Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Do stars actually collide in galactic collisions?

  1. Oct 29, 2011 #1
    Will some of the stars of the Milky Way crash into those of Andromeda if these galaxies merge?

    What are the odds? How large a share of the stars will this happen to?

  2. jcsd
  3. Oct 29, 2011 #2


    User Avatar
    Gold Member

    Writing in the book, "The Chilling Stars", science writer and Cambridge educated physicist Nigel Calder says of galactic collisions,

    "Even with so many billions of stars involved, in each of the participating galaxies, the spaces between stars are so wide that the chances of a direct collision of two stars are slim. Instead, the high-speed encounter of gas carried by the galaxies creates shock waves that compress the gas and provoke its collapse into new born stars..."

    Perhaps the Andromeda Galaxy will collide and merge with the Milky Way in 5 billion years, producing colossal starburst events.

    Respectfully submitted,
  4. Oct 29, 2011 #3


    User Avatar
    Gold Member

    To add slightly to Dotini's excellent answer, I read some where that although the odds of stars actually colliding is essentially zero, the odds of some solar systems being perturbed by near misses is slightly higher (STILL, as I recall, quite small)
  5. Oct 29, 2011 #4
    For some perspective, we can scale stars down to about the size of a BB (or a pea). Assuming sun sized stars, and 4 mm sized BBs, the scale is about 347,750,000,000:1. If we assume 5 LY between stars the distance between our BB sized stars is about 136 km (85 miles). So imagine a cloud of BBs each about a hundred km away from the next nearest one. If two of those clouds 'collided' it would be pretty unlikely for any two BBs to collide. Of course, there are billions of stars/BBs and in the core the density goes up a lot. There will be some collisions, but for any individual star (like the sun) a collision is very unlikely.
  6. Oct 29, 2011 #5
    Ok, thanks.

    What if we leave aside galactic mergers for the moment. Do stars ever collide within galaxies as part of the normal way galaxies evolve? Under what kind of circumstances can this happen?
  7. Oct 29, 2011 #6
    It's all a matter of scale.

    "The core of the Sun is considered to extend from the center to about 20–25% of the solar radius. It has a density of up to 150 g/cm3"

    radius = 6.96E+010 cm
    core radius = 1.39E+010 cm
    core volume = 1.13E+031 cm3
    density = 100 g/cm3
    mass = 1.13E+033 g
    percent H = 50.00%
    moles H = 5.65E+032 mol
    avogadro's number = 6.02E+023
    H atoms = 3.40E+056
    Stars in both galaxies = 1.00E+012
    ratio atoms in sun's core to stars in galaxies = 3.40E+044

    Obviously there's a lot of estimating there. But, even if they are all off by an order of magnitude there is still many times more atoms in a star than stars in a galaxy.
  8. Oct 30, 2011 #7


    User Avatar
    Science Advisor
    Gold Member

    A head on collision would probably produce a few stellar collisions, but, andromeda will probably only rub halos with the milky way. It should still produce an amazing star burst event.
  9. Oct 30, 2011 #8

    Vanadium 50

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    It depends on your definition of "collide". There is often mass transfer between stars in binary systems. This is a process that takes thousands or millions of years. The end result of this may be a single star.

    There exists a category of stars called "blue stragglers", which appear to be formed by this process. They exist only in dense star regions in clusters, which suggests interactions between stars are responsible. There was a debate about whether these were formed by "mass transfer" or by "collisions", but it's now generally recognized that these are statements about the speed of the process, not the nature. If the two stars orbit a million times before merging, it's called "mass transfer". If they orbit ten times, it's a "collision". But it's the same process; the only open issue is how long it takes.

    What evidence he have indicates that blue stragglers form under a variety of time scales.
  10. Oct 30, 2011 #9


    User Avatar
    Gold Member

    Good calculation.

    Stars are governed by gravitational attraction. For any two stars to collide they would have to have a velocity vector directed towards each others centre of mass. Any sideways motion and the 2 stars would swing around each other. Perturbations from another third or more celestial bodie(s) could change the trajectory to a hit or a miss. Chances are that it would be a miss.

    For the core of a sun, the particles are ionized, so a given particule would be subject to attraction and repulsion from other particules. That would change the dynamics to what -more favourable or less favourable for a collision?
  11. Nov 16, 2011 #10
    Stars called blue stragglers that are much younger than the rest of the stars composing the Messier 30 globular cluster are considered remnants of stellar collisions.
  12. Jan 10, 2012 #11
    Dont neutron stars sometimes collide releasing massive amounts of energy?
  13. Jan 13, 2012 #12
    Galactic collisions aren't necessary for two stars to collide. Just as there is the danger of a rogue black hole appearing out of nowhere so too a rogue star might appear out of nowhere and collide with another. In places where distances between stars are closer, such as in the center of globular clusters or galactic hubs, such collisions are more likely.

    These super dense regions also produce the supervelocity rogue stars that are gravitationally hurled out into interstellar or intergalactic space when the gravitational relationship between binaries changes when one of the two goes supernova or gets too close to a black hole and is partially disrupted or gets pulled in.

    The trajectory of the remaining star when its gravitatinl tether "breaks" is unpredictable just as a rogue black hole's trajectory is. Very unlikely? Sure. Utterly impossible? By no means.
  14. Jan 13, 2012 #13


    User Avatar
    Staff Emeritus
    Science Advisor
    2018 Award

    Almost certainly. There are many binary systems composed of double neutron stars that will eventually spiral into each other.
  15. Jan 13, 2012 #14
    Would the velocity vector of star A necessarily be pointed at center mass of star B (and vice versa) for them to collide? depending on the relative speed of galactic collision, might it not be possible that many stars enter degrading binary orbits such that they produce collisions. not necessarily head on collisions, but more of a spiral into each other collision.

    Also, depending on the local stellar distribution density, it may not be the case that there are numerous collisions but rather stars are stripped of their satellites by the pertubations of passing stars.
  16. Jan 13, 2012 #15


    User Avatar
    Staff Emeritus
    Science Advisor
    2018 Award

    It's possible, but usually the case is that the stars are "flung" away from each other instead of becoming a binary system. I believe the two stars would have to have comparable velocities in the same general direction and be slowly moving together for them to enter a stable binary orbit. Not 100% sure on that though, I've never done the math and all.
  17. Jan 13, 2012 #16


    User Avatar
    Gold Member

    I cannot, unfortunately, back this with any reference or hard evidence, but I remember reading somewhere that if two galaxies with a billion stars each collide, the number of stars that actually hit each other would be literally a handfull. That is, something like 5 collisions or so.
  18. Jan 13, 2012 #17
    well if you consider the typical size of a star, compare it to the typical distance between stars in a galaxy, then figure out the average density of stars in the galaxy, would that tell you how likely it would be for stars to collide during a "galaxy interaction event"?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook