Probabilistic neural network identification of an alloy for direct laser deposition

被引：29

作者：

Conduit, B. D. ^{[1
]}

Illston, T. ^{[2
]}

Baker, S. ^{[2
]}

Duggappa, D. Vadegadde ^{[3
]}

Harding, S. ^{[4
]}

Stone, H. J. ^{[5
]}

Conduit, G. J. ^{[6
]}

机构：

[1] Rolls Royce Plc, POB 31, Derby DE24 8BJ, England

[2] Mat Solut, Worcester WR4 9GN, England

[3] Rolls Royce India Private Ltd, Manyata Tech Pk, Bangalore 560024, Karnataka, India

[4] Rolls Royce Plc, POB 3, Bristol BS34 7QE, Avon, England

[5] Univ Cambridge, Dept Mat Sci & Met, 27 Charles Babbage Rd, Cambridge CB3 0FS, England

[6] Univ Cambridge, Cavendish Lab, JJ Thomson Ave, Cambridge CB3 0HE, England

来源：

MATERIALS & DESIGN | 2019年 / 168卷

基金：

英国工程与自然科学研究理事会;

关键词：

Nickel; Direct laser deposition; Alloy; Neural network; CHEMICAL-COMPOSITION; PREDICTIONS; SUPERALLOY; MICROSTRUCTURE; TEMPERATURE; CONFIDENCE; STRENGTH; HARDNESS; DENSITY; DESIGN;

D O I：

10.1016/j.matdes.2019.107644

中图分类号：

T [工业技术];

学科分类号：

08 ;

摘要：

A neural network tool was used to discover a new nickel-base alloy for direct laser deposition most likely to satisfy targets of processability, cost, density, phase stability, creep resistance, oxidation, fatigue life, and resistance to thermal stresses. The neural network tool can learn property-property relationships, which allows it to use a large database of thermal resistance measurements to guide the extrapolation of just ten data entries of alloy processability. The tool was used to propose a new alloy, and experimental testing confirms that the physical properties of the proposed alloy are better tailored to the target application than other available commercial alloys. (c) 2019 The Authors. Published by Elsevier Ltd.

引用

页数：9

共 95 条

[1]

Amin S, 2011, Materials selection in mechanical design, pxi

[2] THERMO-CALC & DICTRA, computational tools for materials science [J].

Andersson, JO ;

Helander, T ;

Höglund, LH ;

Shi, PF ;

Sundman, B .

CALPHAD-COMPUTER COUPLING OF PHASE DIAGRAMS AND THERMOCHEMISTRY, 2002, 26 (02) :273-312

[3]