Abstract

The investigation reported in this paper was an attempt to correlate bending and torsion fatigue results for tempered Swedish valve-spring wire so that the torsion fatigue strength might be inferred within reasonable limits of accuracy from the fatigue values found in the simpler rotating-beam test. The original program called for (a) the development of suitable methods of gripping the wire for the torsion fatigue tests; (b) the determination of bending and torsion fatigue values for one size of wire in the as-received condition, and also after removal of the surface layer in a centerless grinder; (c) the determination of fatigue values for other sizes of wire between ⅛ in. and ¼ as permitted by time limitations; and (d) a survey of the literature for data bearing on the correlation of bending and torsion results. The results of these experiments on straight specimens of tempered Swedish wire in torsion and bending fatigue may be summarized as follows:

1 Methods of gripping were developed which are satisfactory for both bending and torsion fatigue tests with the exception of torsion tests on shot-blasted wire.

2 Zero to maximum torsion fatigue values for straight specimens in the as-received condition were about 70 per cent higher than values estimated from previous tests on carefully coiled springs.

3 The smaller of two sizes of wire tested in the as-received condition had the higher torsion fatigue limit and lower ultimate torsional strength.

4 Removal of the natural surface by grinding increased the bending fatigue strength about 20 per cent, but decreased the torsion fatigue strength about 4 per cent.

5 Shot-blasting increased the bending fatigue strength about 27 per cent and increased the torsion fatigue strength an undetermined amount.

6 The zero-to-maximum torsion fatigue limits of two sizes of wire tested in the as-received condition were 1.5 and 1.6 times their respective bending fatigue limits.

7 For two sizes of wire in the as-received condition the values of 0.84 and 0.90 were obtained for the ratio of reversed torsion fatigue strength to reversed bending fatigue strength.

8 Approximate measurements of surface residual stress were made on the wire in different conditions before and after fatigue testing and before and after coiling, which may have a bearing on the fatigue properties of springs.

This content is only available via PDF.
You do not currently have access to this content.