



I forgot that the local search isn't parallelized (it doesn't really need to be anyway, because it's pretty fast already). I just tested on my laptop with only PSO, and limiting it to 100 iterations so that it would perform the same number in parallel and serial. Here are the results:
Serial: 40 min 14 sec Parallel (two cores): 21 min 56 seconds So pretty near 50% speedup in parallel. I'm pretty happy with that. I haven't tried it with more than two cores, though. Anyway, thanks for your comments. I will try to make those changes to the user guide and upload it today. 



Joined Sep 2015
7 Posts

Hello again Dan,
I have been trying to understand your code to make some modifications but quite honestly my Fortran skills are not very good (I am more a MATLAB guy, don't work with compiled code too much). I am interested in modifying the "airfoil_evaluation" code because I would like to include constraints to set the minimum airfoil thickness and curvature radius in a specific point along the chord (for example, to design a symmetrical airfoil that could act as a fairing for a cylinder). The radius of curvature and thickness values along the chord are already computed for the other constraints during the "airfoil_evaluation" program so it would be just add two penalty contributions and three new inputs (chord position and minimum thickness and curvature radius values at that position). Could you help me? Thanks 



Epico,
The easiest way to do it would be for you to determine the point(s) where you want the constraints specified, and then apply penalty functions there. If you are familiar with Matlab, you should notice that the array syntax is the same in Fortran. I'm at work and don't have the source code with me right now, but basically you will figure out the array indices corresponding to the xlocations where you want to enforce those constraints for the seed airfoil you are working with. I'll give more specific directions when I get home and can look at the source code, but you can start figuring that out now if you want. Dan 


Joined Sep 2015
7 Posts

I got past my initial laziness and after "fully" understanding your code I managed to make the modification to introduce the minimal thickness constraint at a specific point. I've got null experience with Fortran and needed a lot of hours to do this but it has been useful to learn a bit the sintaxis.
I plan to introduce some other modifications (one is to use turbulence intensity to calculate ncrit instead of explicitly defining it) and I will report back. 



Quote:


Thread Tools  
Similar Threads  
Category  Thread  Thread Starter  Forum  Replies  Last Post 
Discussion  Evolutionary airfoil design optimizer available  kcaldwel  Modeling Science  4  Nov 08, 2015 12:04 PM 
Discussion  Airfoil Optimization with XFOIL  Montag DP  Modeling Science  181  Aug 21, 2015 01:47 PM 
Discussion  Airfoil Optimization Using Genetic Algorithms  Ghost_  Modeling Science  38  Apr 24, 2014 07:18 AM 
Profili  Xfoil and reflexed / autostabilizing airfoils  surfimp  Modeling Science  29  Mar 30, 2005 07:32 AM 