Assume just for this question that in one same plane of space, there are the following stars within close proximity of each other:

  • An O-type main-sequence star (18x wider, 15-90x more massive and 40k-one million times brighter than our sun, lifespan of 10 million years)
  • A G0-type main-sequence star (105% wider, 110% more massive and 126% brighter than our sun, lifespan of 10 billion years)
  • A K0-type main-sequence star (85% as wide, 78% as massive and 40% as bright as our sun, lifespan of 20 to 70 billion years)

These three stars don't orbit each other but actually merge together into one new star, one that is bigger, brighter and longer-living than our sun. Could this happen through natural means rather than artificial means?

  • 16
    $\begingroup$ I'd add to the answers this bit - it is impossible for the resulting star to be longer-living than the Sun, since the more massive stars burn up more quickly. $\endgroup$ Dec 27, 2020 at 11:30
  • $\begingroup$ This seems like a better question for astronomy.stackexchange.com. I'm not sure how this question related to worldbuilding, as it's not explicitly stated it's part of a story or game. It's just a direct question about astronomical possibilities. $\endgroup$ Dec 28, 2020 at 20:22
  • $\begingroup$ @computercarguy Because world comes before history, and history comes before story. $\endgroup$ Dec 28, 2020 at 20:53
  • $\begingroup$ @JohnWDailey, that may happen many times, but it's not a rule. I've had many stories come to me in a dream that I'd have to build a world and history around after the fact. $\endgroup$ Dec 28, 2020 at 21:09

4 Answers 4


Yes! Stellar collisions do happen naturally. For collisions to be likely, we need environments where stars are naturally close together - in general, the rate of collisions per unit volume is proportional to the square of the number density. Therefore, high-density star clusters are optimal; in globular clusters, we can have central densities of $\sim10^3$ stars per cubic parsec, roughly four orders of magnitude greater than the local stellar number density. This means that mergers should happen at a correspondingly astounding rate.

In globular clusters, we know of a population of stars known as blue stragglers. These are objects formed by the merger of two stars, which appear as stars that are massive and more luminous than the rest of the stars in the cluster. They should subsequently evolve much as one would expect stars of the appropriate mass and composition.

A triple merger, though, does present some problems, in that even in a globular cluster, it's unlikely that a merger product will undergo a second collision before evolving away. The expected time before a given star undergoes a close encounter is $$\tau\sim10^{11}\left(\frac{n}{10^5\;\text{pc}^{-3}}\right)^{-1}\left(\frac{M}{M_{\odot}}\right)^{-1}\left(\frac{R}{R_{\odot}}\right)^{-1}\left(\frac{v}{10\;\text{km s}^{-1}}\right)\;\text{years}$$ where the factors of $M$ and $R$ arise because the cross-section of the star depends in part on its physical size and in part on gravitational focusing, the effect that increases the star's effective cross-section through its gravitational pull.

Let's say the two lowest mass stars merge. Even assuming no mass is lost in the collision, they should have a combined mass of $M\approx1.9M_{\odot}$; by applying the appropriate mass-radius homology relation, we can assume it has a radius of roughly $R\approx1.5M_{\odot}$. In a cluster, at the lower end, we might expect to see $v\sim10\;\text{km s}^{-1}$. Putting this together, we expect that it will take the merger remnant $\sim3\times10^{12}$ years before it collides with another star, assuming a number density $\sim10^3\;\text{pc}^{-3}$. By this point, it will have evolved far off the main sequence and become a white dwarf.

The same may be true if the collision happens in any other order, e.g. if the O-type star and the G-type star merge. Thanks to the greater mass and radius, the product will have a larger cross-section, but it will also live for much less time. The scenario could be saved if the number density at the core of the cluster is several orders of magnitude higher, though, which could very well be possible in some of the denser globular clusters.

  • $\begingroup$ And when they do collide, do they add to their sizes, luminosities and lifespans? $\endgroup$ Dec 27, 2020 at 2:59
  • 3
    $\begingroup$ @JohnWDailey -- I think, rather by definition, that when two stars collide, their own lifespans have just come to an end! The question then becomes: what are the expected luminosity, size & lifespan of the newly agglomerated star? $\endgroup$
    – elemtilas
    Dec 27, 2020 at 4:35
  • 13
    $\begingroup$ @JohnWDailey The masses are added; the luminosities and lifespans aren't added because luminosity and lifetime aren't linear functions of mass. Rather, you have to recalculate the lifespan and luminosity of the merger product based on the mass/lifespan and mass/luminosity relations. $\endgroup$
    – HDE 226868
    Dec 27, 2020 at 4:47
  • 16
    $\begingroup$ @JohnWDailey - my naive assumption would be that the lifespan of the new star would be dramatically less than that of its components, as larger, higher-mass stars tend to be much more short-lived than smaller, lower-mass stars. $\endgroup$
    – jdunlop
    Dec 27, 2020 at 10:22
  • 3
    $\begingroup$ @jdunlop: Though I wonder how much of the stars' mass would actually merge, and how much would be blown off due to the violence of the merger. IOW, if you have two stars merging, each of mass 1.0, I would think the mass of the resulting star would be something less than 2.0. But how much less? $\endgroup$
    – jamesqf
    Dec 27, 2020 at 18:00

YES, stars can merge

NO, they end result will not be "longer-living than our sun", especially not when the constituent parts include an O-type star.

The much more likely result of an O-type merging with anything else is a short, violent Supernova.

Even if by some miracle you manage the merger without causing core disruption of all the stars involved, adding mass to a star will drastically reduce its remaining lifespan, never lengthen it.

  • 2
    $\begingroup$ I would dispute the claim that the merger would lead to a supernova - unlike in the case of a white dwarf, there's no degenerate matter, so runaway fusion is unlikely. The resulting star would presumably regulate its temperature by briefly expanding to a new equilibrium, but it wouldn't explode. $\endgroup$
    – HDE 226868
    Dec 27, 2020 at 18:17

In addition to the answer by HDE 226868, there is another possible form of stellar collision; the Thorne–Żytkow object (TZO's). TZO's (theoretically) involve the collision of a neutron star with another star, perhaps a red giant or supergiant. The neutron star will migrate to the core position, but will typically be much hotter than the giant star. The result resembles a Wolf-Rayet star with some different chemistry due to higher core temperatures.

There has been only one observed serious candidate for this type of merger (HV 2112), and it is, apparently, low certainty. But TZO's are a currently accepted idea in astrophysics and are definitely an option for you.


Merging two main-sequence stars will result in another (bigger, brighter, shorter-lifespan) main-sequence star.

I am not sure how the transition period will look like, but:

The merger will result in gross turbulence in the star material that will likely reset the resulting star somewhere on the corresponding Hayashi track or Henyey track. After some time, the star will settle on the main sequence, higher and to the left of all the constituents.

There proably will be quite a firework in the period of the physical merger as hotter gas from inside of the stars will be mixed into the outer layers. Some amount of gas may be lost in the process.

The lifespan of the resulting star is determined by the resulting mass and the time each one of the merging stars already spent on the main sequence (i.e. the percent of hydrogen already spent). Bigger stars have shorter lives and the resulting star will live shorter than its constituents' projected lifespans.

Adding a third star into the mix will rather not change the picture much.

The question implies that all 3 stars are main-sequence type. If one of them is a red giant or a white dwarf, the added matter above the already-too-dense core may trigger a supernova explosion of some type.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .