![]() |
|
|||||||||
|
Natural flapping flight is a broadening area of interest for aviators and engineers. In order to study bird and insect flight systems, the PSU universal flapping team created a prototype: a mechanical simulator that could be programmed to replicate various wing motions. Our goal for this project is to design a prototype that researchers and engineers will be able to use in a wind tunnel to collect low Reynolds number data of flapping flight dynamics, kinematics and aerodynamics. Our prototype is a single flapping wing modeled after a hummingbird. The motion of the device is determined by actuators that give 3 degrees of freedom, allowing it to reproduce a variety of flapping flight modes. A rotary actuator called NanoMuscle was used as the driver device. These actuators have sufficient output to drive the wing, are highly efficient, and have a long cycle life. The control of the actuators is accomplished with a HC11 microcontroller. It sends a stream of addresses to each actuator, giving the wing a discretely mapped path to follow. We believe that the experimental data will have an enormous impact on the development of new aviations such as Unmanned Aerial Vehicle (UAV) and Micro Air Vehicle (MAV). |
| A method is needed to simulate flapping wing
flight in a controlled environment to provide reliable data. With
this in mind, the Boeing Company proposed the creation of a universal
flapping system that could be programmed to mimic any desired wing motion.
This device would be used in wind tunnel tests to replicate natural
flight. Boeing partnered with our interdisciplinary team of Penn State engineering students to develop such an electromechanical flapping wing, and required that the device have the following characteristics: Design Criteria: |
| We were given a ready-made solution to the
project at the beginning of the semester, but we wanted to go one step
further and use the latest technology to solve the age-old problem of
obtaining meaningful aerodynamic data. The bulk of the work was
divided equally between the mechanical engineers and the electrical
engineers. The MEs, with experience designing and building, were
responsible for the wing and stand construction, while the EEs were
responsible for the control system hardware and software. The motion of the device is driven by three rotary actuators that manipulate the wing. Two actuators control the horizontal (x) and vertical (y) motion of the wing, while the third controls the pitch (z) of the wing. These degrees of freedom are analogous to those present in a flapping hummingbird wing. We chose to use NanoMuscle for its efficiency and durability; despite its lower frequency response relative to other actuators, it was one of the few that fit in our budget. Pulleys attached to the actuators were connected to the wing's central motion joint via zero-stretch nylon thread. A set of springs were installed opposing the motion of the actuators to create a force balance at the wing's central joint. Rapid prototyping technology allowed us to manufacture our own pulleys, which saved money and time and allowed us the exact shape and dimensions we wanted. The layout of the motion system was optimized to allow for improved control and manipulation of the system, as shown below.
The control system for the universal flapper utilizes an embedded microcontroller to direct the desired wing motion. The chip controls the movement of the wing via two signals: the first causes the actuator to rotate, while the other forces it to contract. By controlling which signal the actuator receives, the direction of motion is also controlled. The two signals are generated by two output compare interrupts that create the pulses needed. The speed of the wing can be adjusted by controlling the duty cycle of the interrupts. As this portion of the program utilizes interrupts, it can work independently of the main program. The main program works by sampling the value from the feedback signal of the actuator and comparing it to a stored value from a table. Depending on whether the table value is smaller or greater that the sampled value, the chip assigns the appropriate signal to the actuator. This will cause the actuator to converge on the correct position. When the two values match, the program updates the pointer to show the next value in the table and the actuator then starts to move to that position.
The outside circuitry that interfaces the HC11 with the actuators, shown above, is fairly simple. The two ‘motion’ signals are connected via a pair of AND gates. When the control signal is high it will activate the rotate signal and vice versa for the low signal. The feedback signal from the actuators varies from 3.46v to 2.67 and the A/D converter requires a voltage from 0 to 5 volts the signal has to be conditioned. This is done via a set of 3 Op-amps that rescale and offset the original voltage. |
| Flapping flight may be the wave of the future for aviation. Throughout our design project, we found that the key challenge in duplicating natural flapping flight is how well the actuator used can simulate bird muscles. Biological muscles are extremely difficult to replicate, and different techniques from a range of disciplines are emerging as attempts to solve this problem. Our design uses a shape memory alloy actuator to model hummingbird motion. This technology does not have the ideal frequency to generate the appropriate flapping speed, so further testing of other actuators is warranted. Ongoing research into other types of actuators includes designs using chemical reactions and other mechanisms to provide realistic muscle movement. Future developers of flapping wing motion devices should assess these technologies before attempting new designs. |
Project Team: Erman Akinci Adam Bussey Adam Hernandez Shinho Kang Jeremy Wray |
Sponsor: The Boeing Company |
Technical Contact: Bill Grauer |
Faculty Advisor: Timothy Wheeler |
|||