking for a thrust device, mainly one that can be used for those Estes mini rockets. Those body tubes are only half an inch across. most of the ones I have seen in the past are for larger rockets though.

idreamz
2007.08.20, 06:51 AM
(...)The tilting experiment can help you find the center of gravity and maybe the coefficient of friction of the tires.
In any case, the 2.6 g sounds a little high. When I did a rough test of the tires, they slip at a little over 1.2 g.
See:
http://mini-z.home.att.net/tires.htm
Firstly I have to compliment on your website, it's propably the best site I've seen explaining RC car handling (other sites are just telling you that yada-yada adds understeer but not really why). For example you had an elegant explanation on how the spring rates affected turn-in which I've never understood before. That tire testing was cool too, but I found it surprising that the rear tires (assuming it's wider than the front) didn't had it's peak G rating at a higher load than the front. Normally, fitting wider tires enables the tire to handle bigger weights.

Back to subject, my model is only intended to predict at what G the car flips over. It assumes there's enough tire grip to do so. It does not give any accurate estimate of the tires friction coefficient due to the fact that the tires are not rolling giving a 90degree slip angle, and the lack of a track-like surface. Because my AWD flips over cornering in track, I deduced it's lateral acceleration at some point must have been high enough to make it flip, which for AWD in left turns is about tan(60degrees)= 1.7G (+-0.1G).

Got to remember the weight of the car is always downward and is spread between left and right sides. When the car is cornering, an extra centrifugal force comes into play. That acts in the horizontal direction on the CG of the car. You have to account for two forces. By tilting the car, you still only have one force, that is the weight of the car. You are just changing the direction of the force, but you are not adding the centrifugal force to it.(...)

For predicting when the car flips over, I think the important thing is the ratio between the lateral force and the downward force. You can get that ratio by obtaining the angle of which the car flips over. Besides forces can be added or divided into components, if you add the "centrifugal force" and the weight you'll end up with a "force" that is pointed obliquely towards one of the outer wheel. If it ponts outside the outer wheel, the car'll flip.

http://img411.imageshack.us/img411/1958/flipcalcnk0.gif

Compare above to below:

http://img66.imageshack.us/img66/6008/flipcalc2bd5.jpg

But if we are to go very advanced, I've been told in school that the centrifugal force doesn't exist really, it's something you "feel" but nothing you can apply in a physics model. So in reality it's a bit more complicated than this of course.

I have tried to find info on the JGTC G-meter without success, the rocket thrust meter sounds interesting. For your electronic experts, is it possible to custom build a piezo device that can sense acceleration and log it? for exampel a hacked helicopter gyro?

Anizzle
2007.09.19, 04:52 PM
I love physics!