How to improve DJI Spark. Part 2. About coal
It was in the evening, there was nothing to do (c)
There was a comment from @RusikR2D2 to the previous article about the propeller:
У вас получился винт, оптимизированный для "нулевой" скорости (висение). Ваш коптер теперь будет медленнее летать, медленнее набирать высоту, скорее всего, "тупее" реагировать на управление и хуже держать ветер. Также могут появиться вибрации (контроллер настроен именно на то соотношение тяги-оборотов, у вас же оно стало другим)
In this article I will try to figure out whether this is actually the case, what minuses and pluses will arise, how these propellers will behave in flight.
The first version of the propellers was 3D printed, which resulted in the blades being extremely flexible. Frankly speaking, I was afraid to carry out tests with such blades on the street because of the loss of the copter. And I decided that for the street I will make the most rigid and shape-holding screws. Of course, I will make them from composite materials, namely epoxy resin reinforced carbon fiber as roving.
But, before you do anything, you need to get equipment. Equipment, in this case – a matrix (form), consisting of two halves. The matrix is designed based on the product to be obtained.
After designing, the matrix can be made in various ways, which I recently wrote in my other article. But, today, the matrix, as well as the prototypes of the screws, I will print on a photopolymer 3D printer. It’s simple, fast, cheap and affordable to play at home, which is exactly what I want to show that using not quite specialized materials you can get a good result.
After printing, the form needs to be processed. The matrices are processed with sandpaper with a grain size of 800 to 1500, about 5 minutes are spent on one half. In the photo below you can see the results of processing.
Because epoxy is a rather sticky substance, it is assumed that the molds need to be protected, if this is not done, then the product will simply not stick to the tooling and the result will be near zero. Therefore, it is necessary to apply a separating layer on the form. There are specialized release systems based on wax, silicones or semi-permanent lubricants, but our task is to make the most of what is at hand. And I have shoe cream on hand. It is necessary to apply it on the form, wait until it dries, then polish and repeat this process 2-3 times. This is especially important for new forms. The second moment – superfluous in the course of processing – not to remove superfluous.
On this, the matrices are completely ready for molding. But we need coal (carbon roving). I went to Avito and contacted the seller about the possibility of acquiring a small amount of the desired thread, the seller went to meet me and sold 5 meters. Epoxy resin is sold in almost any hardware store (ED-20, for example).
You can start shaping. I impregnate the carbon fiber with epoxy resin and lay it in a mold.
After a day, the resin hardens and the product can be removed. In the photo below, the removed product from the mold with flash that needs to be cut off.
After processing, the blades are weighed and assembled into finished propellers. I weigh in order to collect the blades with the closest mass.
Ready homemade screw and its mass. (2.53 g)
Mass of the standard screw. (2.61 g)
For verification, I conducted tests in the room, everything is the same as with the printed blades from the previous article. The increase was just over 8%, which is 1% more than the result on the printed blades. Perhaps this is due to the fact that
And the moment comes for which everything was thought – tests on the street.
First, a copter with standard propellers was sent into flight and it flew away. Yes, yes, he flew away and did not return, he rests somewhere in the forest. Below is his last frame. I did not keep track of the fact that the copter at an altitude of 250 m began to quickly move away from the starting point (I admired the clouds :)). Then loss of signal and yet.
I had to buy the same and repeat the test. The second one went much more smoothly. Due to the loss of the copter, I decided to reduce the maximum climb height from 300 m to 200 m. And the test program is as follows:
Stopwatch start;
Ascent to a height of 200 m;
Descent to a height of 50 m.;
Flying along the road along the route until the moment when Spark wants to fly home;
Arrival at the point home, descent to a height of 1 m, forced retention of the copter in flight;
Landing due to the inability to stay in the air anymore.
Stopwatch stop.
First I fly on regular propellers on the first and second batteries, then on homemade ones.
Test video.
And, of course, the measurement results.
Standard propeller
1 battery 15:10 (910 sec)
2 batteries 15:00 (900 sec)
Homemade propeller
1 battery 16:35 (995 sec)
2 batteries 16:20 (980 sec)
It turns out that when using the first battery, the increase was 9.3%, for the second battery 8.8%, i.e. essentially the same gain as hovering.
But, an attentive reader may notice that the flight time on the spot indoors and outdoors is very different. Indoors, the time did not even reach 15 minutes, and on the street more than 16 minutes, on the street there is a wind that prevents flying.
The secret lies in oblique blowing propeller, due to which the air flow through the propeller increases, and consequently the thrust increases at the same power.
This effect is also used by large helicopters when it is necessary to take off either with an overload or in a rarefied atmosphere (in the mountains), but for this you need to take off by plane. Below is an example of such a takeoff.
Returning to the initial question about what turned out to be an optimized propeller for zero speed. Yes, this is partly true, but due to the fact that the speeds on the propeller, caused by its rotation in the medium, are quite high (the angular velocity of the propeller is 200 rpm), and the speed of the apparatus in the medium is rather small, 2-3 m/s, then and it turns out that there is simply no negative impact of the latter on the final results. Another thing is if an aircraft was involved in the experiments, then yes, the influence would be significant and the efficiency of the propeller at a speed of 0 m/s would be very different from the efficiency during cruising flight.
That’s all, thanks for your attention. Ready to answer questions.
