Goal: Determine which motors to use, measure accurate thrust curves, and measure ignition delay time.
Why?: Measuring the thrust of the motors over time will allows us to accurately simulate future control algorithms.
The results from 3 tests were pretty repeatable, and we were able to obtain an approximate ignition delay time. The actual thrust profile differed significantly from the data from NAR, which also did not include the delay time. The burn time was also more than 1.5 times as long as the expected time. This data will be a good baseline for designing the orientation controller.
Next steps: Can we control the orientation of the rocket with these motors?
Now that we have chosen our method of propulsion, we need to determine which motors to use since hobby rocket engines vary in total impulse and thrust duration. We also need to figure out the variance in ignition delay time and confirm the accuracy of NAR thrust curves. The NAR data only has 32 data points with standard deviations just for total impulse, peak force, and burn time. We want to get a precise set of data with an estimate of the potential error in that data. These thrust curves will be critical to developing the rocket control systems.
Our first consideration was overall burn time. The longer the engine burns, the more time we have to control the orientation and thrust. The settling time of the attitude and altitude controllers have to be shorter if the engine burns quicker, so longer burning motors will provide more flexibility in design.
Overall impulse was also an important factor. We planned to design for lightweight components, but we can only limit the weight of the rocket so much, so we need the motors to provide significant thrust.
With these factors, along with cost and availability in mind, we decided on Estes C6-0 engines. With a nominal burn time of 1.8 seconds and an average thrust of 4.5N, these seemed to be our best bet. Estes engines are easily found in most hobby shops, and the C6-0’s are the best compromise between price and burn time. The motors with the same or longer burn time are more than 3 times the cost of C6’s.
Based on an existing load cell setup, we designed and machined a sleeve to safely hold the motor for the thrust tests.
The load cell used was an Interface Model MB Miniature Beam Load Cell. The load cell was setup to output up to 10V, so we stepped it down to the 5V range of the Arduino’s analog input, and used a voltage follower to ensure a 5V limit. This circuit is seen below.
We also needed circuitry to ignite the motors. We used a transistor so that we could supply the minimum fire current of 2A to the motor igniter. This circuit can be seen below.
The LabVIEW and Arduino code used for data acquisition are provided here [Dropbox]. The Arduino ignites the motor once a signal is sent from LabVIEW. One this signal is received, the Arduino writes the force data from the load cell to LabVIEW, which writes the data to a text file.
The results of the 3 tests are shown in the graph above. The ignition delay time was 584 ms, ±58 ms for these tests. The curves look extremely similar, but offset by slightly different delay times. Our data is significantly different from the NAR data for the same motor, so these tests ended up being very important. We found the average thrust curve by matching up the peaks of the three tests and then offsetting it by the average delay time. Graphing the data with the peaks placed at the same point in time showed that the data was very consistent, since the three curves were right on top of each other with some slight noise variation.
As we expected, the thrust profile was non-linear, with a sharp peak at the start that tapers off to a plateau for the remainder of the burn. The constant portion of the burn could provide a good linear approximation to simplify our controller.
The nominal burn time was 1.8 seconds, but our data shows an average burn time of about 3 seconds, which is good because this means we have more flexibility in our design.
The next step is to start designing the mechanical actuation of the motors. We have to figure out how to control thrust magnitude and direction, and test if this mechanical design can be actuated fast enough with available hardware. Once we establish a mechanical design, we will test to see if this design will be able to reorient the rocket as if it were in freefall.