We attempted a little parametric study to optimize our UAV's flight time using the motors, propellers, and # of batteries as our controllable variables. First we optimized the rotor/motor combinations which is described in another thread. Then we decided to optimize our battery count by incrimentally adding 6000 mAh batteries to our hexacopter, flying it, and calculating flight time based on the current draw and vehicle charge capacity. After doing a few iterations at 1, 4, 5, and 6 batteries, we were able to generate the following curves. Our optimal vehicle weight ended up around 8.5 lbs. We thought this would be interesting and worth sharing. Enjoy!
Is your graph of flight time (red) saying you obtained >40 min with 5 batteries in parallel?
I think that is what your data is saying.
Why not go ahead and do 2 and 3 batteries? (or did you and left out the data points?)
Regardless, thank you for the effort and information.
That is indeed what the graph is saying. We didn't do 2 and 3 batteries simply because we didn't have the time. We were only intersted in finding the optimal point which we were lucky and ended up finding in the first 3 tries! I hope this information is somewhat useful. I should be able to mod the data in such a way that you can optimize your vehicle based on the power densities of the batteries. That may make things a little more generic and useful.
ok, one question though. Did you run the batteries for a full flight as you ramped up more batteries in parallel. Your write indicates you added batteries and then calculated based on current draw. Did you in fact time the full flights? Its hard to tell from your write up whether you actually did the full flights for each battery count to yield the data.
If you did not, then I would suggest you do. The data may not be the same, I have seen as batteries drain the quad has to do more throttle increase to maintain hover, this causes a amp increase and this causes the batteries to drain quicker from mid-flight until the end of the flight. Ultimately this reduces the flight time as you go through mid-flight to end. If you did full flights then this is moot, but if you didn't then your data is not taking into account this situation of throttle increase from mid-flight. All FCs have to do this to compensate to maintain GPS hover or alt-hold hover. When I test my quads and ottos and hexas, i leave it at 50% throttle and put it into one of these modes, this guarantees I will get a good data analysis of flight time and represents what the FC must do to keep the quad in hover and level flight through out the whole flight. Again this is assuming you did not do full flights with each battery setup you reference above in the graph.
For the interest of time, we did not do full duration testing. I think you are 100% correct in your statement. We will do full duration testing at some point and update as needed. Thanks for the input!