There were two main stages involved in building the engine and its systems. The first one was prototype testing and it occurred in May 2019. During this stages, two instances of injector B and two instances of injector C were manufactured (in aluminum) and assembled for cold flow testing at Zucrow Propulsion Labs (Purdue University). In these tests, the injectors were simply connected to a 50 psig water supply and manually operated. After the four injectors were tested, it became evident that injector B (pintle) was not atomizing properly and was in fact not suitable for small diameter, high mixing ratio engines like the P-5. This was expected but it was still worth it to physically demonstrate it since pintle injectors are relatively easy to manufacture and could perhaps be implemented on future different engines.

Figure 9. Cold flow of injector C prototype

The second stage took place from May to the beginning of August and its objective was to recreate what was achieved at Zucrow but using AREX’s own systems in Costa Rica. Starting with the structural system, a test bench was manufactured locally over December 2018, but it required some additional assembly since only the metallic frame had been built. The bench had three shelves: one in the bottom for the carbon dioxide tank and storage of some tools, one in the middle for the electronics (control room), and one in the top for the propellants. These three sections were separated from the platform that held the engine by a ¼” thick metallic sheet.

Figure 10. Metallic frame of the test bench

The electrical system was perhaps one of the most difficult ones to handle, since it had to be built simultaneously with the feed system so that sensors and valves could be tested. Moreover, many sensors had to be calibrated even between tests, and so the software used to control the system (the RETC) had a tab that allowed adjusting the calibration at any time. To calibrate the pressure transducers, regular pressure gauges were used as a reference. To calibrate flow meters, the volumetric flow rate was measured for several tests with a scale and a stopwatch (in the RETC). To calibrate load cells, a pulley system with a spring scale was used. The RETC ran on a Raspberry Pi computer and used a microcontroller (Arduino UNO) to get data from the sensors. The Raspberry Pi allowed to connect remotely through Wi-Fi or an ethernet cable, and so most of testing was done from a laptop. After all the components were tested individually, they were assembled in the test bench and circuit diagram was used as a guide to wire them. To protect the electrical system, the main computer and breakout boards were stored inside a plastic container.

Figure 11. Protective container for the electrical system

The feed system was assembled and disassembled several times since the pressure and instrumentation diagram was updated as the system was being built and tested. High pressure chemical-rated tubbing was used for the main connections while stainless steel fittings (compatible with the hydrogen peroxide) were used in between components. The fuel line had a hydraulic diameter of ¼” while the oxidizer line began with a ¼” diameter that was later replaced by a 3/8” shorter line. The propellant tanks could be refilled (with either water or propellants) by using a pump . For general safety concerns, a relief valve set to 200 psig was added immediately after the CO2 tank, and filters and check valves were put before the injector.

Figure 12. Final version of the feed system

The injection system was characterized by cold flowing the injector multiple times and analyzing the data measured by the electrical system. This allowed designing an upgraded version of the injector and to further modify the feed system for better performance. As mentioned already, injector C was chosen for the final version of the engine. It had nine oxidizer like-doublet elements angled towards the vertical axis, and three fuel like-doublet elements pointing away from it. Primary atomization happened when two jets from one element impinged into each other, and secondary atomization occurred when the resulting fans from adjacent elements intersected. The resulting spraying produced a nearly axial column, which is favorable since it helps avoiding hotspots in the chamber’s walls. If the propellant cloud is ignited before it reaches the converging section of the chamber, combustion should happen. The injector was assembled to the propulsive system and sealed with high temperature O-rings. This assembly was vertically mounted on a platform in which four partially threaded bolts allowed the engine to rise 4 mm such that it pressed against the load cells in the platform when subjected to thrust. In the final version of the combustion chamber, the piece grew in volume, ports for pressure transducers were added, and just like the injector the material was changed to stainless steel 316.

Figure 13. Spraying pattern of final injector

Figure 14. Engine prototype mounted on platform

Figure 15. Final propulsive and injection system

After studying the theoretical framework behind LPRE’s it became evident that the P-5 would require a systematization different from that of more complex high-power rocket engines. The following systems were chosen and characterized separately through MEP’s and testing.

Figure 2. Pie chart showing the six engineering systems

The propulsive system refers to the combustion chamber and nozzle design. This includes choosing the propellants, the chamber pressure, and the desired thrust level. To make these decisions, trade studies were made in July of 2018 with a sample objective in mind: The P-5 engine must be simple enough to be economically and logistically feasible, but complex enough to explore the underlying technicalities involved in low power liquid rocket engines. Therefore, very advanced features such as: cryogenics, hypergolics, turbopumps, thrust vectoring, throttling, high chamber pressures/ thrust levels, etc.; were avoided. Moreover, most of the aerospace rated components are significantly expensive, and therefore it was decided to use commercially available valves and regulators. This set a limit to the maximum operating pressure of the engine, and thus an idea of the mass flow rates that could be achieved (which is proportional to thrust). Perhaps the most difficult decision was the oxidizer choice, since the only non-cryogenic ones (i.e. nitric acid [and derivatives], nitrous oxide, and hydrogen peroxide) are difficult to handle and obtain. Finally, due to its liquid state at STP and commercial availability in Costa Rica, hydrogen peroxide (50% w/w) was chosen. It was clear that the low concentration of the peroxide would lower the engine’s power, but most concerningly, there was a chance the water content would be high enough to not even support combustion. In August 2018, a sample of this oxidizer was bought and mixed with different fuels and was proven to support combustion (further confirmed one year later). The geometry and thermochemistry of the engine was handled with an inhouse software called the Rocket Engine Designer (RED) which links with NASA’s CEARUN to solve for an ideal rocket.

Figure 3. General specs of the P-5 engine original design

Figure 4. Evolution of the propulsive system

The injection system consists of the design and prototyping of the injector which is often considered the heart of the engine. G. S. Gill, and W. H. Nurick [1] present an injector design algorithm for four different injector element types, of which three were selected, designed, and evaluated with a MEP. Afterwards, instances of both injector B and C were manufactured and cold flowed if both Purdue’s Zucrow Propulsion Labs, and in AREX’s Albireo Low Power Propulsion Lab in Costa Rica (see “Testing the Engine”). It was then determined that pintle injectors (B) are not very scalable for engines with a small chamber diameter. Therefore, injector C which was chosen and further improved based on cold flow data.

Figure 5. CADs of injectors considered for the P-5 engine

Table 1. Specs of the injectors

The feed system consists of the tanks, valves, regulators, propellant lines, and all other components required to deliver the propellants to the engine. The system is pressurized with carbon dioxide and in-line regulators allow achieving relatively constant mass flow rates. These regulators and some valves must be manually adjusted, but afterwards, the engine can be controlled electronically at distance. Two stainless steel 2.1-liter tanks were manufactured locally, and the rest of the components were bought from commercial suppliers according to the system’s requirement. The maximum operating (safety) pressure was set to 200 psig.

Figure 6. Final version of the Piping and Instrumentation Diagram of the feed system.

The electrical system includes the wiring and programming of all electronic components. A software called the Rocket Engine Testing Center (RETC) was programmed to control the feed system both manually and automatically, as well as to log data. This software ran on a Raspberry Pi connected to a computer 30 meters away through an ethernet cable. An Arduino UNO was used to record data from sensors such as pressure transducers, thermocouples, and flow meters. The whole system was powered by individual DC power supplies.

Figure 7. A screenshot of the RETC software in manual opperation mode

The structural system refers to the physical structure that holds the engine static, and the space used to arrange all the systems together. A test stand/ bench was designed such that engine could fire vertically, and the components could be stored safely.

Figure 8. CAD design of the test bench.

The cooling system refers to the method used to remove heat from the engine. It was decided to keep the engine uncooled because of the following reasons. First, the theorized temperature of the combustion gases was relatively low compared to other rocket engines (below 2000 K) due to the low concentration of the oxidizer and low chamber pressure. Second, a heat transfer simulation for a three second fire of a much powerful conceptual stainless-steel engine with a 4 mm chamber thickness was made before and proved the engine to withstand much higher temperatures. Third, thanks to the IMEP, the final version of the engine ended up having an immense safety factor due to the chamber thickness going up from 4 mm to 35 mm for more than 62% of the chamber length.