Summary

See the October report for the previous update.

Aerodynamics

The first 10" diameter rotor for the 150g vehicle has been fabricated and tested. The rotor was manufactured using wet-layup of carbon and fiberglass textiles in a double-sided press mold. This technique provides a smooth surface finish and accurate incidence and cross-sectional geometry. Previous blade fabrication has used a single sided mold in conjunction with vacuum bagging. This method can only accurately define a single surface and thickness distribution is entirely dependent on laminate placement and resin flow under vacuum. The two-piece mold represents a minimal increase in the tooling complexity, but is greatly beneficial both in ease of fabrication and in the quality and performance of the final part.

The results of testing show a significant improvement over the WES-Technik 10" dia. rotor in power required at the target thrust value of 50g. The performance of this rotor, compared with the WES-Technik rotor, is shown in the two figures below. Thrust as a function of RPM is very similar for the two rotors, but the power required at the target thrust value is 20% lower for the new rotor design. This increase in efficiency can be largely attributed to the fact that the WES-Technik rotor is designed for forward flight and a positive advance ratio. The new rotor is a point design to maximize static thrust.

The results of the first 3-D CFD rotor analysis for the 4-blade, 2.5cm diameter rotor indicated the possibility of massive separation and unsteady flow. Those results have been invalidated and new calculations have yielded a converged steady-state solution. An error in the boundary condition definition in the first analysis resulted in flow reversal at the outlet plane near the hub. This resulted in converging flows meeting at the center of the blade, causing separation and unsteady behavior. The corrected second calculation was run for 16,000 iterations with convergence in thrust and power required after approximately 12,000 iterations. The predicted thrust at 48,000 RPM is 4.8g requiring 0.52W of power. This is significantly less power than was required in experiment for this RPM and thrust. Input power to the motor during testing was 2.8W. This would require a motor/controller efficiency of 18% to match with the CFD results. The 5mm Smoovy motors are rated to be 30% to 40% efficient depending on loading. A quantitative determination of the manufactured rotors' flying geometry is needed to determine if the discrepancies are due to variations in geometry or due to shortcomings in the flow modeling.

Rotor Mold Fabrication

Efforts in this month are fabricating double-side press molds for 10" diameter rotors. Two sets of molds are required-one for left-handed rotor and the other for right-handed. Modeling procedures in CAD are similar to other rotors. The surfaces of rotor are generated according to the design. Necessary modifications for strengthening the structure were made at areas close to hub. Instead of creating a solid model of rotor, surfaces are used directly for upper and lower molds. The lower mold has cavities of bottom surface geometry, while the upper mold has elevated region with shape of top surface. Machining codes are generated and implemented in 3-axis CNC milling machine. Aluminum 6061-T6 alloy is chosen as mold material. Professional pre-cut blocks are utilized to assure the accuracy of substrate geometry, which can save works in substrate preparation. Alloy dowel pins are used for mold alignment. The pins are press-fitted to the lower mold and can be slightly slid through the upper mold; therefore, pins are anchored and the upper mold can be removed without any problem. Holes for pins should be rimmed exactly so that alignment will be precise. Two grooves from blade tips to the sides of mold are machined for the purpose of mold release. The resultant molds fit each other well, and the grooves helped significantly in mold release.

Systems Integration Testbed

A new version of the PCB flyer was designed. This version incorporates three Murata rate gyros. The output of the rate gyros is band pass filtered between 0.3 Hz and 800 Hz using RC filters. As before, radio communication uses a Linx Technologies 418 MHz receiver. The PIC17C756 microcontroller onboard the flyer reads RF control signals from the ground, samples the rate gyros at 60 Hz, computes stabilization parameters using a PID algorithm, and generates the PWM signals for the four motors.

The new boards have arrived and the code for both the microcontroller in the ground transmitter and the microcontroller in the flyer has been written, debugged and tested in a prototyping setup. The radio communication code has the ability to transmit the PID coefficients from the ground, thus facilitating needed gain adjustments for the flyer since the surface mount PIC on the flyer is one time programmable whereas the PIC on the ground unit is reprogrammable.

The simulator described in recent months was successfully used to simulate stabilization using the rate gyros. This also provided initial values for the coefficients in the PID control of the real flyer. The next step is to complete the assembly of a test flyer for tethered flight tests. Astroflight firefly motors and custom props (see above) are used.

Systems, Applications, and Other Items

Ilan presented papers on current and future mesicopter work at conferences in Florida (on Cooperative Control) and JPL (Biologically-Inspired Engineering for Exploration). These will be available on the web shortly.

Last update: 15-Dec-00 9:39:37 PM
Click the Reload or Refresh button in the browser to be sure you are viewing the updated page.
Click the Submit button to enter changes and view the formatted file.
WebEdit servlet by I. Kroo, Oct. 1999.