In the month of June the Stratos II+ project team carried out four static motor tests of the DHX-200 Aurora, the engine powering the Stratos II+ rocket, at TNO located in Rijswijk. The goal of these tests was not only to gain confidence in the motor design, they were also necessary to gather detailed performance data in order to accurately predict the rocket’s flight, and to analyze the impact of possible failure modes.
All tests were supported by DARE’s new control and data acquisition ground system based on the powerful National Instruments compactRIO and LabVIEW architecture. With these outstanding tools, generously provided by our sponsor National Instruments, the Stratos II+ crew was not only capable of logging data at unprecedented rates but could also control the operations from the same highly integrated system. This added a lot of simplicity and reliability to the operations on all ends.
Test 11, June 9th 2015
The 8th of June was reserved to reinstall the test bench into the facilities of TNO. On the 9th of June we were ready to conduct a motor test. The day started flawlessly and we were able to fill the run tanks with about 82 kg of nitrous oxide and we were ready to start the heating process, which will raise the nitrous oxide vapor pressure to a fixed 60 bars. Due to small electrical problems with a blown fuse in a cable spool, the heating of the tanks was only finished at about 2 o’ clock in the afternoon and the first test was ready to begin.
Shortly after the ignition sequence, about a second into the motor burn a loud bang was heard in the operations room coming from the test cell. A lot of smoke, but no flames were visible on the TV screen. The data connection to the measurement and actuation system had been lost due to a short circuit in the heaters, causing a fuse to blow. The nitrous oxide flow was immediately terminated. A large leak was observed in a hose, which is used for filling the run tanks. This leak slowly drained the tanks until they were empty and the test cell was deemed safe to enter.
Forensic analysis showed that the bang that was heard, was nitrous oxide and possibly fuel vapors suddenly decomposing inside the motor about 1.5 seconds after ignition. As it can be seen in the video, the motor failed to fully ignite and there was a buildup of unburned oxidizer and fuel. This mixture most likely ignited and caused the sudden explosive event. The cause for this failure was traced back to an igniter housing structural failure. The housing had been modified for using DARE’s 3D-printing capability. A 3D printed igniter housing would allow for faster and more reliable production. This test clearly showed that this 3D printed igniter housing design resulted in the unforeseen failure of the motor and the team reverted back to the previous igniter design for all follow-up tests.
Test 12, June 22nd 2015
For test 12 the motor was configured to be similar to test 10, the last test of the previous test campaign and also the previous flight configuration. After the result of last test, the igniter design was also reverted to the igniter design of test 10. The day started with another quick setup of the motor. After a quick check of all the electrical systems we were ready to test the motor at about 1 o’ clock in the afternoon.
After about 10 seconds into the burn, the motor burned through in the pre combustion chamber. The motor was quickly terminated and the remaining nitrous oxide was released out of the system. After a quick inspection of the motor it seemed that the injector plate had failed causing a burn through of the motor at the pre combustion chamber, which was a failure mode similar to test 8. This failure was a big setback for both the propulsion team and many people within the project, since this test would have been the validation of the flight motor design which was set to launch about a year ago. To make the situation even worse, the next test was planned for the 25th of June, so effectively there were only two days to find a solution to the motor failure problem.
In these two hectic days, all the data which was gathered up until the last test was analyzed. During the analyses, the team arrived at an interesting finding. The failure of the injector plate could be linked to longitudinal combustion instabilities in the motor. Thanks to the high sampling rate of the new measurement system, a clear instability was visible in the combustion pressure data at about 450 Hz (also to be heard in the video). This instability most likely caused the injector plate to fail due to excessive heating. The focus now was to reduce these instabilities in order to stop the injector plate from deforming during the burn.
After analyzing previous test results and doing an extensive literature research, it was determined that for the next test the oxidizer mass flow would be lowered from 3.5 kg/s to about 3 kg/s, by reducing the amount of injector orifices. That would lead to a lower oxidizer mass flux (oxidizer mass flow divided by the fuel grain port area) and a lower combustion pressure. Both are strong factors in causing longitudinal combustion instabilities. Moreover, the size of the pre combustion chamber ring was also changed. This ring was made thinner to increase the volume of the pre combustion chamber, which would lower the temperature in the pre combustion chamber and bring the design closer to the pre combustion chamber design of tests 2 till 5.
Test 13, June 25th 2015
After the design changes were implemented into a new motor, the next test could begin. Again the setup of the motor went smoothly. No significant delays were experienced and the motor was ready to be fired at 1 o’ clock on Thursday the 25th of June.
The ignition went smoothly, the observed flame was significantly more stable compared to previous tests. After about 5 seconds into the burn, smoke was visibly coming from the pre chamber and at about 7 seconds the pre chamber burned through. After careful inspection of the test results, it was determined that this was a different failure mode compared to last test. The pressure data showed a strongly reduced longitudinal combustion instability, and the overall roughness of the burn is significantly reduced. Also, the injector plate was still intact, which strongly indicates that the implemented changes in the motor operating conditions were indeed successful.
This being said, the motor still did not work as required and was not producing nearly enough impulse for a decent flight. Therefore the focus was very strong on the outcome of the 14th and last test. The team had just one week to arrive at a solution and to implement this into a new motor before the final deadline.
After another long but productive meeting the outcome was that the material used in the pre combustion chamber is most likely too brittle and not suitable for our needs. A complete redesign of the pre combustion chamber was done in order to find the right combination of design and material which would suit our needs. A pre combustion chamber design similar to the design used in tests 2 till 5, (where the pre combustion chamber did not show any significant problems) was implemented and some other minor modifications to the motor were made.
Test 14, July 2nd 2015
Almost a month after the first test of this test campaign, the last test would prove to be the most crucial of all. This test would determine the future of the project and it would also prove whether the conclusions drawn and changes implemented were correct or not.
As usual, the preparation for the test were without any significant problems. We managed to prepare the motor for testing by noon and the test was ready to commence at 13:15. The motor ignited successfully and the motor burned without problems for over 20 seconds. After around 22 seconds the combustion chamber ruptured at around 200 mm from the injector and the motor was turned off.
After a quick inspection it was found that the motor ruptured because the fuel was depleted. This means that the thermal protection against the hot combustion flames (being the fuel) was no longer there and the temperature of the aluminum casing quickly rose, which caused it to fail. Although the test ended with a failure of the motor, the test itself was a great success! The problems which had haunted the team for many tests seemed to have been resolved and the performance of the motor measured during this test was much better than predicted.
Another important finding in last test was that liquid nitrous oxide was supplied to the motor for much longer into the burn than anticipated. It was expected that after around 18 seconds the liquid phase would stop and only gaseous nitrous oxide would be supplied to the motor. Last test proved that liquid nitrous oxide was supplied for much longer (around 23 seconds). This means that the fuel ran out more quickly, which eventually happened during the test. That being said, liquid nitrous oxide provides much more impulse compared to gaseous nitrous oxide. This means that even though the motor burned for a shorter amount of time, the impulse generated is comparable to a longer burn with a shorter liquid phase.
Conclusion of the test campaign
The test campaign was a mixture of failures and success. All the tests were conducted flawlessly, no major incidents occurred and the test preparation went smooth for every test. The results of test 14 are very promising and the Stratos II+ crew are currently simulating the flight of the rocket using the data of test 14. The fact that the test ended in a failure of the motor is not seen as a major obstacle as the impulse generated before the motor failed would, most likely, be enough to propel the rocket halfway into space.
A short summary to close this blog update. Four motor tests have been performed, the first test failed upon ignition, the second and third test failed around 10 seconds into the burn and the fourth and last test successfully burned for over 20 seconds and ultimately failed due to a fuel shortage. In the last test, a good dataset for the simulation of the rocket has been gathered and the confidence in the motor has greatly increased. Therefore the initial goal for this test campaign has been reached. We are looking forward to the launch this October!
We would like to give a special thanks to TNO especially Wolter Wieling and Jos van den Brand, for their continued support and providing us with the test facilities. Furthermore, we would like to extend our thanks to Air Liquide, for generously providing us with all the nitrous oxide that was necessary for these tests, and to National Instruments, for providing us with the measurement and control equipment that played a key role in this test campaign.