Table of Contents
At first, our team wished to be able to measure the temperature of the internal combustion gasses to measure the oxidizer to fuel ratio and confirm numerous heat transfer analysis. After discussions with Prof. Wellin and Prof. Kempski, it seems as though measuring the extremely hot combustion gasses directly is not very feasible. We could try some other methods such as insulating thermocouples and back-solving temperature data, but it would not be very accurate and may not result in useful data.
Additionally, after speaking with Armour (of SpaceX), it seems as though nobody attempts to measure these temperatures.
After these discussions, we chose to simplify our approach and use the thermocouples as a way to monitor temperatures in the feed system and on the external face of the combustion chamber to monitor how well our insulation was working.
As mentioned above, we simplified our approach to thermocouple selection and decided to order a commonly used product, shown in the link below:
Our team chose to order K-type thermocouples as they are very commonly used and provided a wide temperature sensing range that is expected to be suitable for our needs. Originally, the bare wire leads were chosen for ease of integration into the control system circuitry. After considering the PCB design more, it became clear that miniature TC connectors would be needed and ordered separately. Another choice we had was insulation type. Naturally we chose the glass braided as it had the greatest operating temperature and was not anymore costly that the PFA or Kapton insulation. The last option we had to choose was the wire gauge. Larger wire sizes will result in more robust connections and are less likely to fail, however, smaller wire sizes have a much quicker response time. Originally, we decided that this latter reason was more important to us we chose to purchase 36 AWG. Again, after considering the PCB design, it made more sense to go with the larger 20 AWG.
For the feed system TCs, we found a unique solution that would allow for very easy integration, as seen in the link below:
These TC probes have NPT fittings on them to directly fasten into our feed system lines. The exposed junction type was chosen to ensure the fastest response time.
A great resource for basic thermocouple information can be found here: Using Thermocouples in Temperature Measurement
Resource used for determining approximate response times for a specified wire gauge (see Table 2): Fine Gauge Wire Thermocuples
Resource depicting the difference in junction types for NPT thermocouple probes: Thermocouple Junction Types
In our current feed system design, the controller commands the solenoid valve on the pressurant tank to open or close based on the measured pressures throughout the feed system. In order for this to work reliably, three pressure transducers are needed as shown in the Feed System Schematic. The pressure transducer for the pressurant tank will need to operate at a higher pressure than those for the nitrous oxide tank and injector.
Our team began by making a table of pressure transducers that would work for our system. Both MEs and EEs on team gave input to see what parameters are preferred in selecting this component since we were surprised to find that these components were sold with a wide variety of options. After doing so, many components were dismissed based on preferences and it became clear that the Honeywell MLH Series sensors were best for our application and in our price range. We selected two 0-2000 psi range pressure transducers for the nitrous oxide feed system and one 0-5000 psi range pressure transducer for the pressurant feed system.
The document below outlines which components were dismissed or omitted and for what reason. It includes many notes on which options and parameters were preferred. These preferences drove the selection of the sensors for our application.
When attempting to order our preferred pressure transducers, we were notified that Digikey no longer carried the components we desired. This forced us to go back to our spreadsheet of our preferences and select an in-stock component that still met our primary preferences. This has since caused us to purchase pipe fitting adapters and specialty connectors to integrate them into our system.
The load cell is a very important tool that our team will use to directly measure the thrust of our engine. At first, we only identified the need for one primary load cell, but after beginning small scale testing, we realized that a second would be necessary.
The primary load cell, as previously mentioned, will be used to measure the thrust of the engine. The secondary load cell will be much smaller and will be used to measure the weight of the oxidizer tank. This will help us stay safe in loading and offloading operations as well as provide a method to measure mass flow rate of the oxidizer.
The primary load cell was sourced from PCB Piezotronics as we were instructed that a possible discount or donation may be feasible. After scanning their product options, we settled on the model linked below:
For the secondary load cell, we did not need high performance specifications so we looked to see what DigiKey had to offer. After narrowing the choices based on load range, output, availability, and cost, we settled on the following button type load cell:
Early on in the design phase, we identified the want (not need) for accelerometers mounted on the combustion chamber in an effort to record vibration data. This would potentially be useful in analyzing combustion instabilities and verifying structural vibes data.
Based on this want (not need) for recording accelerometer data, we chose to not select expensive industrial grade accelerometers used in NVH (noise, vibration, harshness) applications. A cheaper solution was sourced from SparkFun that will have easy integration into our DAQ.
Similar to the background of the accelerometers, we identified the want (not need) for microphones to analyze the stability of the burn.
Again, based on this want (not need) for recording noise data, we chose to not select expensive industrial grade microphones used in NVH (noise, vibration, harshness) applications. A cheaper solution was sourced from SparkFun that will have easy integration into our DAQ.
The sensor layouts shown below depict the approximate location of each sensor we plan on using and in what configuration the sensors will be used. It is important to note that the Flight sensors will be designed to be flown with the rocket and integrated into the on-board control system. The Test Stand sensors will be used on the test stand and will be supplementary to the sensors shown in the flight configuration. A combination of all of these sensors will also be used in small scale verification testing as well as subsystem verification testing.
The cost assessment shown below was used in one of our earlier design reviews to estimate and budget the cost of the system and each of the sensors. The items listed have changed since that design review and a finalized table will be uploaded as components are finalized and ordered.