Enabling Conceptual Design and Analysis of Cryogenic In-Space Vehicles through the Development of an Extensible Boil-Off Model

Thumbnail Image
Mendez Ramos, Eugina
Mavris, Dimitri N.
Associated Organization(s)
Supplementary to
This work is motivated by the high degree of uncertainty surrounding early estimates of the propellant losses due to boil-off and the resulting impact to the vehicle design. Typically, the heat entering the propellant tank is assumed to be directly responsible for boil-off of the liquid propellant. This is equivalent to the worst case scenario since in an actual tank only a portion of the incoming heat contributes to the boil-off process. This approach has the potential to significantly overestimate boil-off, and therefore the propellant losses, due to the simplicity and corresponding low fidelity that is used when representing the boil-off phenomenon. This uncertainty in the propellant losses is then propagated throughout the vehicle during the sizing process through the propellant mass requirements. In addition, the above approach ascribes a constant value for the rate at which the propellant losses occur. In reality, boil-off is influenced by a number of factors and changes throughout the mission. Recent studies have shown that cryogenic in-space vehicles have the potential to not only be sensitive to small changes in the boil-off rate, but also to the form of the rate used in the sizing process (a constant rate versus one that changes with time). This has important design consequences to not just to the vehicle mass, but also to the mission duration (via propellant lifetime), as well as the selection of the thermal management approach and pressure control method utilized. To increase the fidelity in the boil-off rate, the heat transfer responsible for the boil-off of the liquid propellant must be determined. This requires modeling of the physical processes that occur within the tank. Modeling and analysis of the propellant tank is generally incorporated once the design has been matured or narrowed down to a handful of designs, since these models are detailed and require longer evaluation times. This presents a gap with respect to the boil-off prediction capability that is available to designers within the space community. The objective of this research is to address this gap by developing a simplified cryogenic propellant tank model capable of simulating the physical processes in the tank, so as to improve the fidelity in boil-off. This task presents a unique set of challenges, since the vast majority of cryogenic propellant tank models available in literature focus primarily on predicting the fluid conditions in the tank, rather than boil-off. Therefore, validation becomes an issue. However, the pressure change in a closed cryogenic tank is directly related to the evaporation process that occurs at the interface, otherwise known as boil-off. This suggests that a model capable of predicting the pressure change with certain fidelity will predict boil-off with a similar degree of fidelity. For the model developed here, it is assumed that during pressurization there is no boiling of liquid propellant or condensation of the propellant vapor; pressure control is achieved with direct venting, which affects the ullage region alone. The ability of the model to predict the conditions in the tank during pressurization is validated using several LH2 self-pressurization experiments from literature. Once the fidelity in the pressurization rate, and thus boil-off, is established, the approach used to represent the venting process is then examined. Once the model evaluation is complete, the propellant losses due to boil-off and subsequent venting are compared with those obtained using the standard approach. The results of the comparison support other observations in literature – that the traditional method utilized during the conceptual design process has the tendency to overestimate boil-off. To demonstrate the benefits of the higher-fidelity boil-off model, the model is implemented in the sizing process of a relevant system – the descent stage of the Human Landing system. Approximately 14,500 candidate designs were evaluated. The higher-fidelity model predicted boil-off losses that were as much as half the losses predicted by the theoretical model. This resulted in lower propellant mass requirements which yielded smaller vehicles. In addition, the lower boil-off predicted by the model significantly increased the loiter time of the vehicle when compared to vehicles sized with the theoretical model. These results demonstrate the severity of the impact to the vehicle when assuming the worst case scenario with respect to boil-off.
Date Issued
Resource Type
Resource Subtype
Rights Statement
Rights URI