Design and analysis of velocimetry for supersonic sled system

Abstract

Knowledge of precise velocity profile of aerospace bodies is essential in research programs-especially now when operating speeds are going much beyond supersonic regimes. Not much research has been reported on velocimetry of such systems. It was therefore considered essential to carry out in-depth study and research of various feasible techniques for high velocity measurements of bodies moving on rail system at supersonic speeds to give fillip to various research programs in defence and space domain. The research centre, where the testing of such systems is carried out, has a precision rail track on which the test articles such as parachutes, unmanned aerial vehicles, missiles and fuses etc are test run at required high speeds. It provides a unique environment for captive flight testing and enables experimentation at high velocities and accelerations for short durations. The research, elaborated in this thesis, was carried out at this facility.

While considering various techniques for determining velocity, it was realized that velocity could not be directly measured by any of the known techniques on the rail system. On the other hand, there were several techniques for velocity derivation from other parameters. Out of these, three such techniques were shortlisted for carrying out in depth studies. The identified techniques were based on different primary measurements of - position, acceleration and pressure-which, finally lead to the derivation of velocity.

Literature published half a century ago points to the use of position based measurements in rocket sled applications on the rails on which magneto inductive sensor were deployed. Thereafter, the field remained un-explored. It was found that magneto-inductive sensor output falls with the increase in velocity and it was not usable at speeds beyond 500 m/s. Simulation of such a system was carried out to understand the reasons for the problem. The outcome of the research was non-obvious. The study suggested reducing the number of turns of the inductive sensor, rather than increasing it, lead to increased amplitude at higher velocities. Based on the studies, the existing sensor was improved accordingly and experiments were conducted to validate the proposition. The experiments were first conducted on a ground based rotor (figure 3.15) and then rail track. Relative error of the order of 0.046% was computed in one of the test configurations. The comparison between existing and proposed sensor, at feasible velocity of 303 m/s, indicated significant improvements in new design. Easy calibration of this system was feasible. The velocity derived from such a distance based techniques acted as a reference for other techniques. It is well known fact that measurement of acceleration can lead to velocity. But its practical implementation leads to host of errors. Research also focused to quantify these errors. Two different solutions were worked out to measure velocity after mitigating these errors. Some of the methods employed included identification of improved method of integration, capping and filtering, bias correction, sensor mounting, reduction in RMS noise, position fix and the like. Under identical conditions, specially developed DSP processor based on-board measurement system included three accelerometers of different makes, which measured linear acceleration in the same direction. The result with one of the sensors was within 1.5% of reference values. These experiments were completed by the year 2014. Thereafter, a method of data fusion was developed which offered similar results. It obviated the use of position fix and the associated hardware. Another technique involved was using pressure as primary measurement. Aircrafts use pressure transducer based Pitot tube to compute velocity. Study of its adaptation to rocket sled motion, associated challenges and its comparative performance was studied. Expression establishing relationship between Mach number or velocity with pressure were derived and were tailored to needs of rocket sled usage. Application specific Pitot tube, acquisition and logging system was developed so as to conduct experiments. Velocity was successfully measured in two different experiments. The challenges encountered in the process have been explained.

It was surmised that there is no single technique which has all the advantages over the rest of the techniques. It was also found that magneto-inductive system offers highest accuracy but its initial cost of installation is high. The acceleration based system is easy to deploy, inherently reduces noise, gives more number of measurement points and is economical. The comparative study of the three techniques is elaborated in the concluding section. The domain of velocimetry is vast and diverse. Challenges have been overcome in adopting technologies to a new application. The findings of this research would improve upon the knowledge base about this ever expanding field.