Engineering steels 247

manufacturing defects which were present at the inner surface of the cylinder,
close to the highly stressed junction where the wall joins the concave base. The
fatigue cracks propagated through to the outside wall of the cylinder, resulting
in leakage of gas. The failed containers had been made by various manufacturers
and conformed to the appropriate specifications at that time. Under the aegis of
the Health and Safety Executive, a committee was established to investigate the
failure mechanism and make appropriate recommendations. The results of this
investigation are the subject of an interesting paper by Harris et al. 21

10 -1

10-2

_r

~ i0-3

2

0

i0-4

R=0.3, sinewave, 1 Hz

9 #

9 9 _x a 4

,,,, ,z;

9 ""i

9 x'r

+ &

A§

+ A

r

10-5 I _ I

10 20 30

X Ir

M

X

150 bar H 2

+

o

o +

9 o

N * oa~+

O & + O

a. 1

Nltm,gen

data

m Old 1CrMo, HB 324

a Old 1CrMo, HB 274

& 34 CrMo4, HB 292

,a New 1CrMo, HB 280

9 Exp. 1CrMo, HB 287

O 35CD4, HB 279

x Exp. C Mn, HB 296

+ 28Mn6, HB 278

.I ....... t t I

4O 6O 80 100

AK (MNm -~)

Figure 3.31 The influence of hydrogen on the rate of fatigue crack growth in gas
container steels (After Harris et al. 21)