Ready to fly....Griffin...

Introduction:

UAVs (Unmanned Aerial Vehicle) has its own category of researches. However, most of them are software focused, which will expand the functionality of the existing platform but will never exist the engineering capacity of the design. In my research, I’m going to design, build, and test a quadcopter that will achieve one simple goal — the fastest vertical ascent from 0 to 100 meters. The measure of success is based on breaking an existing Guinness World Record set by a German Engineer called Dirk Brunner. He designed a quadcopter that can fly to 100 meters above ground from stationary in just 3.871 seconds. My approach of breaking this record is by optimizing my design both in the CFD simulation environment and real world testing senarials. The work will be divided into multiple phases: preliminary drawing, CAD, early assembly, structural strength testing, propulsion system testing and comparison, prototype assembly, prototype test flight, cowling design, CFD simulation approval, final assembly, and test flight, and repeat.

I named my drone “Griffin” and hope it can break the Guinness record. The griffin (Greek) is a legendary creature with the body, tail, and back legs of a lion; the head and wings of an eagle, the griffin is the fastest flying beast in Greek mythology.

Materials and Methods:

After my first few preliminary drawings, I started to model the fame of the quadcopter in Autodesk Fusion 360. I used Fusion instead of AutoCAD because it provides me both a visual representation of my design as well as a 3D file that can then be used in the CFD simulation. For this project, I make a choice of using a “pushing” design so that the motors are actually pushing the frame from below instead of dragging it from above. In this way the motors will be less likely to become lose through time and tear themselves apart. Plus, the fast air flow that is generated by the propellers under the body of the quadcopter will not be blocked by the arms of the quadcopter frame. After I solidified the overall design, I used Autodesk CFD (and simscale.com) to verify my design’s aerodynamic characteristics and to design an outer cowling for the body of the quadcopter so that there will be less resistance at the nose and less vortex will be generated at the tail. Thus less drag. I used the T-spline algorithm inside Fusion so that I am able to modify the design of the shape of the surface precisely according to the CFD results.

RESULTS:

I have gone through a number of iterations before I reached my current design. Most of the changes yet are about structural integrity and aerodynamic characters of the design. Every prototypes that I made are made of plywood, and CFRP will be considered if the design is verified through real-world tests. The starting point of my design (as shown in figure I) appears to have some weak points at the corners. I then modified the ribs that supports the arm so that it is more smooth in shape and is not going to fail on any of the connections during the flight. The built-in FEA simulation inside Fusion was used to simulate the stress on the arms and on the frame as a whole. Parts with small stress were removed to reduce weight. However, since plywood is not a linear material, I was not able to use plywood as the material for the simulation. According to the CFD simulation results, the current design of the frame itself produces around 15.5 newtons of drag when

flying at 50m/s, and the motors that I am using are capable of providing about 600 grams of thrust each with 6-inch propellers. Thus 24 newtons in total. From the trace plot of the CFD result of my current design, a negative-pressurized zone was observed below the frame,(as shown in Figure II) which will cause turbulence and pressure drag. To optimize the drag performance, I tried to design a cowling around the center part of the frame. As the current design goes (as shown in Figure III), this addition will reduce the overall drag by 22.6%. I believe with some more iterations of the shape of the nose and tail cone, we can expect an about 30% reduce in drag.