FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1963
1963 - 0460.PDF
— —^— A DECADE OF MISSILE SIMULATION . . . WGA), Sperry and GEC—which was responsible for the overall development of the Seaslug missile system, was hardening towards the view that it was becoming feasible to build an analogue repre sentation of the missile, its aerodynamics and the associated radar, so complete that it would be possible to carry out in the laboratory assessment of the system previously possible only by means of flight trials and subsequent analysis. As a result of this evolution of thought, Armstrong Whitworth were quick to see the possibilities of electronic machinery which had been developed by EMI, and the pioneering work of these two companies led to the construction of a complete trajectory simula tion of the Seaslug missile, using flight trials primarily to validate this simulation so that detailed studies of performance could be undertaken. Meantime, whilst it was generally accepted that the work at Sperry, based on the comparatively simple concept of fixed-speed fixed-space points, was of considerable value, it was increasingly felt that the limitations of such simulation—particularly when studying near-boundary conditions at altitude—had been reached. Sperry had thus arrived at the third stage of their computational evolution, with the strongly felt need for a complete simulation of the missile in trajectory in order to understand the function and effect of the control in improving overall performance. Conse quently, the decision was taken to install a large analogue computer representing all aspects of the missile likely to be affected by the control system. The philosophy behind the design of this computer is to aim at the compromise between a large-scale special-purpose machine and a general-purpose one. The concept of solving large problems by the use of several general-purpose computers linked together is an attractive one, and to some extent this idea is being realized in the current design. The new simulator consists of 280 drift-corrected amplifiers, 235 multiplier units and 45 diode function generators, and it is housed in five three-bay racks, two of which are shown on p. 437. This machine is capable of representing the aerodynamic equations in a FLIGHT International, 28 March J963 form to include full non-linear representation of all the aerodynamic derivatives, cross-coupling terms and axis transformation of both missile and beam axis in terms of Earth axes. In addition, the rele vant terms are constructed as functions of missile speed and alti tude in order that full trajectory studies can be undertaken. Thus the previous restrictions of having to interpolate between fixed space points have been overcome. Because of the size and complexi'a of the machine, careful consideration has been given to the problems of setting up and checking. Setting up is achieved by the use of patch boards, which are linked for a particular problem and then plugged into the machine. In this way new problems can be set up without interrupting the work in hand, and, since the machine is in five separate bays, it can readily be converted from the full simulation to any combination of the separate bays carrying out less complex studies of more specialized problems. Checking is carried out against check runs, computed digitally each time the problem is changed. The machine is also fitted with digital/analogue input and output equipment in order that actual inputs from flight trials, which are obtained in digital form, can be fed into the simulator as part of J the validation process. In addition, this digital input/output faci lity is invaluable for transferring data from one machine to another for comparison purposes or to obtain particular information from more specialized areas such as, for instance, the guidance equipment. In recounting Sperry's approach to GW control simulation, I have tried also to draw attention to the general trend of develop ment, starting from the Differential Analyser, progressing through the stages of cross-coupling and other added complexities, and arriv- ing at the stage of simulation when optimization studies, tolerance studies and accurate statistical performance data can be obtained by "flying" large numbers of missiles in the laboratory on the mathematical model after a relatively small number of actual flights for validation purposes. This is a remarkable evolutionary process to have been achieved in little more than ten years. With the con tinuing growth of facilities, such as large random noise vibrators for the simulation of flight conditions on hardware, we might over the next decade arrive at the situation where the first missile fired will be a practice shot, rather than one in a research and develop ment programme! 3-GUIDING BLUE STREAK ALTHOUGH Blue Streak is still in the news as part of the ELDO (European Launcher Development Organization) pro ject, its cancellation as an ICBM in April 1960 has made it possible to discuss the advanced inertia! guidance system on which the military vehicle (but not the civil space projects) depended for its navigation and guidance. It represented a great deal of effort by British industry, and by Sperry in particular. As a poten tially extremely successful system, it deserves a place in history, perhaps alongside Britain's (and Europe's) first airborne inertial system; the latter, also of Sperry design, is now in good hands at the Science Museum in London. A major factor in the award of this guidance contract to Sperry London was the British company's long preoccupation with inertial guidance systems. Back in 1949, working to an AM Operational Requirement, the firm developed and tested a fighter system which flew in 1954. The experience thereby gained with accelerometers. mercury-supported gyros and electronic units was invaluable when a contract was issued for the Project 3000 (Blue Streak) inertial navigation and guidance system in November 1955. Military Blue Streak was a strategic ballistic missile, designed for silo firing with tanks full, to fly along a pre-calculated trajectory with the range to impact determined by motor cut-off time. Opera tional range limits would have been 800 to 2,500 n.m. Initial apparent acceleration at launch would have been approximately l.3g, followed by a programmed rise and turn using the two main motors under autopilot control. This initial programme was of approximately 120sec duration. From the end of this programme until main motor cut-off, the thrust vector was maintained nominally in the fixed direction of 36°. subject to control by the guidance system. Apparent accelera tion rose to some 12g as cut-off approached. A typical cut-off time for a 2.500 n.m. target would have been 176sec. After the main motors cut, the warhead separated and "head" motors fired The latter applied a vernier correction velocity to the warhead, which would then have followed a ballistic path with a total fligh: time to 2,500 n.m. of approximately 20min. At any point in space, the future trajectory and the impact range of any missile can be accurately calculated if its position and velo city are known to that point. The range to ground impact for a particular missile can be achieved by any one of an infinite choice of positions, velocities and altitudes at the start of the ballistic phase. It follows that, with the required thrust capacity, and for any random course, missile thrust could be terminated at such a time that a particular range to impact could always be achieved. Con versely, it may be possible to control the missile very accurately to a particular acceleration, velocity and displacement programme computed to achieve a particular range to impact. Such a pro gramme would impose severe requirements on the control system In practice, control was to be exercised to ensure that a nominal trajectory was flown near to the optimum from aerodynamic and fuel considerations. The inertial system was operative throughout the main powerec flight, and was capable of:— (1) Measuring the velocity and position of the missile in a suitable co-ordinate system. (2) Providing a demand signal for yaw steering to a nominal trajectory and later including computed corrections to prevent the line error which would be caused by variations in the time of impact and b; differences between the actual and nominal velocities and positions (3) Providing a pitch steering signal to reduce perturbations in the I pitch directions at cut-off. (4) Initiating an integrating accelerometer in the head when the veloc.i; was a fixed amount short of that for the correct range, alio win;: for variation from the nominal trajectory. (5) Initiating main-motor cut-off when the velocity was a pre-deter-
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
