Airflow modeling software on a small scale

[banshee43]

Hi there, I think this is in the right section if not please inform me and move my thread. I want to design and build a UAV in a copter format whether it be quad, hexa, octa, ect. I want to make a modeling software so that I can design the most efficient blade/motor for my specific needs. I am a Computer Science major in college so the programming is right up my alley. I am however, extremely novice in the aerospace sector. Could you please help me understand what types of relationships, formulas I will need to use?
To boil it all down, in order to model air and its reaction to external factors what does the 'computer and/or I' need to know about the air? Or, if you could point me in the right direction to find these answers?
Thank you for any help you can supply.



Its not the fall that kills you its the F*Δt

 

[RandomGuy88]

This is not a trivial problem. Do you actually want to design the blades or are you just trying to determine things like number blades, rpm, motor torque...

Actually simulating airflow over a body like a propeller blade is no simple matter depending the on the necessary level of fidelity. The most accurate would be computational fluid dynamics but I doubt this is what you are looking for and it is not a great design tool. For this sort of problem you might want to take a look a blade element momentum theory. Also, you can use programs like Xfoil to simulate the flow over 2d airfoils with reasonable accuracy.

 

[boneh3ad]

This depends on how accurate you want your solution to be. Air flow is an extraordinarily complicated process (cf. Navier-Stokes equations or Direct Numerical Simulation) but there are a variety of methods for reducing the complexity with the sacrifice of accuracy. So, in other words, it all depends on what you are looking to achieve.

If you want to program something from scratch then I'd start with a relatively easy concept like a panel method code, which can give you the inviscid flow around a shape (such as an airfoil) and give you an estimate of drag (excluding viscous drag) and lift on that object. You could then move on to more complicated things like a 3-D Euler solver. Then again, you could also go out and just use a code that is already available.

 

[banshee43]

Thanks for the info this gives me a good jumping off point to start to actually understand the different methods. Do you have any suggestions as to a method for pressure differentials or is it included in the methods? It is not necessarily that I want to design the blades but for my understanding of efficiency under different circumstances. There is a fair amount of accuracy that can be sacrificed, seeing as I am not dealing with speeds over 75 mph, so I may steer away from CFD.

 

[boneh3ad]

The speeds have essentially no effect on your solution accuracy. 5 mph is no different than 75 mph is no different than 150 mph when it comes to what method you use, especially when it comes to inviscid solvers. The only time it could have an effect is if you are trying to account for turbulence, where the Reynolds number will affect your accuracy, or once you get fast enough to see the effect of compressibility, which requires more complicated equations (or rather fewer simplifying assumptions on the same equations).

 

[rcgldr]

inviscid solvers.

Would this really be accurate enough to design a propeller? Also propellers on a multi-rotor UAV have to deal with induced flow during hovers, which complicates matters.

Another issue - how do you plan to create a propeller? This can be risky since the propeller can fly apart if not designed and manufactured well.

 

[boneh3ad]

He isn't designing the propeller, but trying to look at the efficiency of a given design... or something like that. It isn't entirely clear what he wants to do, so my answers won't necessarily be 100% applicable to his actual application. An inviscid solver would give a good starting point for the forces on a given design, as it would really only leave off viscous drag. You could get the thrust fairly accurately without that, if I'm not mistaken, so the only remaining issue is if he wanted to look at propeller stall or something else that necessarily needs the concept of boundary layers built-in, in which case he can move on to a more capable solver.

 

[satchurated]

Try ecalc, is not enough reliable because it analyze from an electrical/weight point of view with a database of some motors and propellers you can found in the market. Is a good point, not the bible...

 

[RandomGuy88]

I agree with boneh3ad that an inviscid method is capable of accurate predictions of prop performance assuming the blades are not stalled. The main thing that would be missed is the amount of torque required to turn the prop because some of that torque is due to drag on the blades. However it is possible to predict these effects using something like Propeller blade theory where you can supply the code with real experimental data for lift and drag coefficients.

 

[satchurated]

If you search on the web, there is a paper about a quadcopter and the momentum theory

 

[phys_student1]

I can suggest AVL. it is an open source program from an MIT professor:
http://web.mit.edu/drela/Public/web/avl/