Sage Journals: Discover world-class research

Abstract

In austenitic stainless steel welds it is necessary to control the weld metal composition to promote primary ferritic mode of solidification for minimising the solidification cracking susceptibility of welds and to reduce the amount of slag formation during arc welding. Modern approach for predicting the solidification mode as a function of weld metal composition is by application of artificial neural network (ANN) based model. In the present work, composition only dependent Bayesian classification neural network model for classification of solidification modes is developed. Nickel was found to exhibit a clear pattern in influencing the solidification mode in austenitic stainless steel welds. Analysis of combined effect of nickel and other alloying elements showed that in addition to nickel, chromium, manganese and nitrogen were the other alloying elements whose concentrations determine the solidification mode in austenitic stainless steel welds. There was good agreement between the model predictions and the experimental data and the accuracy of the model predictions on an independent dataset was determined as 81%.

Keywords

ARTIFICIAL NEURAL NETWORK AUSTENITIC STAINLESS STEEL WELDS SOLIDIFICATION MODE

Get full access to this article

View all access options for this article.

References

Suutala

, Takalo

and Moisio

: Metall Trans. A, 1979, 10A, (4), 512–514.

Suutala

, Takalo

and Moisio

: Metall Trans. A, 1979, 10A, (8), 1183–1190.

Takalo

, Suutala

and Moisio

: Metall Trans. A, 1979, 10A, (8), 1173–1181.

Katayama

, Fusimoto

and Matsunawa

: Trans. JWRI, 1985, 14, (1), 123–128.

Lippold

J. C.

and Savage

. F.: Weld J, 1979,58, (12), 362s-374s.

Lippold

J. C.

and Savage

. F.: Weld J, 1980,59, (2), 48s-58s.

David

S. A.

: Weld. 1, 1981,60, (4), 63s-71s.

Brooks

, Thompson

and Williams

: Weld. 1, 1984,63, (2), 71s-83s.

Brooks

J. A.

and Lambert

. J.: Weld. 1, 1978,57, (5), 139s-143s.

10.

Lippold

J. C.

and Savage

. F.: Weld J, 1982,61, (12), 388s-396s.

11.

Cieslak

M. J.

, Ritter

A. M.

and Savage

. F.: Weld. J, 1982,61, (1), 1s — 8s.

12.

Suutala

: Metall. Trans. A, 1983, 14A, 191–197.

13.

Brooks

and Thompson

A. W.

: Int. Mater. Rev., 1991, 36, (1), 1644.

14.

Suutala

and Moisio

: ‘Solidification technology in the foundry and casthouse% 1980, London, The Metals Society.

15.

Bhadeshia

H. K. D. H.

, David

S. A.

and Vitek

. M.: Mater. Sci. Technol, 1991, 7, 50–61.

16.

Elmer

J. W.

, Allen

S. M.

and Eager

. W.: Metall. Mater. Trans. A, 1989, 20A, 2117–2131.

17.

Lippold

J. C.

: Weld. 1, 1994,73, (6), 129s-139s.

18.

Ramoni

and Sebastiani

: Artif. Intell., 2001, 125, 209–226.

19.

MacKay

D. J. C.

: Neural Comput., 1992, 4, (5), 720–736.

20.

Ichikawa

, Bhadeshia

H. K. D. H.

and Mackay

D. J. C.

: Sci. Technot Weld Join.,1996, (1), 43–50.

21.

Suutala

: Metall. Trans. A, 1982, 13A, 2121–2130.

22.

McCowan

C. N.

, Siewert

T. A.

and Olson

. L.: WRC 1989, 342, 1–36.

23.

Osuki

, Yonemura

, Ogawa

, Komizo

and Terasaki

: Sci. Technol. Weld. Join., 2006, 11, (1), 33–42.

24.

Di Schino

, Mecozzi

M. G.

, Barteri

and Kenny

J. M.

: J. Mater. Sci., 2000, 35, 375–380.

Artificial neural network modelling of solidification mode in austenitic stainless steel welds

Abstract

Keywords

Get full access to this article

References