The hot deformation behavior of Al-Zn-Mg-Er-Zr alloy was investigated through an isothermal compression experiment at a strain rate ranging from 0.01 to 10 s-1 and temperature ranging from 350 to 500 °C. The constitutive equation of thermal deformation characteristics based on strain was established, and the microstructure (including grain, substructure and dynamic precipitation) under different deformation conditions was analyzed. It is shown that the steady-state flow stress can be described using the hyperbolic sinusoidal constitutive equation with a deformation activation energy of 160.03 kJ/mol. Two kinds of second phases exist in the deformed alloy; one is the η phase, whose size and quantity changes according to the deformation parameters, and the other is spherical Al3(Er, Zr) particles with good thermal stability. Both kinds of particles pin the dislocation. However, with a decrease in strain rate or increase in temperature, η phases coarsen and their density decreases, and their dislocation locking ability is weakened. However, the size of Al3(Er, Zr) particles does not change with the variation in deformation conditions. So, at higher deformation temperatures, Al3(Er, Zr) particles still pin dislocations and thus refine the subgrain and enhance the strength. Compared with the η phase, Al3(Er, Zr) particles are superior for dislocation locking during hot deformation. A strain rate ranging from 0.1 to 1 s-1 and a deformation temperature ranging from 450 to 500 °C form the safest hot working domain in the processing map.

Flow stress

Deformation mechanism

Isothermal process

Dynamic Recrystallization

10.3390/ma16124404

Cite

Citations (4)

Preparation and conductive properties of double perovskite oxides Ba3Ba1+Ta2-O9-

Ceramics International (2022)

Zezhong Wang Ying Li Wenlong Huang Yushi Ding

10.1016/j.ceramint.2022.01.106

Cite

Citations (4)

Proton conductivity and transport number of complex perovskite barium strontium tantalate

Ceramics International (2020)

Wenlong Huang Yushi Ding Ying Li Zezhong Wang

Tantalate

Barium

10.1016/j.ceramint.2020.09.095

Cite

Citations (5)

Proton Conductivity and Transport Number of Cahf0.9in0.1o2.95

Zezhong Wang Ying Li Wenlong Huang Yushi Ding

Herein, a perovskite-structured proton conductor, i.e., CaHf0.9In0.1O2.95 is fabricated via solid-state reaction, and the proton conductivity and transport number of CaHf0.9In0.1O2.95 are obtained via defect equilibria model. The results reveal that the total conductivity of CaHf0.9In0.1O2.95 reaches 7 × 10−3 S∙cm−1 in a humid atmosphere at 800 °C. Moreover, the conductivities of grain interior are found to be higher than the total sample in the temperature range of 400 to 800 °C. The activation energy of the proton is lower than oxygen vacancies and holes, whereas the activation energy of the grain interior is lower than the total sample. In the temperature range of 400 to 800 °C, transport properties of CaHf0.9In0.1O2.95 are dominated by proton conduction under the given experimental atmosphere, where the transport number is estimated to be 0.497 at 800 °C. Moreover, the proton transport number of grain interior is found to be higher than the total sample. Hence, the grain interior of CaHf0.9In0.1O2.95 can promote proton conduction by obstructing the transport of oxygen vacancies and hole carriers. In summary, these results exhibit that CaHf0.9In0.1O2.95 possesses certain application prospects in electrochemical sensors.

10.2139/ssrn.4379214

Cite

Citations (0)

UniMS-RAG: A Unified Multi-source Retrieval-Augmented Generation for Personalized Dialogue Systems

arXiv (Cornell University) (2024)

Hongru Wang Wenyu Huang Yang Deng Rui Wang Zezhong Wang

Large Language Models (LLMs) has shown exceptional capabilities in many natual language understanding and generation tasks. However, the personalization issue still remains a much-coveted property, especially when it comes to the multiple sources involved in the dialogue system. To better plan and incorporate the use of multiple sources in generating personalized response, we firstly decompose it into three sub-tasks: Knowledge Source Selection, Knowledge Retrieval, and Response Generation. We then propose a novel Unified Multi-Source Retrieval-Augmented Generation system (UniMS-RAG) Specifically, we unify these three sub-tasks with different formulations into the same sequence-to-sequence paradigm during the training, to adaptively retrieve evidences and evaluate the relevance on-demand using special tokens, called acting tokens and evaluation tokens. Enabling language models to generate acting tokens facilitates interaction with various knowledge sources, allowing them to adapt their behavior to diverse task requirements. Meanwhile, evaluation tokens gauge the relevance score between the dialogue context and the retrieved evidence. In addition, we carefully design a self-refinement mechanism to iteratively refine the generated response considering 1) the consistency scores between the generated response and retrieved evidence; and 2) the relevance scores. Experiments on two personalized datasets (DuLeMon and KBP) show that UniMS-RAG achieves state-of-the-art performance on the knowledge source selection and response generation task with itself as a retriever in a unified manner. Extensive analyses and discussions are provided for shedding some new perspectives for personalized dialogue systems.

Relevance

10.48550/arxiv.2401.13256

Cite

Citations (3)

Conduction and Transport Properties of Ba4baxnb2−Xo9−Δ Proton Ceramics

Minghao Yin Zezhong Wang Wenlong Huang Yongtao Huang Jie Zheng

10.2139/ssrn.4940075

Cite

Citations (0)

Conductivities and grain interior transport properties of CaHf0.9In0.1O2.95

Ceramics International (2023)

Zezhong Wang Ying Li Wenlong Huang Yushi Ding