Evaluation of the performance of supersonic exhaust diffuser using scaled down models

Abstract Experiments were carried out on straight cylindrical supersonic exhaust diffusers (SED) using cold nitrogen and hot rocket exhaust gases as driving fluids, in order to evaluate the effects of the ratios of the SED area to rocket nozzle throat area ( A d / A t ), SED area to rocket nozzle exit area ( A d / A e ), SED length to its diameter ( L / D ) and specific heat ratio of the driving gases ( k ) on the minimum starting pressure ratio, ( P o / P a )st, of SED. The rocket nozzle and SED starting transients were also simulated in the models. The study reveals that ( P o / P a )st increases monotonically with increase in ( A d / A t ) and k . One-dimensional normal shock relations were used in predicting the ( P o / P a )st since the compression in long ducts is basically a normal shock process. Predicted values of ( P o / P a )st were validated with experimental data. SED efficiency factors( η ns ) were arrived at based on one-dimensional normal shock relations. η ns goes down at higher values of ( A d / A e ). ( P o / P a )st is lower for lower k values for the same ( A d / A t ). Cylindrical SEDs exhibit no hysteresis. The results of this investigation were utilised in validating the design of high altitude test (HAT) facility for testing the third stage motor (PS-3) of Polar Satellite Launch Vehicle (PSLV). The simulation of starting transients in the model revealed that the HAT facility shall not be operated in the unstarted phase, because the rocket nozzle may fail due to violent oscillations of the vacuum chamber pressure. These experimental data were also utilised for designing a SED for PS-3 sub-scale motor, the results of which are covered in this paper. The accuracy of measurements are within a range of ±0.4%. Error analysis of the data were carried out and are presented in Appendix A .