HEXA MONOBLOCK 814 retractable legs

[tinkerhaus]

Thanks Norbert. I figured it was an antispam measure but was in a rush and did not have time to confirm.

[jurgins]

Hi, this seems a nice build. Is it possible to get 3d files for the 3D printed parts?

[afernan]

Hi jurgins:
maybe you can be interested better in this new version:
viewtopic.php?f=25&t=4760

It´s much optimised mainly for vibrations (that are really important for AQ to get as low as possible).
I can share this design if you want to build it.
Angel

[jurgins]

This one is even better. What is the arm diameter? What 3D software are you using?

[afernan]

@jurgins:
arm diameter is 25mm. This is because vibrations are critical in AQ for hi-quality performances. Maybe you want to read this paper I did about that issue (http://autoquad.org/software-downloads/?did=127)

The use of a large diameter arm ensures high arm stiffness and so, decoupling from excitation frequencies (propellers rotational speed). Please note that AQ controller can´t use any kind of dampers below the board. It´s mandatory to attachach directly the board to the main body. This is because it uses inertial navigation (control) and it´s super-sensible to body accelerations needed to actitud determination.

I use CATIA for design.

See below my last mount with 30mm tubes (0.5 skin thickness) and pushing props . This conf improves in about 300gr the lifting force and drecrease to almost zero the aerodynamic vibrations generated by the arm-propeller interference. It uses ESC32V3 located inside the tube. All parts are 3D printed (excep tubes) in PLA-carbón loaded fillament
I get 45min with a 5000mA, 4S battery. Vibration levels are < 0.5m/2

Angel

[jurgins]

I was going to use 16MM Carbon tube with ~2mm wall thickness. I have them already. You say it would be better to use larger diameter with thicker walls? Im going to use 720KV Tiger motors and to build hexa cause Im making it for areal photography and video.

[afernan]

I recomend you to use ESC32V2 or V3 (much better), and Quatos controller. Performance is much better.
PID is much more sensible to vibrations than Quatos. Also frame is the critical.

The bottom line is to decouple excitation from rpm to natural frequency of the arm (see the paper i´ve reffered). If you don´t know how to do the numbers, give me a few values and i can do it for you. I´d need:
-AUW
-tubes (inner, outer diameter, tube length)
-weight at arm tip (motor+suport+props)
-Propellers you plan to use.

Angel