Cryogenic Bearings Research Articles

A meta-weighted network equipped with uncertainty estimations for remaining useful life prediction of turbopump bearings

Uncertainty quantification and negative transfer characterization are significant challenges in the context of deep learning-based remaining useful life prediction with limited data. Conventional Bayesian methods lack scalability to deep learning algorithms for uncertainty quantification due to intricate network connections and extensive training parameters. Additionally, although transfer learning is a promising way to improve the generalization ability of prediction algorithms, its effectiveness is not always guaranteed when leveraging source data undesirably reduces the prediction performance, which is named negative transfer. This manuscript proposed a meta-weighted neural network equipped with uncertainty estimation to discern in-distribution from out-of-distribution data and an adaptive sample meta re-weighting strategy is designed by specifying the weighting function from prediction loss to sample weight according to the gradient direction. Performance evaluations on cryogenic bearings demonstrate that the proposed algorithm can quantitatively determine the weights of source data based on prediction error, ultimately leading to accurate interval remaining life prediction.

Numerical Investigation of Turbopump Bearing Heat Generation in Liquid Methane

Abstract Liquid methane is one of the most promising candidates for the next-generation rocket propellant because it has an excellent balance between several propulsion performances such as specific impulse, storability, and structural coefficient. JAXA has developed rocket engine technologies for liquid methane engines. In this paper, we investigate the heat generation characteristics of ball bearings used in liquid methane turbopumps. A usual approach to evaluating a cryogenic ball bearing performance is to conduct experimental testing which costs a lot. This research introduces numerical approaches based on physical mechanisms in rotating ball bearings to clarify the heat generation characteristics. The main factors of the heat generation are friction heat generation and fluid heat generation. The friction heat generation is caused by mechanical friction of bearing elements and calculated based on bearing motion analysis and Hertzian contact theory. The fluid heat generation is caused by the fluid drag force on bearing elements and calculated by computational fluid dynamics analysis. The theoretical model is compared with experimental results, showing an excellent agreement. It is clarified that the dominant factor of the bearing heat generation in liquid methane environment is the friction heat generation on races-balls contact at lower velocity condition while it tends to change to the fluid heat generation due to the bearing elements’ rotational motion at higher velocity condition. In addition, cryogenic bearing characteristics which are clarified by theoretical modeling are discussed.

Development of a Cryogenic Tester with Air Bearing to Test Sliding-Rolling Contact Friction

This study aimed to test the friction coefficient of cryogenic bearing lubrication materials. A ball-on-disc type friction tester was developed in our lab using air bearings that could simulate the movement of cryogenic bearings under sliding-rolling contact. The tester is equipped with a temperature-controlled chamber to provide a minimum −175 °C low-temperature environment. Using air bearings is an important technique to reduce the base friction of the tester measurement system and ensure the accuracy of the friction coefficient measurement. The friction coefficients of the Ag coating and the PTFE coating were measured at different sliding-rolling velocities on this tester, and the results showed that the friction coefficient curves agreed well with the Gupta sliding model. The developed tester will provide important data for the dynamic analysis and life evaluation of the cryogenic bearings.

A multi-head attention network with adaptive meta-transfer learning for RUL prediction of rocket engines

Accurate prediction of remaining useful life (RUL) is necessary to ensure stable and safe operations for rocket engines. The paper proposed a multi-head attention network coupled with adaptive meta-transfer learning for RUL prediction. By combining the convolution-based branch with an attention-based branch, the multi-head attention network is proposed for accurate RUL prediction of cryogenic bearings in rocket engines under the steady stage. In addition, an adaptive model-agnostic meta-transfer learning strategy is developed to further improve the performance under small sample circumstances with adaptive hyper-parameters. To demonstrate the superiority, the proposed method is compared with typical benchmark algorithms using real monitoring data from a high-precision cryogenic rocket engine experiment platform. Results indicate that the proposed method achieves better performance compared with existing models under multiple evaluation indexes.

Performance study of preloaded cryogenic bearings in liquid hydrogen pump

As a clean and efficient energy, liquid hydrogen has become an important solution for the energy restructuring in the future. The centrifugal liquid hydrogen pump has been applied in the storage, transportation and distribution of liquid hydrogen. As the support of pump rotor, the performance of cryogenic bearings are essential to the dynamic characteristics and design optimization of liquid hydrogen pump. In this paper, an improved quasi-static model for cryogenic bearing in liquid hydrogen pump was introduced, in which the effects of high-speed centrifugal expansion and shrinkage under cryogenic condition on bearing rings were considered. Based on this model, the effects of rotating speed and axial preload on bearing dynamic characteristics were studied. At the same time, the friction within the cryogenic bearing was analysed and calculated. Finally, the load variation of bearing preload disc spring at cryogenic temperature was calculated and experimentally studied. The results of this paper will provide theoretical suggestions for further engineering design and applications of cryogenic bearings in liquid hydrogen pump.

Optimization of Cryogenic Process for Enhancing the Wear Resistance of Novel HNMS Steel for Cryogenic Bearings of Launch Vehicle

Bearings are highly engineered components that have to tolerate vibratory stresses, bending moments, and high rotational speeds in cryogenic applications. AISI 440C is most widely used for antifriction bearings due to its good strength and high hardness but is prone to carbide banding and failure during service. High-nitrogen martensitic stainless steel (HNMS) was recently developed by partially substituting carbon with nitrogen. This article explores the possible improvement in wear properties by cryotreating HNMS steel. The hardening temperature was varied from 1050 to 1100 °C, cryosoaking period from 0 to 12 h at −196 °C, and tempering temperature from 100 to 300 °C for 1 h. Scanning electron microscopy (SEM) and X-ray diffraction (XRD) analysis revealed the presence of carbides and carbonitrides. The phase fraction and size of the precipitates was found to be influenced by the cryosoaking period. The carbide density variation with the cryosoaking period followed a sigmoidal curve. Eight hours of cryogenic treatment at −196 °C was observed to be the optimal cryosoaking period for HNMS steel. Maximum hardness (671 HV) and minimum wear rate (0.00714 mm3/m) and adhesive wear were observed for the optimum heat treatment condition. Compared to conventional hardening-tempering (HT) treatment, the optimized hardening-cryogenic-tempering (HCT) treatment cycle resulted in a 38% increase in hardness and a 31% decrease in the wear rate. The improved properties can be attributed to the presence of fine and uniform distribution of precipitates in the matrix and transformation of retained austenite to martensite.

극저온 환경용 볼베어링 시험장치 개발 및 터보펌프용 볼베어링 시제품의 성능평가

The turbo pump of a liquid rocket engine is composed of three main parts: the oxidizer pump, fuel pump, and turbine. Liquid oxygen () is the working fluid in the cryogenic environment in the oxidizer pump, but tests are usually performed using liquid nitrogen (), which has a boiling point similar to that of but is comparatively safer and easier to use for the test. In this study, a bearing test rig is developed and its performance is evaluated using a cryogenic ball bearing with as the working fluid. Verifying the performance of the bearing test rig is crucial for ensuring correct working of the turbo pump unit in the liquid rocket engine. A stable test rig for the bearing in a cryogenic environment makes the bearing technology enhance its reliability. The test results show that the system operates stably and the requirement of performance time of 500 s is met. The test results of temperature, motor speed, and torque are discussed. The developed cryogenic bearing test rig is expected to help in widening knowledge and expanding research on ball bearings in the future.

A cryogenic waveplate rotator for polarimetry at mm and submm wavelengths

Mm and sub-mm waves polarimetry is the new frontier of research in Cosmic Microwave Background and Interstellar Dust studies. Polarimeters working in the IR to MM range need to be operated at cryogenic temperatures, to limit the systematic effects related to the emission of the polarization analyzer. In this paper we study the effect of the temperature of the different components of a waveplate polarimeter, and describe a system able to rotate, in a completely automated way, a birefringent crystal at 4K. We simulate the main systematic effects related to the temperature and non-ideality of the optical components in a Stokes polarimeter. To limit these effects, a cryogenic implementation of the polarimeter is mandatory. In our system, the rotation produced by a step motor, running at room temperature, is transmitted down to cryogenic temperatures by means of a long shaft and gears running on custom cryogenic bearings. Our system is able to rotate, in a completely automated way, a birefringent crystal at 4K, dissipating only a few mW in the cold environment. A readout system based on optical fibers allows to control the rotation of the crystal to better than 0.1{\deg}. This device fulfills the stringent requirements for operation in cryogenic space experiments, like the forthcoming PILOT, BOOMERanG and LSPE.

A Fully Coupled Variable Properties Thermohydraulic Model for a Cryogenic Hydrostatic Journal Bearing

The goal set forth here is to continue the work started by Braun et al. [11] and present an integrated analysis of the behavior of the two row, 20 staggered pockets, hydrostatic cryogenic bearing used by the turbopumps of the space shuttle main engine (SSME). The variable properties Reynolds equation is fully coupled with the 2-D fluid film energy equation [η = η(P, T)]. The 3-D equations of the shaft and bushing model the boundary conditions of the fluid film energy equation. The effects of shaft eccentricity, angular velocity and inertia pressure drops at pocket edge are incorporated in the model. Their effects on the bearing fluid properties, load carrying capacity, mass flow, pressure, velocity and temperature form the ultimate object of this paper.