Bring your product to market on time and within budget – join the Eurocircuits Community
As explained in our last blog Torus Blog 1, the Torus team from the Collège Saint-Pierre (Uccle), participated in the CanSat project, organised by the ESA with the support of Eurocircuits for the production of the PCBs. The aim of this contest is for the teams (of six teens) to build a small “satellite” in the volume of a soda can and retrieve various data thanks to it.
We got selected, after two selections rounds based on intermediate reports, to participate in the launch during the finale of the Belgian in Elsenborn on May 8th, 2021, together with 18 other Belgian teams.
Our team was lucky enough to earn the second place, and this is a summary of our results.
Overview of the Flight
During the flight, we gathered 628 records, for a total flight duration of 25 seconds. Our RF-Transceiver lost the can’s signal 17 seconds after the launch because the rocket was out of distance. We received 80% of the data stored onboard while the RF link was active. During the flight, the rocket reached an apex at 329 m from the ground, 11 seconds after the launch.
This a 3D plot of our can’s trajectory using Google Earth.
The yellow trace represents the path of the can during the flight phase. The black marker, which is about 1.5 km away from the launch site, represents the last data point that was successfully received by radio.
A few seconds later, the can got ejected – at only 190 m above the ground which is much lower than the expected 1000m. Finally, in purple, you can see the landing spot.
For the primary mission, we gathered temperature and pressure data. From the pressure data, we could derive the altitude data.
This is the temperature data collected by the BMP sensor and our 3 thermistors. The response time of our thermistors is, as expected, much better than the BMP’s. In less than 3 seconds after ejection the readings of our best thermistor (blue curve), reaches the actual air temperature and then increase again as the can descends.
There were some instabilities. The first group is easily explained by the turbulences when the can casing opens before ejection, the second is less clear (possibly an inhomogeneity in the atmosphere).
Here is the raw pressure data as a function of time as collected every 70 msec by our BMP388.
The pressure first drops as the rocket gains altitude and then lifts again as it should.
At around 28 seconds, there is a disturbance in the pressure records. We assume that these values might be caused by an turbulent airflow when opening the cansat casing at ejection time. It makes those values irrelevant for altitude calculation.
At touchdown, there is a discontinuity in the pressure measurement: our hypothesis is that given our excessive descent velocity (20 m/s), there was a significant dynamic pressure, which cancelled immediately when the can reached the ground.
As you can see on the graph which is based on the pressure data without irrelevant values, the rocket reached an apex of around 329 meters before falling. The reason for the discontinuity at touchdown is that we calculated the altitude directly from the pressure. So, at the touchdown, the excess of pressure due to dynamic pressure cancelled, which results in this apparent change in altitude.
After ejection our descent velocity was stable but out of rage (20 m/s) due to our parachute issue (more on that later).
During the flight, we wanted to see how the parachute shape (which we could change using a servo) influenced the descent velocity. We unfortunately couldn’t collect meaningful data for several reasons:
Firstly, the can was ejected very late during the rocket descent and much closer to the ground than expected;
Secondly, a knot that fastened the parachute to the can went loose, and the seam located between the severed ropes didn’t hold. Although, this is not an excuse, we should note that ejection happened at a velocity of about 360 km/h…
Nonetheless, the can piloted the experiment exactly as expected in those circumstances:
The controller runs the experiment when the can is falling within safety limits (fall velocity between 5 and 16 m/s, altitude above 115m). These conditions were only met twice during this short flight (red areas in the descent velocity graph below).
In those two cases, the servo positionned as expected as shown.
During the rest of the flight, safety conditions where not met and the servo fully deployed the parachute.
When we recovered our can, we saw no part had been damaged. Everything was in a perfect state besides the parachute, and this is clearly partly due to the use of a PCB rather than artisanal circuitry. We are very grateful for Eurocircuits support in this matter.
This project proved that our architecture and design were adequate, but that we still have to improve the quality control of the implementation of our parachute:
The can structure was undamaged and the electronics still worked perfectly after the flight;
Our whole transmission chain worked reliabily from can to ground receiver, real-time processing, transmission to back-end and update of web pages to broadcast our live telemetry.
As for the parachute, we obviously made something wrong in the final assembly because one knot failed. Although we performed the required severe tests (we lifted 54 kg in the parachute fastened to the can), and the constraints on the parachute were significantly lower than in our tests, it still failed.
We’ve also be able to involve many people in our project inside and outside our school, and hope we succeeded communicating and sharing part of our enthusiasm for science projects.