CFD Entry 1: Intro and Simulation Setup
In the process of designing our own aircraft, safety is not just a goal but a requirement. To better understand the properties and behavior of our aircraft we are using computational fluid dynamics or CFD. CFD enables us to simulate our plane in a virtual wind tunnel and gather data on lift, drag, velocity, pressure, and much more.
For our purposes, as a small group with limited funding, a lot of CFD software is out of reach. CFD is something that is generally reserved for professional engineering, as most good CFD software out there has prices suited for companies, not your average user. Currently, we are using SimScale, a free cloud-based simulation suite. However, just like with any free software, SimScale does have its limitations. Limited processing cores and memory hold us back from creating a detailed simulation, but since our airplane design is relatively simple, it does what we need it to do.
Setting up a CFD simulation has a few vital parts,
2. Material Properties of the Fluid
3. Initial conditions
4. Boundary conditions
5. Force Graphs
The first step we took was to bring the CAD model of our wing into the program. From that, we created an enclosure that will define the flow area of the simulation.
A large enclosure is essential to create an accurate simulation because the object you are simulating can have effects far from the object itself. In our simulation, we have the wing flush against a wall where the fuselage would be. The other walls are placed much farther away. The front wall of the enclosure doesn’t have to be so far away because it will act as our inlet. The data we are looking for will be happening around and behind the wing, not far in front of it.
(Note: It may just look like a speck on your screen, but the wing is actually inside the red circle)
Material Properties of the Fluid:
When choosing the properties of our fluid, we looked to see what conditions our plane would be flying in. We won’t be flying very high, or in very extreme climates. This allowed us to choose kinematic viscosity and density values that are around 23 degrees Celsius at sea level. We then assigned this value to the flow region defined by the enclosure.
Initial conditions define the behavior of the fluid at the start of the simulation. The only value we defined was the velocity of the fluid at 23.15 m/s; our cruising speed. Make sure that the direction of the flow is correct.
Boundary conditions define the characteristics of the surfaces of both the enclosure and the wing itself.
With a cartesian box enclosure such as our own, there are 6 enclosure walls to define. The first boundary to define is the one in front of our wing. We defined this as a “velocity inlet”. This means that the air will be rushing into the enclosure from that face at a set speed. We chose the value of 23.15 m/s due to it being our cruising speed. The top, bottom, and side walls of the enclosure are set as “slip surfaces”. A slip surface is one where the friction of the surface and “boundary layer” effect, which causes air to slow down as it approaches a surface, is not taken into account. This allows for the walls of the enclosure to have less effect on the air moving past it. The back face of the enclosure is then defined as a “Pressure outlet”. This means that it will allow air to flow out of the back, but it won’t control the speed at which it does so. This setup would be similar to having a wind tunnel with an open back wall, and a fan positioned at the front.
The wing itself also requires boundary conditions. Every surface on the wing is defined as a “non-slip surface”. This means that the simulation takes the aforementioned friction and “boundary layer” effect into account. This is done to make the simulation of airflow over the wing more realistic.
In order to determine the forces acting on our wing, we used a force and moment plot. This would give us a general idea of how much lift our wing was producing along with the drag forces acting on our surfaces. This step is just a clarification so the computer knows how to format its results. More information about looking at results will be provided in later sections.
Before we solve our simulation, we need to create a mesh. A more accurate mesh will give more accurate results later on. For the larger area of our enclosure, large cell size is adequate. Those are not the places where accuracy is important. Because of this, a coarse fineness is selected.
If we leave the mesh like this, our wing would be blocky and would not represent the actual geometry precisely. To solve this we created a box around the wing. With the area defined, we then used the box as a region refinement and defined the maximum cell size with an edge length of 0.05 meters.
Because of this, the mesh for the wing and the surrounding area is much more accurate.