Study On Test Methods Of High-precision Accelerometers For A Rotating Accelerometer Gravity Gradiometer

The rotating accelerometer gravity gradiometer is used in the measurement of the earth’s gravity gradient field on an airborne or shipborne moving platform,which is of great significance in scientific research,national economy,and military applications.As the core component of this type of gradiometer,the high-precision accelerometer needs to detect the sub-ng acceleration change in the environment with large noise.Therefore,its thorough test is one of the important contents for developing this type of gravity gradiometer.However,with the further improvement of the accuracy of the accelerometer,existing test methods are limited by equipment accuracy,environmental noise,and low-frequency drift of the accelerometer bias,so conventional test methods are faced with many challenges when applied to high-precision accelerometers.It is desired to improve the inadequate methods or develop new ones.Given that situation,this thesis analyzes and summarizes the applicability of the existing test methods to the high-precision accelerometers used in rotating accelerometer gravity gradiometer,and subsequently conducts investigations on the major issues.Based on a series of solutions proposed,detailed error analysis and experimental verification of the test schemes are carried out.The main contents are listed as follows:Firstly,the existing multipoint angular rotation test based on gravity field is systematically summarized,verified,and analysed for errors.The restricted input range of the accelerometer under test lead to a rapid increase in the relative error when the absolute accuracy of the equipment is limited.In view of this,a reference accelerometer with a larger input range is used to deduce the current inclination angle by simultaneously measuring at multiple angles,correcting the inclination errors of the dividing head.The consistency of the scale factor of the high-precision accelerometer improves to 0.5‰,by an order of magnitude,between repeated measurements.Furthermore,a method of multi-position rotation around the sensitive axis of the accelerometer is developed to amplify the acceleration imposed on the cross axis,and the test uncertainty of second-order term coefficients related only to the two cross axes is better than 1 mg/g².Secondly,to get rid of the negative effects of the test environment and instrument drift on the accelerometer’s nonlinear coefficient test,a scheme of continuous rotation modulation combined with gravity component is proposed to improve the measurement accuracy.A systematic solution is given to overall identify various second-order model coefficients.Based on this,a detailed analysis and experimental verification on the cross-coupling coefficient seperation are carried out.In this scheme,the enhanced centripetal acceleration along the cross axis and the gravity components along the sensitive axis are both modulated on a tilted precision single-axis rate table,then the cross-coupling coefficient K_io(or K_ip)is effectively separated from other effects by a series of stepwise-adjusted rotational speeds.After systematically evaluating major errors sources such as tilt tide,rate table wobble,and demoduluaton phase misalignment,the test uncertainty of this coefficient is verified to be better than 0.01 g/g².This uncertainty is an order of magnitude better,compared with the static test which only relies on multipoint rotation in the gravity field.Thirdly,on account that there exist many interference factors in the test of nano-g accelerometer’s temperature model,the requirements for the test scheme are analyed in terms of mounting configuration,test equipment and measurement environment.The experiment proves that the main errors such as fixture thermal deformation in the process of heating and cooling can be suppressed to three orders of magnitude by measuring at symmetrical positions.The experimental results show that the repeatability of temperature coefficient of MEMS accelerometer is better than 0.4μg/℃for bias and 30 ppm/℃for scale factor,respectively.Finally,taking the magnetic field as vectorial,a scheme of testing the magnetic field response is proposed to separate the magnetic sensitivity of each model coefficient along different directions of the magnetic field.By fixing the magnetic field generating device onto the dividing head,the relative direction of the magnetic field is under control,and by modulating the alternating magnetic field,the noise is suppressed.It is experimentally found that the magnetic sensitivity of the model coefficients is dependent on the direction and frequency of the magnetic field.The test uncertainty of the magnetic sensitivity is better than 10 ng/Gs for bias,and less than 6 ppm/Gs for scale factor.Accordign to the test result,combining two types of accelerometers,one based on MEMS technology and the other based on reed spring,clarifying the role of magnetic circuit symmetry,material magnetization property of proof mass,and magnetic induction effect in the generation of magnetic sensitivity.The above work focuses on the test methods for the accelerometer’s response to influential factors such as input acceleration,environmental temperature,and stray magnetic field.Combined with other content inherited or developed from existing test methods,a preliminary test system is established for the accelerometer used in the rotating accelerometer gravity gradiometer.The test method in this paper can also be partially applied to the test of ng resolution accelerometers used in other fields such as aerospace exploration and microseismic detection.