Project Update
Carbon Fiber Reinforced Polymer (CFRP) Drone Frame
ME514
Hongrui Chen
Zeyuan Ma
Shujing Wei
Department of Manufacturing Engineering
University of Wisconsin-Madison
April 5th, 2019
1. Design Explanation
1.1 Overall Design
The concept of our drone design is based on aerial photography and surveillance. Our drone can carry one navigation camera, one multi-function camera and a gimbal for better quality aerial photography. There are a front and rear LED ring to illuminate. This serves as an indication of the facing of the drone so that the operators have a visual clue of the direction.
There will be two versions of the drone. One is optimized for FFF printing with a thicker wall and no lattice structure. The other is CLIP optimized version, which uses more lattice structures to reduce weight and improve structural rigidity. Topology optimization is used as a design reference to improve structure stiffness and also to reduce unnecessary materials.
Figure 1. Rendering of the Drone
Main Cabin Design
The main cabin is used to house the electronics of the drone and also to provide mounting space for the arms. Therefore, the main cabin needs to be spacious and stiff.
In addition, the main cabin provides mounting holes for the gimbal.
Figure 2. Main Cabin Design of the Drone
1.2 Topology Optimization
Topology optimization is used for structural components. With topology optimization, the part can have the benefit of lightweight characteristics and of the improvement in structural rigidity. The topology study is done in Solidworks 2018.
To create a topology optimization study, a rough drawing of the part was used to define and constrain the design geometry. Next, fixtures and load conditions were added to the part. Topology optimization is conducted to the arm, the main cabin, and the rear landing gear. Finally, simulations were run to identify geometrical features. Those key features can then be recreated on the model.
Figure 3. FFF version of the arm design
Figure 4. Topology Optimization of the Arm
Figure 5. Topology Optimization of Rear Landing Gear
2. Results
2.1 Print Statistic
A preliminary print was conducted to evaluate the weight, volume and printed time. By doing so we can find the advantage of different material. The results are shown below.
Part Name | Volume (cm^3) | Weight (g)/From Cura | time(min) | Material |
Main Cabin | 107.88 | 82 | 357 | PLA |
Each Arm | 54.75 | 29 | 144 | T-PLA / CLIP |
Front Cover | 34.51 | 34 | 276 | PLA |
Rear Cover | 16.48 | 16 | 108 | PLA |
Battery Strut | 8.10 | 8 | 47 | PLA |
Battery Hook | 2.36 | 3 | 28 | PLA |
Center Strut | 6.25 | 8 | 43 | PLA |
End Strut | 5.35 | 5 | 24 | PLA |
Front LED Cover | 11.31 | 14 | 69 | PLA |
Rear LED Cover | 5.36 | 7 | 33 | PLA |
Table 1: Printing statistics
2.2 Design challenges in CLIP printing
Original Arm Design
The design of the wing was the major challenge since the CLIP technology requires lattice structure in order to reduce material consumption and improve strength.
The original arm design featured a less dense lattice design to improve airflow through the arm. However, in order to bound carbon fiber to the structure, the part had to be only cured of UV without final curing from heat. After applying the carbon fiber onto the part, the part needed to be placed in a vacuum bag and cured in an oven to fully bound the carbon fiber. With the original design, the resulting part was relatively soft. However, since the original design was too soft, the arm was only half-cured when printed by UV-light. There is a lot of deformation in the lattice structure, resulting in difficulties in applying vacuum. As a result, the original arm printing failed.
The bonding between the arms trial part printed through CLIP technology and carbon fiber is good. This was shown in figure 7 and figure 8. The carbon fiber could attach to the surface of the arm in good quality and stabilize the structure of the arm to some extent. However, during thermal curing, the heat required further decrease the strength of the part, resulting in large deformation of the part.
Figure 6. Original Design of the Arm
Figure 7 and Figure 8. Experimental Printing of Original Design
2.3Design challenges and Changes in FFF printing
Quality
The FFF parts were printed using Ultimaker 3s from Makerspace and Polymer Engineering Center lab. One of the Ultimaker in the PEC lab got a loose belt. The resulted print have poor quality around the edges. 3 pairs of arm were printed in Makerspace. One of the arm suffered from warping around edge. Overall, the quality of the FFF printed part was satisfying.
Tolerance
The drone design involves some part to fit the nut. The M3 nut used in the drone had a minimum radius of 5.44mm. However, in the Solidworks drawings, dimensioning the hole for the nut caused the printed part having difficulty in inserting the nut. By using gradual increment in diameter, additional test print suggested that 5.6mm is the best dimension for inserting the nut.
Tolerance is also considered where two parts mate with each other. In this situation, the less important part reduced dimension in order to fit.
3.Improvement
Improved Arm Design
The new design featured a more dense lattice structure. In addition, it also reduced material use compared to a full lattice design.
A single lattice cell is designed and shown in figure 8. The linear pattern feature in Solidworks is used to create the bulk part of the wing. Some other designs were added to the wing in order to strengthen the structure and provide room for carbon fiber to attach on.
Figure 9. Improved Design of the Arm
Looking forward, we aim to further improve the stiffness of CLIP parts. As for FFF design, we aim to optimize the design so that it will need less support. After testing and verify that all components fit and mates, the FFF version will be updated with CLIP printing in mind. In addition, the front and rear grills can be printed with SLA. Printing in SLA can improve the finish as the grills are harder to print with FFF printers.