The Stratos II propulsion test crew on site.
Between June and August 2014 five static tests (#6 – #10) took place with the DHX-200 Aurora hybrid propulsion system after the motor had been on the test stand five times at the facilities of TNO, Rijswijk in 2013. This was the final set of full scale ground tests to be conducted with this system before the launch of the Stratos II in October. The tests were conducted at the DLR rocket test facilities in Trauen, Germany.
In June this year the propulsion team of Stratos II traveled to Germany to test the rocket motor and to characterize its performance. Three tests were conducted during the first portion of the test campaign. After problems with low combustion efficiency, grain cracking and subsequent exposure of the thin walled aluminium combustion chamber to hot combustion gases resulting in structural failure during the previous series of tests in 2013, the motor had been partially redesigned to ensure structural integrity and higher efficiency.
Test 6, 27. May 2014
Being the first test outside of TNO and the Netherlands altogether, we were accommodated by DLR Trauen. After a few minor operational problems with a faulty solenoid valve the run tanks could only be loaded with enough nitrous oxide for a 12 second (instead of 20s) liquid phase burn. The motor was fired nonetheless and performed admirably. No cracks formed in the grain and the entire outer surface of the combustion chamber stayed within the predicted temperatures and did not lose its load bearing abilities. The detaching flow at the end of the burn only occurs because the nozzle expansion ratio is designed for an altitude of 3000m. During flight the pressure ratio over the nozzle will be much higher and the flow will stay attached.
Test 7, 28. May 2014
After the success with Test 6 the only problem appeared to be to ensure reliable functioning of all valves. All controlled test conditions were virtually identical to the previous test. Although the issues encountered during the tests at TNO last year appeared to be solved, another problem occurred: overheating and failure of the pre-combustion chamber and the injector plate. At the time, it was unclear if the problem was initiated by a failure in the injector, the pre-combustion chamber or that the two problems occurred independend of each other. Furthermore there were worries about the tolerances of the solid fuel grain in the combustion chamber.
Test 8, 3. June 2014
In spite of the structural failure of combustion chamber and injector plate in Test 7 the hardware for Test 8 was put together with hardly any changes as an assembly error, loose fitting parts or fatigue of reuse parts such as injector plate and manifold were identified as the most probable causes (also in light of the very successful Test 6 which was practically identical). The test result was however very similar to the previous one: Structural failure in the pre-combustion chamber due to thermal weakening.
Testing strategy for the remaining tests
Since at least three potential issues had been identified with only two more tests available, a number of counter measures where developed and the two remaing tests were moved to August. First the injector was redesigned. The actual injection pattern and injection orifice L/D ratio where kept constant but the mechanical design was changed to relieve the stress on the injector plate. Second, the pre-combustion chamber ablative ring was outfitted with two o-rings to prevent any circulation around the ring, and hence cancel additional heat transfer to the wall and the injector manifold. Further, the outer surface of the liners where coated in a layer of epoxy to provide both extra strength and a better fit. Finally, an o-ring was placed on top of the pre-chamber ring to provide a tight “crush” fit for the entire stack inside the combustion chamber. High temperature silicone sealing was applied between nozzle and mixer, mixer and grain and between the pre-chamber ring and the injector. The interface between grain and pre-chamber ring was left unsealed to provide pressure equalisation during motor operations, which was identified as a crucial factor during earlier tests.
Test 9, 14. August 2014
The two tests were planned for the 14th and 15th of August. The test on Thursday took place around 12:30 after all auxiliary systems used for preparing and filling the engine worked very well. Preparation for the burn went smooth and without problems, showing that the setup and procedures are now fully mastered by the team. The test itself however did not work out as planned. After about 8 seconds the chamber ruptured near the injector. The failure mode was practically identical to the failures encountered in the previous tests, despite all the changes in the design. It seems that hot gases between the spacer ring and the fuel grain caused quicker erosion of the grain near the top. (The spacer ring is the ring that sits between injector and grain and that defines the size of the pre-combustion chamber.) By design, this connection was not sealed to allow for pressure equalization of air pockets between chamber wall and grain liner. This however also led to quicker erosion of the grain, which allowed hot gases to get to the chamber wall which caused a rupture off the chamber after approximately 8 seconds of burn time. A very positive result of this test was the excellently preserved grain, spacer ring and injector due to the quick shut-off and inert gas purge. It clearly identifies an o-ring failure in the injector manifold as a cause for asymmetrical injection and perfectly axial regression of the grain near the pre-combustion chamber as the cause for combustion chamber wall exposure.
It now seems that the successful Test 6 was a lucky coincidence of the pre-combustion chamber connections sealing themselves up with liquefying fuel.
Test 10, 15. August 2014
To fix the finally identified issues the connection was completely filled up with high-temperature black silicone gel. Also the injector was covered as much as possible with this sealant. Again all subsystems worked flawlessly and a test was performed just before lunch. This time the pre-combustion chamber worked as intended and completely comparable to Test 6. There was a reduction of injector temperature from approximately 300 deg C on all previous tests that led to failure at this location to only 150 deg C. Pressure equalisation over the grain was achieved through holes that were drilled radially in the spacer ring.
After 10 seconds however the chamber ruptured just before the nozzle. The most likely cause for this lies in an assembly error. The nozzle, mixer and fuel grain had been assembled a previous time already, but where taken apart to implement design changes. Post-failure analysis of the chamber showed no traces of sealant material between the mixer and nozzle. Therefore it is expected that the sealant was not applied properly during re-assembly of the motor. The fact that during the previous 8 burns of the motor, the temperature measured at the current failure location never reached critical temperatures suggests that it is not related to the design but rather to the assembly of the motor.
Consequences for the project
On Monday, 18. August 2014, the complete Stratos II team discussed the results of the tests and the consequences they will have for the project. The propulsion team is confident that the pre-combustion problem has been solved and that the nozzle failure can be prevented by proper assembly of the motor. Hence the team is confident that the motor can perform nominal operations in excess of 25 seconds.
The motor has demonstrated successful and stable operations up to 8 seconds, hence tower clearance is not considered to be jeopardised and if a failure occurs, it will happen over open water. Careful analysis will be performed in the coming weeks to investigate the effect that different side forces will have on the trajectory of the rocket. In case that the actual trajectory of the rocket will differ too much from the planned trajectory, a signal will be broadcast by the launch site to terminate motor operations. This system, in combination with the large clearance area provided by the launch site will ensure safe operations. Hence it is decided by the Stratos crew to proceed with the project and attempt to launch Stratos halfway into space in the first week of October.