For this year’s Knights Experimental Rocketry (KXR)’s submission to the International Rocket Engineering Competition (IREC), the aerostructures team designed a full carbon-fiber pre-preg based structure, which provides stronger structural characteristics while also being lighter than previously-used alternatives. The issue with carbon fiber, to the dismay of me and the rest of the telemetry team, is that carbon fiber’s conductive characteristics completely block radio waves.
To allow communication with the rocket avionics system from the ground, I had to come up with a solution to externally mount an antenna tuned to the 70cm (433 MHz) band that could withstand the nearly Mach 2 flight loads and ground impact for full recovery points.
Inspiration
A few months prior to this issue coming up, my Ham Radio friend who develops satellite antennas, Stan Wood, showed me this antenna from an old NASA program he had. Although I don’t have a photo of the original one, this photo is of similar design. These antennas have a sleek, aerodynamic shape, and are made of machined metal and dielectric material (commonly hard plastics or ceramic). I figured this type of design was a good starting place.
Design
To design this antenna, simulations using ANSYS HFSS were done to determine the exact dimensions of each element of the antenna. Without computer simulation, Hundreds of prototypes would have been needed.
Over the span of multiple weeks, I tried multiple different designs with slight differences. I would typically create one base design and slowly modify different dimensions while looking at the S11 return loss to tune it to the desired frequency. Different designs, S11 plots, and far-field gain plots can be seen below.
After coming to a design with return loss and gain I was happy with, I created a CAD model of the ANSYS simulation model using Solidworks. I confirmed the characteristics of the model by bringing it back into ANSYS and simulating it again once the full CAD design was completed. I was incredibly happy with how the design looked, but after exporting it and bringing it to the UCF Machine Shop, the machinist informed me that it would be incredibly expensive to manufacture due the complex geometry machined out of one piece of aluminum. Fortunately, he was willing to mentor me in cost-effective and machinable design and the creation of manufacturing drawings.
Final Manufacturing Design
To decrease machining cost, I modified the design to be made out of three main pieces. The aluminum base and top, and the Delrin middle dielectric.
With the help of the UCF machinists Jim and Arturo, I came up with a design with dowel pins in the aluminum going into slots in the Delrin so it can slide back and forth about 3/4 of an inch after assembly. The reason for this is that the match characteristics are incredibly dependent on the exact location the feed pin contacts the upper piece, but the tuning is also dependent on the Delrin being fully in place. The Delrin being able to slide back and forth allows for the feed point location to be modified slightly using the set screw, and then the Delrin slid back into place to check the match and tuning with a network analyzer. This allowed for easy tuning during assembly. Two bolts in the back permanently secure the Delrin in place once the tuning is completed.
Final Manufacturing Drawings
Thank you to Jim and Arturo for teaching me how to make good, readable, manufacturing drawings.
Result
After Arturo machined the antenna pieces and I assembled the antenna, I found the match and tuning to nearly perfectly line up with the simulated values. I did find that because of the carbon fiber’s effect on the match, I would have to include a metal ground plane sheet curved around the portion of the rocket where the antenna would be located. Although unfortunately the rocket was never able to be built to competition, I do hope to still be able to test my antenna in flight some day.
