Hydrocarbon Processing - May 2021 - 33

Maintenance and Reliability
the tack weld-which joins the weld root
to the reinforcement pad-is present, the
pad is functioning as the backing bar. The
root of the weld comprises this tack weld.
This section (FIG. 5) shows that there is a
sharp corner between the reinforcement
pad and the tack weld.
When a backing bar is used to assist
with the manufacture of a single-sided
butt weld, the resulting weld root often has
a poor profile and a relatively poor fatigue
performance. This weld detail has a Class
F fatigue performance. The fillet weld
around the edge of the reinforcement pad
(joining it to the shell) would be expected
to also have Class F fatigue performance.
In this case, cracking has occurred in
the expected location for a butt weld on a
backing bar, which is at the stress concentration associated with this sharp corner at
the weld root. A large gap exists between
the reinforcement pad and the shell, at the
interface with the nozzle. The tack weld
(which functions as the root of the fullpenetration butt weld) has extended a few
millimeters beyond the shell to fuse to the
reinforcement pad, producing a particularly unfavorable weld root profile. The
applied stresses and actual fatigue life are
not known, but a poor weld root profile results in a more severe stress concentration
and, therefore, produces a lower fatigue
life than would be expected in the absence
of the stress-concentrating feature. This
corroborates with the service condition of
the degassing boot, which is susceptible to
flow-induced vibration and fatigue.
Recommended practices for tack
welds. API 577 (Recommended Prac-

tice-Welding Inspection and Metal-

lurgy) defines a tack weld as " ... a weld
made to hold the parts of a weldment in
proper alignment until the final welds
are made. " Clause 4.3.3.1 mentions quality control items to assess if " tack welds
to be incorporated in the weld are of acceptable quality. "
To improve the fatigue life of a joint
with this fit-up, the welder should not
tack the reinforcement pad to the shell
before welding the shell to the nozzle.
This will avoid fusing the root of the fullpenetration butt weld to the reinforcement pad and will allow a more uniform
root profile to be produced at the shellto-nozzle weld. The welder should take
care to ensure that the shell-to-nozzle
weld is full penetration, but that the root
bead does not protrude very far from the
shell plate. This should allow the weld's
Class F fatigue weld to be achieved.
An inlet riser deflector was previously
welded to the inner side of the shell. After this failure had occurred, the deflector plate was removed to be reused during repair work. This deflector plate was
removed by gouging, which explains the
two parallel gouge marks. As this was removed after the failure, it has not had any
effect on the failure.

* The component has failed due to
fatigue crack propagation through
the full thickness of the shell.
* Cracking had initiated at a poor
weld root profile, caused by the
presence of the tack weld.
The investigation ultimately yielded
these recommendations:
* It may not be possible to avoid
tack welds during manufacturing
or repair activities. To improve
the fatigue life of this component,
the reinforcement pad should not
be tacked to the shell during the
manufacture of the nozzle weld.
This would allow a Class F fatigue
weld to be achieved.
* Attention should be provided to
the detrimental effects arising out
of tack welds, particularly where
fatigue loading is envisaged.
* Welders and inspectors must
be more cautious and vigilant
regarding tack welds, fit-ups and
weld quality to improve the fatigue
life of the components.
* Welds should be considered as
a checkpoint in the inspection
checklist during manufacturing/
repair activities.

Conclusion and recommendations.

ACKNOWLEDGMENTS
The authors thankfully acknowledge the support of Kuwait Oil Co. and the management of The
Welding Institute in the publication of this article.

The following conclusions can be drawn
from this investigation:
TABLE 3. Summary of tensile properties

Proof stress Rp 0.2%, MPa

Tensile strength, MPa

Shell material, Specimen 1

300

535

Shell material, Specimen 2

317

530

ASTM A516M-17 Grade 60

220

415-550

TABLE 4. Summary of chemical compositions compared to ASTM A516M-17 Grade 60 and ASTM A106-18 Grade B speciﬁcations
Element, wt%
Material

ASTM A516M-17
Grade 60*

0.21
(max.)

0.79-
1.3

0.025
(max.)

0.025
(min.)

ASTM A106-18
Grade B*

0.3
(max.)

0.29-
1.06

0.035
(max.)

0.1
(min.)

0.4
(max.)

0.15
(max.)

0.4
(max.)

0.012

Shell parent

0.16

1.18

0.011

0.002

0.32

0.034

0.004

0.19

0.036 < 0.004 0.005

0.03

0.012

0.003

0.002

0.0017

Nozzle parent

0.18

0.66

0.012

0.011

0.18

0.058

0.008

0.036

0.022 < 0.004 0.005

0.067 < 0.002 0.006

0.001

0.004

0.0023

Reinforcement
pad parent

0.16

1.16

0.017

0.003

0.35

0.036

0.006

0.19

0.044 < 0.004 0.006

0.025

0.015

0.017

0.006

0.003

0.0022

Weld 1

0.06

1.05

0.016

0.008

0.62

0.049

0.009

0.034 < 0.003 0.005

0.006

0.05

0.003

0.005

0.011

0.008

0.0003

Weld 2

0.07

1.16

0.017

0.009

0.68

0.038

0.006

0.04

0.004 < 0.004 0.006

0.039

0.003

0.005

0.015

0.011

0.0004

*Based on product analysis:
* Boron, lead and zirconium were below 0.005 wt% for the compositions measured.
* Tungstun was below 0.01 wt% for the composition measured.
* Cerium and antimony were below 0.002 wt% for the compositions measured.

Hydrocarbon Processing | MAY 2021

Hydrocarbon Processing - May 2021

Table of Contents for the Digital Edition of Hydrocarbon Processing - May 2021