Endothermic hydrocarbon fuel has many advantages as a cooling medium in hypersonic aircraft; however substantial application has been greatly hindered due to easy coking, uncontrollable gas products, and low heat sink issues. Catalytic cracking of hydrocarbon fuel has been regarded as one of the most effective ways to improve the heat sink via modulating the cracking pathway. The incipient-wetness impregnation method was selected to load Co salts in commercial ZSM-5 in a large-scale manner; subsequent calcination in Ar gas resulted in Co 3 O 4 nanosheet wrapped HZSM-5 composites. Salinization treatment was adopted to facilitate the dispersion of catalyst in a nonpolar solvent. The ratio of Bronsted/Lewis acidity decreased from 8.00 to 2.79 after modification by Co 3 O 4 nanosheets. The synergistically catalytic effect between Co 3 O 4 nanosheets and ZSM-5 was beneficial to the generation of a larger gas production rate with higher content of alkene, and thus resulting in a higher heat sink than benchmarked fuels. Catalytic cracking of n-decane (C10) in the presence of 0.1 wt % Co 3 O 4 nanosheets@ZSM-5 could yield a heat sink as high as 4.64 MJ/kg at 758 °C, much higher than those of bare ZSM-5 (2.99 MJ/kg at 687 °C) and thermal cracking of C10 (3.77 MJ/kg at 728 °C). Meanwhile, the smart combination of Co 3 O 4 nanosheets and commercial ZSM-5 could effectively suppress the coke deposition on the external surface of composites, thus resulting in efficient catalytic cracking at elevated temperatures for obtaining a higher heat sink.
Adding an initiator is an effective method
of promoting hydrocarbon pyrolysis and improving the heat sink of
fuels. Nitropropane was proposed as an initiator with good performance,
owing to its lower reaction activation energy for C–N bond
cleavage. To study the effects of this initiator on hydrocarbon pyrolysis,
a miniature tube reactor that can simulate a real heating procedure
in an aeroengine was used to investigate the n-decane
pyrolysis with and without nitropropane under experimental supercritical
conditions. The results demonstrate that the nitropropane initiator
promotes the pyrolysis of fuel as it flows through a tube with a large
length–diameter ratio within a certain temperature range. The
initial decomposition temperature of n-decane is
reduced by approximately 100 K, and the increase in the conversion
leads to a higher heat sink for n-decane, which can
result in decreases in the fuel and reactor temperatures under the
same heating condition and within the effective temperature range.
A stronger promoting effect can be achieved by increasing the concentration
of the nitropropane initiator. The variation laws for the n-decane pyrolysis reaction rate along the flow reactor
are changed by the initiator, the presence of nitropropane greatly
accelerates the pyrolysis reaction of fuel at a lower temperature,
and the opposite tendency appears as the fuel temperature increases,
which is caused by the consumption of the initiator. In addition,
the selectivity of methane, propane, and alkenes, especially ethylene,
increases because of the propyl radical generated by the C–N
dissociation of nitropropane before the initiator is consumed.
In-context learning of GPT-like models has been recognized as fragile across different hand-crafted templates, and demonstration permutations. In this work, we propose prototypical calibration to adaptively learn a more robust decision boundary for zero-and few-shot classification, instead of greedy decoding. Concretely, our method first adopts Gaussian mixture distribution to estimate the prototypical clusters for all categories. Then we assign each cluster to the corresponding label by solving a weighted bipartite matching problem. Given an example, its prediction is calibrated by the likelihood of prototypical clusters. Experimental results show that prototypical calibration yields a 15% absolute improvement on a diverse set of tasks. Extensive analysis across different scales also indicates that our method calibrates the decision boundary as expected, greatly improving the robustness of GPT to templates, permutations, and class imbalance.
In this study, experimental methods are used to investigate the effects of
different vibration and pressure parameters on heat transfer performance are
analyzed. The results show that at a subcritical pressure(0.1 Mpa), the
external vibration begins to affect the heat transfer when the fuel passes
the phase-change point and becomes gaseous.; at a near-critical pressure(3
Mpa), the external vibration deteriorates the heat transfer of fuel across
the critical-temperature zone; at the supercritical pressure(5Mpa), the
external vibration intensifies the heat transfer in the hot end of the
channel only when the fuel is below the critical temperature and the
internal wall is above the critical pressure. Combined with data analysis, it
can be seen that the external vibration mainly acts on the temperature
boundary layer of the fuel oil to affect the wall temperature and heat
transfer coefficient.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.