Going Further 10.5: Gravity Probe B
Gravity Probe B (GPB) was an experiment developed at Stanford University and flown by NASA in 2004. However, nearly 50 years of planning and development had gone into creating the spacecraft that ultimately made the measurement of frame-dragging around Earth.
The experiment was designed to test a prediction of general relativity that was first calculated by two Austrian physicists, Josef Lense and Hans Thirring, in 1918; the effect is often called the Lense-Thirring effect to honor the scientists who first proposed it. GPB tested the effect with four sensitive gyroscopes, each composed of rapidly spinning 1.5-inch-diameter spheres made of fused quartz coated with a thin layer of niobium. The spheres were the most precisely produced in history, being perfectly spherical to 3 parts in 10 million. To put this in perspective, the maximum difference in the distance from the center of any of the spheres to any point on its surface was less than a few atomic diameters.
The spherical gyroscopes were suspended inside chambers and spun up to high spin rates by gently blowing on them with streams of helium gas. They were prevented from hitting the sides of the chambers by electrical fields that repelled them. These precautions were required because any contact with the walls of the chambers or with a mechanical device would destroy their sphericity, thus nullifying their use in the experiment. For the same reason, no mechanical bearings were used to support the spheres, as is typically done with gyroscopes. To make this measurement, the gyroscopes had to be completely isolated from the spacecraft, orbiting Earth independently within the enclosure of the spacecraft.
The measurement was made by aligning each of the four gyroscopes such that its spin axis pointed to a particular star; IM Pegasi was used for this experiment. The spacecraft also had a tracking telescope that was trained on the star and used to maintain the orientation of the spacecraft itself. As GPB orbited Earth in a polar-aligned orbit, the very weak frame-dragging effect gently tugged on each gyroscope, causing the orientation of its spin axis to slowly drift off of the star. After a year running the experiment, and more than 5000 orbits of Earth, the very small deviation of the spin axis could be measured. It amounted to a deflection rate of only about 37 milli-arcseconds per year (mas/yr), with an uncertainty of 7 mas/yr. The prediction of general relativity is 39 mas/yr, well within the range of uncertainty of the experiment.
Gravity Probe B measured a second prediction of general relativity simultaneously to measuring frame-dragging. This effect, called the geodetic effect, is due to the warping of spacetime by Earth’s mass and is distinct from the wrapping effect caused by Earth’s spin. The geodetic effect would be present even if Earth were not spinning. General relativity predicts that the geodetic effect should cause each gyroscope to be deflected by about 6606 mas/yr in a direction perpendicular to that of frame-dragging. The deflection measured was 6601 mas/yr with an uncertainty of 18 mas/yr. Again, the agreement between experiment and theory is quite good. This measurement basically confirmed earlier measurements related to the gravitational redshift: an earlier mission, Gravity Probe A in 1976, had used atomic clocks, two on the ground and one placed into orbit, to measure the difference in the flow of time predicted by general relativity. Recall that the curvature of time is quite similar to the curvature of space, and so GPB’s result for the geodetic effect was fully expected.
However, the frame-dragging effect had never been measured. While it is very small for objects the size of Earth, or even the Sun, for massive objects like black holes it will be extremely powerful. Efforts to measure spacetime effects like frame-dragging very close to black holes are underway. The measurement has to be done remotely using x-ray telescopes to observe the emission from gas very close to the black hole, but to date our instruments are still not sensitive enough to carry out these observations successfully.
