G measurements with time-of-swing method at HUST

Phil. Trans. R. Soc. A 2014 372, 20140141

Qing Li1, Jian-Ping Liu1, Hui-Hui Zhao1, Shan-Qing Yang1, Liang-Cheng Tu1, Qi Liu1, Cheng-Gang Shao1, Zhong-Kun Hu1, Vadim Milyukov2 and Jun Luo1

¹MOE Key Laboratory of Fundamental Physical Quantities Measurements, School of Physics, Huazhong University of Science and Technology, Wuhan, Hubei 430074, People's Republic of China

²Moscow State University, Sternberg Astronomical Institute, Moscow 119992, Russia

e-mail: junluo@mail.hust.edu.cn

We review the G measurements with time-of-swing method at HUST. Two independent experiments have been completed and an improved experiment is in progress. The first G value was determined as 6.6699(7)×10⁻¹¹ m³ kg⁻¹ s⁻² with a relative standard uncertainty (u_r) of 105 ppm by using a long period torsion pendulum and two cylindrical source masses. Later, this result was corrected to be 6.6723(9)×10⁻¹¹ m³ kg⁻¹ s⁻² with u_r=130 ppm after considering the density distribution of the cylinders and the air buoyancy, which was 360 ppm larger than the previous value. In 2009, a new experiment by using a simple block pendulum and spherical source masses with more homogeneous density was carried out. A series of improvements were performed, and the G value was determined to be 6.67349(18)×10⁻¹¹ m³ kg⁻¹ s⁻² with u_r=26 ppm. To reduce the anelasticity of the torsion fibre, fused silica fibres with Q's of approximately 5×10⁴ are used to measure G in the ongoing experiment. These fibres are coated with thin layers of germanium and bismuth in turn to reduce the electrostatic effect. Some other improvements include the gravity compensation, reduction of the coating layer effect, etc. The prospective uncertainty of the next G value is 20 ppm or lower.