Showing papers by "Defence Research and Development Laboratory published in 2003"

Seeker Based Optimal Guidance Law, A Few Issues

Abstract: Optimal Guidance Law (OGL) dynamically eliminates guidance time constant over homing duration and improves miss distance compared to conventional PN/APN guidance with accurate estimation of time-togo and target acceleration. In other words OGL attempts to make guidance system with time lag to appear as zero-lag guidance system over duration of homing by computing commanded acceleration on the basis of correct information of system dynamics, target maneuver and time-to-go. Due to un-modeled part of system dynamics such as body rate coupling, radome error coupling among others, OGL behaves unsatisfactorily when applied to Homing Guidance System with RF Seeker even when time-to-go & target maneuver are estimated accurately. If Inner Guidance Loop (IGL) is designed on basis of stability condition, some canonical form of guidance will give rise to unstable IGL. For normal PN/APN based system, a slightly modified formula guidance loop design based on minimum stability margin is presented. Moreover for sensitive scheme like OGL, some extra margin is required for guidance system to perform satisfactorily. In this paper, Zarchan’s formula of just stable IGL is modified for minimum control margin of IGL. An approximate measure of extra guidance time constant required for RF Seeker based OGL is also presented. Moreover IGL transfer function is derived and analyzed for OGL law. It is found that while stability of IGL for OGL based homing guidance with time constant selected based on stability margin formula of PN/APN based homing loop for 3 ' = N , is nearly retained; miss distance vs. flight time curves show oscillatory behavior because of un-modeled dynamics. One way to overcome this effect is to have longer range RF seeker or more stringent radome slope error specification. It is also shown that if time constants for OGL is selected with stability margin formula based on higher navigation gain, somewhat longer seeker range will assure guidance performance without making radome slope specification too stringent to be peacticable. Obviously, OGL based law can be easily applied to IR seeker based system as radome slope error is not present.

Knowledge management in long gestation projects

Abstract: When compared to short term, dynamic, profit-driven projects like those in IT, fashion, infotainment and the service industries, long gestation complex high-technology projects pose different kinds of challenges in Knowledge Management. Typical of such programs are defence programmes, infrastructure projects and drug development. These projects run for more than a decade and there is a frequent turnover of specialists. In this paper, issues related to Knowledge Management throughout the various activities of such long gestation programs has been addressed so that the project is brought to successful completion irrespective of the continued non-availability of the specialists throughout the programme. An example of a surface-to-air missile system development project has been taken where the tools and techniques of Knowledge Management have been applied to ensure the success of this long gestation project.

Development of A 3-D Laminar Compressible Navier-Stokes Solver

Abstract: It is conventional to apply upwind methods for inviscid terms and central differencing schemes for the viscous terms for numerical simulation of compressible viscous flows. Differing from this, the KFVS [1] based discretization procedure in the form of Chapman-Enskog fluxes [2], treats both the inviscid and viscous terms by upwinding. Recently, a 2-D flowsolver has been developed and validated [3]. In the present work, the above approach has been extended to develop a 3-D Navier-Stokes solver which has been validated for a blunt cone-cylinder-flare configuration and the results are compared with results of another viscous code MB-EURANIUM [6]. The flowsolver CERANS works in conjunction with a versatile preprocessor which provides the necessary data to the solver in an unstructured format. Higher order spatial accuracy has been obtained using linear reconstruction of primitive variables. Convergence acceleration is obtained using LU-SGS [5].