System to Remotely Transport and Deploy an Unmanned Helicopter
Submitted to: Dr. Paul Oh
The Senior Design Project Committee
Mechanical Engineering and Mechanics Department
Drexel University
Team Number: MEM-10
Team Members:
Jason Collins (MEM)
Michael Perreca (ECE)
Caitlyn Worthington-Kirsch (MEM)
Submitted in partial fulfillment of the requirements for the Senior Design Project
May 19, 2008
Abstract
Despite the recent increase in the use and design of unmanned aerial vehicles (UAVs), methods of remotely transporting and protecting these valuable pieces of equipment has gone largely untouched. We have designed and built an enclosedtrailer system that can be used to transport and aide in the launching of UAVs. Combining suspension, leveling and actuation systems, our design can be incorporated into previously designed autonomous vehicles and allow for safe transport of a UAV to a specified site while protecting it from outside dangers, as well as maintaining a level base for safe takeoffs and landings. This system is designed to be user friendly and is easy to actuate locally or remotely. Our project was culminated in fully successful testing of the prototype trailer. Videos of the test launch may be viewed at http://dasl.mem.drexel.edu/SeniorDesign07%2D08/public/.
Table of Contents
2
Abstract 2
Introduction 4
Problem Description 4
Constraints on Solution 5
Design Decisions 6
Prototype Fabrication 9
Data Acquisition and Actuation Electronics 14
Testing Results 19
Economic Analysis 21
Project Timeline 22
Teamwork Analysis 22
Social, Ethical, and Environmental Impacts 22
Conclusions 23
Appendix A: Leveling System Test Data 24
Appendix B: Approved Purchase List 25
Appendix C: Gantt Chart 26
2
List of Figures
2
Figure 1: The vibration control system 8
Figure 2: Gimbal brakes during testing 10
Figure 3: The completed gimbal, without the counterweight 11
Figure 4: The helicopter on the gimbal, stabilized 11
Figure 5: The gimbal and suspension system mounted on the trailer 12
Figure 6: The latch and linkage mounted to the gimbal 13
Figure 7: The completed enclosure frame, open 13
Figure 8: the completed trailer, with the enclosure frame closed 13
Figure 9: The fully enclosed trailer 14
Figure 10: Two Bosch-Style SPDT Relays Wired as an H-Bridge 15
Figure 11: LabVIEW Graphical User Interface 17
Figure 12: A local user using the break out box 19
2
List of Tables
Table 1: Threshold and Objective Requirements 5
Table 2: Trade Study for Vibration Control System 7
Table 3: Trade Study for Controller System 8
Table 4: Trade study for Enclosure Design 9
Table 5: Cost Breakdown 21
Introduction
Our goal was to design and build a trailer capable of safely carrying a UAV (Unmanned Aerial Vehicle) over rough terrain to the designated launch point, and then safely releasing the helicopter for launch. This would allow workers to effectively use a UAV helicopter without endangering themselves. In the fall quarter, we designed the trailer, including a gimbal, a suspension system, an enclosure, and a remote actuation system for the trailer. We also built two proof-of-concept scale models of parts of the planned trailer, in order to reduce the risk of going forward with the project. At the beginning of the winter quarter we assembled a purchase list. We have since built the trailer system, encountering few major setbacks, and keeping the project under budget. We have done a successful test of the system, by using the trailer to carry the UAV to the launch site, remotely preparing for launch, and remotely launching the UAV helicopter.
Problem Description
In an emergency, it may be vital to gather information in an environment that is dangerous to human rescue workers. An unmanned aircraft, such as a UAV (Unmanned Aerial Vehicle) helicopter, is increasingly valuable for a search and rescue operation: sensors on the aircraft can locate trapped victims, or find the extent or source of the damage, or determine the stability and strength of involved buildings, or locate obstacles and paths around them, all without exposing emergency workers to dangerous environmental conditions. Armed with the information provided by the unmanned aircraft, emergency workers can work efficiently to rescue victims and contain the emergency, without spending unnecessary time in dangerous conditions, and with a reduced risk of being trapped or stopped by obstacles.
However, a UAV helicopter must be launched. At present, all solutions available on the open market require the operator to be at the launch site in order for the helicopter to be launched. This means that the helicopter must either be launched outside the dangerous area, which may limit the distance it can cover and the quality of information it can supply, or the operator must venture inside the dangerous area in order to launch the UAV helicopter. Neither solution is optimal.
We have therefore designed and built a method of remotely launching a UAV helicopter, and built a prototype of our design. The launching apparatus and helicopter will be towed to the launch site by a remote-controlled ATV, such as D.I.A.S I or II (Drexel Integrated ATV System) and the launching apparatus will protect the helicopter while it is being towed. Once at the launch site, the barrier protecting the helicopter will be automatically removed, the launch pad will move to a level position and be clamped in place, the helicopter will then be ready to launch.
Constraints on Solution
Constraints for this project include the need for the trailer to fit into our available workspace, but accommodate and support a UAV helicopter carrying its full capacity of observational equipment. The helicopter must be protected during transit over any terrain accessible to the ATV used to pull it, and the platform supporting the helicopter must be level in order for the helicopter to be launched. The weight of the platform along with the weight of the helicopter must be an acceptable payload for the ATV that tows the apparatus. Table 1, below, details the threshold and objective goals for the project.
The towing strength of the 90cc DIAS2 could not be determined from the available ATV literature. We therefore tested DIAS2 towing various weights on a variety of slopes. We determined that DIAS2 could tow 850 pounds on flat ground.
Size / fit through double doors / Fit into standard storage pod
Minimum towing vehicle / 90 cc ATV / 350 cc ATV
Protect UAV during transport / Dirt road / Off road
Launch prep time / 2 Minutes / 1 Minute
Weather protection / Shield contents from light precipitation / Shed steady rain
Launch angle of UAV / Safe angle for human pilot
+/- 5 degrees / +/- 2 degrees
Able to carry / 3 foot rotor, 15 lbs / 6 foot rotor, 35 lbs
Table 1: Threshold and Objective Requirements
The trailer must have both manual and remote operation controls. The trailer must protect the helicopter from rain or low tree branches, as well as from vibrations and instability caused by transit over rough terrain and other obstacles. The helicopter must be supported so that it does not move within the trailer during transit. The trailer must be able to accommodate the size and safely support the weight of the fully-loaded helicopter. The loaded trailer must not be too heavy for the ATV to easily pull.
The helicopter platform must remain level independent of the trailer and within the launch tolerance of the helicopter. The UAV must be deployable within 2 minutes after reaching the launch site. The controls must be easily understood by rescue workers, easily manipulated, andlegible in low visibility conditions.
The trailer must be able to survive use through rough terrain and expected obstacles. The manufacture cost must be less than the cost of the helicopter. The trailer and controls must be field serviceable if the equipment breaks down.
Design Decisions
In order to design the trailer we needed to determine the best way to protect the helicopter from vibration, to keep the helicopter level, to protect the helicopter from damage during transit, and to control the trailer. We also needed to determine the best source for the trailer.
We chose to use a gimbal and an innovative vibration control system, consisting of a compressible ball between two bowls, to protect the helicopter from vibration and movement. An outer cover made of strong fabric would protect the helicopter during transit, and then fold back to allow the helicopter to launch. A National Instruments Compact RIO would be used to control and monitor various functions and sensors of the trailer system. The trailer would be a 56" x 55" deck over a trailer base sourced from a local manufacturer.
In order to keep the helicopter level, we designed a gimbal with a counterweight to fit the helicopter landing gear. A gimbal is a proven method for allowing free rotation of an object, and the counterweight would ensure that the helicopter remained level. The inner frame of the gimbal, with the helicopter mounted on it, would move independently of the outer frame and the trailer, protecting the helicopter from tilting if the trailer encountered hills or obstacles.
The outer frame of the gimbal would be mounted on a vibration control system, in order to protect it from vibration. At first we considered a classic spring-dashpot system, or a simpler system with springs but no damping, to reduce vibration. However, it was also necessary to protect the helicopter from sideways sliding or twisting movements, and to reduce complexity and size we chose to combine these functions. A separate device to allow sideways movements could make the overall structure very heavy and far more complicated. This led to the idea of allowing the spring and dashpot to pivot and shift horizontally, allowing for sideways and twisting movement. Another alternative would have been multiple spring-dashpot systems, one for each axis. However, the system would have a lower part count and be far more stable if we used a spherical system: a compressible rubber ball. In order to force the ball to return to the center of the system, we made the plates above and below it into curved bowls. This final design would, in theory, allow the system to adapt to more movements than a spring-dashpot system could. In order to be certain this would work, we built and tested a proof-of-concept demonstrator.
Allows for expected movement: 30% / 5% / 30% / 5%
Ease of implementation and tuning: 20% / 15% / 10% / 15%
Limits Vibration: 50% / 50% / 40% / 40%
Total / 70% / 80% / 60%
Table 2: Trade Study for Vibration Control System
Table 2 shows the trade study to determine the vibration control system. We considered three systems: a classic spring-dashpot damper, a spring damper without a dashpot, and an innovative design consisting of a rubber ball between two bowls, mounted at the four corners of the helicopter platform. The bows would be further connected by springs or cordsto preventthem from separating and allowing the ball to escape. This last system would be harder to calibrate and is associated with greater risk, but it would allow for more twisting and turning movement than the more traditional solutions. In order to reduce risk further, we have assembled a proof-of-concept model of the system, to demonstrate that the design does work. Having tested the model, we determined that it transmits approximately 5% of ground vibration, a substantial reduction which will protect the helicopter during transit. Figure 1, below, shows the system moving both horizontally and vertically.
2
Figure 1: The vibration control system at rest, shifted horizontally, and compressed vertically
While the vibration control system was being designed, it was also necessary to consider what system to use to control the trailer. The system would need to control the opening and closing of the trailer enclosure, and to control the restraints that hold the helicopter in place during transit. Several controllers were suggested, the IFI Controller, the Lynxmotion, the Pololu, and the NI Compact RIO.
IFI Controller / Lynxmotion / Pololu / NI Compact RIOAvailability (20%) / 20% / 10% / 15% / 17%
Voltage System (5%) / 5% / 5% / 5% / 2%
Learning Curve: Software (15%) / 10% / 10% / 5% / 5%
Learning Curve: Hardware (10%) / 10% / 7% / 7% / 10%
Cost (10%) / 10% / 9% / 10% / 5%
Communications (20%) / 10% / 10% / 10% / 20%
Known Issues (20%) / 5% / 10% / 10% / 15%
Total (100%) / 70% / 61% / 62% / 74%
Table 3: Trade Study for Controller System
Table 3 documents the alternative controller systems we considered. The readily available controllers were evaluated for availability, the voltage system each used, their usability, cost, and the presence of any known issues, including size, limited inputs and outputs, and the likelihood of overheating problems. In the end, we determined that the NI Compact RIO would be the best choice for a controller.
Choosing the trailer was also a major decision. The trailer had to be narrow enough to fit through the doors of DASL, and had to support the weight of the enclosure and helicopter pad.
Price (15%) / 10% / 5% / 10%
Weight (10%) / 5% / 5% / 10%
Complexity (15%) / 10% / 5% / 15%
Protection from environment (30%) / 30% / 30% / 20%
Protection from UAV hitting cover (30%) / 10% / 20% / 30%
Total (100%) / 65% / 65% / 85%
Table 4: Trade study for Enclosure Design