Fuel Tanks

When dealing with internal combustion engines, suitable fuel tanks are, of course, required. These are either sourced externally and mounted within the fuselage (typically injection molded nylon with simple screw-top fittings) or form part of the aircraft structure itself - we have tried both approaches. Currently, we tend to design integral fuel tanks built into the SLS nylon structure near the aerodynamic center of the aircraft. Tanks formed this way need to be thoroughly cleaned after manufacture and suitably sealed, but they work extremely well. The use of SLS nylon permits a host of desirable features such as internal baffling, fittings for caps, fuel supply lines and level sensors, and so on. Figure 2.6 shows a cut-away view of the integral SLS nylon fuel tank from the SPOTTER aircraft (note the corrugated internal baffle that also adds structural rigidity and support for the payload which is slung below the tank).

Control Systems

As UAV systems become more capable and need greater resilience, so the on-board command and control systems tend to grow in complexity. At the most basic level, a simple aero-modeler-grade receiver coupled to the control servos and supplied by a precharged battery is all that is needed. Even so, we find that the various makes of equipment vary in their robustness and capabilities: we tend to adopt Futaba systems and have had good experience

A typical integral fuel tank

Figure 2.6 A typical integral fuel tank.

with them. They will allow control of the aircraft within an operational radius dictated by the range of the transmitter, the pilot’s ability to monitor the aircraft, and any airspace regulations in force. Few aero-modeler systems have ranges beyond 1500 m, and even at this range it is extremely difficult for a pilot to observe the behavior of the aircraft as it is beyond visual line of sight (BVLOS). The alternative is long-range transmitters, but even these do not deal with the difficulty of feedback to the pilot. Air regulations typically prohibit normal aero-modeler flying beyond a radius of 500 m.

The next step up in complexity requires some form of on-board autopilot. At the most basic level, autopilots need to be able to fly an aircraft from place to place while maintaining speed and height. This necessarily requires that they can override any instructions coming from the receiver normally controlled by the pilot. The pilot does, of course, need some method of switching between manual control and autopilot control. Obviously, it is also necessary to be able to upload way-point instructions to the autopilot, either before takeoff or during flight. Any flights out to locations beyond the manual control transmitter range will rely solely on the autopilot to fly the aircraft, without any prospect of direct intervention. To enable the autopilot to decide where the aircraft is, it is normal to rely on global positioning systems (GPS), sometimes backed up by estimated positions using dead reckoning, given speed and compass heading. GPS vary in accuracy but even the most straightforward will give locations to within a few tens of meters. Provided the final way-point of a mission is within, say 500 m of the pilot, it is then possible for the pilot to take back direct control and land the aircraft. Most autopilots will also provide telemetry data via some form of radio downlink or store such data on board for subsequent analysis, see Figure 2.7.

Further to such basic operations, it is possible for the autopilot to conduct taxi-out, take-off, mission, and landing maneuvers without pilot intervention - the SkyCircuits system often used by the team at Southampton is capable of all these functions (we have achieved startlingly good repeatability in landings using this system coupled with a laser-based height sensor, see Figure 2.8). It also provides downlink telemetry of various health monitors on the aircraft such as fuel remaining, engine temperatures, and so on.[1]

Typical telemetry data recorded by an autopilot. Note occasional loss of contact with the ground station recording the data, which causes the signals to drop to zero

Figure 2.7 Typical telemetry data recorded by an autopilot. Note occasional loss of contact with the ground station recording the data, which causes the signals to drop to zero.

Flight tracks of the SPOTTER aircraft while carrying out automated takeoff and landing tests. A total of 23 fully automated flights totaling 55 km of flying is shown

Figure 2.8 Flight tracks of the SPOTTER aircraft while carrying out automated takeoff and landing tests. A total of 23 fully automated flights totaling 55 km of flying is shown.

In addition to an autopilot, more complex UAVs may carry some means of trying to avoid collisions with other aircraft. At their simplest, these are transponders that allow aircraft to be alerted to the presence of each other and their intended direction, speed, and altitude. Mode C transponders transmit a four-digit code and pressure altitude. Mode S transponders transmit additional information such as the aircraft identity, direction, speed, and so on. In some areas, additional data channels, known as automatic dependent surveillance - broadcast (ADSB), allow a full picture of nearby aircraft including height and direction information to be plotted. The main problems with such equipment are size, weight, and cost - this tends to restrict their use to larger and more expensive UAVs. In addition to transponders, research is progressing on the so-called “sense-and-avoid” systems that take active steps to change the UAV’s flight path when a possible collision is detected. Such systems are still in their infancy at the time of writing, but rapid developments can be expected, linked to cameras, radars, and transponder systems.

Beyond this, the degree of complexity on board is practically limitless given sufficient computing power. However, in our work the only other non-payload avionics we have seen necessary to fit are concerned with battery charging and health monitoring, so that mission endurance is not limited by finite battery capacity. Even so, the main wiring diagram for an aircraft with twin engines and multiple redundant receivers and charging circuits with a range of control surfaces can quickly become quite involved. Figure 2.9 shows the main wiring diagram from the SPOTTER aircraft (more detailed views are given in later chapters), and Figure 2.10 shows the SkyCircuits autopilot we typically use.

  • [1] http://www.skycircuits.com/.
< Prev   CONTENTS   Source   Next >