calculation of the car cg

[dhingra]

if we are starting off with the suspension geometry design then how should the cg of the car be decided.we have honda cbr 600.and a mild steel chassis.

[dhingra]

if we are starting off with the suspension geometry design then how should the cg of the car be decided.we have honda cbr 600.and a mild steel chassis.

[Chris Lane]

Read Chapter 18 of 'Race Car Vehicle Dynamics' by Milliken and Milliken.

That should get you started!

[Kurt Bilinski]

If you're asking "how" to find the CG, I created a spreadsheet with the x,y location of every component (within reason) and its weight. The spreadsheet does a balancing act to find both the height and front/back location. After the car was built, the spreadsheet agreed to well within 1% of actual.

[vandit]

in RCVD , it's nicely explained how to calculate the cg.......but i think that's more applicable when you have the whole car ready.......but to use a number for all calculation , anything between 10" to 12" inches above ground can be approximated.....i know it's very wide range , but you can iterate and keep changing things to work out optimized setup and numbers.....try searching similar topics on this forum and you will get some numbers from other teams.....that should give you start ...

[Chris Lane]

You really need to sit down with you're team's sub-system leaders and estimate where your CoG is going to lie on a big whiteboard.

Estimate each components' weight and where it's CoG will lie relative to your datum in 3 dimensions.

At the end of the (lengthy) exercise, you'll have a total car weight estimate, a CoG estimate, and a list of stuff you need to design and make or buy!

[rjwoods77]

I cant remember which book it is in but there is a diagram that shows how to use two scales under the rear wheels and then raising the front end up a known amount and you can caluclate cg. It is stupid easy and works real well from my experience. It cam be done to any 2 or 4 wheel car. If you are crafty enough you can do it to find the lateral cg as well.

Pretty sure it is in this one.

http://www.amazon.com/gp/reader/1557883661/ref=sib_dp_p...-4824433#reader-link (http://www.amazon.com/gp/reader/1557883661/ref=sib_dp_pt/105-6467630-4824433#reader-link)

or check this one

http://www.amazon.com/gp/reader/0912656468/ref=sib_dp_p...-4824433#reader-link (http://www.amazon.com/gp/reader/0912656468/ref=sib_dp_pt/105-6467630-4824433#reader-link)

[mtg]

I've tried the scales under the wheels and tipping the car up on the side. Rotating the car up around the rear axle axis with scales under the wheels had terrible accuracy. You really need to tilt the car up to the balance point and measure the angle relative to horizontal to get a good result.

If you rotate up to the balance point about the rear axle axis, when you get up to the balance point, the person in the car would have to apply the brakes to keep the car from rolling and falling down. That will make it kind of tricky to measure the balance point angle, because the point of contact on the floor moves as the car rotates.

If you tip the car up along a longitudinal axis (where the car is rotates up on the tire sidewalls), the rotation axis is much more constant (especially if you use angle iron pieces around the tires), making it easier to get an accurate measurement of the balance angle.

At one point in time, I made a set of Cg lift rigs that were used on several large cars, yet were designed to accept cars down to the size of an FSAE car. They made the balance angle really easy to measure and gave two very convenient hoisting points. There was an accompanying spreadsheet that factored out the Cg of the lift rigs themselves to give you the car's Cg. If you entered the corner weights also, you got the Cg in 3D coordinates.

[Christopher Catto]

put everything as low as you can without handicapping your ability to replace/tune things at the track. so no dampers that will scrape the concrete when you apply the brakes or stuff where you cant reach the bolts with a tool.

then build the car and tip it on its side like Matt said. Lock the steering wheel, lock the brakes and put dummy dampers as for the car ride-height. The rest is simple maths.

oh. and do it SAFELY! dont save money on hoists etc. use proper tools and procedures.

[Bill_Murray]

We use the scales under wheels approach and have had very consistent results. might be worse if you don't have good scales and a scale jig that holds the scales under a axle easily. If you do it this way locking the brakes and steering is not necessary. Dummy dampers improves accuracy but is not a big effect if you keep the pitch angle of the car low.

Tilt method should work too, I just don't have a really good way to measure the angle. It's also hard to have a driver hold posture during the tilt test.

[Buckingham]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">We use the scales under wheels approach and have had very consistent results </div></BLOCKQUOTE>

Repeatability and measurement uncertainty are not the same. How do you know that the measurement you are taking, and the math you are using is correct? Just because you get the same answer 3 times, doesn't mean the answer is correct, it might just be repeatably incorrect. I strongly suggest trying all methods, and estimating total uncertainties for each method.

I would also caution against using small angles if you use the scales-under-wheels approach. The fact that not having dead-shocks installed shows negligible changes at small angles should be raising flags about using small angles, not making you comfortable about leaving the shocks in.

[mtg]

Here's what you measure the balance angle with:

Digital Inclinometer (http://www1.mscdirect.com/CGI/NNSRIT?PMPXNO=9317885&PMT4NO=30023786)

And take Donavan's advice on the small angle limitation. Any engineer should understand why calculating the Cg position by only rotating the car by small angles is bad.

[RStory]

Hey, Bill Murray!


I have had okay results using the scales under the back wheels, lift the car a small angle (~10 deg) method. I don't even lock down the suspension. But I am working with roughly 3000 lbs cars, and am willing to accept some error.

On a car that size I figured five pounds of error in the scales would result in about an inch of error in the CG calculation. I would do similar calculations for your situation, and see if the amount of uncertainty is acceptable.

I use one of those digital inclinometers for this method too, rather than trying to measure the height I've lifted the front wheels.

[Bill_Murray]

Hey Robert!

I think you where around when we did the scales method in 06'. We tried it with a wide range of angles and heights, with and without hard links, and with and without drivers. Once we did get our math right (forgot it was rotating on the spindles, not the tires...results in low cg) we were coming out with really consistent results between all the different tests.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Repeatability and measurement uncertainty are not the same. How do you know that the measurement you are taking, and the math you are using is correct? Just because you get the same answer 3 times, doesn't mean the answer is correct, it might just be repeatably incorrect. I strongly suggest trying all methods, and estimating total uncertainties for each method. </div></BLOCKQUOTE>

I understand that there is measurement uncertainty in all these methods and would like to sit down and figure out the sensitivity to the error is as you have suggested. I was merely trying to say that the math is really very easy, and throughout our tests, (up to about 45 degrees pitch) we saw consistent results. could have to do with the fact our cg is really close to the center of our wheels, aka we see very little weight transfer from the cg in this test.

PS: We have one of those digital level gauges, it is just very slow to settle on a value within +-.3 degrees so it may be hard to get a good measure with it at the balance point. But for our car, it looks like that would only make a +-0.15 inch cg height variation.

I am curious what you guys think about the sidewall deflection affecting track width in this tilt test and the fact that the car will be rolled up onto the sidewall (eg. the CG will be above some point on the sidewall, not the edge of the tire) It seems like if the cg is centered half and inch up the sidewall, your cg will be off by over a half an inch. Or do you try to estimate and measure and account for these factors as well?

[RStory]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Bill_Murray:
I think you where around when we did the scales method in 06'. </div></BLOCKQUOTE>

I was around, didn't participate though. I do remember trying to figure out why the height was coming out so low.

For the sideways tilt test you should probably have the corner of the tires in a piece of angle iron, so you know the CG is above the edge of the iron on the ground.