I wanted to make an improvement to the quadcopter since the timber arms, although cheap to replace, tend to break too easily. The main problem is that on impact the force tends to split the timber where it mounts into the hub.
I came up with a quick printed arm design and then did some testing. This also gave me a chance to compare the real life results with the FEA simulation. Here are the results. The print settings were two shells, 5% fill, 0.2mm layer height and took a little over 2 hours to print using slow enough print speeds to get a nice print in ABS.
The testing was using a small digital kitchen scales with the printed arm held in an adjustable wrench, producing a load of ~800gms.
As can be seen the displacement from an “upward” force (lift force or downward impact with the ground) is excessive. The simulation sees 16mm at the tip which is close to what is seen in the real life test. The arm was definitely too flexible in the up/down direction with too much stress & strain in the part.
The part performed much better in the side loading where the trellis structure comes into play.
The arm was much stiffer with the side loading test ( similar to a forward impact or hitting the ground at speed).
A little bit of flex is wanted to absorb the impact forces and dampen vibrations but if the airframe is too flexible it will lead to problems with the stabilisation computer and poor responsiveness from the quadcopter in flight.
With a bit of work a much better design could be made which would perform well in all directions while also increasing the stiffness. There is also a new filament for FDM printers which has chopped strands of carbon fibre in the plastic. The fibre filled filament could make a big improvement to the strength of the part while also keeping weight to a minimum.