Yo, analyze this...
In my last post I spent some time going over some of my design decisions, which ones were based on engineering reasoning and which ones not so much. This installment is going to cover some of the analysis I did early on and the application I found for this task.
When I first started talking to Adam about designing a wing he asked me if I was using an aerodynamic analysis software. It was at that point that I wanted to engineer the shit out of this thing. I started looking a round for software packages that would give me some insight into how well my design was working. I first looked at SimScale. SimScale is a very capable FEA and CFD analysis package that is cloud based. I have used it in the past to analyze some of my 3D printing designs. It is a sister company to OnShape, the 3D drafting package I use for my printing designs. The problem that i ran into is that SimScale is very powerful and it was overkill for this project. It was just to cumbersome to configure for what I needed.
I came across another package called JavaFoil. JavaFoil is a java application that is used to design and analyze airfoils. This is a great little package for airfoil analysis. This was a great starting point because I could compare different airfoils and look for the one that best suited my design requirements. It will let you look lift, moment and drag across a range of Reynolds numbers. The downside to this application is that it doesn't tell me anything about my wing design as a whole. For that I would need to step up to a more advance software.
I read a discussion post somewhere that mention a software called XFLR5, an airplane design package that seemed popular with glider pilots and designers. This is a fantastic application that will let you dig way down into your models design. You start a design by importing the airfoils you wan to use. If you are going to use a different airfoil for root and tip then you will need to import both of them. The airfoils coordinates can be downloaded from several different online databases or you could create your own. Once you have imported your airfoils you will need to run an airfoil analysis and generate your Polar graphs. These graphs are the input data the application will use to do the analysis out on the design. The next step is to build your wing design in the application. This is pretty straight forward and is the basis for the analysis. A nice feature is that you can build multiple version of your design, each with a modification so you can compare results. I build several version to test things like root and tip cords, different airfoils, washout or twist angles, different loading and CG points.
I tried several iterations and and used the applications different graphs to compare the results. The image to the right shows my wing design with a color overlay of pressure and the green lift distribution. Basically I wanted to look at lift vs drag, efficiency, pitching moment and lift vs alpha (angle of attack). The graphs below show the results of several runs. Some are different CG points and some are different wing designs.
I will briefly run through findings but I won't go into that much detail. There is a lot of information on the internet on how to best design a wing and interpret results. Firstly efficiency, my wing design has its maximum efficiency between 35 and 40 mph and is petty efficient right through 60 - 65 mph. This was one of my original goals. The upper left graph shows lift vs AOA. This shows a very linear graph through the AOA range I would expect to see. This airfoil's lift drops off pretty fast after 10-20 degrees. This is probably out of the range of AOA that I would expect to see with a wing. It is pretty hard to stall a wing because pitch is so shortly coupled. The elevons with such a short moment arm don't usually have enough authority to achieve an AOA that would completely stall the wing. Plus, as the wing angle increases the elevons pitch authority is decreased because they are "shaded" from the airflow by the wing itself.
The last graph, the one on the bottom left shows pitching moment vs lift. This is used to judge the relative stability of the wing. I used this graph to help nail down the CG and make sure that the wing would be stable when hit by wind and such. The rising blue line was a design that was not stable and would not return to a stable state after being upset. In the end this is what I came up with:
Wingspan: 42"
Root Chord: 15"
Tip Chord: 9"
Washout: 2 deg. (more on that in the next post)
Sweep Angle: 30 deg.
All Up Weight: 3-3.5 lbs
Batter Capacity: 10,000 mAh
Motor: Sunny Sky 2814 1250 kv. 400 watts 3s, 1100 watts 4s
ESC: 70 -80 A
Cruising Speed: 38-40 mph. lots of fun at 65 mph
CG: 6"-6.5" from the nose
I am going to dedicate the next post to a discussion on washout, or twist. This is a pretty important consideration and one that is often overlooked by scratch builders and manufactures alike. I am by no means an expert but I will share my understanding of the concept.
When I first started talking to Adam about designing a wing he asked me if I was using an aerodynamic analysis software. It was at that point that I wanted to engineer the shit out of this thing. I started looking a round for software packages that would give me some insight into how well my design was working. I first looked at SimScale. SimScale is a very capable FEA and CFD analysis package that is cloud based. I have used it in the past to analyze some of my 3D printing designs. It is a sister company to OnShape, the 3D drafting package I use for my printing designs. The problem that i ran into is that SimScale is very powerful and it was overkill for this project. It was just to cumbersome to configure for what I needed.
I came across another package called JavaFoil. JavaFoil is a java application that is used to design and analyze airfoils. This is a great little package for airfoil analysis. This was a great starting point because I could compare different airfoils and look for the one that best suited my design requirements. It will let you look lift, moment and drag across a range of Reynolds numbers. The downside to this application is that it doesn't tell me anything about my wing design as a whole. For that I would need to step up to a more advance software. 
I read a discussion post somewhere that mention a software called XFLR5, an airplane design package that seemed popular with glider pilots and designers. This is a fantastic application that will let you dig way down into your models design. You start a design by importing the airfoils you wan to use. If you are going to use a different airfoil for root and tip then you will need to import both of them. The airfoils coordinates can be downloaded from several different online databases or you could create your own. Once you have imported your airfoils you will need to run an airfoil analysis and generate your Polar graphs. These graphs are the input data the application will use to do the analysis out on the design. The next step is to build your wing design in the application. This is pretty straight forward and is the basis for the analysis. A nice feature is that you can build multiple version of your design, each with a modification so you can compare results. I build several version to test things like root and tip cords, different airfoils, washout or twist angles, different loading and CG points.
I tried several iterations and and used the applications different graphs to compare the results. The image to the right shows my wing design with a color overlay of pressure and the green lift distribution. Basically I wanted to look at lift vs drag, efficiency, pitching moment and lift vs alpha (angle of attack). The graphs below show the results of several runs. Some are different CG points and some are different wing designs.
I will briefly run through findings but I won't go into that much detail. There is a lot of information on the internet on how to best design a wing and interpret results. Firstly efficiency, my wing design has its maximum efficiency between 35 and 40 mph and is petty efficient right through 60 - 65 mph. This was one of my original goals. The upper left graph shows lift vs AOA. This shows a very linear graph through the AOA range I would expect to see. This airfoil's lift drops off pretty fast after 10-20 degrees. This is probably out of the range of AOA that I would expect to see with a wing. It is pretty hard to stall a wing because pitch is so shortly coupled. The elevons with such a short moment arm don't usually have enough authority to achieve an AOA that would completely stall the wing. Plus, as the wing angle increases the elevons pitch authority is decreased because they are "shaded" from the airflow by the wing itself.
The last graph, the one on the bottom left shows pitching moment vs lift. This is used to judge the relative stability of the wing. I used this graph to help nail down the CG and make sure that the wing would be stable when hit by wind and such. The rising blue line was a design that was not stable and would not return to a stable state after being upset. In the end this is what I came up with:
Wingspan: 42"
Root Chord: 15"
Tip Chord: 9"
Washout: 2 deg. (more on that in the next post)
Sweep Angle: 30 deg.
All Up Weight: 3-3.5 lbs
Batter Capacity: 10,000 mAh
Motor: Sunny Sky 2814 1250 kv. 400 watts 3s, 1100 watts 4s
ESC: 70 -80 A
Cruising Speed: 38-40 mph. lots of fun at 65 mph
CG: 6"-6.5" from the nose
I am going to dedicate the next post to a discussion on washout, or twist. This is a pretty important consideration and one that is often overlooked by scratch builders and manufactures alike. I am by no means an expert but I will share my understanding of the concept.
Enjoying the read muchly :)
ReplyDeleteThanks! I am enjoying writing it. I have had this stuff rattling around in my head for months now and it feels good to get it you. Please ask questions or tell me when I get something wrong. It's all a learning experience
DeleteVery helpful!
ReplyDeleteGlad I could help!
Delete