# Center of Mass Calculation: x-bar, y-bar

1. Nov 9, 2012

### aaronfue

1. The problem statement, all variables and given/known data

I need to find the center of mass of the given figure attached. I am given the density (7850 kg/m^3) and a thickness of 0.3 cm (not sure where this goes)

2. Relevant equations

Just some things I need to verify:

When finding the x-bar of the function y=x3 using the equation:

$\bar{x}$= $\frac{∫ \tilde{x} dm}{∫dm}$,

Is my $\bar{x}$ going to be the x distance(which will just be "x") times x3dm?

Also, will x3 be in the denominator just before "dm"?

If I am also given a density ρ, will that stay in front of each integrand or in front of ($\frac{∫ \tilde{x} dm}{∫dm}$)? And would I calculate the thickness with the $\bar{z}$?

Is the same process is done for the $\tilde{y}$ or are there differences?

I know this may be a lot but I'd like to make sure of the method.
Thanks!
1. The problem statement, all variables and given/known data

2. Relevant equations

3. The attempt at a solution

#### Attached Files:

• ###### xid-5701632_2.jpg
File size:
7.6 KB
Views:
274
Last edited: Nov 9, 2012
2. Nov 9, 2012

### Simon Bridge

dm as the amount of mass in a volume element at point (x,y) with width dx, height dy, and thickness t ... that would be $dm = \rho(x,y) dV = \rho tdxdy$
...that help clarify things?

Of course you can simplify the calculation a lot by understanding it.

Last edited: Nov 9, 2012
3. Nov 10, 2012

### rude man

I suggest paying very little attention to the (uniform) density and not too much time on the (uniform) thickness ...

4. Nov 10, 2012

### Simon Bridge

Well, me too ... though it comes out in the wash :) It's part of the "you can simplify the calculation a lot by understanding it" thing.

It is also easier to do the integration against y instead of x and part of the integration can be done just by knowing the formula for the area of a triangle - but I don't know how much aaronfue understands vs memorizes equations. Do need to hear from him before we can continue methinks.

5. Nov 11, 2012

### rude man

Big 10-4, Simon. Not having my textbook at the ready right now, I would approach the problem as finding the axis of zero net torque under gravity and arbitrary orientation, in which case integration of a differential wedge dθ might make sense. As an EE I'll probably pass on doing the actual work, so talk's cheap!

6. Nov 11, 2012

### Simon Bridge

Who needs text books when you have google ;)

7. Nov 11, 2012

### rude man

True, true. So chalk it up to indolence on my part ...

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook