Development of a System to Quantify Coking in Rocket Nozzle Cooling Channels

Sammanfattning

Liquid methane is becoming an increasingly attractive rocket propellant due to its high performance characteristics and potential to support in-situ resource utilisation. Methane, however, when heated, can thermally decompose in a process known as pyrolysis. In regeneratively cooled rocket engines, the solid carbon products from the pyrolysis reactions are deposited on the walls of the cooling channels. This increases the thermal resistance of the channel walls, resulting in higher wall temperatures. In turn, this can facilitate cracking and crack propagation, presenting a potential problem in rockets, especially for future reusable designs. It will therefore be necessary to inspect the state of the cooling channels between flights. The carbon layer also changes the catalytic properties of the surface, affecting the onset temperature of methane pyrolysis, and thus impacting the pyrolysis behaviour during subsequent flights. It is possible to clean the channel using a mixture of gaseous oxygen and ozone, however, preliminary testing has indicated that not all the carbon is removed within a reasonable time frame. An experimental facility exists which can control the thermal and flow conditions in straight test channels to replicate the conditions seen in methane rocket nozzle cooling channels. The purpose of this project is to develop a system to quantitatively assess the amount of carbon deposition in these test channels after methane pyrolysis has occurred within them, and following ozone cleaning. The developed system is an optical method which uses a borescope to capture images within the coked channel. These images are then run through bespoke image processing software to determine the proportion of the inner channel wall that is coked. The software has been developed and a provisional mechanical setup has been designed. Initial validation tests have been conducted to assess the accuracy of the software used in conjunction with the borescope and camera. The results indicate that the system is capable of quantifying coke in a metal channel with an error of 1.489%±0.232% or less.