Satellites in orbit

Two identical satellites A A and B B are launched in equatorial circular orbits around Earth with the same period of revolution of 4.8 hours. Satellite A A rotates west to east (with Earth's rotation) while satellite B B rotates east to west (against Earth's rotation).

On a particular day, both satellites were seen directly overhead by an observer standing at point P P on Earth’s equator. How long (in hours) will it be until both satellites are seen directly overhead again from the same point P ? P?


Details and Assumptions:

  • The orbits are located on nearly the same circle, with just enough separation to avoid collision.


The answer is 12.00.

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.

8 solutions

Once in space, the motion of the satellites is quite independent of the earth's daily rotation. There is no need to find their speeds relative to point P P .

Due to the symmetry of the situation, the satellites meet each other twice during each revolution, on opposite sides of their orbits. Let's call the original meeting point X X and the point at the opposite side, Y Y . The satellites meet { at point X t = 2.4 n for even n , at point Y t = 2.4 n for odd n . \begin{cases} \text{at point}\ X & t = 2.4n\ \ \text{for even}\ n, \\ \text{at point}\ Y & t = 2.4n\ \ \text{for odd}\ n.\end{cases} The observer can only see the satellites overhead if he is directly below either point X X or point Y Y . He is { below X t = 12 m for even m , below Y t = 12 m for odd m . \begin{cases} \text{below}\ X & t = 12m\ \ \text{for even}\ m, \\ \text{below}\ Y & t = 12m\ \ \text{for odd}\ m.\end{cases} Thus we must solve 2.4 n = 12 m for n , m > 0 both even or both odd . 2.4n = 12m\ \ \ \text{for}\ n,m > 0\ \text{both even or both odd}. This reduces to n = 5 m for n , m > 0 both even or both odd , n = 5m\ \ \ \text{for}\ n,m > 0\ \text{both even or both odd}, which is obviously true for m = 1 , 2 , 3 , m = 1, 2, 3, \dots and n = 5 , 10 , 15 , n = 5, 10, 15, \dots .

The smallest solution is m = 1 m = 1 , n = 5 n = 5 , ant t = 12 hours t = \boxed{12}\ \text{hours} .

Generalization: Let the periods of the satellites be T 1 < T 2 |T_1| < |T_2| ; use negative signs for direction against earth's rotation. They will meet each other periodically, with period T T' , where 1 T = 1 T 1 1 T 2 . \frac 1 {T'} = \frac 1 {T_1} - \frac 1 {T_2}.

At the first new encouter each satellite will have traveled p i = T / T i p_i = T'/T_i revolutions. At the n n th encounter this is n T / T i nT'/T_i . Meanwhile, if the earth's rotational period is T 0 T_0 , the earth will have rotated over p = n T / T 0 p = nT'/T_0 . The satellite is directly above the observer if p p i mod 1 p \equiv p_i\ \text{mod 1} , that is, n ( T / T 0 T / T 1 ) = k (an integer) . n(T'/T_0 - T'/T_1) = k\ \ \ \text{(an integer)}. (It doesn't matter whether we use T 1 T_1 or T 2 T_2 , since T / T 1 T / T 2 = 1 T'/T_1 - T'/T_2 = 1 ).

Calculate therefore k n = T T 0 T T 1 = T 2 ( T 1 T 0 ) T 0 ( T 1 T 2 ) . \frac k n = \frac{T'}{T_0} - \frac{T'}{T_1} = \frac{T_2(T_1 - T_0)}{T_0(T_1-T_2)}. Reduce this to the lowest terms with n > 0 n > 0 an integer. Then the desired time is t = n T = n T 1 T 2 T 1 T 2 = T 0 T 1 T 2 d , t = |nT'| = \left|\frac{nT_1T_2}{T_1-T_2}\right|=\left|\frac{T_0T_1T_2}d\right|, where d d is the greatest common divisor of T 2 ( T 1 T 0 ) T_2(T_1-T_0) and T 0 ( T 1 T 2 ) T_0(T_1-T_2) .


In this situation, T 1 = 4.8 ; T 2 = 4.8 ; T 0 = 24 T_1 = 4.8;\ T_2 = -4.8;\ T_0 = 24 . We get T 2 ( T 1 T 0 ) = 92.16 ; T 0 ( T 1 T 2 ) = 230.4. T_2(T_1-T_0) = 92.16;\ \ \ T_0(T_1-T_2) = 230.4. Their greatest common divisor is d = 46.08 d = 46.08 ; their quotient is k / n = 2 / 5 k/n = 2/5 . We find t = 24 4.8 ( 4.8 ) 46.08 = 12 hours . t = \left|\frac{24\cdot 4.8\cdot (-4.8)}{46.08}\right| = 12\ \text{hours}.

The calculation becomes a little more elegant if we write T 1 , T 2 = ± 1 5 T 0 T_1,T_2 = \pm \tfrac15T_0 . Then T 2 ( T 1 T 0 ) = 4 25 T 0 2 , T 0 ( T 1 T 2 ) = 2 5 T 0 2 , T_2(T_1-T_0) = \tfrac4{25}T_0^2,\ \ \ \ T_0(T_1 - T_2) = \tfrac25T_0^2, with g.c.d. d = 2 25 d = \tfrac 2{25} . Then t = 1 5 ( 1 5 ) 2 25 T 0 = 1 2 T 0 . t = \left|\frac{\tfrac15\cdot(-\tfrac15)}{\tfrac 2{25}}\right|T_0 = \tfrac12 T_0.

To interpret a little more: In this case T = 2.4 T' = 2.4 hours, which is the time between two encounters of the satellites. And n = 5 n = 5 because it takes 5 of these encounters before the satellites are again overhead.

Arjen Vreugdenhil - 4 years, 1 month ago

Log in to reply

This is a great explanation. Thanks.

Rohit Gupta - 4 years, 1 month ago

We should have been told to assume the Earth revolves on its axis in exactly 24 hours!

Geoff Collinson - 4 years, 1 month ago

Log in to reply

You mean be told how long a day is? Fair point (the math shouldn't require astronomy knowledge... or assume the person is from Earth ;-)

Jeff Millard - 4 years, 1 month ago

Log in to reply

24 hours is actually incorrect - the relevant figure here is more like 23 hours 56 minutes. 24 hours is how long a day appears to an observer on the surface of the planet, based on observation of the sun; it is slightly longer than the true rotation period of the planet, as the planet is also moving around the sun. So in a way specifying this assumption is very relevant!

George Foot - 4 years, 1 month ago

A day is not exactly 24 hours in duration, only assumed so to make life easy!

Geoff Collinson - 4 years, 1 month ago

Because in fact, the earth does NOT revolve on its axis in exactly 24 hours. 24 is the average time between solar noons, but meanwhile the earth has progressed in its orbit around the sun, needing a few minutes less than 24 hours to complete a full spin, and the rest to catch up with the new apparent position of the sun.

We should NOT be required to assume a falsehood on our own to achieve the "correct" false answer within a few decimal places!

So this question should be omitted from the official scoring, and so should a bunch of other questions!

Give us an easy way to claim that, without burying it in the comments!

Wes Pinchot - 3 years, 6 months ago

Log in to reply

And if we ARE ever supposed to look up and calculate with some irrational number with many different values, estimated to different numbers of decimal places, please tell us how close to the best known answer we have to get!

Probably better to just tell us what values and assumptions to use.

Wes Pinchot - 3 years, 6 months ago

I completely agree! Everyone who is a bit clever knows that it is not 24 h and that the problem starts to be more complicated. E.g. see https://en.wikipedia.org/wiki/Sidereal_time

Karel Kolář - 3 years, 1 month ago

Correct! And if we take the annual rotation to be exactly 365 days then T 0 = 24 364 365 T_0 = 24\cdot\frac{364}{365} and we find k / n = 291 / 728 k/n = 291/728 (see my generalization below) so it will take 1747.2 1747.2 hours (1/5 of a year) before the satellites are together overhead.

Or more precisely, if T 0 = 24 364 1 4 365 1 4 T_0 = 24\cdot\frac{364\tfrac14}{365\tfrac14} then k / n = 2912 / 7285 k/n = 2912/7285 so it will take 17484 17484 hours (i.e. about two years) before the satellites are together overhead.

Arjen Vreugdenhil - 4 years, 1 month ago

Log in to reply

The annual rotation isn't exactly 365 days. Remember why we have leap years?

Richard Desper - 4 years ago

Log in to reply

@Richard Desper That's why I put in the second part ("Or more precisely, ...") in my comment. It accounts for leap year every fourth year (but not the skipped leap year 3 out of 4 century years).

Arjen Vreugdenhil - 4 years ago

This is a nice idea to treat the motions of the satellite and the Earth independently. Can we generalize this result by taking the time period of revolutions of the satellites as T 1 T_1 and T 2 T_2 and of Earth as T T ?

Rohit Gupta - 4 years, 1 month ago

Why is it t = 12 m t = 12m ?

Fidel Simanjuntak - 4 years, 1 month ago

why you didn't consider earth spin ? The point P will be moving relatively !

Vishnu Vardhan V - 4 years, 1 month ago

Log in to reply

But I did... only after figuring out when and were the satellites will meet. The number "12" in my solution is based on the earth's rotation.

Arjen Vreugdenhil - 4 years, 1 month ago

Log in to reply

you are right

Vishnu Vardhan V - 4 years ago

If satellites have to keep up with Earth's rotation momentum, they why the heck do planes have no problem with that? #FlatEarth

Victor Shirosaki - 4 years, 1 month ago

Log in to reply

Satellites have no problem keeping up with earth's rotation. Once they have the proper rotational speed they will maintain that speed, because of the principle of inertia and because of the lack of air resistance above the atmosphere. Satellites do not propel themselves, except to make small corrections to their path.

The challenge is to launch satellites with sufficient speed to stay in orbit. That speed must be greater than that of the rotating earth. (7900 mph vs. 3500 mph) Can you think of a reason why satellites are launched to the east?

Arjen Vreugdenhil - 4 years, 1 month ago

Log in to reply

I was referring from an earthly point of view. My bad. Geometrically speaking, there is a difference. But planes, still, have to encounter the momentum parameter.

Victor Shirosaki - 4 years, 1 month ago
Li Cheung
May 9, 2017

Satellites A and B crosses above point P and will cross again there and at the opposite side (180 degrees) of their circular orbits, indefinitely. They will cross every 2.4 hours with each satellite traveling half the orbit. It will take point P 12 hours to make it to the other side of the circle where the satellites pass one another for the 5th time since the first observation.

Yes, the motion of the satellite and the Earth are independent of each other and can be considered separately.

Rohit Gupta - 4 years, 1 month ago

Log in to reply

Or put another way, the satellites are 5 times faster than the earth in terms of rpm. Since we know the satellites pass every 2.4hrs we know it will take 5x the time for the earth observer to rotate around so it is under the first absolute meeting place. 5 x 2.4hrs = 12 hrs. Or the 3rd time the satellites cross paths over that same absolute position in space.

Jeff Millard - 4 years, 1 month ago

First, we have to convert the periods into speeds. To make things easy, we can use units of spins per hour.

Obviously the Earth spins one time in a day, so its angular velocity is ω E = 1 / 24 \omega_E = 1/24 spins per hour counterclockwise. Similarly, the angular velocities of satellites are ω A = 1 / 4.8 \omega_A = 1/4.8 spins per hour counterclockwise and ω B = 1 / 4.8 \omega_B = 1/4.8 spins per hour clockwise.

We have to find out how fast these satellites move relative the person standing at P . P. We can do this by simply addition and subtraction, a Galilean transformation into the frame of the person at P . P. Relative to the person standing at point P , P, satellite A has angular speed ω A = ω A ω E = 1 / 6 \omega_A^\prime = \omega_A - \omega_E = 1/6 spins per hour counterclockwise. Similarly, satellite B has angular speed ω B = ω B + ω E = 1 / 4 \omega_B^\prime = \omega_B + \omega_E = 1/4 spins per hour clockwise.

Thus, in the reference frame of the person at P , P, their periods are T A = 1 / ω A = 6 h T_A = 1/\omega_A = \SI{6}{\hour} and T B = 1 / ω B = 4 h . T_B = 1/\omega_B = \SI{4}{\hour}. Obviously, it will take 12 h \SI{12}{\hour} for these satellites to overlap again above the person standing at P . P.

I like the idea of taking the units of the angular velocity as spins per hour to get rid of the factor of 2 π 2\pi .

Rohit Gupta - 4 years, 1 month ago

this all makes sense to me up until you show the periods T. how do you arrive at twelve from periods 6 and 4?

Matthew Agona - 4 years ago

Log in to reply

@Matthew Agona , sorry I didn't see this until now. The satellite with period 6 goes around twice in 12 hours. The satellite with period 4 goes around three times in 12 hours. So they're both back at the starting point at the same time at 12 hrs.

muntasir sengupta - 4 years ago
Jayraj Inamdar
May 9, 2017
  • Earth completes one revolution in 23.92 hours (sidereal time). Which means that it is able to cover 15 degrees in an hour.
  • Both satellites can cover 75 degrees in an hour.
  • Sat A, as seen by an observer on Earth, is able to cover (75-15) degrees per hour.
  • Sat B, as seen by observer, covers (75+15) degrees per hour.
  • To the observer, the sat A will take 6 hours to complete one revolution and sat B will take 4 hours. (Hint: One revolution(=360 deg)/Degrees covered per hour will give the time taken for one revolution.)
  • The LCM of 4 and 6 is 12. So there is your answer!

PS: No complex relations involved. Just pure logic and basic math required.

Where will they meet in space? Will they meet at the same place as they were at the beginning?

Rohit Gupta - 4 years, 1 month ago

Log in to reply

Yes they will meet in the same place as they started. An implicit assumption is to consider that the satellite were at the same point in space as seen by the observer.

Jayraj Inamdar - 4 years, 1 month ago

Log in to reply

But, in 12 hours the Earth will not come to the same point, then how can they meet at the same point in space? Shouldn't meet at the diametrically opposite point?

Rohit Gupta - 4 years, 1 month ago
Jonathan Quarrie
May 14, 2017

If the satellites are orbiting at the same speed, they cross paths every 2.4 hours - 0.5 of a complete orbit ( O O ).

Because the satellites are moving at the same speed, point P P can only observe a potential crossing every 0.5 of a complete rotation of Earth ( R R ) - every 12 hours.

We need to find a multiple of 2.4 that is a multiple of 12, but also the first such instance where the progress of the satellites within a given orbit ( O O ) matches the progress of P P within a given rotation of Earth ( R R ) - I.e. Where the satellites & P P are specifically on the same side of Earth.

O ( m o d 1 ) = R ( m o d 1 ) O\pmod{1} = R\pmod{1}

Orbits ( O O ) O ( m o d 1 ) O\pmod{1} Rotations ( R R ) R ( m o d 1 ) R\pmod{1} Hours
0.5 0.5 \color{#D61F06}0.5 0.1 0.1 0.1 \color{#D61F06}0.1 2.4
1.0 0.0 \color{#D61F06}0.0 0.2 0.2 0.2 \color{#D61F06}0.2 4.8
1.5 0.5 \color{#D61F06}0.5 0.3 0.3 0.3 \color{#D61F06}0.3 7.2
2.0 0.0 \color{#D61F06}0.0 0.4 0.4 0.4 \color{#D61F06}0.4 9.6
2.5 0.5 \color{#20A900}0.5 0.5 0.5 0.5 \color{#20A900}0.5 12 \large \boxed{12}

Thanks goes to @Pranshu Gaba for pointing out that we should check that the crossing of the orbits and point P P are on the same side of Earth.

For example, with an orbit time of 9 hours, the first multiple of 4.5 that is a multiple of 12 is 36, but while the timing of a crossing and a 12 hour interval has been met, point P would be on the opposite side of Earth to the satellites (as Earth would be half way through its 2nd rotation, while the satellites complete their 4th full orbit).

Nicely presented :) There still remains a small part to check, that P P is facing where the satellites are meeting, and not the opposite side.

Pranshu Gaba - 4 years ago

Log in to reply

That's very true. Good spot! Thank you for correcting me.

We can check that the number of complete orbits of the satellites O O and the number of complete rotations of the earth R R are aligned when O ( m o d 1 ) = R ( m o d 1 ) O\pmod{1}=R\pmod{1} .

I think that's the correct LaTeX for what I mean.

Do you mind if I add this to my solution?

Jonathan Quarrie - 4 years ago

Log in to reply

Yes, the LaTeX looks right. Please go ahead and add it to your solution! Another way of putting this is:

We want to find the first odd multiple of 2.4 that is also an odd multiple of 12, or the first even multiple of 2.4 that is also an even multiple of 12, whichever is smaller.

Pranshu Gaba - 4 years ago
Ayon Ghosh
Apr 29, 2017

Relevant wiki: Uniform Circular Motion

We have to also take into account rotation of the earth.

So for one satellite ω r e l a t i v e ω_{relative} = = ω E a r t h ω_{Earth} + + ω s a t e l l i t e ω_{satellite}

Whereas for the other satellite ω r e l a t i v e ω_{relative} = = ω s a t e l l i t e ω_{satellite} - ω E a r t h ω_{Earth}

We know, ω E a r t h ω_{Earth} = = 2 π / T 2π / T = = 7.27 1 0 5 r a d s 1 7.27 * 10^{-5} rad s^{-1}

(where T = 24 h o u r s = 86400 s T = 24 hours = 86400 s )

Also ω s a t e l l i t e ω_{satellite} = = 2 π / T 2π / T = = 3.63 1 0 4 r a d s 1 3.63 * 10^{-4} rad s^{-1}

(where T = 4.8 h o u r s = 17280 s T = 4.8 hours = 17280 s )

Calculating,

ω s a t e l l i t e A ω_{satellite A} = = 2.903 1 0 4 r a d s 1 2.903 * 10^{-4} rad s^{-1}

ω s a t e l l i t e B ω_{satellite B} = = 4.357 1 0 4 r a d s 1 4.357 * 10^{-4} rad s^{-1}

Converting their angular velocities into Time periods,

T A T_{A} = = 2.164 1 0 4 s 2.164 * 10^4 s = = 6.011 h o u r s 6.011 hours = = 6 h o u r s 6 hours (estimation)

T B T_{B} = = 1.442 1 0 4 s 1.442 * 10^4 s = = 4.005 h o u r s 4.005 hours = = 4 h o u r s 4 hours . (estimation)

The time after which they meet again is just their LCM which is 12 12 h o u r s hours .

Hence the satellites will cross each other at 00 : 00 00:00 h o u r s hours .

Unfortunately, because of your estimation, they will not be exactly overhead in 12 hours. They will just be "pretty closely overhead".

Calvin Lin Staff - 4 years, 1 month ago

Log in to reply

Alright, I just determined that you didn't need to do the approximations. The "apparent" time period is indeed 4 and 6 hours exactly, and you were off because you approximated in your solution.

Calvin Lin Staff - 4 years, 1 month ago

@Calvin Lin Thanks a lot for the necessary corrections.

Ayon Ghosh - 4 years, 1 month ago
Yuchang Liu
May 10, 2017

A,B相遇一次所需时间:2.4h
A,P相遇一次所需时间:3h
B,P相遇一次所需时间:2h
取最小公倍数:12h


English translation:

A, B meet once required time: 2.4h

A, P meet time: 3h

B, P meet once required time: 2h

Take the least common multiple: 12h

How do you find the meeting time of A, P and B, P?

Rohit Gupta - 4 years, 1 month ago

Log in to reply

A,P :T(AP)=2pi/(2pi/T-2pi/T(P))=3h
B,P :T(BP)=2pi/(2pi/T+2pi/T(P))=2h 其中T=4.8h,T(P)=24h

Yuchang Liu - 4 years, 1 month ago

Si marcamos la velocidad angular por Revoluciones/hora a partir del periodo de los satelites y de la tierra, y sabiendo que son movimientos no inerciales, podemos marcar una velocidad relativa de los satelites respecto la tierra restando la velocidad angular de esta a la de los satelites. Si metemos esto en ecuaciones de ángulo recorrido y sumamos una vuelta al satelite A, dado que este tiene que recorrer un periode antes de volverse a encontrar con el punto P, la variable de tiempo se despeja como 12 h.

English translation:

If we mark the angular velocity by Revolutions/hour from the period of the satellites and the Earth and knowing that they are non-inertial movements, we can mark a relative speed of the satellites with respect to the Earth subtracting the angular velocity of this one to one of the Satellites. If we put this in angle equations and add a return to satellite A, since it has to travel a period before returning to meet the point P, the time variable is cleared as 12 h.

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...