Runner Alignments on Circular Tracks: Meeting Times and Speed Combinations

  • Context: Graduate 
  • Thread starter Thread starter Gerenuk
  • Start date Start date
Click For Summary

Discussion Overview

The discussion revolves around the behavior of multiple runners on circular tracks with varying speeds, specifically whether they will eventually align closely together over time. Participants explore the mathematical implications of their speeds, particularly focusing on conditions under which they might meet and the challenges posed by rational and irrational speeds.

Discussion Character

  • Exploratory
  • Technical explanation
  • Debate/contested
  • Mathematical reasoning

Main Points Raised

  • Some participants propose that for n runners with different speeds, there exists a time t such that they can be arbitrarily close together, expressed mathematically as ∀i, ε>0, ∃t: [t·s_i]<ε.
  • Others suggest simplifying the problem by considering whole numbers for speeds that are coprime, asserting that they will align radially again at some time t.
  • A participant expresses concern about the meeting times of three runners compared to a fourth, questioning whether the group of three could separate before meeting the fourth runner.
  • One participant discusses the period of each runner and the possibility of finding positive integers that satisfy alignment conditions, noting the complexity introduced by rational and irrational periods.
  • Another participant acknowledges the difficulty of achieving an exact match in alignment due to the nature of rational approximations and the potential for significant deviations over time.
  • Some participants clarify that while rational numbers can be approximated, the same approach may not hold for irrational numbers, complicating the proof of alignment.
  • There is a discussion about the impracticality of seeking exact matches after a certain number of revolutions, with some emphasizing the importance of approximate matches instead.
  • A later reply corrects an earlier assumption about the nature of meeting times, emphasizing the need for clarity on whether the runners meet within a specified distance rather than at exact points.

Areas of Agreement / Disagreement

Participants express a range of views on the conditions under which runners will meet, with some agreeing on the feasibility of alignment for rational speeds while others highlight the complications introduced by irrational speeds. The discussion remains unresolved regarding the exact conditions and implications of these meeting times.

Contextual Notes

Limitations include the dependence on the definitions of speeds (rational vs. irrational) and the unresolved nature of the mathematical steps required to prove alignment under various conditions.

Gerenuk
Messages
1,027
Reaction score
5
If I have n runners on a circular tracks at different speeds [itex]s_i[/itex], will they always meet up arbitrarily close together in a group?

So does there always exist a time t such that
[tex] \forall i,\varepsilon>0 ,\exists t: [t\cdot s_i]<\varepsilon[/tex]
where the bracket denote the fractional (non-integer) part.

And if that time always exists, what is the combination of runner speeds such that it takes them longest time to meet up again?
 
Physics news on Phys.org
Gerenuk said:
If I have n runners on a circular tracks at different speeds [itex]s_i[/itex], will they always meet up arbitrarily close together in a group?

So does there always exist a time t such that
[tex] \forall i,\varepsilon>0 ,\exists t: [t\cdot s_i]<\varepsilon[/tex]
where the bracket denote the fractional (non-integer) part.

And if that time always exists, what is the combination of runner speeds such that it takes them longest time to meet up again?

Lets make this into a simple problem using whole numbers to specify the number of revolutions per minute A1,A2,A3,...A(n) all "n" whole numbers being coprime to each other. If they all start up on the same radius, at a time t they will indeed all be radially aligned again. Can you prove this?
 
ramsey2879 said:
Lets make this into a simple problem using whole numbers to specify the number of revolutions per minute A1,A2,A3,...A(n) all "n" whole numbers being coprime to each other. If they all start up on the same radius, at a time t they will indeed all be radially aligned again. Can you prove this?
All rationals is easy to prove I guess, so I'm more worried about algebraic independent numbers (if that's the correct term). And it's easy to show that 3 runners will meet.
But I'm not sure about more of them, since it might take so long before this 3 group meets the 4th runner, so that the 3 group might split by that time.
I found http://eom.springer.de/k/k055910.htm but that doesn't help either?
 
Here is my take on the problem. I hope I'm not wrong bigtime. : ) -- Edit P.S.: oh well, this was the "easy" part that Gerenuk referred to.

Let [itex]T_i = 2 \pi R / s_i[/itex] be the period (time to complete one revolution) of each runner.

Then you look for positive integers [itex]n_i[/itex], such that all [itex]n_i T_i[/itex] are equal.

Since some [itex]T_i[/itex] can be rational and some irrational, this won't be generally possible. But you can always find rational approximations to your [itex]T_i[/itex] (f.i. by a bisection method), as close to the originals as you wish.

So assume now that all [itex]T_i[/itex] are expressed as the reduced fraction [itex]a_i / b_i[/itex], with [itex]a_i, b_i[/itex] coprime integers.

Let [itex]L[/itex] be the least common multiple of all [itex]b_i[/itex]. The integer [itex]c_i = a_i L / b_i[/itex] is the numerator of each fraction, when expressed with [itex]L[/itex] as the (common) denominator. Let [itex]C[/itex] be the lcm of all [itex]c_i[/itex]; now each [itex]n_i[/itex] will be [itex]C / c_i[/itex]. As we have taken the least common multiple on each step, there are no smaller integers satisfying this condition.
 
Dodo said:
But you can always find rational approximations to your [itex]T_i[/itex] (f.i. by a bisection method), as close to the originals as you wish.
As you noticed there is a tricky part :smile:

The rational approximation might need such long times to meet up, so that the small deviation from rationals isn't small anymore.
 
Oh, I was wrong. That guy Kronecker did actually prove it and the link I mentioned contains the statement.
 
I am pretty much lost in that topology link, but I do think there is a problem with my argument. Looking for an *exact* match after n_i revolutions each, given rational approximations of the periods, is exceptionally impractical. What you were asking is if the runners *approximately* meet after some revolutions, which is a completely different question. My apologies for the pointless ornamentation. : )
 
Dodo said:
Looking for an *exact* match after n_i revolutions each, given rational approximations of the periods, is exceptionally impractical. What you were asking is if the runners *approximately* meet after some revolutions, which is a completely different question.
Your result for rational numbers solves the problem. An exact match is of course also an approximate match.
Yet, for irrational numbers this proof doesn't work. In the problem I pose I'm looking for an "arbitrary close" match (not exact though), but it's not that trivial to show that this occurs. You can of course approximate all number by rational, but this still existing little deviation might be blown up, once the number is multiplied by a large number of revolutions.
 
Gerenuk said:
Your result for rational numbers solves the problem. An exact match is of course also an approximate match.
Not really. Two runners with periods 333/1000 and 667/1000 will meet at 222111 turns of the first, while what you probably wanted to know is that they nearly meet every other turn at the beginning.

(Edit: sorry, I meant 667 turns of the first, not 222111. But you get the idea.)
 
  • #10
You were right in the first place with your derivation.
I actually have "no time" limit. I wanted to know if they ever meet within any given distance.
 

Similar threads

  • · Replies 12 ·
Replies
12
Views
19K
  • · Replies 15 ·
Replies
15
Views
1K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 61 ·
3
Replies
61
Views
13K
Replies
11
Views
5K
  • · Replies 4 ·
Replies
4
Views
4K
  • · Replies 16 ·
Replies
16
Views
7K
  • · Replies 28 ·
Replies
28
Views
7K
  • · Replies 175 ·
6
Replies
175
Views
28K
  • · Replies 46 ·
2
Replies
46
Views
6K