Orbit 9-12-2020

A particle of mass M = 10 M = 10 is fixed at the origin. At time t = 0 t = 0 , another particle of mass m = 1 m = 1 begins with position ( x , y ) = ( 1 , 0 ) (x,y) = (1,0) and velocity ( x ˙ , y ˙ ) = ( 2 , 2 ) (\dot{x},\dot{y}) = (-2,2) . The universal gravitational constant G = 1 G = 1 .

At time t = 5 t = 5 , how far away is the moving particle from the origin?

Note: This one is surprisingly badly behaved, so make sure to do proper time-step convergence testing if solving numerically


The answer is 0.788.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

2 solutions

Krishna Karthik
Sep 12, 2020

I solved the problem using a Lagrangian approach. The Newtonian approach would be less straightforward and possibly harder. With this, I just yielded a system of equations, and numerically solved.

In Hossam's solution, a first order equation was produced using a lot of mathematical manipulation. In the end, Hossam's equation of motion is far better behaved and easier to numerically solve than the coupled system below.

With the system, there are 2 generalised coordinates, hence two coupled differential equations will result.

Later, I might try using generalised polar coordinates r r and θ \theta , and solve the Euler-Lagrange equation for them.

It may lead to a first order equation, actually.

So, the Euler-Lagrange equations for this system is given by:

d d t ( L x ˙ ) = L x \displaystyle \frac{d}{dt} \left( \frac{\partial L}{\partial\dot{x}} \right)= \frac{\partial L}{\partial x}

d d t ( L y ˙ ) = L y \displaystyle \frac{d}{dt} \left( \frac{\partial L}{\partial\dot{y}} \right)= \frac{\partial L}{\partial y}

The Lagrangian for the system can be found by finding the kinetic energy (the easy part) and the potential energy (the slightly harder part).

The gravitational potential energy can be found by:

U g = G M m r \displaystyle U_g = -\frac{GMm}{r}

Where r r is the distance from the origin. This distance can be found with a bit of Pythagoras:

r = x 2 + y 2 r = \sqrt{x^2 + y^2}

The Lagrangian:

L = x ˙ 2 2 + y ˙ 2 2 10 x 2 + y 2 \displaystyle L = \frac{\dot{x}^2}{2} + \frac{\dot{y}^2}{2} -\frac{10}{\sqrt{x^2 + y^2}}

Solving the Euler-Lagrange equation will give:

x ¨ = 10 x ( x 2 + y 2 ) 3 2 \displaystyle \ddot{x} = -\frac{10x}{(x^2+y^2)^{\frac{3}{2}}}

y ¨ = 10 y ( x 2 + y 2 ) 3 2 \displaystyle \ddot{y} = -\frac{10y}{(x^2+y^2)^{\frac{3}{2}}}

The system is kind of badly behaved, even when you solve with a small timestep. But a timestep of 1 0 6 10^{-6} is enough to give you a good result.

Here's code if you want to see how I solved the coupled equation @Lil Doug

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
import math

time = 0
deltaT = 10**-6

#initial values
x = 1
y = 0

xDot = -2
yDot = 2

while time <= 5:
  #the acceleration in x and y; coupled second order non-linear differential equations
  xDotDot = -10*x/(x**2+y**2)**(1.5)
  yDotDot = -10*y/(x**2+y**2)**(1.5)


  #Euler's method
  xDot += xDotDot*deltaT
  yDot += yDotDot*deltaT

  x += xDot*deltaT
  y += yDot*deltaT

  #updating time
  time += deltaT

print(math.hypot(x,y))

@Krishna Karthik very nice solution.
Is the solution incomplete with that above image?

Talulah Riley - 9 months ago

Log in to reply

Yes, it is incomplete without the guy. Michael Rosen's a legend.

"Hot food!"

Krishna Karthik - 9 months ago

@Lil Doug btw I thought you had a date? Lol

Krishna Karthik - 9 months ago

Log in to reply

@Krishna Karthik she is busy. We have decided it to tomorrow.

Talulah Riley - 9 months ago

Log in to reply

@Talulah Riley Lol. How did you start dating her? Did you just ask her out like that?

Krishna Karthik - 9 months ago

Log in to reply

@Krishna Karthik @Krishna Karthik we both decided.

Talulah Riley - 9 months ago

@Lil Doug btw did you solve the problem? How did you solve the differential equation?

Krishna Karthik - 9 months ago

Log in to reply

@Krishna Karthik i reached that two coupled differential equation after that I am unable to solve them.

Talulah Riley - 9 months ago

Log in to reply

@Talulah Riley But how did you get the answer then? Who is the second solver, if that is the case that you didn't solve it?

Krishna Karthik - 9 months ago

Log in to reply

@Krishna Karthik @Krishna Karthik i don't know. Maybe Karan sir the ultimate Legend.

Talulah Riley - 9 months ago

Log in to reply

@Talulah Riley Yeah. He's always the first solver, isn't he? Lol.

Krishna Karthik - 9 months ago

Log in to reply

@Krishna Karthik @Krishna Karthik in 90% cases he is the 1st solver.

Talulah Riley - 9 months ago

Log in to reply

@Talulah Riley 7% is you. 3% is me.🤣

Krishna Karthik - 9 months ago

@Lil Doug

Coupled equations are too hard to solve by hand. Better if you learn how to do it using code.

Krishna Karthik - 9 months ago

Well done! There is a typo in your equations of motion. A negative sign is missing.

Karan Chatrath - 9 months ago

Log in to reply

Oh yeah; thanks for noticing. Cheers mate 😁

Krishna Karthik - 9 months ago
Hosam Hajjir
Sep 12, 2020

Using polar coordinates, we can write

r ( t ) = ( x , y ) = r ( cos θ , sin θ ) \vec{r}(t) = (x, y) = r ( \cos \theta , \sin \theta )

Differentiating with respect to time t t ,

r ˙ = r ˙ ( cos θ , sin θ ) + r ( sin θ , cos θ ) θ ˙ \dot{\vec{r}} = \dot{r} (\cos \theta, \sin \theta) + r ( - \sin \theta, \cos \theta ) \dot{\theta}

Differentiating again,

r ¨ = r ¨ ( cos θ , sin θ ) + 2 r ˙ ( sin θ , cos θ ) θ ˙ + r ( cos θ , sin θ ) ( θ ˙ ) 2 + r ( sin θ , cos θ ) θ ¨ \ddot{\vec{r}} = \ddot{r} (\cos \theta, \sin \theta )+ 2 \dot{r} (- \sin \theta , \cos \theta) \dot{\theta} + r ( - \cos \theta, - \sin \theta) (\dot{\theta})^2 + r (-\sin \theta , \cos \theta ) \ddot{\theta}

The force acting on the particle is F = 10 r 2 ( cos θ , sin θ ) = m r ¨ \vec{F} = \dfrac{-10}{r^2} (\cos \theta , \sin \theta) = m \ddot{\vec{r}}

Equating radial and traverse components of the vectors, on both sides of the equation, we deduce that,

r ¨ r ( θ ˙ ) 2 = 10 r 2 ( 1 ) \ddot{r} - r (\dot{\theta})^2 = \dfrac{-10}{r^2} \hspace{20pt}(1)

and

2 r ˙ θ ˙ + r θ ¨ = 0 ( 2 ) 2 \dot{r} \dot{\theta} + r \ddot{\theta} = 0 \hspace{20pt} (2)

Multiplying the second equation by r r , we get,

2 r r ˙ θ ˙ + r 2 θ ¨ = 0 ( 3 ) 2 r \dot{r} \dot{\theta} + r^2 \ddot{\theta} = 0 \hspace{20pt} (3)

The left hand side is just the derivative of r 2 θ ˙ r^2 \dot{\theta} , therefore,

r 2 θ ˙ = K ( 4 ) r^2 \dot{\theta} = K \hspace{20pt}(4)

To determine K, we use the initial conditions, we're given that r ( 0 ) = 1 , θ ( 0 ) = 0 , r ˙ = ( 2 , 2 ) r(0) = 1 , \theta(0) = 0 , \dot{\vec{r} } = (-2,2) , Using the expression for velocity developed above, we have,

r ˙ = r ˙ ( cos θ , sin θ ) + r ( sin θ , cos θ ) θ ˙ \dot{\vec{r}} = \dot{r} (\cos \theta, \sin \theta) + r ( - \sin \theta, \cos \theta ) \dot{\theta}

so that,

( 2 , 2 ) = r ˙ ( 1 , 0 ) + ( 1 ) ( 0 , 1 ) θ ˙ (-2, 2) = \dot{r} (1, 0) + (1) (0, 1) \dot{\theta}

Therefore, r ˙ ( 0 ) = 2 \dot{r}(0) = -2 , and θ ˙ ( 0 ) = 2 \dot{\theta}(0) = 2 , and it follows from this that the constant K = 2 K = 2 . Plugging in equation (4) into (1),

r ¨ 4 r 3 = 10 r 2 ( 5 ) \ddot{r} - \dfrac{4}{r^3} = -\dfrac{10}{r^2} \hspace{20pt}(5)

By relating the time-derivative of r r to the θ \theta -derivative, equation (5) becomes,

4 r r 4 8 r 2 r 5 4 r 3 = 10 r 2 ( 6 ) \dfrac{4 r''}{ r^4} - \dfrac{8 r'^2}{ r^5} - \dfrac{4 }{ r^3} = -\dfrac{10}{r^2}\hspace{20pt}(6)

where r = d r d θ r' = \dfrac{dr}{d\theta} and r = d 2 r d θ 2 r'' = \dfrac{d^2 r}{d\theta^2} .

Now, we will use the change of variables, u = 1 r u = \dfrac{1}{r} . This will change equation ( 6 ) (6) into,

4 u + 4 u 10 = 0 ( 7 ) 4 u'' + 4 u - 10 = 0 \hspace{20pt}(7)

where u = d 2 u d θ 2 u'' = \dfrac{d^2 u}{d\theta^2} . Equation ( 7 ) (7) is a linear differential equation with constant coefficients. Dividing by 4,

u + u 5 2 = 0 ( 8 ) u'' + u - \frac{5}{2} = 0 \hspace{20 pt} (8)

The solution of this differential equation is:

u = 5 2 + A cos θ + B sin θ u = \frac{5}{2} + A \cos \theta + B \sin \theta

Since r ( 0 ) = 1 r(0) = 1 , then , u ( 0 ) = 1 u(0) = 1 , and therefore, A = 3 2 A = - \dfrac{3}{2} , and one can show that u ( 0 ) = 1 u(0) = 1 , hence B = 1 B = 1 . And therefore,

r ( θ ) = 1 5 2 3 2 cos θ + sin θ ( 9 ) r(\theta) = \dfrac{1}{\frac{5}{2} - \frac{3}{2} \cos \theta + \sin \theta } \hspace{20pt} (9)

This is an equation of an ellipse. Now from equation (4), we have,

d θ d t = θ ˙ = 2 r 2 = 2 ( 5 2 3 2 cos θ + sin θ ) 2 \dfrac{d \theta}{dt} = \dot{\theta} = \dfrac{2}{r^2} = 2 (\frac{5}{2} - \frac{3}{2} \cos \theta + \sin \theta )^2

And this is a first-order differential equation in θ \theta , which we want to integrate over [ 0 , 5 ] [0, 5] . For this I used the Runge-Kutta 4th-order method, with a step-size of h = 1 0 4 h = 10^{-4} , I got θ ( 5 ) = 23.725 \theta(5) = 23.725 radians. Plugging this value for θ \theta into equation ( 9 ) (9) , gives r = 0.7876 r = 0.7876 .

@Hosam Hajjir
You are on fire.
Very very very nice solution.
I was searching for this only. Upvoted.


Talulah Riley - 9 months ago

Log in to reply

Yeah. See, if you want to avoid a coupled system, this is the way to go @Lil Doug .

Nice solution.

Krishna Karthik - 9 months ago

Thanks for the complement.

Hosam Hajjir - 9 months ago

@Hosam Hajjir

You're first solver, eh? I like how you used a different method to get an easier differential equation. Brilliant indeed.

Krishna Karthik - 9 months ago

Log in to reply

Thanks for the complement.

Hosam Hajjir - 9 months ago

@Krishna Karthik Hosam sir always rocks with his sttuborn anaylitical solutions.

Talulah Riley - 9 months ago

Log in to reply

@Lil Doug Wait a minute... I just read that he used Runge-Kutta 4. Eitherway, he reduced a coupled system into a first order differential equation, which is still admirable.

It's still not fully analytical, though. I'm sure Mark Hennings will come up with something lol

Krishna Karthik - 9 months ago

Log in to reply

@Krishna Karthik yeah it is still admirable. Bye

Talulah Riley - 9 months ago

Log in to reply

@Talulah Riley Aight, see ya.

Krishna Karthik - 9 months ago

@Hosam Hajjir

You just proved Kepler's law in your solution!

Krishna Karthik - 9 months ago

@Hosam Hajjir i think you deleted the problem Ladder dynamics.

Talulah Riley - 8 months, 3 weeks ago

I've reposted it now.

Hosam Hajjir - 8 months, 3 weeks ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...