Aerospace Systems Technology and Rocket Operations – BOOM 2004

Background

Created in the Fall of 2000 as a research and design project for students interested in aerospace engineering, the original goal of the Aerospace Systems Technology and Rocket Operations (ASTRO) team was to design an autonomous landing vehicle similar in design to the Mars Rover lander - able to land safely using rocket engines to slow its descent. After development and initial testing, the autonomous landing vehicle (ALV) would be dropped from a height sufficient to allow the vehicle to achieve terminal velocity. This ALV would then land itself under its own power and without the aid of parachutes or stabilizing fins, making it operational in a generic environment – an important part of the project.

This year, ASTRO’s goal is to complete one of many steps which will take place before completion of our ultimate goal: developing a vertical-takeoff, vertical-landing (VTVL) vehicle which will initially rise a few feet of the ground under its own power, autonomously hover and demonstrate its stability over the course of a few seconds, and then land safely and without damage. After this goal is achieved, the current design will be improved to either be able to translate laterally while retaining stability, and ultimately to achieve the goals set out for the original ALV concept. A model of the original ALV design concept can be seen below:

Work in Previous Semesters

The project began with determining overall system performance requirements and the testing of a commercially available liquid-fueled rocket motor. The original rockets used ran on gasoline and hydrogen peroxide, two propellants readily available to us, but there were many problems with their design. Sealing poppets would never seat, presenting a safety hazard; the engines weren’t throttleable, and they weren’t designed to burn for more than ten seconds. By ASTRO’s second year, design had begun on our own rocket engines, which were meant to burn for a minute and operate on hydrogen peroxide and gasoline, producing a target thrust of 100 lbs. At the same time, our electrical group was trying to determine what type of control system we wanted to use for autonomous control of the ALV. Neural networks were initially considered, and a PID controller was developed for a 3 degree of freedom test-bed, which was created to obtain practice in developing a sample control system on a simpler, lower cost dynamical system.

During the 2001-2002 academic year, ASTRO’s test site was also developed, located on a plot of land owned by the University. This land was first cleared out, and a testing region built to establish a permanent testing site for the ALV rocket engines. In addition, test stands were designed and built to work with the commercially purchased gasoline-peroxide engines being tested at that time.

The team then spent a year developing and testing the next generation of engines, with limited successes. These engines had a stainless steel body with ceramic lining to prevent overheating, and a titanium throat plug. Due to new problems obtaining hydrogen peroxide caused by health and safety concerns as well as complications with its storage, the propellants were switched to propane and Nitrous Oxide. The problem with these engines turned out to be the injector-manifold system, an overly complex design which was unable to provide the required flow rate to achieve ignition with our new gaseous propellants. The test stand was also redesigned to work with the new engine design.

Last year, the electrical group did a lot of control system development with the previously mentioned test-bed, and also developed a 6-degree of freedom helicopter system which allowed us to test on a system similar to the ALV. By the end of the spring semester, the electrical group had the helicopter operational, and had developed a control system for the VTVL in VX Works, a small, quick operating system. Since then, work has begun on second- and third-generation control systems for the VTVL.

The VTVL structure was also built last year, designed to mount the three rocket engines, fuel tanks, the onboard electronics, and onboard computer and sensors rigidly and with low weight. A picture of the structure can be seen below (without mounted components):

2004 – Progress and Goals

This semester’s goal for the team is to develop the control and propulsion subsystems to a level such that a first liftoff in late April of this year occurs. Progress this semester has been very strong, with design, manufacture and testing of subsystems going smoothly. I’ll outline the basic work done up to this point.

After redesigning the rocket manifold system to two simple impinging inlets, checking that we could obtain high enough mass flow rates with our pressure-fed fuel feed system, and relocating the spark plug to directly between the Nitrous and propane inlets, we tested this system earlier this semester and had our first successful firing. A picture of the engine firing (with a clear combustion chamber for purposes of verifying ignition) is shown below.

The thrust obtained from this test is not known, due to the absence of load cells on this previous-generation test stand. Since this test, our test stand has been redesigned to test three engines at once, obtain thrust and temperature data, and hold a data acquisition board. A close-up of the 2^nd generation Nitrous Oxide-propane engine with old test stand at the test-site, after testing:

In addition to getting our 2^nd generation engines to fire, we developed a new set of engines this semester which are designed to get higher thrust and better combustion, and were designed to burn longer. Analysis was performed on these engines using ANSYS and Fluent for determining fluid flow and heat transfer in the combustion chamber and nozzle. The combustion chamber for the first version of this engine has now been manufactured, and the engine as a whole will be ready for testing within a few weeks. Our current limitation in testing is the winter weather.

On the controls system side of things, a few mechanical engineers worked with the electrical group to form a control systems group during the fall semester. This group concentrated on the dynamics of the VTVL, and creating a realistic dynamical simulation in Matlab which takes into account thrust misalignment, wind forces, 6 DOF accelerometer sensor signal noise, and other sources of disturbances. With this simulation complete, a next-generation control system for the VTVL is currently being designed by undergraduate and graduate students in ECE and CS. We are also in the process of designing electronic circuits for support of data acquisition during testing.

The VTVL’s onboard power system is also currently being developed – powering the spark plugs and solenoid valves is currently being done with car batteries during testing, which are not ideal for use on the VTVL due to their weight.

Looking ahead, we expect to complete testing on the 3^rd generation engines by the middle of this semester and determine the thrust response we can expect for given pulse-width modulation (PWM) inputs to the fuel-feed system by early April. This data will be used with the control system algorithm to allow for variable thrust output, either using PWM of our on-off solenoid valves, or variable area flow control valves. Our algorithms will also be translated from Matlab to C, so that they can be run inside the VX-Works operating system which will be used on the VTVL’s onboard computer. All sub-teams are currently working on systems integration while developing their own subsystems in preparation for first liftoff in April. Pending a successful liftoff, we may choose to develop a wireless communication system for the VTVL for transmitting telemetry data to the computer controlling the test procedure.

This semester is an exciting time in ASTRO’s development of an autonomous, rocket-controlled Vertical-Takeoff, Vertical-Landing Vehicle. If you’d like to get involved, contact a sub-team leader today and let us know your area of interest – we’re always looking for a few good engineers.