Slowing Down Stars


One of the long standing challenges in stellar astronomy, is explaining why stars rotate so slowly. Given their large masses, as they collapsed to form, they should spin up to the point of flying apart, preventing them from ever reaching the point that they could ignite fusion. To explain this rotational braking, astronomers have invoked an interaction between the forming star’s magnetic field, and forming accretion disc. This interaction would slow the star allowing for further collapse to take place. This explanation is now over 40 years old, but how has it held up as it has aged?

One of the greatest challenges to testing this theory is for it to make predictions that are directly testable. Until very recently, astronomers were unable to directly observe circumstellar discs around newly formed stars. In order to get around this, astronomers have used statistical surveys, looking for the presence of these discs indirectly. Since dust discs will be warmed by the forming star, systems with these discs will have extra emission in the infrared portion of the spectra. According to the magnetic braking theory, young stars with discs should rotate more slowly than those without. This prediction was confirmed in 1993 by a team of astronomers led by Suzan Edwards at the University of Massachusetts, Amherst. Numerous other studies confirmed these general findings but added a further layer to the picture; stars are slowed by their discs to a period of ~8 days, but as the discs dissipate, the stars continue to collapse, spinning up to a period of 1-2 days.

Another interesting finding from these studies is that the effects seem to be most pronounced for stars of higher mass. When similar studies were conducted on young stars in the Orion and Eagle nebulae, researchers found that there was no sharp distinction between stars with or without disks for low mass stars. Findings such as these have caused astronomers to begin questioning how universal the magnetic disc braking is.

One of the other pieces of information with which astronomers could work was the realization around 1970 that there was a sharp divide in rotational speeds between high mass stars and lower mass ones at around the F spectral class. This phenomenon had been anticipated nearly a decade earlier when Evry Schatzman proposed that the stellar wind would interact with the star’s own magnetic field to create drag. Since these later spectral class stars tended to have more active magnetic fields, the braking effect would be more important for these stars.

Thus astronomers now had two effects which could serve to slow rotation rates of stars. Given the firm theoretical and observational evidence for each, they were both likely “right”, so the question became which was dominant in which circumstance. This question is one with which astronomers are still struggling.

To help answer the question, astronomers will need to gather a better understanding of how much each effect is at work in individual stars instead of simply large population surveys, but doing so is tricky. The main method employed to examine disc locking is to examine whether the inner edge of the disc is similar to the radius at which an object in a Keplarian orbit would have a similar angular velocity to the star. If so, it would imply that the star is fully locked with the disc’s inner edge. However, measuring these two values is easier said than done. To compare the values, astronomers must construct thousands of potential star/disc models against which to compare the observations.

In one recent paper astronomers used this technique on IC 348, a young open cluster. Their analysis showed that ~70% of stars were magnetically locked with the disc. However, the remaining 30% were suspected to have inner disc radii beyond the reach of the magnetic field and thus, unavailable for disc braking. However, these results are somewhat ambiguous. While the strong number of stars tied to their discs does support the disc braking as an important component of the rotational evolution of the stars, it does not distinguish whether it is presently a dominant feature. As previously stated, many of the stars could be in the process of evaporating the discs, allowing the star to again spin-up. It is also not clear if the 30% of stars without evidence of disc locking were locked in the past.

Research like this is only one piece to a larger puzzle. Although the details of it aren’t fully fleshed out, it is readily apparent that these magnetic braking effects, both with discs and stellar winds, play a significant effect on slowing the angular speed of stars. This runs completely contrary to the frequent Creationist claim that “[t]here is no know [sic] mechanical process which could accomplinsh [sic] this transfer of momentum”.

27 Replies to “Slowing Down Stars”

  1. Jon,

    In paragraph 1 line 4 you might mean braking as in putting the brakes on rather than breaking, as in ‘a dish falls on the floor and never be the same again’, or breaking up — is hard to do.

    Also paragraph 2, line 6, and paragraph 3 line 5, — same incorrect word choice.

    You have it correct in the 4th paragraph line 6, and paragraph 7, lines 4 & 6, and paragraph 8, line 2.


      1. Carpenters? I thought that Mary was referring to increasing entropy ! 😉

    1. Yo Mary, you didn’t spot the missing possessive apostrophe in “stars” in the fifth line of the fourth paragraph: “… the stars own magnetic field…”

  2. This runs completely contrary to the frequent Creationist claim that “[t]here is no know [sic] mechanical process which could accomplinsh [sic] this transfer of momentum”.

    Doz Creationists Haz Bad Edumahkashun!

  3. It is a complete riddle to me why all heavenly bodies rotate in the first place. If a nonrotating dust cloud collapses into a denser body the angular momentum shopuld be zero! So where does all this angular momentum come from in the first place?

    1. Takojohn
      A sphere of dust collapsing under its own gravity is never going to undergo absolutely uniform collapse so is bound to move in some direction, the slightest random rotation will get accentuated as the cloud collapses, like a skater pulling in outstretched arms so the laws of the universe make it a dead cert. Anything that doesn’t spin would be the exception.

    2. Oops – clicked ‘like’ when I wanted to click reply.

      But no clouds of dust in space are perfectly non-rotating! Ultimately, we don’t know what the net ang. mom. of the universe is. It could be zero – it may not be. But as far as an individual cloud of dust being perfectly non-rotating, it is so unlikely as to effectively be impossible. It’s like balancing a pen on its tip. You could in theory do this, but there are far more states available to the system in which it is not perfectly balanced than the one in which it is. Couple with that the fact that both systems are unstable and will actively move further from the ‘balanced’ situation as they evolve to lower energy states, and we can pretty much say that all astrophysical systems will rotate to some degree.

      1. Oops – clicked ‘like’ when I wanted to click reply.

        In that case, just click on “Like” again to ‘unlike’.

  4. There might be another way of looking at it.

    I always thought that a star with an accretion disk should spin up if it is drawing in mass (ice skater pulls their arms in and spins faster). I wouldn’t have thought magnetic braking on the disk could overcome that. Really big stars may also unload energy via polar jets but eventually limit their own growth when they reach the Eddington limit (of radiation pressure). These reasons may be adequate to explain why a star never spins up to the extent that it can spin itself apart.

    The accretion disk eventually dissipates – either due to depletion or due to radiation pressure after the star goes main sequence. From there, the star’s spin slows due to magnetic braking on its own stellar wind particles. There is good correlation between the age of stars and their rate of spin (i.e. old equals slow) – but that can change if they begin to accrete new material again and spin up.

  5. this topic has my head spinning.
    how do we explain the generation of magnetic fields in matter of “other” states of collaps ?
    the crab pulsar obviously has one and the jets of black holes indicate association also.
    what is the mechanism?
    these objects are supposedly entirely composed of neutrons.

  6. this topic has my head spinning.
    how do we explain the generation of magnetic fields in matter of “other” states of collaps ?
    the crab pulsar obviously has one and the jets of black holes indicate association also.
    what is the mechanism?
    these objects are supposedly entirely composed of neutrons.

    1. 1) A neutron star may consist of only neutrons, but neutrons are made of quarks (one up- and 2 down-quarks), and these are charged. It is also possible that a neutron star does not consist of “solid” neutrons below its surface, but rather a neutron or even a quark soup, since the pressure is really extremely high.
      The magnetic field is also a relic of its time as a normal star. Due to the collapse it gets amplified to tremendous values, sometimes exceeding 10^9 T.

      2) A black hole on itself does not have a magnetic field, since is does not contain a single charge. The magnetic field of a black hole that causes jets, is anchored in the accretion disk surrounding the hole.

      1. With regards to magnetic fields and black holes, an uncharged black hole will not hold a magnetic field. The transverse modes of particles near the horizon are dilated and increased. Therefore the field amplitudes cover the entire event horizon. As a result any charge distribution or separation in the material which enters a black hole gets “mixed” as it is frozen onto the horizon as seen from the exterior. As a result there are no currents and charge separations which can be observed, and consequently no magnetic field.

        The Reissner-Nordstrom metric involves an electrically charged black hole. If in addition this black hole is rotating there will then be a magnetic field. The electric charge defines a BPS black hole, which in many ways is similar to an elementary particle. In fact for quantum black holes they are in a way elementary particles. So just as an electron with a unit charge and spin has a magnetic dipole field, so too will a charge black hole that is also spinning.

        In practical terms an astrophysical black hole, one of 5M_{sol} or greater is not going to have much of an electric charge. It will largely be neutral, and so has a negligible magnetic field.


  7. It is known since 2005 that that Earth’s inner core rotates faster than the rest of the planet. What about a faster spinning inner core of the Sun?

  8. A simple physics demonstration is a metal disk on an axle with the strong N and S poles of an electromagnet that enclose a part of it. If you try to turn the disk with the electromagnet on it resists turning. Then if you turn the electromagnet off the wheel spins freely. The electromagnetic flux ? = ??B*da, B = magnetic field, da = unit of area the field enters sweeps around the disk as it turns. The unit area the field enters changes. Thus we have

    d?/dt = B*(da,/dt)

    where we assume a constant B field in a patch of area a, say the area of the electromagnetic poles. If the wheel is rotating with angular velocity ? this is then

    d?/dt = ?Ba

    This also gives the electromotive potential E = d?/dt. This induces a current which opposes the force that generates it. This is sometimes called Weber’s law.

    Essentially the same is going on there, but the magnetic field source is rotating. The field lines induce currents in the protoplanetary disk which oppose the pondermotive force which generates the field. This is then the braking action on the rotating star.


    1. A simple physics demonstration is a metal disk on an axle with the strong N and S poles of an electromagnet that enclose a part of it.

      I presume that you’re referring to a homopolar generator (a.k.a. Faraday disc; demonstrated here)?

      1. That is it. In this case the currents or EMF induced on the disk is tapped to run the light.


  9. French astrophysicist Évry Schatzman‘s 1962 paper (mentioned in the article) writes (page 5): “2. Electromagnetic Coupling [..] The idea is the following: we suppose that when ejection of matter occurs at the surface of a star which has a magnetic field, the ejected matter is forced to turn with the star up to the critical distance beyond which it is no longer dragged by the magnetic field. Since the distance is much larger than the radius of the star, the loss of angular momentum per unit mass is large”.

    Stephen G. Brush writes: “In 1942 [Hannes] Alfvén showed that an ionized gas surrounding a rotating magnetized sphere will trap magnetic field lines, acquire rotation, and thereby slow down the rotation of the sphere [9]. Ferraro [l0] had obtained this result earlier but did not suggest its possible use in cosmogony. Alfvén [ll], [12] proposed that the early sun had a strong magnetic field, and that its radiation ionized a cloud of dust and gas, which then trapped the magnetic field lines and acquired most of the sun’s original angular momentum.”

    A possible answer to Takojohn regarding rotation may be found in Alfvén’s 1942 paper (page 5) where he writes: “Moreover, in a recent paper(4) it has been shown that if ionized matter is brought into the neighbourhood of the sun, electromagnetic forces will tend to make it take part in the solar rotation. This effect is caused by the polarisation due to the magnetic field of the sun. As is usually the case when a conducting body moves in a magnetic field, a system of currents is produced, which tends to brake the relative motion.”

    Alfvén also notes in the same paper (same page) that “Even at the distance of Pluto a proton moving with the same velocity as Pluto is much more affected (250 times) by the magnetic force than by solar gravitation! [..] Consequently there seem to be strong indications that at the genesis of the planetary system electromagnetic forces have been more important than mechanical forces.” (emphasis in original)

    1. If you set structure formation as ordering importance, I would think it is gravitational forces (system aggregation) > mechanical (disk matter distribution) > nuclear (stellar fusion) > EM (disk momentum distribution). Alfvén did love his plasma physics.

      1. I think it depends on which structure, the constituents, and the time during its formation, and the order does not apply to everything at every time.

      2. Surely it depends on (a) the type of structure (b) it’s constituents (c) where it is in its evolution. Gravity dominates sometimes, EM dominates sometimes.

  10. Steve Nerlich wrote: “I always thought that a star with an accretion disk should spin up if it is drawing in mass (ice skater pulls their arms in and spins faster). I wouldn’t have thought magnetic braking on the disk could overcome that.”

    I believe that astronomer Vincenzo Ferraro first investigated this in a 1937 paper, resulting in “The Ferraro Co-Rotation Theorem” and “Ferraro Isorotation Law”.

    These are described in the NASA document “Transfer of Angular Momentum and Condensation of Grains>” which quantifies how gravity is balanced by both centrifugal and electromagnetic forces in a rotating mass, concluding that the gravitational force is balanced, 2/3 by the centrifugal force and 1/3 by the electromagnetic force.

  11. I’ve just stumbled upon this paper (PDF) with regard to molecular cloud rotation and magnetic braking:
    Evolution of Rotating Molecular Cloud Core with Oblique Magnetic Field.

    That paper has some nice illustrations, too!

  12. As I understand it, these discs are already truncated with an inner edge. As such, the parent star is no longer accreting matter which would spin it up. Hence, if a disc is present, but not accreting, it is a potential source of braking.

  13. The magnetic force transfers angular momentum. This will happen until the magnetic field induced currents on the disk are at a minimum or zero. So the star’s angular momentum is reduced and transferred to the disk.


Comments are closed.