For over 20 years, the binary star DI Herculis has been measured to have anomalous precession, inconsistent with the predictions of general relativity, in the limit of two spherical masses orbiting with the measured orbital parameters.
So, either there were some other classical torques in the system, or general relativity was wrong.
DI Herculis is a close B star binary, with 10 day orbital period and eccentricity of ~0.5, that is orbiting near edge-on to our line of sight, and thus the stars eclipse each other.
The stars are hot and massive - 5.15 and 4.52 solar masses respectively, with orbital radii of about 2.5 times that of the Sun. System age is likely few tens of millions of years.
The orbit of the stars precesses, and is expected to undergo measurable relativistic precession, analogous to the orbit of Mercury around the Sun, but when measured, the actual precession was signficantly larger than expected from relativistic effects alone.
As you can imagine, this 'caused some flights of fancy among theorists, though to be fair most noted that a number of classical perturbations could provide a torque of the necessary magnitude (and sign) to drive the anomalous precession, and the bet has mostly been that we'd eventually nail this sucker down.
Question was which additional physics was there.
Well, now we know: "Misaligned spin and orbital axes cause the anomalous precession of DI Herculis" by Albrecht et al, just out in Nature this week (Nature letter (sub. required)).
The stars are not spin aligned, and the spins are not aligned with the binary orbital axis but are near orthogonal.
Basic technique is the Rossiter-McLaughlin effect, where spectral lines in eclipsing binaries are distorted due to the projected spin of the stars - basically there is differential doppler shifting of different parts of the stellar surface due to the projected rotation.
Deconvolving this is hard, and requires good measurements of detailed spectral profiles during eclipses, combined with lots of fits to models. Which the authors did, using high resolution, high S/N repeated measurements of a magnesium line at primary and secondary eclipse, using a 2m telescope.
DI Her seems to have stars rotating with inclination angles of ~70 and ~80 degrees respectively, relative to the orbital axis.
That will do it.
Nice result.
Now we just have to figure out how the hell the stars ended up this way...
- Log in to post comments
More like this
No fair, I was going to blog about that.
I think this paper is a lovely piece of research.
The fact that they could use a 2m telescope to work out this particular problem is great - you don't always need an 8 meter/space-based telescope to do great astronomy.
The figures in the paper are also lovely work - they have the line profile data, and the two theoretical cases for rotation axes parallel to the orbital axis, and their best fit model. And the best fit model agrees beautifully with the data!
It's great that the figures pretty much tell the whole story in one reading, and it's a confirmation of relativity to boot. Great stuff.
Hah! It is mine now, forever more... Hah!
Or you could, like, blog on it.
It deserves some air.