"
MPC – A Council of the WRC, the Materials Properties Council is dedicated to providing
industry with the best technology and the best data that can be obtained on the properties of
materials to help meet today’s most advanced concepts in design and service, life assessment,
fitness-for-service, and reliability and safety.
PVRC – A Council of the WRC, the goal of the Pressure Vessel Research Council is to
encourage, promote and conduct research in the field of pressure vessels and related pressure
equipment technologies, including evaluation of materials, design, fabrication, inspection and
testing.
WRC Bulletins contain final reports from projects sponsored by the Welding Research Council, important
papers presented before engineering societies and other reports of current interest.
No warranty of any kind expressed or implied, respecting of data, analyses, graphs or any other
information provided in this publication is made by the Welding Research Council, and the use of any
such information is at the user’s sole risk.
All rights are reserved and no part of this publication may be reproduced, downloaded, disseminated, or
otherwise transferred in any form or by any means, including photocopying, without the express written
consent of WRC.
ISSN 0043-2326
Library of Congress Catalog Number: 85-647116
Library of Congress
Catalog Card Number: 85-647116
Copyright © 2009 by
Welding Research Council, Inc.
All Rights Reserved
Printed in U.S. A
FOREWORD
In recent years there has been increasing need to acquire technology for the safe handling
of hydrogen for petroleum and other energy related applications. The pace of research into
metal/hydrogen interactions accelerated at the start of the space age due to the hydrogen's
efficiency per unit weight as a rocket fuel. Practical applications called for hydrogen to be
transported and used at high pressures thereby increasing its hazard potential. Exotic and
expensive alloys could be used to put astronauts into space because cost was of little concern
where safety was an issue. For most conventional applications steel is still the practical choice.
It has been long recognized that at ordinary and elevated temperatures hydrogen can be
destructive to steels, not just to high-strength steels, but also to steels of ordinary strength levels.
Hydrogen's capability to enter and then diffuse through the metallic lattice, accelerated by stress
gradients and seeking out points of weakness where it can concentrate or react, renders it capable
of destroying pressure retaining metals from the inside where damage defies detection and until
it is too late and the pressure containment has failed or is no longer safe.
This Welding Research Council (WRC) Bulletin is part of a series that captures the
essential studies of the interaction of steel-hydrogen interactions in recent years. Topics include
Modern Vanadium Steels for High Temperature Petroleum Reactors ( # 524), Fabrication and Repair of
Low Alloy Steel Pressure Equipment ( # 525), Performance of Steels in Hydrogen Charging
Environments ( # 526), Practical Aspects of Hydrogen Attack ( # 527), Test Methods for Hydrogen
Induced Cracking ( # 530), Metallurgical Studies of Steels for Sour Service Environments ( # 532),
Studies of Cladding and Overlay for Pressure Vessel Service ( # 534), and Toughness, Fracture and
Fitness for Hydrogen Service ( # 535).
Martin Prager
Executive Director, WRC
Table of Contents
S. Dittrich
Thyssen Schweisstechnik GmbH, Hamm, Germany
G. Grote
Kloeckner Wilhelmsberger, Osnabrueck, Germany
A. Heuser
BASF, Ludwigshafen, Germany
Dedicated to
Prof. Dr. Dietman Werner
BASF,Ludwigshafen,Germany
To evaluate the potential risk of hydrogen induced fabricated with a high grade of basicity including
weld cracking it is necessary to know the different the necessary nucleation aids. The objective of this
factors that influence the weld besides high measure is getting a weld metal structure with a
stresses. The most important role is played by the smaller grain size which results in a reduced
microstructure of the weld metal and the actual amount of precipitation along the grain boundaries
hydrogen content in this area. The first is directly because of the wider distribution. This lowers the
coupled to the preheating/interpass temperature, sensitivity to temper embrittlement and increases
the second is generally unknown. This report will the toughness properties [Ref. 5].
show actual figures of the hydrogen content as
measured within the welded joint after applying A typical flux of the agglomerated type is i. a. UV
different dehydrogenation temperatures. 420 TTR which has been successfully used for
more than a decade worldwide for submerged arc
2. Critera for the selection of flux welding of heavy wall 2 114 Cr - 1 Mo vessels up to
a thickness of approximately 300 mm. This flux,
Besides the chemistry of the wire and its which is mostly used outside Japan for the above
cleanliness regarding detrimental tramp elements purpose has a basicity index of 3. 5 according to the
the right selection of flux is of great importance in Boniszewski formula [Ref. 6]. The index is in this
order to meet all requirements which are typical for case based on Mol %. The moisture content after
2 1/,i Cr - l Mo weldments. Principally there are proper redrying (for 2 h at 300-350°C) is
two types of fluxes available, either fused or approximately 0. 04 % which corresponds with a
agglomerated. diffusible hydrogen content of 7 - 8 ppm according
to ISO 3690. This is approximately 100 % more
2.1 Fused fluxes than the hydrogen content of most fused fluxes.
The special advantage of the fused flux type is its Nevertheless, despite their higher hydrogen content
extremely low moisture content which results in a agglomerated fluxes are normally used worldwide
correspondingly low hydrogen content according to for SAW of 2 114 Cr - l Mo material because of the
ISO 3690, which determines the maximum already mentioned advantages. Principally it is
hydrogen content immediately after welding of one possible to reduce the hydrogen content of
single bead. Fused fluxes have a lower grade of agglomerated fluxes below 5 ppm by higher baking
basicity than high basic agglomerated fluxes and temperatures during manufacturing. However,
they do not contain any microalloying elements when using a flux like UV 420 TTR this is not
which should be present in order to support the recommendable, since this would destroy
nucleus formation in the solidifying weld metal. microalloying elements which are needed for the
Both are important factors when trying to achieve initiation of the nucleation process.
good toughness properties. If fused fluxes are used,
it is more difficult to guarantee the high toughness 3. Objectives of the performed tests
properties in the low temperature range after ,,step
cooling" which are required for most of the Experience gained over the last decade has shown,
concerned projects. that no delayed hydrogen induced cold cracks were
observed after using an agglomerated flux (UV 420
2.2 Agglomerated fluxes TTR) when suitable welding and soaking
precautions were observed during fabrication of
Agglomerated fluxes, which are sometimes called heavy wall 2 114 Cr - l Mo weldments. This fact can
bonded fluxes, give more metallurgical possibilities be explained by the high preheating and interpass
when optimizing the weld metal properties and are temperature which is required for the welding of
highly recommended. The expression ,,bonded" is heavy wall 2 114 Cr - l Mo vessels. The long time
due to the liquid silicate binder (sodium or welding at a consistent interpass temperature of
potassium silicate or a mixture of both) which approximately 250°C has the same effect as
facilitates the agglomeration or granulation process continuous soaking. Final soaking at an elevated
during the flux manufacturing. Such fluxes can be
temperature in the range of 280° - 350°C in Table 2. Preheating and interpass temperatures
guarantees a sufficient reduction of hydrogen in the were based on specific fabrication experience of
upper part of the joint, which has a shorter time KlOckner Wilhelmsburger GmbH with this type of
exposure to the interpass temperature than the steel. Preheating was done electrically with
lower part of the weld joint. resistance elements which were uniformly
distributed below the weld area.
Despite these observations there are still some
doubts expressed in the literature regarding the The flux was intentionally redried at the lower
efficiency and reliablility of soaking treatments. As temperature limit of the manufacturer
a result a few specifications require an intermediate recommendation (300°C) and then kept at 150°C
stress relieving immediately after welding instead until used.
of the more economicai dehydrogenation by
soaking the weld joint. Up to now there are no Temperatures were monitored using thermocouples
actual measured hydrogen figures that have been fixed in drilled holes in the wall center (Figure 1).
reported in order to compare the real difference There were 120 beads used to finish the weld and it
between stress relieved and long time soaked took 42.5 hours to complete. The long welding
weldments. Therefore, the objective of this paper time was caused by interruptions to avoid
was to perform such actual measurements under temperature rise above the specified maximum
simulated shop floor conditions. The kind of interpass temperature but agreed with welding
performance and the test results are reported in the times of a real pressure vessel.
following sections.
4.3 Heat treatment
4. Experimental details
Immediately after completion of the joint, the weld
4.1 Materials plate was cut into four test pieces without allowing
cool down. The four test pieces were then subjected
For the tests a 120 mm thick plate of 2 114 Cr - l to different heat treatments shown in Figure 2. The
Mo steel was used, quenched and tempered to 355 treatments B and C were performed with electric
N/mm2 minimum yield strength. Table l shows the resistance elements without moving the test pieces;
chemical composition of the test plate. The test the intermediate PWHT D at 620°C was done in an
plate was cut into 8 pieces of 280 mm length and electric furnace.
165 mm width each, which were welded together
by SAW process according to Figure 1. Treatment A (as welded without any subsequent
dehydrogenation) was chosen for comparison
In order to allow 4 test samples to be cut without reasons, only. Treatment B (280°C/2 h) comes very
heat influence on the weld area after completion of close to the specified maximum interpass
the joint, 20 mm wide pieces of 2 V. Cr - 1 Mo temperature which ensures that the final weld
were arranged between the test plates to support beads deposited can dehydrogenate sufficiently. It
the molten pool. The narrow gap joint geometry represents the lower level of the known specified
used for the test samples corresponds with the soaking treatments. Treatment C (350°C/4 h)
bevel design which is normally used for heavy wall represents the upper level of the given information
pressure vessels. about soaking, Treatment D represents the
sometimes specified intermediate stress relieving.
The SAW wire was Thyssen S l CrMo 2 (EB 3)
and the flux used was UV 420 TTR. Weld metal 4.4 Specimen preparation for hydrogen
chemistry is also shown in Table 1. determination
measurement. For this purpose, specimens of 6 mm A. HAZ hardness is about 20 HV 10 lower. The
diameter were drilled out of each part of the joint hardness values after treatment B and C can be
at five different locations. The appertaining details expected in the same range as after treatment A.
are shown in Figure 3. The center of the weld was
chosen for the hydrogen determination because this After a soaking treatment at 620°C/l h (sample D)
location would contain the maximum amount of weld metal and HAZ show a decrease in strength.
hydrogen due to the thermal exposure during Both areas almost meet the well known hardness
welding. limit of maximum 240 HV l 0 which has been
specified in order to avoid hydrogen embrittlement
The drilling process was performed under during service or fabrication.
continuous cooling. After drilling, the individual
specimens were stored in liquid nitrogen until the 5.2 Hydrogen distribution
final determination of hydrogen was performed by
three different laboratories. Figures 6 to 9 summarize the results of the
measured hydrogen distribution along the weld
4.5 Determination of hydrogen metal thickness. The absolute hydrogen contents
are average values out of 3 individual
From each test plate and locations, the ,,total" determinations. In all cases the minimum values
(residual plus diffusible) and ,,diffusible" hydrogen were found in the root area. For treatment A the
content was analysed. For the measurement of the total hydrogen content of the top surface area drops
total hydrogen the ,,Total Hydrogen Determinator" from 1.5/1.8 to 0.75 ppm in the root area (Figure
of LECO Corporation was used and the analyses 6). When compared to treatment B it gives a small
were performed by two different laboratories using reduction within the top surface area of about 0.10
different models of the LECO instrument. to 0.35 ppm but the maximum values in the 1/4 t-
Laboratory A applied a hot extraction temperature position are hardly influenced (Figure 7). The
of 860°C, being in accordance with the steel mill diffusible hydrogen content is influenced in a
specification (Method A). For calibration purposes similar way.
the Brammer standard was applied (2.10 ppm with
± 0.23 ppm standard deviation). Treatment C gives a significant reduction with
maximum values of about 1.0 ppm total and 0.2
Additionally, laboratory B extracted the hydrogen ppm diffusible hydrogen (Figure 8). After 1 hour
at 1800-2000°C, and used two standards for intermediate stress relieving at 620°C (treatment D
calibration, 1.1 ± 0.3 and 4.8 ± 0.4 ppm (Method in Figure 9) the total hydrogen content is on a
B). much lower level and the diffusible hydrogen has
completely disappeared.
The diffusible hydrogen was determined by
Laboratory C using a gas chromatograph (Type Figure 10 represents the diffusible hydrogen values
Oerlikon/Yanaco) and an extraction treatment of 6 determined in the root and top surface area for the
hours at 150°C, which corresponds with AWS different treatments. The typical distribution of
Standard A4.3-86 and gives results comparing to high hydrogen values near the top surface and the
ISO 3690 (72 hours at 25°C) [Ref. 7]. low values in the root area is due to the soaking
effect during welding. All values are below l ppm
5. Test results even for the as welded condition which can already
be regarded as far below the critical limit. Still
5.1 Hardness values more on the safer side is the soaking treatment at
350°C. There must be an answer to the following
Figures 4 and 5 show the hardness distribution in question: How meaningful is a hydrogen content of
the welded joint after treatment A and D. The 7 - 8 ppm according to ISO 3690 given for this
maximum hardness values of about 340 HV 10 type of flux under the conditions of continuous
were found within the weld metal area of sample welding of heavy wall plates?
Element
Base metal Weld metal
{weight%)
c 0,11 0,08
Si 0,21 0,23
Mn 0,60 0,84
p 0,005 0,007
s 0,002 0,004
Cu 0,07 0,07
Sn 0,008 0,004
Al 0,02 0,019
Cr 2,28 2,37
Mo 1,04 0,99
Ni 0,09 0,05
v 0,01 0,01
As 0,012 0,005
Welding Wire size Re-drying Preheat and max. Current Voltage Speed
consumables treatment of flux lnterpass temp.
(mm) (•C) (•C) (A) (V) (mm/min)
Thyssen S 1 CrMo 2 4,0 300°C/2 hrs, then 250 I 290 550 :!:50 30 I 20 550 :!:50
EB 3 1so•c until use (DC/electrode
positive)
(wire)
UV 420 TTR
(flux)
thermocouple
42.5 h
~ 250 - 290 °C
welding
air
24 °C/h
RT
43 h. 2h
20 °C/h
~ 62 °Cih
250 - 290
welding
°c
RT
4h
45 h' 23 °Cih
~ 62 °C/h air
~h
250 - 290 °c
welding
RT
1h
45 h*
/\ Hydrogen samples
\
~. ~ sample a
\
sample b
E
E
0
N
l
Fig. 3 Location of the Hydrogen Test Specimens
Cf}J as welded
340 ..--~~~~.....,...~~.....,...~~~~~
320
300
> 280
6
CJ)
CJ)
Q)
260
c
~ 240
Ctl
.s:::.
220~........o;;;;u"'~~~-t:-~~-----...;....;;~-t-~-b'~~~~=i1
200 1--~~~~-+-~~--..........~~~~---~..........~~-+~~~~~~
180
base metal HAZ weld metal HAZ base metal
~ 620°C/1 hair
340 .--~~~~-.-~~-.---~--------~--~~..........~~~~~,
320 1--~~~~-t-~~-t-_;..----~~~~~~~~-+~~~~~,
300 1--~~~~-r-~~-t-~~------_,_..........---...........~~-+~~~~~,
2oo'c::::::O::~..,.f_.~~-+-~~~~~~...J-~~~---CJ~=-l
180 1..-~~~~-+u~~-t-~--~~~~----+~~--t~~~~~~
~ aswelded
O total hydrogen method B
A diffusible hydrogen
rface area A
.7 ..,'
./
v/
--
~ - - ./.
;
I~
~
-I'
/
/
7
A
/~
'.J
,,/' -
....
rea .....
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
hydrogen [ppm]
~ 280°C/2hair
O total hydrogen method B
A diffusible hydrogen
.. J-
rface area
- I'
- I
I -,
.... / ,
I - .I
v
A
.......
f
I I//
v . .-. . -
... /
[7'
A/ -
,.,,..,,.
rea ...... ......... -
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
hydrogen [ppm]
~ 350°C/4hair
O total hydrogen method B
h. diffusible hydrogen
...-
J
-
ii.
I - . "'
-A
-
.... .I I '""'
I -
... -. I
·-
~
...J
...
-
_.
- -
//
I
I
I
1-
'"""
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
hydrogen [ppm]
~ 620°C/1 hair
O total hydrogen method B
A diffusible hydrogen
urface area
- --
- --
.-..I
I
I
..-
- I
..... ....
,_I -
- 1"7
I
-
--
~
rea - ........
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
hydrogen [ppm]
§ 620°C/ 1 hair-
\.\' .
L£I'.J 350 °C I 4 h air -
~, - ........
"' ..........
..........
L!J) 280 °C / 2 h air - \.
-\
\
-
••A
\
\
I
~ as welded - - I
-1
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
hydrogen [ppm]
200 0
E
; 150
l
E
~
~
"'e-
•
:su" 100 0
c:
"'.,
iV
.c
e
•• Crack region
•' •
a.. 50 06> 0 0
Ph.Bourges *, A.Cheviet **
* Head of the Welding department
Centre for Research on Materials, Creusot-Loire Industrie,
56 rue Clemenceau, B.P. 56, F-71202 Le Creusot Cedex (France)
** Senior Product Manager
Creusot-Loire Industrie
B.P. 68, F-42800 Rive de Gier (France)
treatments :
HAZ = Heat Affected Zone (Temperature achieved Location in SK
during welding :?: AC3) thicknesi +SC
SCHAZ = Sub Critical HAZ (Tempering Temperature ::;; Top -l l 7°C -107°C
Temperature achieved during welding::;; ACl) 1/4 -l 18°C -108°C
Fused Metal 1/2 -115°C -l 14°C
SR= Stress Relieving Treatment (690°C-8h)
3/4 -120°C -ll3°C
Q+T =Quality treatment (Quenching & Tempering)
Root -117°C -111°C
SC = Step Cooling Heat Treatment (embrittling treat- average -117'C . -111 'C
ment) Top -125°C -98°C
1/4 -122°C -98°C
2-3-1- Horizontal weld 100 mm thick. Table 7 Fusion Line 1/2 -109°C -87°C
summarizes the CVN curves (TK54J values) for different I 3/4 -l 13°C -95°C
locations. Root -123°C -104°C
average -118 'C -96'C
Top -123°C -99°C
Location in SR Q+T+SR 114 -120°C -99°C
thicknes~ +SC +SC HAZ 1/2 -127°C -98°C
Top -140°C -120°c 3/4 -114°C -114°C
114 -128°C -123°C Root -ll5°C -95°C
Base 1/2 -126°C -121°C average -120'C -101 'C
Metal 3/4 -130°C -l l6°C Top -96°C -89°C
Root -140°C -123°C 1/4 -105°C -92°C
average .133 'C -121 'C SCHAZ 1/2 -101°C -76°C
Top -ll3°C -100°C -129°C -120°C 3/4 -102°C -92°C
1/4 -l 12°C -99°C -127°C -ll5°C Root -lOl°C -90°C
Fused 1/2 -110°c -99°C -126°C -l17°C average -JOJ'C . -88'C
Metal 3/4 -1os 0 c -100°c -125°C -ll4°C Table 8
Root -115°C -94°C -132°C -121°C
average -112'C -98'C -128'C -117'C
Top -ll8°C -101°C
1/4 -107°C -89°C 2-3-3- Horizontal weld 200 mm thick. Table 9
Fusion 112 -110°c -76°C summarizes the TK54J values obtained in the fused
Line 3/4 -l 14°C -94°C metal and in the SCHAZ.
Root -l 13°C -101°C
average -112 'C -92'C
Top -113°C -l06°C Location in SR
114 -111°C -98°C thickness +SC
HAZ 1/2 -I 18°C -95°C Top -99°C -82°C
3/4 -122°C -103°C 1/4 -100°C -77°C
Root -117°C -97°C Fused Metal 1/2 -102°C -80°C
average -116'C -JOO'C Root -99°C -76°C
Top -97°C -85°C average -JOO'C -79'C
114 -88°C -84°C Top -87°C -80°C
SCHAZ 112 -94°C -76°C 1/4 -78°C -63°C
3/4 -101°C -78°C SCHAZ 112 -74°C -63°C
Root -100°c -83°C Root -80°C -77°C
average -96'C -81°C average -84'C -72'C
Table 7 Table 9
Figure 2
~
"'0 200
'3
0
0
;;.., 2-3-4-4- The effect of a step-cooling treatment is the
8l
0 same in all points of the weld metal and the heat affected
w= 100
zone, due to the fact that its effects are linked to the
amount of impurities in the tested metal.
Table 10
25
,....__ G
rri
0 20
<l.)
.....,
'"5 200 'tj"
0 V)
t;., ~ 15
;>., ~
OJ) <l
....
<l.)
i:: 100 10
~ HorizlOO
5 10 15
Figure 5
-150 -100 -50 0 50
Temperature (QC)
Figure 3
Horizontal weld 100 mm thick
Influence of the location of the notch on the TK54J
G
'2._,
-HorizlOO -100
~
"""""'Horiz200 If)
~
f....
0
-150 -100 -50 0 50 -125
Temperature (QC)
Figure4 -150
Figure 6
3- COMPARISON WITH SUBMERGED ARC
WELDING
perature : 250QC ; PWHT : 680QC-12h
The comparison between the conventional Submerged
Arc Welding and Electron Beam Welding for A387 3-1- Chemical compositions
Gr.22 Cl.2 is based mainly on the properties of the depo- Typically, with the better wire and flux couple, the
sited or fused metal. chemical composition obtained in the deposited metal is
compared in table 12 with the horizontal weldment 100
For instance a weldment using SAW process has been mm thick.
made with the following parameters :
Preheat : 225QC ; Heat Input : 2.5 kJ/mm ; Interpass tern-
% c Si Mn Cr Mo 02
0,079 0,290 0,340 2,370 1,020 0,0349
EBW
SAW
Horiz.100 0,134 0,207 0,509 2,269 1,000 0,0018 300
% p Sn As Sb x
SAW 0,0071 0,005 0,008 0,0006 9,80 ,....._
(/J
creep behaviour. The specimens were centered on the in accordance also with DEVILLERS [3].
fusion line in order that, parts of the fused metal, HAZ
and base metal will be in the gauge length. Results are
given in table 14 and figure 8. The rupture is always
located in the base metal near the HAZ, exactly in the Horizontal weld 100 mm thick - CVN curves for
SCHAZ. Fused Metal, FL & HAZ at 1/4 Thickness
o....._.-===~------.a.-~--.....a.
-160 -120 -80 -40
Temperature (°C)
• EBW 0 SAW
Figure 9
400
0 5- CONCLUSIONS
Crj'
0.
300 •-o
6 I~ The Electron Beam Welding of A387 Gr22 Cl2 has
"'"' D
~
Ci) 200 • 0
0
shown good properties which may satisfy the require-
ments of the petrochemical industry.
o..,._.
It appears that it can provide in welds the high quality
100 of today's base metals. Submerged Arc Welding has
16 17 18 19 transition temperatures at least 60°C higher. Creep beha-
viours are very close.
Larson-Miller Parameter [ T(20+1ogt)]
REFERENCES
Silvano Bertolini
N uovo Pignone, Massa Plant
Massa, Italy
a) management and handling of the welding consumables, a) direct supervision of the weld process at the shop floor
b) welding parameters automatic calibration system, by technical experts,
c) advanced NDT procedures.
b) electrical preheating and post heating on complex
Results are highlighted by the data showing mechanical geometry,
properties of the welded joints as well as overlay deposits.
c) ultra-narrow gap beveling,
Experience and qualification in the construction of heavy
reactors with the recently codified Cr-Mo Vanadium d) ergonomic organization of the work station.
modified steels is now being obtained. Main mechanical
properties of the weld are defined.
Direct supervision on the shop floor
by welding technology experts
INTRODUCTION
Experience suggests that a good performance may be
The iatest refinery processes mainly used for enhanced achieved by a direct supervision activity on the shop floor by
production of synthetic fuels and heavy oil have led the way a branch of the welding service. Dedicated operators should
to the development of Cr-Mo steels featuring advanced show particular aptitudes and should be educated in
characteristics. The two main targets are: metallurgy and welding by proper training at a welding
institute. They are selected among the best welders after 10-
refinement of existing materials to improve their 12 years of experience and have demonstrated a natural
fabrication and low temperature properties; tendency to handle related problems.
Closed cycle of the flux handling complex geometries like non-radial nozzles and heads in
pieces. Also, in this case, the metal temperature is
The handling of the fluxes for SAW and overlay is also automatically controlled through a sensor that in turn is ruling
a very important activity for the successful results of the the heating power. An operator is charged with the overall
welding with reference to the H 2 problems. control and the welder utilizes temperature sticks or portable
thermocouples for a full supervision.
Today, the welding consumables for traditional low alloy
Cr steels capable of meeting the mechanical properties
required by the users are readily obtained, One of the main Ultra-narrow gap beveling
problems is the hydrogen control which must start at the
receiving of the consumables by a check of the total water For many years, the SAW Tandem submerged arc
content of the flux and the chemical composition of the wires. welding has been successfully endorsed. In Fig. 1, the narrow
gap bevel evolution is shown. This evolution is possible
Total water content of 0.02 to 0.05% in weight are because of the following improvements:
generally acceptable figures for high quality consumables.
availability of fluxes with easy slug removal,
To avoid further hydrogen sources that may be
encountered at the manufacturer shop by careless handling, self-regulation of the stick out and the arc position with
there must be implemented a handling procedure called respect to the bevel walls by a home designed
"closed cycle of flux handing" for total control of the welding positioning sensor,
parameters.
anti-drift system implemented on rollers that stops any
From the drying oven located inside the storehouse the shifting of the shell during welding,
fluxes are transferred to the weld station in the shop through
portable facilities equipped with air filters, air dryers, electric machining of the bevels normally applied today to forged
heaters and flux feeders. The complete system is so arranged rings as well as to rolled plates.
that the contact with the environment is practically
impossible and the flux is continuously maintained hot to
avoid possible H 2 absorption. Ergonomic arrangement of the work station
Current figures of total diffusible H 2 measured on actual An ergonomic arrangement of the work station that
jobs according to BS 6694, part 4, are from 3 to 5 ml/I 00 gr includes refrigerated platforms, proper installation of heating
with random values as high as 7 ml/I 00 gr. These are in line burners and a camera for the remote control of the welding
with the generally accepted levels of H 2 and are considered point on a monitor may concentrate the operator's activity to
acceptable for the H 2 associated problems. the follow-up and corrective actions when required.
Measurement of residual diffusible Hi show figures of l The weld inspection by advanced UT procedures allows
to 1.5 ml/I 00 gr after the total weld cycle including\reheating the recording of the weld status, repetition of the inspection
at 250°C postheating at 350°C with 2 hours holding time during the life of the equipment and the control of the
before final PWHT on 2'/.iCr-lMo enhanced material as per evolution o~ ~y existing fljwf as well as the possibility of
ASME Code Case 1960-1, thickness 250 mm. new ones ansrng. !J
Preheating, interpass and post heating temperature The important mechanical properties obtained on actual
(250°C, 300°C and 350°C, in procedures for 2'/.iCr-IMo jobs for SAW and SAW tandem arc processes are presented
materials) must be ruled by an automated systemlffle fuel in Table l.
used is natural gas. The actual temperature of the metal to be
welded is read by an optical system which, in turn, is
governing the check valve of the fuel flow. The temperature DISBONDING TEST
is so maintained within the allowable range.
It must be emphasized that no domestic or international
High performance electric heaters are applied to standard yet governs the specimen preparation and the test
-,
20
lT
5 0
90° 90°
.,.,0
N,
R30 I.11.5
conditions for the disbonding so that results coming from CHEMICAL COMPOSITION OF
different sources cannot be directly compared. Typical results JCr-1Mo-l/4V-Ti-B
from the following procedure by an independent European PER ASME CODE CASE 1961-1
laboratory on weld overlayed material are in Table 2. The
data are in line with the literature and the soundness is
confirmed by experience over the past 30 years. 3Cr-1Mo-lf4V-
Ti-B
EXPERIENCES ON CR-MO VANADIUM per ASME Code Deposited
MODIFIED STEELS Case 1960-1 Weld Metal
c 0.14 0.13
The increase in refinery process severity demanded a
new material development. Currently the qualification of Mn 0.44 0.46
Vanadium modified steels of 2lf4Cr-1Mo and 3Cr-1Mo
compositions is underway. The relevant welding processes, p 0.007 0.007
i.e., SAW, SAW Tandem, and SMAW in accordance with
ASME Code and major user requirements are being studied.
s 0.005 0.004
The main properties of the base metal and welding are as Si 0.059 0.08
follows:
Cr 2.97 2.95
Mo 0.94 I
CONCLUSION
Cu 0.041 0.09
High performance vessels require quality assurance
programs to obtain Ni 0.041 0.06
Sn 0.003 0.003
As 0.001 0.0061
Sb 0.0009 0.0005
A387 278 586 461 465 410 24 26 - - 179-196- 6901 178 173
22CL2 198 26hrs 159 147
166 162
KV-29 ocJ
A336
F22
250 641 - 534 - 24 l- 68 - 191-220-
223
6601
42hrs
201
202
137
88
ENQ. 271 117
CASE
73
BS5500
KV-40 oc J
All Temperatures = °C
100 °C/Hr 0
200 °C/Hr 15
675 °C/Hr 31
Y.S. (MPa) 582 @ RT - 496 @ 450°C 560 @ RT - 485 @ 450° 615@RT- 480@450°C
Impact Toughness, KV -18° C (J) 93/79/113 W.M. 68/94/87 W.M. 95/102/133 W.M.
116/134/140 HAZ 106/98/112 HAZ 160/159/115 HAZ
120
110
fli" 100
LU
....I
::J 90 WELD METAL PWHT
0
:J,.
LU 80
:::l
~
t-
70
~a.. 60
~ 50
40
30
TEMPERATURE(DEG.CENT.)
100
90
er
UJ 80
..J
::J
0 70
::::-;.
UJ
::J 60
~
I- 50
~
Cl.. WELD PWHT + STEP C
-1-~~~~~-7,£-~~~
~ 40
30
20
-60 -50 -40 -30 -20 -10
TEMPERATURE(DEG.CENT.)
90
80
er
UJ 70 WELD METAL PWHT
..J
::J
0 60
::::-;.
UJ
::J 50
~
I- 40
~
Cl..
~ 30
20
10
-60 -50 -40 -30 -20 -10
RESTORATION OF DAMAGED
2~ Cr - 1 Mo VESSEL
J.S. Lee
Chicago Bridge & Iron Company
Houston, TX, USA
C.D. Lundin
University of Tennessee
Knoxville, TN, USA
ABSTRACT INTRODUCTION
This paper describes the diagnostic evaluation and As a result of an external fire, a 2114 Cr - 1 Mo
repair of a quenched and tempered, stainless steel weld (quenched and tempered) Hot Separator Vessel
overlaid 2114 Cr - 1 Mo pressure vessel that was damaged incurred substantial metallurgical damage. The fire
in an external fire. Items presented include the review of heated the vessel to a high enough temperature to alter
vessel history, diagnostic evaluation, which included the mechanical properties of the vessel material. It is
significant metallurgical evaluation of samples removed estimated that some areas of the vessel base material
from the vessel, and the subsequent of the vessel. and welds reached a temperature of above or near Ac3
The objective of the metallurgical evaluation, in during the fire, and then were either slow cooled in air
conjunction with the inform.1tion gained from the review or rapidly quenched with water used to put out the fire.
of vessel history, was to determine the repair methods
needed to restore the vessel to the required condition The vessel is 84" ID x 24'-0" (tangent length) (2134
for continued Thfa included quenching and x 7315 mm) with hemispherical heads on both ends. The
tempering of several vessel parts with the existing shell is 5-5/64" (129 mm) thick and the heads are 3"
deposited stainless steel overlay. The vessel was repaired (76 mm) thick, as shown in Figure 1. The vessel is weld
by cutting the vessel into the needed component sizes overlaid with a special 1/8" (3.2 mm) minimum
for the quenching and tempering heat treatment, thickness Type 309 overlay. It was manufactured in
welding back together, and final postweld heat treating accordance with ASME Section VIII, Division 2 rules
the vessel. and designed for a maximum operating pressure of 2425
psi (170 kg/ cm 2) at a temperature of 810° F ( 432° C).
After visual observation, dimension checks and a ring and several small nozzle forgings in the vessel shell
certain amount of hardness checks, the vessel was and top head. AU nozzle forgings were supplied in the
shipped to a heavy wall vessel manufacturing facility for quenched and tempered condition, austenitized at 1652 -
repair. 1688° F (900 - 920° C) and tempered at 1330° F (720° C)
with test coupons stress relieved at 1310° F (710° C) for
10.5 hours.
REVIEW OF THE ORIGINAL VESSEL
MANUFACTURING The actual heat treatments for the shell plates and
nozzle forgings are included in Table 1. It also lists
Before undertaking the repairs, it was important to tensile properties of the test coupons and surface
review the vessel design conditions, the materials and hardnesses (HBN) after the maximum PWHT on test
heat treatments given to the base metal and test coupons.
coupons, and the fabrication procedures and heat treat
procedures used by the vessel manufacturer. Original Vessel Manufacturer Heat Treat Procedures
The vessel was manufactured in 1978 to the The actual heat treat temperatures and times were
following requirements: not available; however, the vessel owner provided the
following heat treat procedures:
Design Code: ASME Section VIII, Division 2
1. Vessel Shell and Heads
Shell Plates, Heat Plates and Upper Part of the Skirt:
SA 387, Gr. 22, CL 2 Hot form at 1688° F + 38° F (920° C ..±_20° C) for 3.8
hours, austenitize at 1688° F ..±..18° F (920° C ..±_10° C)
Vessel Nozzle Forgings: SA 336, Gr. F22 followed by accelerated cooling at 36° F /min.
(20° C/min.) minimum and tempered.
Hardness: 225 HBN Maximum in Base Metal, Weld
Metal and Heat Affected Zones The vessel shell was to be subjected to 7.8 hours
and the heads to 3.0 hours of intermediate PWHT
1/8" minimum thickness of Type 309 (single layer) weld at 1202° F ..±..38° F (650° C ..±..20° C) during
overlay conforming to the following requirements: Cr ..2:.. fabrication.
22.0%, Ni ..2:.. 8.0% and C -=::;. 0.10%. Ferrite content
limited to 8% when measured by the Schael:1er diagram 2. Final PWHT of the Vessel
and 10% when measured by the Severn gage.
The completed vessel was given a final PWHT at
1299°F + 54°F - 0°F (704°C + 30°C - 0°C) for
The Original Vessel Materials and Heat Treatments 5.25 hours.
The vessel shell consisted of two shell rings, each The above heat treatments resulted in a maximum
ring made of two plates around. Both vessel heads were total tempering parameter, LMP, of about 37.0 x Hf,
formed from plates. The nozzle forgings were integral assuming a nominal final PWHT temperature of about
type with radiographable butt welds. 1326°F (719°C) as compared to the total LMP of about
37.25 x la3 for test coupons for the vessel shell and
The shell plates and head plates were furnished in heads. This resulted in about 8 hrs. avaiiable PWHT
the normalized and tempered condition, for later heat time at 1275°F (the nominal PWHT temperature for the
treatment by the original vessel manufacturer, with test vessel repair) for undamaged vessel shell and heads.
coupons heat treated to simulate the heat treatments
during fabrication. The test coupons were austenitized Welding Procedures
at 1688° F (920° C), accelerated cooled at 36° F (20° C)
per minute, given an intermediate PWHT at 1202° F The vessel drawings indicated that the following
(650°C) and a final PWHT for 10.5 hours at 1310°F procedures were used for the original butt welds in the
(710°C). vessel:
The vessel has a manway (Ml) in the top head, a • Shielded metal arc welding (SMAW) for head seams
bottom nozzle (Nl), a nozzle (N2) in the upper shell and nozzle forgings
" Narrow groove submerged arc welding (SAW) for Soft Zones and Hard Zones
circumferential shell seams
Soft zones were identified as those areas with
• Narrow groove gas metal arc welding (GMAW) for surface hardnesses less than 155 HBN (which
longitudinal shell seams corresponds to about 75 ksi). It was desirable, however,
to have a somewhat higher hardness than 155 HBN
(about 160 HBN) to allow for possible loss of strength
PRELIMINARY INVESTIGATION AND MAJOR due to the additional PWHT after repairs (5% hours
AREAS OF CONCERN minimum at 1275° F); therefore, those areas with 155-
160 hardness were identified as low hardness zones.
Visual observations, some NDE and hardness
readings were performed on the vessel at the refinery to The method of repair of the soft zones depended on
obtain information on the general condition of the the evaluation of the extent (size and depth) of the low
vessel. This provided the following information: hardness areas and metallurgical assessment of the
material. The material would need to be replaced or
• There was no noticeable damage or distortions to reheat treated (quenched and tempered, or normalized
the vessel, except for the skirt support. and tempered) if the softened zone extended through
the full thickness of the shell, or if the metallurgical
The vessel skirt was distorted and would have to be evaluation indicated that material was damaged from the
replaced. fire (even if it had acceptable hardnesses).
There were several "soft" zones (below 155 HBN) If the softened zone extended only to a partial depth
and "hard" zones (above 225 HBN) on the outer (with undamaged adjacent material), the repair could
surfaces of both vessel shell rings. High hardnesses possibly be accomplished by removing the softened
were also found on outside surface of nozzle N2 and zones by arc gouging or grinding and repair welding
on the upper part of the skirt support. Low hardness these areas.
zones were also observed in the top head. This
indicated that some parts of the vessel have been Hard zones were considered those with hardnesses
exposed to excessive heat and cooled slowly (which exceeding 225 HBN. Metallurgical and tempering studies
resulted in low hardnesses) and some parts were would need to be performed to determine whether the
subjected to rapid cooling (water quench) after material can be tempered back to 225 HBN maximum
being exposed to high temperatures (which resulted with adequate microstructures. If so, there would be no
in hard zones). need to reheat treat the material.
• Spot UT of the weld overlay did not detect weld Microstructures and Response to Heat Treatments
overlay disbonding, and spot PT did not detect any
cracks in the overlay. There was no information on the actual
temperatures and exposure times during the fire. The
The preliminary information obtained on this vessel high hardnesses on the vessel shell, nozzle N2 and the
provided enough confidence that the vessel could be skirt transition piece may have resulted from the
repaired, but there was a need for a detailed diagnostic material approaching or exceeding the upper critical
evaluation at a manufacturing facility to further evaluate transformation temperature (Ac3) and subsequent water
the present condition of the vessel and determine the quench.
best and most economical method of repair for each
part of the vessel. It was important to know that the heat affected
areas have adequate hardnesses/strength and suitable
The major areas of concern were the following: microstructures after repair, since this may have an
effect on elevated temperature strength and on
• Soft zones and hard zones in the vessel material. toughness properties. The same hardness can result
from different microstructures; however, the type and
., Response to heat treatment (plates, forgings, welds). distribution of carbides may have a significant effect on
elevated temperature properties.
Condition of the weld overlay.
Since there was no material readily available from 3. Hardness Readings on the Vessel Surfaces. The
the heat damaged areas of the vessel (e.g., cutouts, core hardness readings taken on the outside of the vessel at
samples) for evaluation of microstructures and response the refinery provided general information on the
to various heat treatments, it would be important to condition of the outside surface of the vessel. However,
obtain samples from the soft zones, hard zones and additional hardness readings were needed to fully
unaffected zones for metallurgical evaluation and study evaluate the extent of the high hardness and low
of response to various heat treatments. It was decided to hardness areas on inside and outside surfaces of the
obtain boat samples from the vessel for this purpose. vessel. On the outside surfaces, hardness readings wer
Figure 2 shows the general configuration and size of the e taken from high and low hardness zones on
typical boat sample. approximately 12" (305 mm) grid lines. For testing on
the inside surface, the weld overlay was stripped off at
each hardness test location. Approximately 200 hardness
Stainless Weld Overlay readings were planned from the locations listed in Table
3.
Sigmatized weld overlay can be embrittled and result
in cracking when subjected to high tensile residual 4. Metallurgical Evaluation of Boat Samples From
stresses. It was, therefore, important to know: Outside and Inside Surfaces (Including Weld
Overlay).Boat samples were selected from 20 locations,
• The ferrite content and ductility to insure that no including soft zones, hard zones in the shell plate, two
cracking would occur during repair welding. unaffected (i.e., sound) areas, soft areas ground out to
about 160-165 HBN hardness (where the hardnesses
• The effect of various heat treatments (quenching increased with the depth of grinding), and from the shell
and tempering, PWHT) on the weld overlay butt welds in the areas exposed to the fire. Boat samples
properties and disbonding characteristics. were needed for the following evaluations:
A plan was prepared for detailed evaluation of the • Time/temperature response of the damaged
present condition of the various parts of the vessel, for material, to confirm the heat treatments needed for
study of response to heat treatments for restoration and restoring materials properties from hard zones and
the ability to weld repair and heat treat the repaired soft zones to the desired properties.
vessel. Milestones were established to evaluate the
results and arrive at the most desirable and economical • Tempering studies on boat samples from soft zones,
method of repair. The plan included the following: hard zones and unaffected zones from base metal
and weld metal (longitudinal and girth welds) to
1. Dimensional Measurements. The non-uniform evaluate hardness and microstructure after one
heating and cooling from the fire can cause severe PWHT cycle of 5114 hours PWHT at 1275° F
thermal gradients and thermal stresses which can cause (690°C).
distortions. n was, therefore, important to perform
dimensional checks for straightness, bulges, H was imoo1rta:nl to determine the effects the fire on
and to assure that there were no local distortions which the materials properties at soft zones and hard zones
needed to be considered in the repair. and the ability to restore the materials properties
elevated temperature with
2. NDE of the Vessel After Shipping it to the subsequent heat treatments and to evaluate the
Fabrication Plant Extensive NDE was required to characteristics of the weld after the various heat
assure the vessel base material had not been physically treatments.
damaged by the fire and that there was no cracking or
disbanding in the stainles& weld overlay. Table 2 " Electron microscopy of carbides on extraction
summarizes the NDE requirements at various stages of replicas on material from soft zones ( < 155 HBN),
the repair. low hardness zones (about 155-160 HBN) and high
hardness zones ( > 225
6. Testing of Large Samples of the Actual Vessel with evaluations of different types being completed
Material. It was important to perform tests on the actual concurrently to establish the extent of vessel damage.
(undamaged) vessel material to insure a successful
repair of the vessel which would result in the desired External attachments were removed from the vessel,
properties for continued service. It was possible to and the vessel was sandblasted to remove rust to
reduce the length of the vessel to permit removing a 7" improve access and sensitivity for the examinations.
(175 mm) wide band from one end of the upper shell
ring for the following testing: To reduce the risk of residual stresses causing
distortions during disassembly of the vessel (if needed
a. Tests on the prolongations of the longitudinal vessel for the repair), the vessel was postweld heat treated for
shell welds and base material to confirm that these 2 hours at 1200°F (650°C).
welds will meet the required mechanical properties
after all restoration heat treatments, including Dimensional measurements were performed after
quenching and tempering. These welds had lower the initial PWHT. All measurements were within ASME
carbon contents (about 0.10%) than the base metal Code and owner specification requirements except for
and there was concern that these welds may respond the roundness. Both shell rings were found to be out of
to the heat treatments in a different manner than round at the center circumferential seam of the vessel.
the shell plates, which may necessitate removing the The difference between the maximum and minimum
original weld metal and rewelding the longitudinal diameters was over two (2) times the ASME Code
seams. allowable of 1% of the nominal diameter for this vessel.
The correction of the out of roundness condition is
b. Weld tests on the overlay to confirm that the described in the Vessel Repair section of this paper.
overlay could be welded without cracking due to
sigma phase embrittlement. The non-destructive examinations performed on the
vessel showed no physical damage or cracking in the
7. Engineering Evaluation of Top Head. The vessel materials, welds or the stainless weld overlay as a
preliminary hardness readings at the refinery indicated result of the external fire to the vessel or rapidly cooling
several low hardness zones on outside of the top head. the vessel from extinguishing the fire.
Since there was excess thickness of material in the
heads, engineering calculations were performed to Hardness readings were taken on outside surfaces
determine if the material could be reclassified as the from the locations in high and low hardness areas on
lower strength SA 387 Gr. 22 CL 1 material. This about 12" (305 mm) grid lines. Several hardness readings
approach would only be accepted if the metallurgical were also taken on the inside surfaces of the vessel,
evaluation on boat samples indicated that the heat opposite the hard zones and soft zones on outside
affected material is suitable for continued operation in surfaces. The weld overlay was stripped off for about 3"
elevated temperature service. x 8" area to obtain the hardness readings on inside
surfaces. Table 4 summarizes the results of the hardness
8.Information From the Original Materials Supplier. readings. Areas were found that were above and below
Although metallurgical studies on boat samples and tests the maximum and minimum hardnesses of 225 and 155
would provide much of the necessary information on the HBN.
vessel material, it was also be important to obtain as
much information as possible from the original material Based on the results of the hardness tests, twenty
supplier to establish the proper heat treatments for (20) boat samples were cut from hard zones, soft zones
restoration of the materials properties. The information and unaffected zones to evaluate the metallurgical
needed was mainly the optimum austenitizing properties. Most boat samples were taken from the
temperature, minimum cooling rate and tempering outside surfaces of the vessel. Several boat samples were
curves to successfully accomplish the restoration heat cut also from the inside surface of the vessel, opposite
treatment. either the hard zones or soft zones on the outside
surfaces. Metallurgical evaluations included Rockwell
Vessel Evaluation at the Manufacturing Facility HRB or HRC hardness testing and optical and
SEM/TEM microscopy evaluation of selected samples.
After the vessel was received at the vessel Evaluations included testing. after the quenching and
manufacturing facility, the diagnostic work for evaluating tempering heat treatments and final PWHT. Table 5
the damaged vessel proceeded as rapidly as possible shows the details on boat sample location and
identification. Most of this work was completed at the a 30 HRC (285 HBN) hardness on the surface and 90
University of Tennessee with additional work by the HRB (185 HBN) at a depth of about 1" (25.4 mm). The
repair organization at its welding facility in Houston, microstructure clearly shows that the material was
Texas. heated to above the critical temperature during the fire.
Thus, the shell ring had been re-austenitized at this
Metallurgical Evaluations location and subsequently quenched during fire
extinguishing operations. A boat sample from this
The metallurgical evaluation of the fire damage location was given a simulated PWHT (tempered) for 5
vessel was conducted contemporaneously with the repair hours at 1275°F (690°C), and the resulting
efforts since the importance of the metallurgical microstructure is shown in Figure 5. This is clearly a
determinations were critical to the decisions to reheat tempered structure, and the hardness of 85 HRB (165
treat portions of the vessel to insure that the vessel met HBN) demonstrates how the hardness decreased upon
the Code requirements. simulated PWHT. This demonstrated that the fire
hardened region effectively can be softened by the
Since the vessel was involved in a fire, some areas of simulated PWHT. However, in some regions (adjacent
the vessel were heated to high temperature and to the material which was re-austenitized and quenched)
subsequently quenched during fire extinguishing the shell ring was "overtempered" by being subcritically
operations, the metallurgical condition of the vessel heated in the fire and the hardness was reduced below
varied with location. The hardness survey revealed that indicative of Class 2 properties. Such reduced
significant differences in hardness particularly on the hardness (low strength) regions cannot be restored
outside surface of the vessel. Boat samples were without a complete reheat treatment involving
removed for metallurgical evaluation and to determine austenitizing, quenching and tempering. Thus, it was
the efficiency of the heat treatment procedures selected necessary to reheat treat the shell rings and a nozzle
for optimum repair. The majority of the boat samples forging. Figure 6 shows the microstructure of the base
were obtained after the vessel had been given 1200° F metal after the reheat treatment (Q&T).
(649° C) /2 hours stabilizing/ dehydrogenation removal
heat treatment. The shell rings were joined by GMA welding in the
original construction. Because the carbon contents of
The metallurgical evaluation involved determination weld metal tends to be less than that of base metal,
of the hardness distribution through the depth of the particular attention was paid to the heat treatment
boat samples by micro and macro hardness methods and response of the weld seams during the complete reheat
the evaluation of the optically resolved microstructure. treatment of the shell. The mechanical properties
A number of the boat samples were given additional determined from coupons heat treated along with the
heat treatments to determine the heat treatment shell and the weld microstructure revealed that these
response or to place the material in a simulated final weld regions would meet Class 2 properties.
fabrication condition. These included (1) simulated
PWHT, (2) quenching and tempering, and (3) Similar studies were accomplished on the heads,
quenching, tempering and PWHT. A selected set of nozzles, shell skirt stub and piping, in addition to the
samples was evaluated for carbide status by STEM and stainless steel weld overlay, to insure that the repair
TEM metallographic techniques in order to determine methods selected would result in a vessel with a suitable
the final fine scale microstructural morphology. metallurgical condition.
The following examples of the metallurgical The inside surface of the vessel was weld overlaid
evaluation illustrate the work accomplished to validate with a 309 stainless steel deposited by a fusion welding
the repair methods. The typical quenched and tempered process. The original vessel specification required an
2% Cr - 1 Mo base metal shell ring microstructure is upper ferrite limit of 8FN; however, the measured
shown in Figure 3 at 400X. The material clearly exhibits ferrite number was 1.2 which reflected the original
a quenched and tempered structure, and the hardness of vessel PWHT. The decrease in ferrite number is due to
85 HRB (165 HBN) reflects the original undamaged the formation of carbide and sigma from the ferrite
microstructure. In contrast, the OLM (optical low during PWHT. Samples which received the 120(f F
magnification) shell plate microstructure from a fire (649° C) /2 hours stabilization and H 2 removal treatment
damaged area is shown in Figure 4. The hardness showed a ferrite level of 0.2-0.7FN. This lower ferrite
distribution in this region shbwed a gradient from the level is a res uh of the additional time at 1200° F
surface to a significant depth(> 1") (> 25.4 mm) having (649° C). After austenization and quenching, the ferrite
rose slightly to 2FN; but after tempering (1250°F microstructural assessment showed favorable carbide
(677"C)/5 hours) and PWtlT (1275°F (690°C)/6 types and distributions in all low hardness regions
hours), the ferrite was gain reduced to 0.2FN. The examined.
typical microstructure for a "sample exposed to the 6
hour /1275" F PWHT is shown in Figure 7 etched with The skirt stub showed an adequate response to
modified Murakami's reagent. The microstructures reduction in hardness after the fire exposure and was
reveals the remnants of delta ferrite which was suitable for the re-attachment of a new skirt
substantially transformed to both carbide and sigma.
The ferrite/transformation products are not fully The stainless steel weld overlay responded as
interconnected, as would be expected from the anticipated because of its low initial ferrite level (FN 7-
calculated ferrite content for the original weld deposit. 8). The transformation to carbide and sigma would not
The interface between the weld overlay and the base cause further changes in the properties over that for the
metal reveals the normal structure for a weld overlay original vessel.
surface.
In total, the decisions to undertake the more
In order to determine the original ferrite level of the extensive re-austenitizing heat treatments for the shell
weld overlay and to determine if the weld overlay was and nozzle N2, and the decisions to utilize heads, skirt
susceptible to cracking in the HAZ caused by repair stub and external piping in the as-existing (and PWHT)
welding, several GTA spot welds were made on the weld condition were well supported by the metallurgical work.
overlay surface. The remelted weld overlay, using a
GTA spot, had a ferrite number of 7. No cracking was
seen in the fused zone or in the surrounding HAZ. CONCLUSIONS FROM DIAGNOSTIC WORK
Thus, the original weld overlay was expected to exhibit AND EVALUATIONS
suitable weldability for the vessel repair. The fact that
the carbide/sigma product created from ferrite during Shell Rings
PWHT is not interconnected indicated that the weld
overlay will perform as it had during the vessel 1. Brinel! hardness readings and metallurgical
operation prior to the fire. evaluations on the OD and ID surfaces indicated
that the shell rings were affected by the fire,
The decision to reheat treat the shell rings and particularly the 180° circumferential segment which
nozzle N2 was supported by the metallurgical included the N2 nozzle of the vessel. The vessel
evaluations. Mechanical tests performed on the test material, to a depth from the surface to about 1"
coupons demonstrated that the vessel shell and the (25.4 mm), had been heated at various locations to
nozzle responded well to the reheat treatment resulting near or above the upper critical temperature and
in "as-new" material for these parts of the vessel. The either quenched by fire water (producing hard
tensile specimens, after reheat treatment, showed zone), or air cooled producing a soft zone. Test
fracture in the base metal, and the Charpy impact tests results showed that the required properties for the
met the required values. The weld metal responded 21/4 Cr - 1 Mo material and the damaged
favorably to reheat treatment primarily because of its microstructure could not be resorted by postweid
favorable carbon content (0.10% ). The nozzle material, heat treatment alone.
while not available for testiAg after reheat treatment,
was predicted to be similar in behavior because of a 2. Boat samples were heat treated to determine that
similar chemical composition. This was also Class 2 mechanical properties of the 21/4 Cr - 1 Mo
demonstrated by the general microstructure and suitable base material could be restored by re-austenitizing,
types and distribution of cartides. water quenching and tempering. These test results
also indicated that the material would meet
Engineering calculations on the top head showed mechanical property requirements after 11 hours of
that it could be utilized with Class 1 properties and that postweld heat treatment at 1275° F (690" C). Thus,
any fire affected material in this area would be suitably the PWHT time was sufficient for two (2) final
tempered to reflect hardness ievels less than 225 BHN. postweld heat treatments (one reserved in case of
Further, fire exposed material which experienced repair). H was concluded that the lower and upper
subcritical thermal exposure was not reduced in shell rings of the vessel had to be austenitized,
hardness and strength sufficiently to cause a property quenched and tempered to restore mechanical
decrease below Class 1 properties. Also, the properties.
3. Each shell ring also included two longitudinal seams at 1275°F (690°C). Based on these results, the stub was
that were welded by GMAW. Boat sample test data considered acceptable for continued service.
revealed that the welds could be austenitized,
quenched and tempered in the same manner as the Stainless Steel Weld Overlay
base material, resulting in mechanical properties
that were adequate after PWHT of 1275°F (690°C) Evaluation of the type 309 weld overlay indicated
for 11 hours. that the weld overlay was acceptable after addition
reheat treatment (Q&T) and PWHT. The evaluation
Shell Nozzles also showed that the overlay would not be significantly
embrittled by the vessel heat treatment.
Of the two nozzles located in the shell rings of the
vessel, the small N4-2 nozzle (3"Q,>) (see Figure 1) was VESSEL REPAIR
not affected by the fire. The evaluations indicated that
the properties would be adequate after final repair Renair Plan
PWHT. However, the large nozzle N2 (20"Q>) in the shell
was in an area that was affected by the fire, as indicated After completion of the diagnostic work and
by the high hardness readings and the metallurgical evaluation of the results, the next step was to prepare a
evaluation of the boat samples. This nozzle required detailed plan for the restoration of the vessel. The plan
austenitizing, and quenching and tempering to needed to address a number of issues and be in strict
adequately restore its properties. accordance with the requirements of National Board
Inspection Code, the ASME Boiler and Pressure Vessel
Ton and Bottom Heads Code, and good engineering practice.
Testing indicated that the top and bottom heads The repair plan included the following major items:
were unaffected by the fire, and the material would be
metallurgically acceptable after 11 hours of PWHT at " Identify the components which needed to be
1275°F (690°C). repaired or reheat treated.
Restoration vs. As-ls Justification vessel components after cutting it apart, the vessel was
postweld heat treated at 1200°F (649"C) for 2 hours
It was important to insure that the repaired vessel prior to disassembly.
exhibited properties which were in complete Code
compliance and be satisfactory in terms of fitness for For disassembly, the vessel was set on weld rolls for
service. However, to minimize the amount of repair handling purposes. The vessel was disassembled as
work and meet the required reinstallation schedule, it follows:
was important to assess each item with regard to restore
or reuse in as-is condition. 1) The top and bottom heads were removed by oxy-
fuel cutting. A preheat of 300° F (150° C) minimum
The results of the di~gnostic work described was applied prior to all thermal cutting. The cuts
previously placed each component in one of the were typically completed through the center of the
following four categories, based on the action required weld. Essentially all of the old weld metal was
to make the component acceptable for continued removed as part of the weld bevel operation for re-
service: assembly in a later stage of the project. After the
heads were removed, a stainless steel baffle (the
1) Components which needed Quenching and only internal in the vessel) was removed by carbon
Tempering to restore the material's strength and/or air arc gouging.
metallurgical properties.
2) The lower shell ring was then separated from the
2) Components that no longer had Class 2 material upper shell ring using the oxy-fuel cutting process.
strength (75 ksi minimum ultimate tensile strength), This was completed using a "Saw Tooth Burning
but had acceptable metallurgical properties, and Technique", where the vessel was shortened by 7
could be shown by calculations to be acceptable inches (178 mm) to provide test coupons for the
based on Class 1 material strength (60 ksi minimum quenching and tempering heat treatment and the
UTS). subsequent postweld heat treatments. The cuts were
made in a manner where essentially all of the
3) Components that still retained original Class 2 coupon test material was removed from the upper
strength and had satisfactory metallurgical shell ring. The test material included two
properties. longitudinal GMAW welds so that they could also
be mechanically tested after the quench and temper
4) Components which needed to be completely operation and simulated PWHT.
replaced.
Figure 8 shows the shell rings as cut apart for the
Table 6 lists the components of the vessel and the Q&T heat treatment. The cut line leaves a pair of
required remedial action for each component. prolongations for test coupons in each ring. One
prolongation in each ring contains a portion of the
Based on the diagnostic investigation, both shell longitudinal weld. Figure 9 shows the size and
rings and nozzle N2 needed to be quenched and location of the test material (coupons) attached to
tempered to restore their original properties. A repair a shell ring. Each prolongation was buffered to
sequence was then generated that involved cutting insure that its cooling rate was representative of the
the vessel apart to the degree to accomplish shell ring.
the reheat treatment (Q&T) of the three cornpc)ne.nts
and then reassembling the vessel. 3) The two shell nozzles (N2 and N4) were removed
from the upper and lower shell rings, respectively,
by the cutting process. The N2 nozzle
would reheat treatment and the N4 nozzle
When sufficient diagnostic work was completed on would only require to be welded back into the lower
the vessel to determine the basis of the repair, shell ring after heat treatment (Q&T) of the shell
disassembly of the vessel for repair was started. The rings.
external fire and the non-uniform heating and cooling of
the vessel introduced thermal gradients and residual
stresses in the vessel shell. In order to reduce the risk of
residual stresses in the vessel causing distortion to the
Heat Treatment of Fire Damaged Comnonents second shell ring was then placed back in the furnace
and held at the austenitizing temperature for
As stated previously, the upper and lower shell rings approximately one (1) hour prior to water quenching.
as well as the large N2 nozzle required quench and The aim temperature at immersion was 1680°F (916°C).
temper heat treatments in order to restore the The shell rings were maintained at 100° F minimum
mechanical and metallurgical properties of the damaged temperature until the start of the tempering heat
2% Cr - 1 Mo material. treatment. This was to minimize the possibility of
quench cracking due to the hard untempered
The two shell rings were separated from one microstructure in the 2% Cr - 1 Mo material.
another to allow them to fit the furnace and in the Thermocouples were attached to the shell rings to insure
quench tank in an upright position and permit an that proper temperatures were maintained during the
effective quench from both sides. To avoid potential austenitizing and the water quenching process.
distortion of the shell rings due to uneven cool down
during the water quench, all nozzle forgings were Ranges of acceptable hardnesses were developed
removed from the shell rings and the cutouts sealed with after each step in the Q&T process to monitor the
temporary baffles thus making the rings as axisymetric response of the material to heat treatment. The desired
as possible. A small hole (about 1/4" in diameter) was minimum hardness after quenching was about 280 HBN.
placed in one baffle of each nozzle to avoid damage of The average measured hardnesses were about 300 HBN.
the baffle from heating and expansion of air in the air
space between the baffles. The quenching operation was followed by tempering.
The shell rings were initially tempered at 1200°F for 5
To minimize the risk of quench cracking during the hours and cooled in water. The hardness after this
water quench from austenitizing, all cut edges were operation was found to be harder than 230 HBN
blend ground to remove any burning serrations and (equivalent to approximately 112 KSI) which would
notches, and all corners were ground to a 1/4" (6 mm) make it difficult to fit and weld. The shell rings were
minimum radius. given a second temper at 1250° F (677° C) for 3 hours to
lower the hardness below 230 HBN, prior to proceeding
The furnace used for austenitizing the shell rings is with testing of the test coupons and further fabrication
a large natural gas fired furnace of approximately 35 ft. of the shell rings.
(10.7 M) x 35 ft. (10.7 M) x 100 ft. (30.5 M) long. The
water quench tank was 16 ft. (4.9 M) diameter x 16 ft. The N2 nozzle was quenched and tempered
(4.9 M) in height. The quench tank had a special following similar procedures in a smaller furnace and
arrangement of internal piping to provide proper water quench tank, except that a different procedure was used
circulation in the tank for adequate cooling of the shell for test coupon since no test material was available from
rings during the quenching operation and to maintain this forging. The decision was made to use a piece of
contact with the shell rings without forming steam the shell ring test band attached to the forging and
pockets. A special 3 point ring handling device was built buffered. The forging and plate material were
for the project for handling the rings from the flat sufficiently similar in composition to justify this
bottom furnace car to the water quench tank. The approach.
handling of the shell rings fro:.n the time of opening the
furnace door to lifting into the quench tank was
rehearsed prior to the actual austenitizing heat
treatment of the shell rings. It was determined that the
maximum transfer time from opening the furnace door Test Coupons
to submersing a shell ring into the quench tank was 3
minutes to avoid cooling the rings below the desired
minimum quench temperature. Figure 10 provides
details of the quenching operation. Test coupons from the base metal and the narrow
groove GMAW welds were tested after the quenching
Both shell rings were charged into the austenitizing and tempering of the shell rings and nozzle N2 prior to
furnace concurrently. The shell rings were austenitized welding the components back together. Coupons were
at 1750°F .±_25°F (954°C .±_14°C) and held for 5 hours tested at the repair organization's Corporate Welding
minimum. The furnace was opened and the first shell Laboratory.
ring was lifted and submerged into the quench tank. The
Prior to mechanical testing, each test coupon was typically met a 40 ft-lb mllllll1um absorbed energy
stress relieved at 1200° F (649" C) for 10 hours and at transition temperature below -20°F (-29°C).
1275°F (690°C) for 10.5 hours simulating the maximum
amount PWHT the vessel could receive during the 2) Stainless steel overlay welding was completed with
repair cycle. This provided sufficient time on the the SMAW process. E309L electrodes were used for
coupons for the final PWHT and a repair cycle. all weld layers. The first layer E309L electrodes had
Mechanical tests included tensile tests (at 1/4t depth, a "high" ferrite content (9 - 14%) to compensate for
0.500" diameter type per ASTM A370) and full size dilution of the base material in the first layer. Type
Charpy V-Notch toughness specimens (also at 1/4t E309L electrodes used for the second and remaining
depth) tested at +20°F (-7°C) and at -20°F (-29°C). layers were purchased with a ferrite content of 5 to
All specimens were taken from a distance at least one 9%. The desired result was to have ferrite content
(1) t from any quenched edge. in the deposited overlay of 10% maximum as
determined with a Severn Gage to meet owner's
Tensile tests included the base material and requirements and minimize the amount of
transverse tensile specimens of the GMAW welds. CVN transformation of the ferrite to the brittle sigma
testing included the base material, the weld metal and phase from PWHT.
the heat affected zone of the welds. All tests results
were acceptable with an ultimate tensile strength in the
range of 79 to 82 KSI (545 to 565 MPa) and a yield Preheat and PWHT
strength of 59 to 64 KSI (407 to 441 MPa) for both base
metal and weld metal. All CVN results exceeded the A preheat of 300°F (149°C) minimum was used for
owners requirements of 40 ft-lbs average/35 ft-lbs welding with 2% Cr - 1 Mo welding consumables.
minimum (for one specimen) at +20°F (-7°C). Base Preheat was maintained from the start to finish of the
material specimens exceeded 100 ft-lbs when tested at welding operation and until the start of intermediate
both +20°F (-7°C) and -20°F (-29°C). postweld heat treatment (IPWHT). However, the
preheat was lowered to 200°F (93°C) minimum for all
welding with type E309L welding consumables. IPWHT
Welding Processes and Filler Metals at 1200° F (649° C) was completed prior to any
volumetric inspection of the completed weld joints. After
Welding Procedures Specifications to ASME all welding was completed on the vessel, it was given a
Section IX and owner requirements were completed final postweld heat treatment at 1275°F (690°C).
concurrently with the diagnostic testing and evaluation
of the vessel materials. The following welding processes
and filler metals were used: Nondestructive Examination
1) The SMAW process was used for base metal welds Heavy wall, 21/4 Cr - Y2 Mo, ASME VIII, Division 2
joining 21/4 Cr - 1 Mo together and for repairs to the Vessels require extensive NDE. The repaired vessel
2% Cr - 1 Mo base material. The SAW process was received even more NDE due to concerns arising from
used for welding the thre9 (3) circumferential seams its exposure to fire.
on the vessel. E9015-B3 :electrodes were used with
the SMAW process and EB-3 electrodes were used ASME Code requires 100% radiography (RT) of
with an agglomerated basic fluoride type flux for the butt welds and 100% magnetic particle (MT)
SAW process. Welding filler metals were selected examination of all repair welds. The additional NDE for
that have low amounts of impurities (residual this vessel included 100% MT of all root passes and
elements) to provide high resistance to temper backgouged surfaces, 100% ultrasonic examination (UT)
embrittlement All 2% Cr - 1 Mo materials were of butt welds both after the intermediate PWHT and
batch tested after minimum PWHT of 2 hours at after final PWHT, 100% MT of all outside weld surfaces
1275°F (690°C) and minimum PWHT plus step- before and after final PWHT, and 100% liquid
cooling. penetrant test of all inside stainless overlay weld surface
after the initial PWHT at 1200° F.
The consumables tested exceeded the owner's
requirements of 40/35 ft-lbs at +20°F (-7°C) after Table 2 provides a summary of all the NDE
minimum PWHT and 40/35 ft-lbs at +50°F after performed on this vessel at the vessel repair contractor's
minimum PWHT plus step-cooling. The consumables facility.
An alternate approach was adopted when it was 4) Final PWHT the vessel at 1275°F (690°C) and
determined that the size of jacking equipment required complete the required NDE. Complete surface
to bring the shell rings within the 1% criteria would be hardness checks to confirm the required minimum
excessive. ASME VIII, Division 2, Article AF-130.3 and maximum hardness levels have been achieved.
permits out-of-roundness greater than 1% if stress
analysis is performed on the out-of-round geometry to 5) Hydrostatically test the vessel and complete the final
demonstrate that stresses in the out-of-round vessel still NDE. Figure 11 shows a photograph of the vessel
conform to all Code allowable stresses. The final prepared for hydrotesting.
diameter difference after re-assembly of the vessel was
a maximum of 15/16" (24 mm). Stress analysis showed
that the remaining out-of-roundness was acceptable. Documentation
Membrane hoop stress due to internal pressure was
determined to be 20.1 ksi (130 MPa). The additional The final component of repair was the proper
bending stress due to out-of-roundness resulted in a documentation of the repair. Planning this activity began
total maximum surface stress of only 30.4 ksi (210 MPa). even before the physical repair work commenced. Initial
repair plans were written and submitted to the
Since the bending stress is classified as secondary, responsible Authorized Inspection Agency based on the
the Code permits total stress intensity as high as three initial diagnostic work done at site. The initial repair
times the basic Code allowable (3 x Sm)· The resulting plan was refined as more detailed information became
allowable stress intensity is 3 x 22.28, 66.8 ksi ( 461 available and was eventually submitted to the
MPa) at the maximum design of 810° F Authorized Inspector in the form of a formal
(432°C). Plan in conformance with NBIC Chapter
Supplement 4, Section C. The Repair Plan document
included description of diagnostic work, diagrams
documenting the data collected from hardness
and boat testing, damage assessment, the
The re-assembly of the vessel started immediately sequence, detailed drawings for the work, and detailed
after completion of testing of the test coupons from the procedures for welding, heat treatment, NDE, and
reheat treated shell rings and the N2 shell nozzle. material verification sampling and testing.
The vessel was final assembled in the following steps Two supporting documents were referenced in the
and shipped to the owner. Each welding operation was Repair Plan, namely the metallurgical reports by the
followed by IPWHT and NDE examination. contractor and by the University of Tennessee,
WRCTu~etin 525
40 Fabrication and Repair of Low Alloy Steel Pressure Equipment
The Welding Research Council, Inc.
FINAL RESULTS AND CONCLUSIONS 4. Wada, Isuguyasu and Eldis, G.T., "Transformation
Characteristics of 2 1/4 Cr - 1 Mo Steel •, Application
1) The 2% Cr - 1 Mo vessel was successfully repaired. of 2 1/4 Cr - 1 Mo Steel for Thick-Wall Pressure
The complete repair sequence, diagnostic work Vessels, ASTM STP 755, G.S. Sangdahl and M.
through hydrotest and shipping took slightly less Semchyshen, Eds, ASTM, 1982, pp. 255-274.
than 3 months to complete.
5. Lundin, C.D. and Wang, Y., •Repair Welding of
2) The repair of the fire damaged vessel demonstrated Service Exposed Cr - Mo Steel Weldments•, WRC
the foll~wing evaluation and repair techniques: Bulletin 346, November 1989.
" The use of boat samples can be used to 6. Guidelines for Assessing Fire and Explosion Damage,
evaluate the metallurgical condition of the MTI Publication No. 30, 1990 (Materials Technology
material and even more importantly, to predict Institute of Chemical Process Industries, Inc., St. Louis,
the effect of heat treatments that Will be Missouri).
required during repair. The boat samples also
provide material for optical, SEM and TEM 7. Sumitomo Steel Corp., Private Communication.
microscopy and for hardness testing of the boat
samples in various heat treatment conditions.
_J
w
(/)
(/)
w
>
w
a:
::::>
0 (/)
z (/)
i cc
w
CV :l
w
J:
(/)
c:
a:
a..
w
5: C\J
0
d
J: ~
I- ()
(!)
z
w
...J
I-
(!)
z
a: C\J-
z
w
C\J
(!)
~ ClVlO ·a·1
(WWl>f:~G)
,r:r,L
'E
E
'E O>
E
IO ....
C\I
c;:; ~
~ ~
IO
9 IO
~
""
C\I
8
z
cc
...J '$
...J 0C\I
w
J: c(i
(/)
a: z
w~
w ...J E
0...
fl. j::j E
2. o~
C\I z !£?,.
'!!::
CJ
z
a:
/~l
/ e·1 OXY-F\JEL /
CUTUNES ~
/'
_J
. ·10=J
U~ I
RING 12 (UPPER SHELL RING)
O!
I 1 - - - - - - - - - - - - - -.......... :......................Rl....N.. G#l (LOWERSHEllRING)
MANWAY Ml-18' 0
(457mm)
NOZZLE N2-20' 0
(506mm)
T- OF SHELL RING
11 Photograph of
77mm Head 'I 688°F(920°C); 1202°F (650°C) 1310°F{710°C) 37.25x10 3 77.9 159 Unavailable
Plates 0. 75 hrs/in. (Cool for 10.4 hrs. for 10.5 hrs. {537 MPa)
36°F {20°C)/min. (Cool at 100°F (Cool at 100°F
above 752°F (55°C)/hr. above (55°C)/hr. above
(400°C)) 800°F (426°C)) 800°F (426°C))
129mm Shell 1688°F(920°C); 1 202°F (650°C) 1 31 0°F(710°C) 37.25x10 3 75.7-77.7 155-161 Unavailable
Plates ·1 hr/in. (Cool for 10.4 hrs. for 10.5 hrs. (522-536
36°F (20°C)/min. (Cool at 100°F (Cool at 100°F MPa)
above 752°F (50°C)/hr. above {55°C)/hr. above
(400°C)l 800°F (426°C)) 800°F (426°C))
Required By:
Time of Exam Area to be Examined Type of
Exam
Code Fab Other
• Two small low hardness areas on the bottom head (outside and inside)
• Two areas (hard zones and soft zones) on the lower shell ring (outside and inside)
• Three areas (hard zones and soft zones) on the upper shell ring (outside and inside)
• Outside of Nozzle M 1 in the area of the lowest hardness readings found at the vessel site
• Welded joints (weld metal and HAZ) in the areas exposed to fire
• Unaffected areas on the outside surfaces of shell plates, nozzle forgings and welds to establish
the normal hardness range (and strength) after the vessel fabrication
\.·.·< >.·······•··.·.
.· ...·
·.·iv·•· <···. <.···.
•..••• ·.. ·.······.::: .:::> :·:·:.:.·:: ··.:··, ·:::·.:
<t >······· ·.< •.•.•••.•.•..•••.•...•.•. MAXIMUM MINIMUM
. ·•·
Shell Ring 1 (Lower Shell Ring) Outside (0-270-180 Deg.) 343 183
Shell Ring 1 (Lower Shell Ring) Inside (0-270-180 Deg.) 188 188
Shell Ring 1 (Lower Shell Ring) Outside (0-90-180 Deg.) 222 193
Shell Ring 2 (Upper Shell Ring) Outside (0-270-180 Deg.) 310 152
Shell Ring 2 (Upper Shell Ring) Inside (0-270-180 Deg.) 180 187
Shell Ring 2 (Upper Shell Ring) Outside (0-90-180 Deg.) 265 187
TABLE 5
LOCATION ID DESCRIPTION
17-2 (Out) No. 1 Shell Ring Outside (Hard Spot) (Near 330°) Location)
17-2 (In) No. 1 Shell Ring Inside (Opposite 17-2 Outside)
23-5 (Out) No. 1 Shell Ring Outside (Soft Spot) (Near 300° Location)
23-5 (In) No. 1 Shell Ring Inside (Opposite 23-5 Outside)
23-16 (Out) No. 1 Shell ring Outside (Normal Surface Hardness) (Near 225° Location)
18-21 (Out) Vert Weld No. 1 Shell Ring (Outside) (60° Location)
18-21 (In) Vert Weld No. 1 Shell Ring (Inside)
4-2 (Out) No. 2 Shell Ring Outside (Soft Spot) (Near 330° Location)
4-2 (In) No. 2 Shell Ring Inside (Opposite 4-2 Outside)
7-11 (Out) No. 2 Shell Ring Outside (Near N2 Nozzle) (Hard Spot)
11-22 (Out) No. 2 Shell Ring Outside (Unaffected) (Near 45° Location)
N2 (Out) N2 Nozzle - Outside Bore - (Hard Spot)
N2 (In) N2 Nozzle - Inside Vessel (Soft Spot)
T22-2 (Out) Top Head Outside (Low Hardness (Near 45° Location)
T22-2 (In) Top Head Inside (Opposite T22-2 Outside)
B7-2 (Out) Bottom Head Outside (Low Hardness) (Near 270° Location)
B8-3 (In) Bottom Head Inside (Near B7-2 Outside) (Near 255° Location)
B-1 Bottom Head Girth Weld (Contains Shell Base Metal) (Near 0° Location)
N1 P (Out) Bottom Piping Near Stainless Steel Pipe End
S1-4 Skirt (Hard Spot) (Near 300° Location)
B. Judgement on fire affected/unaffected areas was based on surface hardness and information
on the fire
N2 (20" nozzle in upper ring) Damaged Microstructure Class 2 properties required Q&T
Soft zones and hard zones
M 1 (Manway in top head Acceptable, Class 1 properties acceptable Revise design calculation to
Class 1 strength show Class 1 properties to be
acceptable.
N4-1 (3" nozzle in top head) Acceptable, Class 2 properties acceptable PWHT after rewelding
Class 2 strength
N4-2 (3" nozzle in lower ring) Acceptable, Class 2 properties acceptable PWHT after rewelding
Class 2 strength
N1 (10" nozzle in bottom head) Acceptable, Class 2 properties acceptable Nothing required. Final PWHT
Class 2 strength with vessel
C. Lundin
K. Khan
P. Liu
C. Qiao
G. Zhou
The University of Tennessee
M. Prager
The Materials Properties Council, Inc.
would cover the majority of materials currently in use which employs an actual weld deposit This test
or those which are now offered in the market place overcomes an additional attribute of the Gleeble and
for the construction of petroleum/refinery vessels. A Spiral Notch tests in that the stress relaxes as a
total of 17 lots representing 17 different heats of function of time at the test temperature in much the
l 1/4Cr-1/:;? Mo materials were acquired from both same manner as the residual stresses relax in an
foreign and domestic sources. The time span of the actual weldment. Furthermore, the test incorporates
production of the materials was greater than 25 years. natural weld contour conditions which introduce
It was originally intended that materials from vessels realistic stress concentrations for the initiation of
which had experienced cracking problems would reheat/PWHT cracking.
become available during the conduct of the program.
Unfortunately, no such materials were made a) The Gleeble stress rupture test employing a
available. Nevertheless, it is felt that the range of simulated HAZ has been used for over 15 years
compositions and behavior obtained within the 17 in assessing HAZ cracking during PWHT. The
materials tested is sufficient to formulate approaches details of the test methodology have been defined
to vessel fabrication and repair that are meaningful by the work at The University of Tennessee and
and technically sound. other researchers [2]. Criteria have been
established and significant correlations have been
The service exposed materials were made with service performance. The test has
re-normalized at 1650°F for 1 hour and tempered at been used as a basis for the determination of
1150° F for one hour before being tested. This welding conditions necessary to minimize the
re-virgining of materials which were in the service occurrence of reheat/PWHT cracking, the rank-
exposed condition was considered important even ing of a materials sensitivity to reheat/PWHT
though the microstructure in the HAZ is changed cracking and for fundamental studies of the
(transformed and homogenized) during both the mechanisms of reheat/PWHT cracking.
simulation and actual weld tests (it was found that
the HAZ weld thermal cycle exposure may not be The Gleeble tests were carried out using energy
sufficient to erase all prior metallurgical changes inputs of 12 and 120Kj/in. to span the range of
which occur during service exposure). energy inputs utilized for a wide range of welding
procedures. The testing was carried out as a
The chemistry of all materials studied was function of stress at a temperature of 1150°F and
obtained from the same laboratories and analysis for as a function of PWHT over the range of
23 elements was made. The chemistry for the 17 1150-1350°F. The service exposed and the
heats is presented in Table 1. re-normalized conditions were used for the
ex-service materials and the virgin condition was
used for the heats of new materials. The samples
REHEAT/PWHT CRACKING ASSESSMENT were used for fractographic studies and for the
determination of the hardness of the HAZ and to
In order to determine the reheat/PWHT cracking define the hardness changes as a function of
response of the 17 heats of 1%Cr-V2Mo materials three PWHT. In addition to the ductility
test methods were employed. One of the tests, the measurements, which allow a definition of the
Gleeble HAZ simulation test [2], has been used for materials sensitivity to reheat/PWHT, cracking,
many years in assessing reheat/PWHT cracking three additional important findings were: (1) the
ter1mmc:y while the other two tests were developed at of PWHT below
The University of Tennessee to answer special needs 1250° F has only a minimal effect on the rupture
in reheat/PWHT cracking assessments. The Spiral ductility indicating that the material retains a
Notch test was developed in order to assess which sensitivity to coarse grained HAZ cracking unless
zone in the HAZ is most susceptible to a PWHT in excess of 1250°F is erntpl«::>ye:ct
reheat/PWHT cracking and to define the effect of fabrication (a PWHT greater than 1300" F
different weld metal techniques on the sigjnificar1Uy improves rupture ductility), (2) the
sensitivity of the HAZ to reheat/PWHT cracking. hardness of the coarse HAZ does not
The PREVEW (Petroleum REfinery Vessel Evaluation fall rapidly with time for temperatures to 1200"F
of Weldability) test was developed in the current and above 1200" F the hardness falls at rate
program so that the reheat/PWHT cracking considered to be slower than anticipated. This
se11s1t1v1ty could be determined from a test hardness retention was also noted in the on
the creep behavior of the coarse grained region in production, in addition to testing simulated
as a function of PWHT temperature and (3) a HAZ's for cracking sensitivity.
high energy input places the coarse grained HAZ
in a condition which is more susceptible to c) The development of the PREVEW test was
reheat than a low heat undertaken as a part of this program to provide
a correlation with the Gleeble and Spiral Notch
b) The Spiral Notch test was introduced in order to tests using a full scale weldment. This test has
more rapidly and to better define the region of the advantage of the utilization of an actual weld
the HAZ which is most susceptible to with residual stress relaxation during the PWHT
reheat/PWHT cracking in actual welds and to duplication and is thus very similar to actual
define the conditions which mitigate the weldment behavior. The stainless steel test
occurrence of reheat/PWHT cracking. The test fixture, shown schematically in Figure la with a
is conducted at 1150° F at a constant stress level sample in place pictorially in Figure lb, is simply
indicative of the residual stress in a welded constructed and no specialized equipment, except
structure. The location of the cracking and for a suitable furnace, is required to conduct the
rupture in the notched span defines the HAZ test. The specimen size, 10" x 4" x V2", shown in
microstructure most sensitive to reheat/PWHT Figure 2a, allows for actual welds using
cracking and the stress range over which coarse appropriate consumable and the employment of
grained cracking occurs is a secondary indication NDE methods for a rapid assessment of the test
of a materials reheat/PWHT cracking sensitivity. results. The slot in the through plate under the
The Spiral Notch test results on materials stem plate causes a stress distribution as
UT1-UT11 were evaluated to compare with and indicated in Figure 2b, when strained to the base
verify the Gleeble assessment. There was a metal yield strength in three point bending. The
one-to-one correlation between the data from tests were conducted on all materials whose
both tests. configuration permitted the extraction of samples
(thus not all materials were tested in this
A significant amount of work was conducted in a manner). Dye penetrant NDE is used after
PVRC study (1) on the efficacy of the use of welding, after application of the test strain and
controlled deposition procedures in mitigating after testing at elevated temperature to insure
reheat/PWHT cracking on the material that the cracking found is representative of true
designated UT6 in this study (it is to be noted reheat/PWHT cracking. The degree of cracking
that the UT6 material is the most sensitive to in a test sample is indicative of the relative
reheat/PWHT cracking of all of the materials sensitivity of the material. The appearance of
examined in this study). The PVRC study defined cracking in a fillet weld test coupon is shown in
the conditions necessary for the mitigation of Figures 3a and b. The cracks are dearly noted
reheat/PWHT cracking by controlled deposition emanating from the weld toe into the coarse
techniques. This PVRC study revealed that if the grained HAZ. The extent of cracking is
coarse grained region of the HAZ is eliminated determined from crack depth and crack length.
by grain refinement from successive weld passes The tests also showed that the hardness of the
the reheat/PWHT cracking sensitivity essentially base metal (original plate condition) was
disappears. The Spiral Notch test was important to the extent of reheat/PWHT
instrumental in this behavior for 10 full cracking, in that the harder the base
scale weldments the UT6 material. This material the more extensive the which
same elimination of reheat/PWHT cracking occurred in the coarse grained HAZ. This is due
controlled deposition coarse grained HAZ to the fact that a softer base material allows for
refinement was also shown for a low carbon more relaxation of stresses and thus
pre~c11nta1ho1n h;ardlenmg steel by the utilization of reduces the of stresses as a
the Notch test. the Notch function of PWHT time in the HAZ. The
test was used, in an other study, to define the location of the cracking clearly showed the
differences in HAZ structure for down-hand and influence of the condition of the weld toes in
out-of-position welds. Thus, the Notch test exacerbating the initiation of cracking in sensitive
has an inherent utility in the testing of full scale material. The more abrupt the fillet weld contour
weldments duplicating the procedures to be used the greater the cracking tendency (the change
from the base metal to weld metal should be 2) a tramp element function encompassing the
smooth and gradual). following:
Tramp= 2[4.3(Sn+As) + 150Sb+Cu+50(P-0.01)],
It is felt that the development of the PREVEW 3) a sulfur function: Sfn = 1+30(S-0.02Tramp);
Test is significant, as a verification test, as a test that For Sfn < 1, Sfn = 1,
can be utilized for the screening of materials, and for 4) an aluminum function: Alfn = 1+15(Al-0.015);
the evaluation or selection of welding procedures For Alfn < 1, Alfn = 1.
which influence reheat/PWHT cracking.
MPC Factor-5 = [Cfn(Tramp+Sfn)Alfn]-1
The reheat/PWHT cracking sensitivity ranking of
the program materials using the Gleeble and MPC Factor-7 (Figure 2) utilizes multiple re-
PREVEW test is shown in Table 2. The data is gression analysis with empirical cut-off limits placed
presented in three categories; High, Intermediate and on some elements with the contributions combined in
Low. The correlations are considered quite adequate an additive fashion.
and while more materials were rated with the
Gleeble test than the PREVEW test, the rankings are MPC Factor-7 =
consistent. The Gleeble rankings are given separately 2(C-0.12)-0.25(Mn-0.6) + 150Nb + 15V + 15Ti-100B +
for the low and high heat input evaluations and this 40(P-0.010) + 5(S-0.015) + lO(Al-0.010) + 20(Cu/100 +
illustrates that at the high energy input conditions no Sn/3+As/3+3Sb)
material ranks in the "Low" category. This ranking is
of significant value when the assessment of the It is felt that the sensitivity to reheat/PWHT
sensitivity, based on chemistry of the base material, cracking is reasonably assessed by either of these
is considered in the following section. factors and a good correlation exists with fabrication
behavior. Using MPC Factor-5, the limiting value for
the onset of a potential for reheat/PWHT cracking is
DEVELOPMENT OF FACTORS TO PREDICT 2.0 and for MPC Factor-7 the limiting value is 0.5.
REHEAT/PWHT CRACKING Thus, the evaluation of a material in regard to its
reheat/PWHT cracking behavior can be made
The literature presents at least 10 factors relating utilizing either of these Factors. These Factors can be
the behavior of Cr-Mo materials to heat chemistry. used to assess the chemistry effects on cracking
At least 8 of these factors are related, in some way, potential and specifications for material chemistry
to reheat/PWHT cracking. These factors are can be derived from these Factors.
presented in Table 3. The assessment of these factors
for the 17 heats showed that the extent of scatter was Correlations between the MPC Factors and the
too great to utilize any of the literature factors as a PREVEW test results are illustrated in the
definitive index of reheat/PWHT cracking sensitivity. presentation in Figure 6, which shows the maximum
Thus, a series of chemical element factors based on crack depth in the PREVEW test as a function of
multiple regression analysis and mechanistically MPC Factor-7. Clearly, the materials which fall to
related criteria was undertaken. From these efforts a the left of the 0.5 cut-off for Factor-7 exhibit no
series of Material Performance factors based on cracking, while those to the higher Factor values
Composition, MPC Factors, for predicting behavior show cracking in the PREVEW Test The same
for reheat/PWHT cracking sensitivity evolved. Two correlations can be shown for MPC Factor-5.
of the Factors offer the best assessment and are
shown in Figures 4 and 5, using the average Gleeble Unfortunately, no "problem" materials were
reheat cracking ductility as a function of the Factor received from cracked vessels. Literature descriptions
(Factors 5 and 7). [6,7] of four reactors which experienced HAZ
cracking were used to provide a check on the factors
MPC Factor-5 contains four groupings of the described above. The elemental analysis for 4
compositional elements in regard to mechanistically "problem" materials is shown in Table 4. The entire
related functions: spectrum of elements was not available in the
literature reporting and for the elements that were
1) a carbon function which includes the basic not reported a value representing the lowest
strengthening elements in addition to carbon: detectable limit was used in calculating the factor
Cfn = ( 5C + lOOONb + lOOV + 50Ti-0.5) + 1, (shown in Table 4). The "problem" materials are
shown at an arbitrary selected ductility of 3% as "X's" Heat inputs studied were 45, 80 and 120 Kj/in
on Figures 4 and 5. It is dear that these materials while PWHTs evaiuated ranged from 1150 to i375°F.
fall to the high side of the factors describing The materials used were UT2, UT3, UT4, UT5 and
reheat/PWHT cracking behavior. Indeed, if the UT8. These ranged from excellent (UT2) to poor
actual chemistry was available for the unreported (UT3 and UT4) in weldability and included marginal
elements in these "problem" heats, the points would materials as well (UT5 and UT8).
fall at higher Factor levels. Thus, the selected factors,
MPC-5 and MPC-7 appear to be relevant to reported The smooth/notch bar study corroborated the
in-service cracking occurrences. conclusions of the reheat cracking study and provided
insight into some cause and effect relations. At the
test stresses and temperatures the materials displayed
NOTCH BAR AND SMOOTH BAR CREEP very low ductility, despite tempering. The effect of
TESTING OF THE COARSE GRAINED HAZ tempering on creep rate and life was very small.
Notch sensitivity appeared to decline as ductility
The basis for the smooth and notched bar increased and hardness decreased, the superior
stress-rupture study was the concept that long term ductility of low heat input materials was attributed to
cracking in service was a manifestation of notch differences in transformation products. It is apparent
sensitivity (shortened life with stress concentration). that for a given material, a low heat input weld
Comparisons were to be made of the relative lives of provides higher ductility and creep rates. The effect
smooth and notch specimens at a selected on smooth bar life is small, but notch bar life is
temperature and stress which would prove to be improved.
discriminatory.
The extraordinarily high creep rate for UT4 in
Utilizing MPC's specially designed Gleeble the high heat input condition and the similarity to
specimens [5] valid elongations and creep rates could low heat input rates suggests unusual transformation
be obtained from the smooth bars to help understand behavior leading to similar transformation products
behavior. Then the effects of composition, heat input irrespective of heat input. This material was noted to
and PWHT might be systematically studied. Earlier have an abnormally high manganese content which
work by MPC [5] on one of the steels studied showed could favor martensite.
that the simulated HAZs displayed the classic
behavior of short notch bar life at low stresses and Ductility may also be influenced by impurities.
low creep temperatures. Heats with excessive Sn, Cu (correlated variables),
etc., tended to display low ductility after PWHT. The
For this study a number of test conditions were combination of high creep rates and low ductility
considered. It is generally recognized that notch would lead to very short life.
sensitivity decreases with increasing temperature and
increasing stress. Tests at design stresses and Tempering tended to have beneficial effects on
temperatures, however would be unduly long. The notch ratios and ductility. The behavior of materials
compromise condition selected for most tests was tempered below 1275"F was notably inferior. It is
1025°F and 20 ksi. The resulting range of test expected that markedly superior results (ratios)
durations, about 400 to 5000 hours, was practical yet would be obtained if the hardness of the HAZ could
the conditions were not far removed from those be reduced by tempering to values closer to that of
which might be encountered in petroleum industry the base metal. This appeared to occur most readily
service and lead to failure in a time frame of 5 to 10 with the low carbon material UT2 and when low heat
years. For example, on a Larson-Miller Parameter input was used.
basis (using C = 20), 5000 hours at 1025° F equals
about 10 years at 950°F. Slightly higher and lower The reheat/PWHT cracking and creep rupture
stresses (25 and 16 ksi) and higher temperatures were evaluations dearly showed that the high energy weld
examined to determine if there was sensitivity to heat inputs resulted in lower rupture ductilities than
stress or temperature. It was concluded that the iow heat input conditions. It was aiso found from
increasing stress or temperature provided less the creep testing of the coarse grained HAZ that
discrimination among the heats while decreasing higher creep rates were evident for the low energy
stress unacceptably prolonged tests for some input conditions even though the initial HAZ
materials without changing the ranking. hardnesses were high (indicating a stronger material).
There are literature reports that indicate that higher the weld heat affected zones of 11/.o1Cr-Y2Mo steel
temperature transformation products such as bai.nite respond relatively slowly to PWHT and may display
result in a stronger material in the creep range than low ductility at elevated temperatures. Ductility
the lower temperature transformation product depends on material composition, weld heat input
martensite. Extensive OLM, SEM, STEM and TEM and PWHT conditions. While these qualitative
work on the high and low heat input samples as a characteristics were not surprising the quantitative
function of PWHT temperature clearly showed that details which emerged from the study were. For
the carbide evolutionary sequence is distinctly example:
different for the two microstructures. For a
martensitic structure the carbides evolve to the more a) HAZ creep ductilities among the materials varied
stable forms such as M 7 C3, M 23 C6 and M 6C more by a factor of ten.
rapidly as a function of time at equivalent
temperature than do the initially bainitic HAZs b) Coarse grain HAZs of high carbon materials
characteristic of high heat input weld conditions. The tended to display low creep ductility, perhaps
high heat input weld conditions result in the only a fraction of 1% to failure, even after
formation of M 3 C and M 2C type carbides which PWHT.
persist for long times and thus resulting in
strengthening of the material in the elevated c) For a given heat input and hardness, creep rates
temperature regime. The decrease in hardness, upon of coarse grain HAZs varied by as much as a
PWHT, for the coarse grained HAZ microstructures factor of 10 depending on composition
is less than anticipated from studies of normally heat (transformation microstructure).
treated base metal and thus the maintenance of
strength at higher temperatures speaks to the need to d) The creep ductilities of some heats were
PWHT at high temperatures to place the material in improved significantly by heat treatment while
a more creep ductile condition. These observations in others reached a plateau and remained relatively
several materials with widely differing reheat/PWHT notch sensitive.
cracking response are not in conflict with the
traditional assessment of reheat cracking tendency. e) Smooth bar and notched bar stress-rupture lives
The cause of reheat/PWHT cracking most probably of the materials were found to vary by as much
does not lie in just only one aspect of material as a factor of ten.
behavior and thus the approach to the MPC Factors
with the mechanistic approach, incorporating a f) Creep rates and smooth bar rupture lives of
variety of effects, including strengthening, tramp simulated HAZs were only slightly affected by
elements, effect of all elements on hardenability, tempering temperatures from 1250°F to 1350°F.
carbide prec1p1tation kinetics and creep rate However, the ratios of notch to smooth bar lives
differences are all important in the ultimate and ductilities tended to improve.
causation.
g) The PWHT temperature necessary to
Elemental species such as manganese and boron significantly reduce creep notch sensitivity in the
appear to influence creep behavior (increase creep HAZs varied among the heats by as much as
rate) by enhancing hardenability and martensite 100°F. Fabrication and procedures should
formation in the HAZ for equivalent levels for other take this into account
elements. This effect on creep rate may be an
important ancillary feature that leads to reduction of h) Controlled deposition techniques and low carbon
residual stress without the strain accumulation and filler metals may be used to reduce the tendency
,..,,.,1,,.""' of boundaries which are the sites for for heat treatment and service.
sej.!~rei;i,at10n of the P and S.
i) Materials which were found to display
CONCLUSIONS PWHT /reheat cracking susceptibility tend to
indicate poorer expectations for service.
A total of seventeen commercial heats were
obtained and information on others was utilized. j) The simulated HAZ (Gleeble) reheat/PWHT
Based on analysis of the behavior of more than 17 cracking test, the Spiral Notch test and the newly
heats it has been concluded that the hardest areas in developed PREVEW test can be in a
complementary manner to assess weldability and [6] Nomura, T. et. al., "Creep Embrittlement of
potentiai for in-service elevated temperature Structurai Components in Catalytic Reformer
cracking of Cr-Mo steels. Reactor," Trans. Japan Soc. of Mechanical
Engineers, 1993-9, pp. 2066-2073
k) The effects of material composition on weld
HAZ behavior during fabrication and in elevated [7] Cantwell, J., "Private Communication to M.
temperature service can be assessed on the basis Prager of MPC" May 1993.
of the testing conducted herein.
[8] Prager, M., "Development of the MPC OMEGA
1) As a result of this work the compositional factors Method for Life Assessment in the Creep
MPC-5 and MPC-7 are considered useful for Range," PVP Vol. 288, ASME Pressure Vessel &
material selection. Piping Conference, June 1994.
REFERENCES
MATERIALS c Mn p s Si Ni Cr Mo v Nb Ti
UT1 0.15 0.58 0.006 0.022 0. 71 0.13 1 .27 0.58 0.002 0.00005 0.00035
UT2 0.10 0.45 0.010 0.013 0.65 0.18 1 .30 0.53 0.004 0.00007 0.001 73
UT3 0.15 0.50 0.007 0.017 0.61 0.21 1.27 0.4 7 0.010 0.00025 0.00238
UT4 0.18 0. 79 0.009 0.003 0.52 0.20 1 .33 0.53 0.005 0.00006 0.00149
UTS 0.17 0.50 0.008 0.004 0.64 0.08 1 .41 0.50 0.004 0.00062 0.00250
UT6 0.103 0.53 0.013 0.013 0.58 0.25 1 .31 0.48 0.014 0.00084 0.00122
UT7 0.14 0.62 0.013 0.011 0.64 0.26 1.25 0.50 0.004 0.00008 0.00031
UT8 0.13 0.42 0.010 0.015 0.62 0.17 1.22 0.58 0.003 0.00005 0.00313
UT9 0.14 0.41 0.012 0.025 0.58 0.06 1 .25 0.61 0.003 <0.00001 0.00017
UT10 0.12 0.54 0.011 0.019 0. 75 0.07 1.01 0.56 0.001 <0.00001 0.00010
UT11 0.086 0.42 0.010 0.016 0.49 0.01 1 .04 0.54 0.003 <0.00001 0.00039
UT12 0.10 0.56 0.004 0.001 0.57 0.12 1 .46 0.55 0.003 0.00001 0.00357
UT13 0.12 0.45 0.004 0.024 0.59 0.29 1.27 0.49 0.004 <0.00001 0.00015
UT14 0.17 0.51 0.015 0.012 0.66 0.09 1.16 0.49 0.002 0.00004 0.00137
UT15 0.096 0.46 0.010 0.027 0.62 0.01 1.27 0.54 0.005 <0.00001 0.00064
UT16 0.11 0.47 0.008 0.020 0.56 0.16 1 .16 0.4 7 0.004 <0.00001 <0.00001
UT17 0.19 0.58 0.009 0.003 0.63 0.05 1 .05 0.48 0.006 0.00011 0.00211
MATERIAL~ Co Cu Al B w As Sn Zr Sb Ca N 0
UT1 0.013 0.13 0.005 <0.00001 <0.01 0.009 0.010 <0.001 0.00075 0.0005 0.0089 0.0094
UT2 0.010 0.05 0.004 <0.00001 0.008 0.007 0.007 <0.001 0.00041 0.0016 0.0155 0.0058
UT3 0.012 0.16 0.003 0.00003 <0.01 0.004 0.015 <0.001 0.00095 0.0011 0.0116 0.0048
UT4 0.015 0.15 0.017 <0.00001 0.001 0.016 0.014 0.005 0.00168 0.0001 0.0112 0.001
UTS 0.007 0.09 0.021 <0.00001 0.001 0.007 0.009 0.002 0.00055 0.0014 0.0083 0.002
UT6 0.011 0.15 0.031 <0.00001 <0.001 0.005 0.014 0.003 0.0008E 0.0009 0.009 o.·oo3
UT7 0.006 0.03 0.003 <0.00001 0.01 0.003 0.003 <0.001 0.0013C 0.0009 0.008 0.0055
UTB 0.010 0.11 0.026 <0.00001 <0.01 0.008 0.012 <0.001 0.0014€ 0.0013 0.012 0.005
UT9 0.012 0.16 0.003 <0.00001 <0.01 0.005 0.013 <0.001 0.0012! 0.0008 0.012 0.016
UT10 0.010 0.08 0.003 <0.00001 <0.01 0.008 0.009 <0.001 0.001 oc 0.0007 0.009 0.010
UT11 0.006 0.02 0.002 <0.00001 <0.01 0.004 0.002 <0.001 0.0001 E 0.0010 0.011 0.011
UT12 0.010 0.07 0.017 0.00005 <0.01 0.010 0.005 <0.001 0.0007( 0.0012 0.007 0.002
UT13 0.014 0.17 0.002 <0.00001 <0.01 0.007 0.015 <0.001 0.0019~ 0.0010 0.008 0.007
UT14 0.009 0.13 0.003 0.00002 <0.01 0.007 0.012 <0.001 0.001 o: 0.0014 0.013 0.008
UT15 0.006 0.04 0.003 0.00001 <0.01 0.006 0.004 <0.001 0.0003; 0.0002 0.014 0.010
UT16 0.012 0.17 0.001 <0.00001 <0.01 0.009 0.015 <0.001 0.0017( 0.0013 0.007 0.003
UT17 0.011 0.01 0.018 0.00028 <0.01 0.003 0.003 <0.002 0.0003< 0.0016 0.006 0.001
PS R = Cr + Cu + 2 Mo + 10 V + 7 Nb + 5 Ti - 2
MCF = Si + 2 Cu + 2 P + 10 As + 15 Sn + 20 Sb
MPF ::::: C + 10 V + Al
TABLE 3. Continued. Numerical values of literature chemical factors for i7 heats studied.
Gleeble-s1·
MATERIALS J x 21G PSR CERL CERL+Cr MCF AG1 R MPF 120 12
KJ/IN KJ/IN
UT1 206.4 11.3 1.200 0.582 0.079 1.349 1.237 2.700 0.069 0.175 SS NS
UT2 187.0 13.7 1.081 0.459 0.053 1.353 0.953 2.081 0.055 0.144 SS NS
UT3 244.2 13.9 0.902 0.483 0.085 1.355 1.228 2.402 0.078 0.253 HS HS
UT4 301.3 17.0 1.120 0.598 0.100 1.430 1.242 2.920 0.112 0.247 HS SS
UT5 193.8 12.6 1.092 0.556 0.059 1.469 1.052 2.792 0.062 0.231 HS SS
UT6 299.7 19.5 1.007 0.572 0.087 1.397 1.183 2.037 0.082 0.274 HS HS
UT7 201.6 15.15 0.932 0.322 0.038 1.288 0.827 2.332 0.042 0.183 HS NS
UT8 228.8 16.33 1.158 0.536 0.079 1.299 1.149 2.458 0.084 0.186 ES SS
UT9 247.5 18.3 1.287 0.661 0.092 1.342 1.194 2.687 0.081 0.173 HS SS
UT10 258.0 15.9 0.866 0.221 0.069 1.079 1.167 2.066 0.071 0.133 SS NS
UT11 109.2 11.3 0.846 0.172 0.032 1.072 0.623 1.706 0.028 0.118 SS NS
UT12 101.7 7.4 1.299 0.678 0.048 1.508 0.907 2.299 0.052 0.147 HS NS
UT13 197.6 11.7 0.919 0.461 0.095 1.365 1.271 2.119 0.090 0.162 HS NS
UT14 315.9 21.0 0.793 0.297 0.085 1.245 1.221 2.493 0.083 0.193 ES HS
UT15 151.2 12.4 1.093 0.443 0.049 1.319 0.846 2.053 0.041 0.149 HS SS
UT16 236.9 15.75 0.743 0.310 0.100 1.260 1.265 1.843 0.097 0.151 HS NS
UT17 145.2 10.66 0.683 0.091 0.024 1.074 0.749 2.583 0.029 0.268 SS SS
• SI = Susceptibility Index •• Heat input: 30 KJ/in.
%RA< 5% · EX1remely Susceptible (ES); %RA between 5 · 10% · Highly Susceptible (HS);
%RA between 10 · 15% ·Slightly Susceptible (SS); %RA> 20% • Not Susceptible (NS).
Material c Si Mn p s Ni Cr Mo Cu Al As Sn Sb N 0
PM-1 0.15 0.67 0.57 0.014 0.012 0.15 1.28 0.53 0.25 0.018 0.024 0.025 NR* NR NR
PM-2 0.14 0.59 0.60 0.017 0.008 0.18 1.38 0.51 0.24 0.006 0.030 0.032 NR NR NR
PM-3 0.15 0.67 0.58 0.015 0.009 0.17 1.21 0.53 0.23 0.015 0.032 0.028 NR NR NR
I PM-4 0.15 0.51 0.61 0.018 0.019 0.21 1.15 0.49 0.21 0.005 0.018 0.020 .0045 .0111 .0035
E
E
LOADING BOLT
• - - - - - - - 3 2 0 mm (12.6")-------lliioF"
FIGURE 1. (a) Schematic of PREVEW test device, (b) Pictorial of PREVEW test device with sample.
FIGURE 2. (a) Schematic of PREVEW test specimen, (b) stress distribution in test specimen.
Ill Ill
FIGURE 4. Average reduction in area in Gleeble reheat cracking test as a function of MPG Factor-5.
30
11110 MPC FACTOR-7
=2(C-0.12)-0.25(Mn-0.6)+150Nb+15V
+15Ti-1OOB+40(P-0.010)+5(S-0.015)
25 Ill 2 f!-1O(Al-0.010)+20(Cu/1 OO+Sn/3+As/3+3Sb
5- 11114 Ill 6
x xx x
0
0 0.2 0.4 0.6 0.8 1.2
MPC FACTOR-7
FIGURE 5.
4
NO CRACKING >'<I
CRACKING
•
-E
-E
:::c:
I-
3
D..
w
c
~ 2
0
<
a:
0
::
:::>
:: 1 •
x
<
:: •
0
•
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
FIGURE 6. Maximum crack depth in PREVEW tested specimens as a function of MPC Factor-7.
C. Lundin
The University of Tennessee
Knoxville, TN, USA
fyl. Prager
The Materials Properties Council, Inc.
New York, NY, USA
The basic findings reached in the industry survey and FABRICATION GUIDELINES
assessment showed that the preponderance of vessel cracking
was in the coarse grained region of the weld HAZ. The The fob1ication guidelines, derived from the evaluation of
location of cracking was both metallurgically related and, in 17 commercial heats of l 1/.Cr-1hMo material, are presented in
significant measure, due to stress concentrations at the fom1 of a flow chaii (Figure I) which will direct the user
geometrical "notches". Cracking may have been present from to the considerations necessary for successful fabrication of
the fabrication stages (cold cracking and/or reheat cracking) vessels and components of l Y.Cr-1hMo steel for use at
which was later revealed by crack extension in service. elevated temperatures, into the creep temperature regime
Thermal stresses due to the changes in cross section and (>825 °F). The fabrication guideline recommends that the
start-up and shut-down were also considered as exacerbating users first establish the composition of the material of
factors. The experimental program which followed this initial
construction and consider the strength level (Class) to be 2) a tramp element function encompassing the following:
employed. Tramp= 2(4.3(Sn+As)+l50Sb+cu+50(P-O.Ol)],
Guidelines are offered for both the Class l (60-85 ksi) or 3) a sulfur function: Sfn = l +30(S-0.02Tramp);
Class 2 (7 5-100 ksi) strength levels. The initial material For Sfn<l, Sfn=l, and
strength (hardness) during fabrication will be dictated by the
final strength or Class desired in the vessel or component. The 4) an allllTiinlllTI function: Alfn 1+15(Al-0.015); For
material must be purchased at a specific strength/hardness Alfn<l, Alfn=l.
level so that the application of the required PWHT schedules
does not reduce the base metal strength below that for the MPC Factor 5 = [Cfn(Tramp+Sfn)Alfn]-1
design Class desired. For example, a quenching and tempering
(Q&T) operation may be required to maintain Class 2 strength MPC Factor 7 considers all elements statistically regressed
ru.'ter the desired PWHT ex-posure. It is evident from the testing (with empirical cut-off values on some elements) to determine
accomplished, that the higher the initial strength of the the coefficients.
material the more likely is the occurrence of reheat/PWHT
cracking in a sensitive material and if the material enters MPC Factor 7 =
service with the HAZ only moderately tempered to preserve 2(C-O. l 2)-0.25(Mn-0.6)+ l 50Nb+ l 5V+ l 5Ti-100B+
the high end of the strength level, in-service HAZ cracking 40(P-O.O l 0)+5(S-O.O 15)+ l O(Al-0.010)+20(Cu/l 00
is also more likely. The research also strongly suggests that +Sn/3+As/3+3Sb)
the vessel or component should be PWHT high in the PWHT
range and certainly above 1250°F in any event in order to Satisfactory performance cut-off for the factors are: 2.0
impart good HAZ ductility. PWHT at high temperatures will for MPC Factor 5 and 0.5 for MPC Factor 7. These factors
more rapidly reduce strength levels and may cause loss in assesses the sensitivity toward reheat/PWHT cracking and
toughness, bringing further considerations to bear in tenns of in-service sensitivity to low ductility HAZ cracking. It is also
properties. The study of the coarse grained HAZ creep important to consider the carbon content of the material
properties also showed that the strength of the HAZ, for a because of its major effect on hardenability (ability to attain a
given material, was not significantly affected by the degree of specific strength {hardness} level during the initial heat
PWHT but notch sensitivity was greatly increased by treatTtent; e.g., Q&T or N&T). The carbon content also plays
relatively low PWHT. Fmiher, in the considerations for design a major role in the level of toughness which can be maintained
it must be recognized that when time-dependent properties after PWHT. The higher carbon content materials tend to
govern, the ASME allowable stresses are substantially the experience a greater loss in toughness with a given PWHT.
same for both Classes of material (above 900°F). Thus, it
would be prudent to select a Class l strength level for these When the strength Class, carbon content and MPC Factor
applications. have been detennined, the flow charts may be entered in the
appropriate column to dete1mine the basic considerations in
The composition of the material should be considered fabrication which will aid in avoiding PWHT as well as
from several standpoints. Today, most 1'!.Cr-Y:zMo steel is in-service cracking and property degradation.
furnished to relatively low J and X-bar factors (J factors of<
200 andX-bar levels less than 15) which are readily attainable
with current steel making practice. However, to avoid crack GUIDELINE DETAILS (See Figure 1)
susceptible heats a MPC compositional Factor has been
defined. Of the several factors evaluated, MPC Factor Sor 7
should be employed as they provide the optimum sensitivity.
These factors, deiived from the eXJJer1m1ental The Class 2 column which is
study of cracking sensitivity, are described in a companion characterized by a carbon content 20.15 % and a high MPC
paper in these proceedings ("Specifying l Y..Cr-Y:zMo Steels for Factor represents the most c1itical material for fabrication. The
Freedom from in Hydroprocessing Service"). initial direction calls for a "Refined Joint which
MPC Factor-5 contains four groupings of the operational stresses and should invoke the requirement that an
compositional elements in regard to mechanistically related joints be full penetration welded and utilize the least amount
fur1ctions: of filler metal. The weld crovvns should be removed and
ground flush with the plate surface for butt welds and the
l) a carbon function which includes the basic strengthening contour of the fillet welds should be ground such as to provide
elements in addition to carbon: a smooth transition to the base material (blend ground). The
Cfn = (5C+lOOONb+lOOV+50Ti-0.5)+l, grinding scratches should be transverse to the weld axis and
the final grinding should be done with fine wheels. The The high carbon and high MPC Factor conditions are the
sequencing of welds shouid be carefully considered so as to most critical in the fabrication of a vessel and component and
minimize long range residual stresses. will cause the most carefully considered fabrication steps to be
invoked.
The weld metal carbon content should be aimed at the
0.06-0.08 % range as this carbon content range provides CASE 2. If the MPC Factor is low with a high carbon
sufficient elevated temperature strength without the potential content or a high MPC Factor with the carbon content in the
for excessive hardness in the weld deposit. Extensive Pressure recommended range of 0.11-0.14 % certain relaxation in
Vessel Research Council research has showed that little fabrication procedures can be invoked and these are basically
benefit is gained in terms of elevated temperature strength as that controlled deposition welding procedures are probably
the weld metal carbon content is raised much above the not necessary and the PWHT temperature can be reduced to
0.06-0.08 % level. Preheat should be 300°F minimum, in 1300°F mino The reduction in PWHT temperature will no
accordance with ASME recommendations. A postweld hold doubt improve toughness response upon PWtlT and the base
at the preheat temperature is advisable when the section size metal and weld metal tensile and creep strength will most
exceeds 'h". Low weld heat input should be utilized, likely have a better match. The degree of in-service inspection
employing small passes which induce low temperature can probably be relaxed after the initial start-up, although the
transformation products to yield a better ductility response attention to joint design details and the contouring of the welds
during PWHT and higher creep rates to re-distribute the should be rigorously followed.
strains attendant with reduction in welding residual stresses.
Small passes also provide for maximum overlap in the base CASE 3. With the optimum range of carbon content and
metal HAZ and thus will limit the extent of the coarse grained a low MPC Factor or a low carbon (s0.10 %) content and a
region remaining after welding is complete (it is the coarse high MPC Factor a relaxation to conventional joint designs
grained region which has the maximum potential for and normal treatment of weld contours and grinding can be
cracking). As an example, the number of passes to insure low accomplished. Moderate heat input welding can be employed
heat input should exceed approximately 16 for a l" thick plate. and the PWHT temperature reduced to 1275°F min. Better
Controlled deposition procedures should be strongly retention of toughness will be affected. However, at the low
considered dependent upon the MPC factor and carbon carbon level, the tensile strength may be reduced to values
content. The employment of controlled deposition procedures approaching the lower limit of the Class 2 properties if
can be used with success in refining the base metal HAZ extended PWHT times are employed.
microstructure thus minimizing the potential occurrence of
reheat/PWHT cracking or cracking in the HAZ in-service. CASE 4. For a carbon content (sO. l 0 %) and a low MPC
Factor the chance of reheat/PWHT or in-service HAZ
PWHT should be performed at a minimum of 1325°F cracking is minimal and the PWHT temperature can be
with controls placed on the heat-up rate and the b,. T within the reduced to l250°F min. However this condition, of low
vessel during the PWHT treatment to minimize stresses. The carbon and low MPC Factor, may require a Q&T heat
high PWHT will greatly "de-sensitize" the HAZ tendency treatment for the base metal prior to fabrication with a
toward low ductility cracking in-service. controlled tempering temperature to maintain Class 2
properties.
Several issues must be addressed for the fabrication of the
high carbon and high MPC Factor materials. Loss in base
metal toughness, upon the high temperature PWHT, is a
concern and efforts to combat this loss may involve limited The columns representing Class l properties in the
heat treatment and before fabrication in the attempt Fabrication Guidelines are less restrictive in fabrication
to maintain toughness after completion of fabrication. procedures than those for Class 2 properties basically because
However the the strength of the initial base material the lower strength (hardness) of the base material provides for
the higher will be the potential for cracking problems. High a greater degree of stress relaxation during PWHT and for
carbon content materials with high levels of carbide fonning better redistribution of stresses during transient conditions
elements (which contribute to the MPC factor) may cause the in-service and this tends to reduce the potential for coarse
retention of high base metal HAZ hardness after fabrication grainedHAZ cracking during PWHT or as a function of time
and influence certain applications for which the hardness level for in-service exposure.
is a consideration. The high hardness HAZ's are also more
sensitive to HAZ cracking in-service and will cause Conventional joint designs and the weld metal in
consideration for a significant level of in-service inspection of 0.06-0.08 % carbon content range are recommended with
the welds, especially early-on during operation. alteration of welding and PWHT procedures depending on the
base metal carbon content and MPC Factor.
For the most restrictive case of the high carbon content If PWHT is to be employed the MPC Factor should be
(>0.15 %) and high MPC Factor, low weld heat input is detern1ined and if the Factor is high, controlled deposition
recommended with controlled deposition welding procedures weld procedures should be used togeilier wiili PWHT of
employed in critical weld regions. Control of the PWHT 1325°F minimum. If the MPC Factor is low, normal welding
process should be Ul)dertaken with careful control of thermal practices can be employed and PWHT should be
gradients (~ T) and PWHT should be accomplished at accomplished at 1275°F minimum. These precautions will
1325°F. As with the Class 2 considerations, the base metal mitigate ilie cracking potential and provide adequate creep
toughness may be significantly reduced by the more severe behavior for the less than 2 year anticipated additional life.
PWHT schedule and this should be attended to by the overall
heat treatment schedule for the base material. The election to repair wiiliout PWHT involves many
factors, but significant considerations should be given to tills
If the MPC Factor is reduced at the high carbon level or option if the component to be used at elevated temperatures,
the carbon content is reduced to the recommended range of especially in the creep regime, since PWHT related
0.11-0.14 % with a high MPC Factor, the PWHT temperature metallurgical changes will occur as a function of operational
may be reduced and a relaxation in the employment of time; fuus, softening the repair and relaxing the residual
controiled deposition welding procedures may be affected. stresses. If the decision is to repair without the application of
Again, a concern for a reduced toughness at the 1300°F PWHT, controlled deposition procedures should be used
PWHT temperature may be evident with the high carbon together with close process control to minimize thermal
material. strains during welding and equipment start-up after repair.
The elimination of the coarse grained HAZ by use of the
As the carbon content is further reduced and in controlled deposition procedures will result in the virtual
combination with the MPC Factor as shown in the Guidelines, elimination of the type of cracking responsible for most
the PWHT may be reduced to 127 5 °F minimum. problems in the ! 1!.Cr-1/iMo system. The issue of hard HAZ's
must be addressed in terms of the service environment for the
With a low carbon content and low MPC Factor the component or vessel.
chance for HAZ cracking, during either PWHT or in-service,
is virtually eliminated and the PWHT temperature may be Repairs, by their very nature, are critical operations for
reduced to 1250°F. Normalization will most likely be which ail precautions and controls should be invoked and
adequate to retain the Class l properties even at the low considered before the on-set of repair. The repair cavity size
carbon level. and configuration should be chosen so that ilie controlled
deposition procedures can be properly implemented and low
Repair Guidelines heat input welding techniques always should be used.
The considerations for repair are presented in the Repair The Guidelines presented for boili fabrication and repair
Guidelines flow chart (Figure 2) which is divided into two cover the full range of material strength and composition and
categories depending on the required life of the vessel thus encompass the entire gamut of fabrication and service
subsequent to repair. conditions that may be encountered in practice. The
Guidelines were constituted to remove ilie need for overly
The long term repair scenario (greater than 2 years) conservative fabrication procedures and critical operational
invokes considerations similar to the original fabrication and modes governing pressure vessels for all construction to
requires controlled deposition welding procedures on Class 2 mitigate cracking and enhance service life by consideration of
components. The column in the Fabrication Guidelines the interplay of strength, composition and fabrication
appropriate to the carbon content and MPC Factor should be techniques.
employed for long term repairs.
Short term repairs (less than 2 years) are governed by the REFERENCES
considerations in the Repair Guidelines and it is recomn1ended
that a low carbon weld metal (0.025-0.035 % carbon deposit [l] Lundin, C. D. and Wang, Y., "Half-Beadffemper-
chemistry), E8018-B2L, be considered. The use of this filler Bead/Controlled Deposition Techniques for Improvement
metal will aid in mitigating weld cracking problems (including of Fabrication and Service Performance of Cr-Mo
cold cracking) and reduce the level of residual stresses that Steels," Draft Final Report Submitted to ilie Committee
will occur as a result of localized repair. Furthennore, the on Welds of the Pressure Vessel Research Council,
creep properties of the low carbon weld metal will be December 1993.
adequate for short term service after repair.
[2] Lundin, C.D. and Khan, K.K., "Fundamental Studies of [5] Hashimoto, K., Tokuno, K, Hamada, K, Tsuchida, Y. and
the Metallurgical Causes for Reheat Cracking in 2'/4Cr Prager, M., "Creep Embrittlement of V-Modified
-IMo, l '!.Cr-Y:zMo and Copper Precipitation Hardenable 2V.Cr-1Mo Steels, ASME PVP 1992, Serviceability of
Steels and Problem Mitigation," Final Report to the Petroleum, Process and Power Equipment, PVP-Vol.
Weldability Committee of Welding Research Council, 239, MPC-Vol. 33, pp. 81-87.
January 1993.
[6] Nomura, T. et. al., "Creep Embrittlement of Structural
[3] Lundin, C.D., Khan, K.K., Zhou, G. and Al-Ejel, K.A., Components in Catalytic Reformer Reactor," Trans.
"The Efficacy of the Utilization of Low Carbon Cr-Mo Japan Soc. of Mechanical Engineers, 1993-9, pp.
Weld Metal for Repairs in Cr-Mo Vessels and Piping," 2066-2073.
Progress Report Submitted to the Committee on Welds of
PVRC, January 1994. [7] Cantwell, J., "Private Communication to M. Prager of
MPC" May 1993.
[4] Lundin, C.D., Khan, K.K., Zhou, G. and Liu, P., "The
Efficacy of the Utilization of Low Carbon Cr-Mo Weld [8] Prager, M., "Development of the MPC OMEGA Method
Metal for Repairs in Cr-Mo Vessels and Piping," for Life Assessment in the Creep Range," PVP Vol. 288,
Progress Report Submitted to the Committee on Welds of ASME Pressure Vessel & Piping Conference, June 1994.
PVRC, May 1994.
Class 2
0.06-0.08%C 0.06-0.08%C
0.06-0.08%C 0.06-0.08%C 0.06-0.08%C 0.06-0.08%C 0.06-0.08%C 0.06-0.08%C
Deposit Deposit
Deposit Deposit Deposit Deposit Deposit Deposit
Low Heal Input Low Heat Input
Low Heal lnpu Moderate lllloderate Low Heat Moderate Moderate
Con. Deposition Con. Dep for
Heat Input Heat Input Input Heal Input Heal Input
Avoid Restrain! Critical Areas
Control [ Control
0
~---p-~,..T_H_T--~ ~-1-3-0p_o_;;_H_T_m_in~ ~-12_7_~__~_H_T_m_in~ 25~;;
PWHT PWHT PWHT
PWHT 1300°F min 1275°F min 1250°F min
HTm in 1H
~ ~ ~ ~
Improved Possible Low Possible Low Toughness May
PWHT PWHT Be low for
Toughness Tensile Tensile
1325°F min 1325°F min High Carbon Base
Strength Strength
Maleh ol with with Metal
BM/WM Low Carbon Low Carbon
Tensile + Base Metal Base Metal ISSUES
low BM Toughness Creep Toughness May
High HAZ Hardness Be low
In-Service HAZ
Inspection
Potential Low Ductility
The Welding Research Council, Inc.
ure 2
Follow Fabrication
Yes Guidelines Except Use
Controlled Deposition on
All Class 2 Compoments
Use Controlled
Deposition
Determine
See Appendix I
MPC Factor
Close Process
Control
Minimize Thermal
Strains
ISSUES
• Hard HAZ
Use Controlled Usual Good
Practices
PWHT
1325°F Minimum PWHT
1275°F Minimum
F. Faure
FRAMATOME, Materials Department
92084 Paris la Defense, France
TABLE l : CHEMICAL COMPOSITION OF BASE Two kinds of stick electrodes have been used for these tests :
.MATERIALS AND CONSUMABLES (Wt %) - stainless steel (309 L)
- nickel base alloy (E Ni Cr Fe 3)
c s p Si Mn Ni Cr Mo Nb
Typical chemical compositions of base material and consumables
are given in table 1.
WEilllNG BASE
Prior to welding, the electrodes were first dried for 1 hr at 300°C,
A 508 Cl3 0.16l 0.01( 0.009 0.22 1.35 0.71 0.24 0.51 I
then re-humidified at 20°C in 90 % humidity (holding time between
TESf MATERIAL
48 and 96 hrs). Previous studies [2] had shown that this treatment
yielded reproducible high hydrogen content in the welds. These
309 L O.OH 0.014 0.013 0.78 1.26 12.9 24.1 0.037 I levels have been measured for all tests, by degassing of welded
samples at 1050°C under vacuum, this method is described in ref
I I [2].
Ni base
0.025 0.005 0.010 0.68 5.90 65.0 16.2 0.251 2.5
consumables alloy Welding parameters have been kept constant for all types of
consumables (V == 30 ± 2 V, I= 140 ± 10 A, travel speed 30 ± 5
cm/min) to have consistent thermal welding cycles, that have been
measured in HAZ in separate experiments, to be able to relate
Low alloy
hydrogen fluxes and temperature.
0.05 0.00' 0.009 0.29 1.7 0.8 0.05 0.49 I
steel
The effect of preheating at 175°C was also investigated.
Weld Bead
Spectrometer
""
Specific location I
where hydrogen
motion is analysed I
I
/Waive
Turbomolecular !======~[)<:)
Pump Calibrated leak
0 0 0 0
F
Detection of hydrogen flux under beads
d.e1poo11ted. on square blocks
Quadrupole gaz Electronics
analyser sensor
The implant test is a well known method for studying cold cracking
...,.. Diffusion Pump
[7] and a modified technique has been proposed for studying
interfaces in welds between dissimilar metals [8J. A small notched
Mas_5 snectrometor
cylindrical specimen (the implant) is inserted into a hole in a support
plate. A weld bead is deposited on top of the implant, so that the
notch is located in the HAZ and a stress is applied to simulate
welding stresses toward the end of the thermal cycle. The effect of
various parameters (e.g. welding energy, preheating, stress,. .. ) on
cracking and/or rupture, to assess the risk of cold cracking, can thus
be studied.
TABLE2
EXPERIMENTAL CONDITIONS
FOR THE TESTS PERFORMED
(*) Z =distance between fusion line and : - bottom of hydrogen collection hole for implant specimens
- highest part of the collecting hole for square block specimens
Mass
Spectrometer
H.A.Z
Weld Bead
I ..
3 -RESULTS
3.1 - Hydrogen jn HAZ on smiare blocks
Experiment n° 1
Results are given in fig. 5. In HAZ under 309 L deposited metal,
hydrogen evolution curves are quite similar for the 2 experiments
performed. Flux increases after the arc has gone over the hole, then
a maximum is observed after 15 to 20 seconds, then the hydrogen
flux decreases and increases again, and last starts decreasing slowly
- Rates of changes are somewhat slower with tests performed with
preheating.
Under the Ni base alloy weld bead, only one peak is observed 17
sec after the arc has gone over the collecting hole.
The reasons for the differences will be discussed later but it must be
observed that for experiments with a 309 L electrode, the collecting
hole was fully located in the HAZ while for the experiment with a
Ni base alloy electrode, the collecting hole was only partly in the
Experiment n° 3
HAZ. (Fig. 6).
Experiment n° 4 Experiment n° 7
309 L Ni base alloy
--- ...
-.- pnh••tiog
I1/
. Estimated thermal cycle - no preheating
T (OC)
1000
';'
e
'3 1--------1---------'<--------4 T (oc)
~~
.,.,,,.,-~-. ____ ~
./ -........ .£
.........
500 f---,.-----+-1--------------"'"-=::----!800
,,
::
(\"
l--+:_'~,~-.,Lil-----------------1'00
··. __J
·r-· . . . . . . Est!m:;ted thermal cycle
time (s)
. --------L .. ______ .,
tillle (0)
Figure 8 : Hydrogen flux during implant tests under Figure JO : Evolution of hydrogen flux during
309 L stainless steel deposited metal flame-cutting
The second maximum in welding tests (and the maximum observed used in many other applications where hydrogen transport is the
during the flame cutting test) occurs at low temperature and can be controlling factor, and where it is important to know hydrogen
explained by a competition between several mechanisms : motions.
Both fig. 5 (curve 2) and 8 (curve 5) illustrate the effect of pre and
post heating on the hydrogen diffusion curve - It is shifted to longer
time, in general agreement with slower cooling cycles. Also if
hydrogen flux is integrated over time, it is clear that more hydrogen
is degassed with pre and post heating, as was already well known -
The effect of the thermal cycle on hydrogen evolution is confirmed
by the much slower phenomenon observed during flame cutting.
x200
While in welding hydrogen comes mostly from hydrogen adsorbed ® Low alloy steel Z = 0,85
or combined in the coating of the electrodes, during flame cutting it
is probably formed by dissociation of hydrocarbon and/or water
molecules in the flame. Figure 11 : Influence of deposit metal on the
microstmcture of the HAZ near the bottom
of the hole (Implant tests)
CONCLUSIONS
More work is needed for a full analysis of the data, in combination [4] D.G. HOWDEN
with other techniques (such as numerical analysis or critical Welding Research supplement, vol 103 (1982) p 108 s.
concentration studies), since this technique does not provide
concentration data but fluxes. However its efficiency could also be
[6] D.NOBLE
Metal Construction, October 1985
[7] M. DADIAN
Annual Assembly of the International Institute of welding,
1982
Document IIS!IIW IX - 1240-82.
C.CASCIARO, P. MARINELLI
Belleli Ricerche Scpa
TARANTO, ITALY.
3
0.12
Mn(%) 0.55 0.53 0.97
Si(%) 0.17 0.09 0.13
p (%) 0.004 0.007 0.003
s (%) 0.004 0.008 0.003
Ni(%) 0.055 0.030 Fig. 1 Locations of specimen used for hydrogen
Cr(%) 2.25 2.58 2.43 measurements
After welding and heat treatment, specimens for Extensive measurements have been carried out by an
hydrogen measurements (15xl5x30 mm) have been cut electrochemical method based on a new equipment called
from weld metal of test coupons and stored in liquid S.M. Hydrogen Monitoring System. This method is
nitrogen to prevent hydrogen loss. The locations of based on the hydrogen release in a gel filled probe of
specimen used for hydrogen measurements is shown in 0.2N NaOH with traces of Na2[Pd(N02)4]; the
Fig.I. specimens become working electrodes held at +200m V
SCE by an electronic device and the system is totally
computer assisted. The anodic current is related to specimens were subjected to static stress at various
hydrogen content and the experimental values were levels; these levels were definited as a percentage of the
interpolated by numerical code and analytical function[5]. ultimate tensile stress of uncharged material.
The H-induced cracking test is based on ASTM STP
962 to determine the hydrogen critical concentration
(CCR) of the steels in the absence of applied stresses [6]. RESULTS
These tests were conducted electrochemically on a
cylindrical sample ( lOmm, l= 60 mm cut across the The hydrogen measurements by electrochemical system
weld) of the material in a solution of H2S04 1 N mean values, are shown in Fig.4. Some specimens have bee1
containing 5 mg/l of As203 and few drops of CS2. The checked by both (A WS and electrochemical) systems; th
cell is arranged so that the sample is surrounded by an results are very similar with a mean difference less then 15%.
anode of Pt; the current density used is 50 mA/cm2. In
order to ensure a known hydrogen concentration profile
varying from a maximum value on the surface to zero on
its internal axis [7], the hydrogen introduction time tc (s)
is calculated on the basis of the following relation [8]:
•Bi
Fia
Fib
tc =0.04·-
D
r2 6'
-
0
0
;:::::,
3
•m F2
Fia LM
where r = radius of the cylinder in metres and D = ::i:::
hydrogen diffusion coefficient in m2/s
After the test the cylinders are cut perpendicularly to
the axis in order to observe the geometry and the position
of the cracks. Considering the hydrogen distribution
profile[7,8], and on the base of the metallographic
examination, the critical concentration range of crack
nucleation without applied stress, is calculated.
In addition, mechanical tests have been carried out
using a cylindrical specimens (<\> = 8, l = 110 mm)
according to ASTM F519 (Fig.3): the notch was located
as close as possible to the fusion line.
AW DHT ISR
0.9 ---calc.
0.8 - I I - Meas.DHT
0.7 --3'- Meas.I SR
0.6
0.5
-
0.4
u 0.3
ro 0.2
u 0.1
0
0 2 3 4 5 6 7 8 9 10
De-hydrogenation factor;D*t [cm2] Fig. 7 Hydrogen assisted microcraks without external
load (160x)
ACKNOWLEDGEMENTS
Authors thank Mr. A.Bertoni and Mr F.Fusari of
BELLELI Spa for the usefull discussion on metallurgical
and procedural problems of the research.
REFERENCES
1. N. Yurioka, H. Suzuki: Intern. Mater. Reviews,
Vol.35, 1990, p. 217
2. M. W. Mallet, P.J. Rieppel: Weld. J., Vol.25, 1946,
p.748
3. J.M. Sawhill, A.W. Dix, W.F. Sawage: Weld. J.,
Vol.53, 1974, p.554
4. T.A. Siewert: : "Testing of welding electrodes for
diffusible hydrogen and coating moisture", ASTM STP
962, L. Raymond, Ed ASTM, Philadelphia, 1988,
pp.238-246
5. R. Valleggi, D.M. De Micheli, L. Bosio: "Numerical
code for S.M. hydrogen monitor system", Scienzia
Machinale Thecnical report, 1993
6. G.M. Pressouyre, Faure F.M. : "Quantitative
Analysis of Critical Concentrations for Hydrogen-