Sunteți pe pagina 1din 8

Proceedings of the ASME 2010 Pressure Vessels & Piping Division Conference

PVP2010
July 18-22, 2010, Bellevue, Washington, USA
PVP201 0-2527 4
POTENTIAL DETRIMENTAL CONSEQUENCES OF EXCESSIVE PWHT ON
PRESSURE VESSEL STEEL PROPERTIES
Cedric Chauvy
ArcelorMitlal lndusteel
Chil.teauneuf, France
Lionel Coudreuse
ArcelorMitlal lndusteel
Chateauneuf, France
Patrick Toussaint
ArcelorMitlal lndusteel
Charleroi, Belgium
ABSTRACT
During fabrication of Pressure Vessels, steels undergo
several heat treatments that aim to confer the required
properties on the entire equipment, including welds and base
metal. Indeed, the Quality heat treatment of the base material,
which leads to achieve the target properties, is most of the time
followed by Post \Veld Heat Treatment (PWHT). The aim of
such treatments is to insure a good behaviour of the welded
zones in terms of residual stresses and obviously properties
such as toughness. Generally, many simulated PWHT (up to 4
or more) are required for the testing of the base material, which
can affect its properties and even lead to non acceptable results.
In some cases for fabrication purposes an intermediate Stress
relieving treatment can be required
Special attention is paid on CHMn steels (e.g. SA/A516
from ASME BPV Code) with the effect of thickness and Ceq
(11\V Carbon equivalent fonnula: see page 3) requirements on
the final compromise between properties and heat treatments. In
particular, toughness and UTS are the critical parameters that
will limit the acceptance of too high P\VHT. Although
alloying is a mean to increase the resistance to P\VHT, this
leads to difficulties in softening the heat affected zones. This
solution is therefore not the best one considering the whole
equipment optimisation. Finally, the manufacturing process can
play a major role when specifications are stringent. Quenching
and tempering can indeed provide better flexibility in terms of
PWHT and improved toughness for given Ceq and thickness.
The case of steels, which are widely used in
the energy industry, is also addressed. Indeed, P\VHT
requirements for increasing the toughness in the weld metal can
lead to decrease the base metal properties below the
specification limits. For example, the case of SA/A387grll is
very typical of metallurgical changes that can occur during
these high P\VHT leading to a degradation of toughness in the
base metal. Another focus is made on the Vanadium Cr-Mo
grade SA/A542D that must withstand very high PWHT (705C
and even 7l0C) because of welds toughness issues.
Optimisation has therefore to be done to increase the resistance
to softening and to guarantee acceptable microstructure,
especially in the case of thick wall vessels.
Some ways for improvement are proposed on the basis of
the equivalent LMP tempering parameter concept. The basic
philosophy is to fulfil the need for discussion between
companies involved in pressure vessels fabrication so that the
best compromise can be found to ensure the best and safest
behaviour of the equipment as a whole.
INTRODUCTION
Plates and other components of a pressure vessel usually
undergo various heat treatments in order to make them meet the
requirements, which may be either customer specifications or
intemational standards. The base metal before welding can thus
be normalized, normalized and tempered or quenched and
tempered.
During vessel fabrication, the manufacturer performs
several thermal cycles: preheating, post heating, DHT (De-
Hydrogenation Treatment), ISR (Intermediate Stress Relieving),
PWHT (Post \Veld Heat Treatment). Due to the temperature
range, only JSR and PWHT can be considered as real heat
treatments.
Usually, it is considered that tempering confers the
mechanical properties on the base metal and the subsequent
heat treatments confer the properties on the weld area. In
particular in case of repairs, it is not unusual to perform several
cycles of PWHT: up to 3 or 4. Actually, all these thermal cycles
will affect more or less the properties of the base metal. This
becomes stili more critical for thick wall vessels for which long
PWHT duration can be necessary.
In order to minimize the effect of P\VHT on the base
metal mechanical properties, the steelmakers are frequently
Copyright 2010 by ASME
asked to perform the tempering (if required) at higher
temperature than the PWIIT, which focuses on the weld metal.
Moreover, the mechanical tests have to be carried out after
simulated P\VHT (Temperature above 454C according to
ASME or other code) in order to prove that the base metal can
withstand these thennal cycles.
In many cases, the PWI-IT has a real influence on the base
metal properties and the stcehnakers have to take it into account
while defining the grade and the manufacturing route. For
instance, in case of thick wall reactors (> lOOmm) which require
PWHT at high temperature during long periods, the steel
supplier must sometimes decline the tempering temperatures
required by the customer specification. Indeed, it is sometimes
not possible to guarantee the required properties after PWHT if
using a too high tempering treatment for a given grade.
Genera11y, the most convenient way to quantify the effect
of several thermal cycles on the properties is to use time-
temperature equivalent laws or tempering parameters. For
pressure vessel steels, the parameter commonly used is the
Larson Miller Parameter LMP (also called Hollomon-Jaffee
Parameter).
The goal of this paper, after a brief rehearsal on the
tempering parameter, is to show the limits in term of tempering
and PWHT of different steel grades. Especially in case of heavy
gauges, the final properties arc not only given by the tempering
step but by the whole thermal history that has to be applied to
the material.
EQUIVALENT TEMPERING PARAMETER: LARSON
MILLER PARAMETER
The influence of time and temperature on mechanical
properties is well known. Figure I gives an illustration of
typical curves that can be drawn from experience (for a 2.25Cr-
lMo steel grade in that case). Nevertheless, it is difficult to
compare the effect of different heat treatments, made at
different temperatures and different durations, and it is also
impossible to determine the cumulative effect of two or more
heat treatments at different temperatures.
600
700
;urs {TeniP-63-S'C) UTS (Temp 665-cf UTS (r.imP 690'C)'
oYS{Temp6SS"C) t.YS(Temp690"C) j
soo+---

" 500 {} ... -
- "' ... .
' 400
300
10 100
Tlme(h)
Figure I: Influence of temperature and time on tensile
properties (grade A387gr22 cl2).
Therefore, the usc of an equivalence parameter, which
involves both time and temperature, is of great interest. The
most used tempering parameter is called the Larson-Miller
Parameter LMP (or the Hollomon-Jaffee parameter).
The Larson-Miller Parameter formula is given hereafter for a
single heat treatment:
LMP=B(20+1og(t)) (I)
with: 8 temperature (K) and t time (hours)
It is also possible to extend this relationship to a multiple
treatment. In that case, the method is to choose one temperature
8 (usually the one used for the first treatment) and to calculate
for each treatment an equivalent duration teq at temperature 8,
by equalizing the LMP between the set of conditions ( 8, teq) and
(8,, t;).
These durations have then to be used in formula (1) and
this leads to the following general expression:
LMP,
1
= 8 [ 20+ log(

J] (3)
Bis the chosen reference temperature (K)
t .. qi is the equivalent duration at 8 for a treatment i made at B,
during t
1
(hours)
Figure 2 below is an example of the relevance of such a
parameter. It shows a lot of data obtained with a single
treatment (blue dots) but also some results from multiple
treatments (red and green dots within red circles).
[.-uTsonH)'o;.\e YSonec,ode --- - 'uTi(sOO:.C.shw+69o:lOC:i4hJ
1
t 690'C-2-il\) UTS(710'C-6h + 690'C-33h) o VS (710'C-61>+ 690'C-33h) !
2
800

600
a. .
:a: 5oo
400
300 +-----------
19000 19500 20500 21000 21500
LMP
Figure 2: Mechanical properties as a function of LMP (grade
A387gr22 cl2)
Once this parameter is defined, it is possible to study the
Copyright 20 I 0 by ASME
effect of cumulative heat treatments, which can be tempering,
ISR and P\VHT, on the different properties required for the
application (mainly tensile characteristics and Charpy V -Notch
toughness).
C-MN STEELS FOR PETROCHEMICAL APPLICATIONS
The case ofC-Mn steels such as A516 is very interesting
in that sense that many parameters must be taken into account
when considering the effects of P\VHT on materials properties.
First, the composition is often controlled through the use of a
maximum Carbon Equivalent Ceq. This parameter aims to
ensure a good weldability and is commonly defined by using the
ll\V formula given hereafter:
(4)
There are therefore a lot of clements that are not included
in this concept but allowed in the composition according to
A516 or A20. \Ve will see later how some other clements can
be used and their limitations.
Then the target is to obtain an acceptable microstructure,
matching the properties requirements, with a given chemical
composition. For a given manufacturing route, the cooling rate
for higher thicknesses will be lower than for the thinnest ones.
In some cases, particularly for heavy gauges, ensuring that the
properties are met implies to adapt either the Ceq value or the
cooling process.
Finally, depending on fabrication conditions, more or less
strong PWHT can be required leading to potential problems
regarding minimum values of mechanical properties.
In few words, for a given thickness, A516 must
simultaneously satisfy requirements of:
Ceq
PWHT
UTS and YS
Charpy V -Notch impact values
Most of the time, A516 is delivered in normalized
condition. For thickness above lOOmm, this can lead to
undermatch the target values for mechanical properties in case
of too high P\VHT. Then the only alternative, for a given
manufacturing route, is to increase the level of Ceq as shown in
figure 3.
This leads to PWHT lintitation of typically 610"C+/-IO"C
for 15h. In other words, guaranteeing tensile properties in
conformity with the A516gr70 standard implies to limit the
tempering parameter value at approximately 18500 and 18900,
for a Ceq of 0.43 and 0.45 respectively.
Influence of tempering parameter on UTS values
550 .-------
540 t-------
530
520
'i 510
e 5oo
5 : ~
470
460+----
450
17500 18000 18500
LMP
19000 19500
Figure 3: Influence of tempering parameter on UTS values for
A5!6gr70.
It is also important to note that in the case of C-Mn steels,
too strong PWHT leads to deterioration of both the toughness
and the tensile properties. Indeed, toughness is decreasing for
normalized C-Mn steels when increasing the global tempering
parameter and/or increasing thickness as shown in figure 4.
Depending on targeted values and temperature, this can
sometimes give the limitation in terms of acceptable PWHT.
200.0 ,------- -
175.0
150.0
v 125.0
~
~ 100.0
75.0
50.0 t-----=-<.<
25.0
0.0 - - - - ~ - - ~ - - ~ - - ~ - ~ - - - <
50 75 100 125 150 175 200
Thickness (mm)
Figure 4: Influence of increasing thickness on Charpy values for
Normalized C-Mn steels at given LMP value.
As previously said, for special applications, it is frequently
asked for A516gr70 to limit Ceq to 0.43 for weldability reasons,
or even to fulfill the requirements of hardness in HAZ (for
example NACE recommendation is maximum 248Hv /22HRC)
for use in H
2
S environments.
In order to satisfy these requirements an alternative would
be to add a small amount of Nb to achieve the required tensile
properties after P\VHT, without any effect on Ceq value.
However adding Nb has an influence on HAZ behaviour, as
shown on curves of figures 5 and 6 that compare the softening
occurring in the Heat Affected Zone of grade A516gr70 and of
another steel corresponding to tl}e standard BSI501-225 (which
contents 0.035%Nb).
3 Copyright 20 I 0 by ASME
Steel without tlb /V (CarEisoe 70)


'

24BHv
'-";
r-"6
8
180
""\'J ,99 kJAnm -er 2, 78 kJtnm
---_:- 1,98 kJ/mrn- -- Maximum a/leAved
Figure 5: Influence ofPWHT on hardness values for A516gr70
without micro-alloying elements.
The curves of figure 5 show that, for grade A516gr70, one
can guarantee hardness in HAZ lower than 248Hv with a stress-
relieving heat treatment at low temperatures (600C). It is to
note that the recommendation of NACE, which indicates that
the PWHT should be performed above 6200C, is not justified
for C-Mn steels. In other words, the minimum temperature to be
applied is the one allowing respecting the 248Hv maximum
Heat Affected Zone Hardness while respecting the minimum
stress relieving temperature as per the construction code, which
is 595C (1100F) according to ASME VIII div I and div2. H
2
S
resistance, which is the concern of the NACE recommendation,
is already obtained using 600C.
Steel with Nb /V (type BS1501.225}
1340


'
1:
""
..c
"-.

\%:'

cr

-g260
E./j
248Hv
-g220
:::>
180

ou
"
8 8 0

8 il
b

"' "'
"'
-er 1,35 kJ/mm v- 2, 7 kJAnm
1,85 kJ/mrn- Maximum af/CAv
Figure 6: Influence of PWHT on hardness values for C-Mn
steel containing micro-alloying elements.
To the contrary, for micro-alloyed steel with Nb it is
necessary to use strong tempering parameters to satisfy the
requirements of hardness in HAZ (figure 6). This evolution
requires P\VHT at very high temperature, which can deteriorate
base metal toughness. This example illustrates the "perverse"
effect of some requirements. Indeed, it would be possible to
match the mechanical requirements and to get good toughness
with non micro-alloyed steel with a lower PWHT temperature.
For instance, it would be possible, according to figure 3, to
meet the required mechanical properties of A516grade70, with
a Ceq of 0.43 and a P\VHT limited to 600C.
As demonstrated, with the aim of accepting more stringent
conditions with higher thicknesses (typically higher than
150nmi), playing with chemical composition has limitations.
One can increase Ceq but this could deteriorate weldability.
One can add micro-alloying elements but this wili create Heat
Affected Zone difficult to soften during PWHT and this can
lead to worse toughness. Another solution is to use different
cooling rates that can lead to better microstructure without
having to adapt the chemical composition.
As allowed by 5.4 of SA/A516 "cooling rates faster
than those obtained by cooling in air are permissible"
provided that a suitable tempering treatment is subsequently
applied. For example water quenching of a 200mm thick plate
will give at mid thickness approximately the same cooling rate
than normalization of a 30mm thick plate. Therefore quenched
and tempered steels can be used, meaning no need to increase
Ceq because of higher cooling rates. It will lead to slightly
higher mechanical properties, significantly improved toughness
and enhanced resistance to strong PWHT. It appears then that
this solution exhibits many advantages over normalizing when
high thickness, low temperature Charpy values and strong
PWHT are required. Obviously, this does not allow to use
much higher temperature PWHT but for example to add one
cycle.
Figure 7 shows how mechanical properties behave, for
Ceq of 0.43%, when increasing the global tempering parameter
LMP as a function of the manufacturing of the plates (N or
Q&T). It appears clearly that the Q&T steel can withstand
higher PWHT for a given level of performance
4
600
UTS Q&T
550
UTS Normalize
"---- --fl_.___ J.

0
-.:-- ---- ---1-ys-nn- ----+-----j

400

-'-=-.o-.-.
I
15000 16000 17000 16000 19000 20000
Larson-Miller Parameter
Figure 7: Evolution of mechanical properties for A5l6gr70
(Ceq 0.43%) as a function of heat treatment and LMP.
Copyright 20 I 0 by ASME
Moreover the Q&T plates show significantly better
toughness and even a better evolution of this toughness with
increasing LMP. Figure 8 here below illustrates this behaviour.
160
-- -- --------
I
I,

CVN Q&T

I

t I .

g_.-. -
-":-._o_ lo
t o

0
-- I oo_o -
-J
--
--
140
120
100

d so
60
40
20
15000 16000 17000 18000 19000
Parameter
Figure 8: Evolution of CVN toughness at 0C for A516gr70
(Ceq 0.43%) as a function of heat treatment and LMP.
As a conclusion, more and more stringent requirements
show up; leading to the need of finding always more accurate
compromise between properties and fabrication. There arc
several parameters to be used but each of them has its own
limitation regarding the application. For the thickest range of
products (>l50mm), modem stringent specifications can lead to
usc quenched and tempered steels that show higher performance
levels, especially in terms of toughness.
Whatever the application, discussions are always needed
between the actors involved in the fabrication chain, in order to
best define the good compromise for these high quality C-Mn
steels.
1%CR - Y,MO STEEL FOR HIGH TEMPERATURE
REACTORS
This CrMo low alloy grade (A387grll) is widely used in
manufacturing high temperature reactors for refining I
petrochemical applications. Indeed, it exhibits good mechanical
properties for operating in the typical range of 350-500C.
Chemical compositiOn does not confer enough
hardenability on the grade. For thin plates, normalizing can be
sufficient to achieve suitable microstructure whereas quenching
and tempering is often mandatory when increasing the
thickness. This means that for thick products (typically
thickness >80mm) a significant amount of Pearlite and Ferrite
can be created even for quenched and tempered material, and
can be detrimental for toughness.
Moreover, after long exposure at temperature above
650C, there can be a modification of carbides structure that
may decrease both strength and toughness. This is linked to
globalization and coarsening of existing carbides, which is
called Upper Nose embrittlement and is known since the 50s [1-
3]. Therefore one has to be very cautious because the
combination of tempering with Post Weld Heat Treatment may
lead to this situation. In general, the H4Cr- Y2Mo grade is to be
used very carefully when targeted thickness is above 75mm. In
particular, it can be helpful to reduce the tempering in order to
get a larger margin for PWHT that aims to confer toughness on
welds.
Figures 9 and 10 illustrate respectively the evolutions of
the tensile and toughness properties as a function of the
tempering parameter LMP for A387grll. These results have
been achieved from SO and 143 mm thick plates with
comparable chemical analysis. The tempering parameters have
been calculated for different combinations of temperatures and
holding time. (temperature from 660 to 740C; holding time
between 30 min and 38h).
[ .. urS-sOmm thick "YS SOmm thick ur$ 143mm thlc_k __
700
. .. -
!!'.""'
" 450 },:
400

"" +------------
19000 19500 ,0000
"""' ""' """
LMP
Figure 9: Evolution of tensile properties of A387gde II cl2 as a
function of the tempering parameter.
Figure 9 shows that the tensile properties required for the
grade can be obtained for tempering parameters up to 21000.
There is no significant difference between both thicknesses. For
that LMP range, tensile properties are not the limiting element
for that grade.
f" 'Kv (:19-.C)_ SOmm thlck o Kv lhlcij
400 -------------------
"'
"' -1-----'--

e;roo
g
w "'
100
'

' '
50 +----------- -\
'"""
19500 201100
""' """ ""'
LMP
Figure 10: Evolution of toughness (Kv -29C) of A387gdellcl2
as a function of the tempering parameter.
5 Copyright 2010 by ASME
On the other hand, figure 10 shows a different behavior
for toughness properties at -29C, which get worse beyond a
given tempering parameter value. Moreover, the thickness plays
a really detrimental role on the results and have thus to be taken
into account. It appears that impact properties and thickness
constitute the limiting clements to the tempering parameter for
E4Cr- 'llMo steel. As a matter of fact, the more important the
thickness is, the less there wi11 be possibilities to perform
tempering treatments and PWHT at high temperatures.
Then, for the thickest products, steehnakers are often
conducted to decline customer specification requirements.
\Vhen the tempering parameter calculated following the
requirements does not allow guaranteeing impact properties,
there is never a single answer that can be made. Therefore, there
is a need of discussion between the different concerned parties
in order to define the best adapted solution in tenus of
compromise between PWHT, LMP, tempering and Charpy
V-notch requirements. If ever the requirements are mismatching
the material feasibility it is then necessary to choose another
material. For thick vessel, when low temperature toughness
properties are actually required (-l8C; -29C) with the use of
strong PWHT, the use of A387gde22 c12 is a safer alternative.
2'14CR - 1MO STEEL FOR HIGH TEMPERATURE
REACTORS
2%Cr- I Mo type steel is also used for petrochemical
reactors in the same temperature range as for A387grll.
To the contrary of JlACr- lhMo steel, the hardenability is
large enough to obtain a suitable microstructure (that is to say
without ferrite) for thickness up to more than 200mm as far as a
quench and temper process is used. This kind of grade is
therefore more comfortable with regard to PWHT/toughness
combination and toughness is indeed not a usual issue. Strength
will thus be the limiting property when increasing PWHT.
Figure 11 illustrates the influence of the tempering
parameter on tensile strength of 21;4Cr-1Mo steel (A387gr22),
for plates up to 250 mm thick. On this figure, the lower
thresholds of UTS requirements are specified according to
different standards of 2
1
4Cr-1Mo steels (ASTM and EN).
Chemical compositions for these different standards are the
same; that is only by modifying the heat treatment that it is
possible to have an effect on the mechanical properties.
2-25Cf11.lo sloe! so!t..-.!ngwrw
larson Miller Parameler
Figure 11: Influence of tempering parameter on tensile strength
of 21;4Cr-1 Mo steel.
ln order to guarantee the properties, a maximum value of
the tempering parameter can be determined depending on the
corresponding standard. The table below gives the maximum
tempering parameters for satisfying the tensile strength
requirements from the different standards; the maximum
tempering temperatures and PWHT conditions (for 200mm
thick plates) arc also reported. It appears that the choice of the
standard will strongly affect the amount of PWHT allowable on
the grade. The lower the minimum UTS, the higher could be the
Larson Miller Parameter and therefore the more there will be
possibilities for increasing the total amount of PWHT.
Standards
LMP
Tempering PWiiT
maxi
-
A387 gde22 c/2 20850 710"C 690"C -33h
ENI00282
20550 660"C 680"C- 33h
12CrMo910
A542B c/4 20000 650"C 655"C -33h
Table 1: Examples of heat treatment hmllations as a function of
the 2%Cr-l Mo standard used.
The tempering parameters given in the above table have
been calculated for tempering times leading to a homogeneous
temperature through the whole thickness of the product. In fact,
there is often no tempering times requirements. Therefore, it
could be possible to reduce the tempering parameter by
reducing the holding time. However from a metallurgical point
of view, this can produce heterogeneities of properties though
the thickness that can have consequences during the fabrication
of the vessel.
For example, tempering at 720C during I h leads to a
tempering parameter of 20,000. The same tempering parameter
can be obtained with a tempering at 690C during 5h30.
Whereas, for a 50 mm thick plate, there is no problem to
perform a tempering treatment at 720C for 1 h, it is not the case
for a 250 nun thick plate, for which it is better to apply a
6 Copyright 2010 by ASME
tempering of 690C for 5h30 in order to homogenize the
thermal effect through the thickness. Therefore, the LMP
approach must not be applied without taking into account the
specificity of thick plates. There must be production rules that
ensure thermal homogeneity of the products and thus give the
boundaries to respect in terms of heating rate and holding time.
2Y,CR- 1MO- V.V STEEL FOR HIGH TEMPERATURE
REACTORS
21,4Cr-1Mo-JAV steels (SA 542DCI4a or equivalent) arc
more and more used in refining installations for high hydrogen
pressure and high temperature ranges. Indeed, using a 21.4Cr-
1Mo-%V grade instead of 2%Cr-1Mo allows for a significant
reduction of thickness, especially since the new 2007 ASME
VIII division 2 issue, which increased the a11owable stresses.
The interest has thus grown with the size of the newly designed
reactors. Another important advantage for the final users is weld
overlay hydrogen induced disbanding resistance. Indeed, this
Vanadium enhanced grade does not exhibit some hydrogen
induced disbanding problems at the plate-overlay interface.
However, once again, in order to obtain the required
properties in the welds, especially the Charpy-V impact
toughness, the vessel fabricator needs to perform PWHT at high
temperatures (typica11y 705C and probably 71 0C in the near
future). This leads that grade to be able to withstand such high
temperature PWHT during sometimes more than 30h.
Considering the trend to increase these PWHT requirements,
some optimization studies were necessary in order to improve
the resistance to softening. There are two ways to reach that
goal. The first one is to play on the microstructure itself
whereas the second one is to add some chemical elements that
will create stable precipitates resisting to softening.
Figure 12 illustrates the effect of different alloying
elements on the softening curve and shows how it is possible to
guarantee tensile strength properties of A542D cl4a for very
thick products (>l50mm), and for high tempering parameters
by using micro-alloying. On one hand, Boron promotes
hardenability and therefore improves the microstructure of the
steel, especially for thick products. A full bainitic
microstructure can be achieved at mid thickness up to 300nun
thick product, allowing to keep tensile strength properties at
mid-thickness for higher tempering parameters. On the other
hand, an optimization of Vanadium and Niobium contents
reduces the softening during tempering. This creates stable
carbides known to be resistant to softening. In any cases, the
amounts of B and Nb are kept at a very low level.
700
.lt,<UTS :tK'J?a)--- ------------- --------.------ ------------
20000 20200 21).100 20600 20800 21000 21200 21400
LMP
Figure 12: Influence of alloying on UTS for 2%Cr-lMo-1.4 V
(thk > 150mm, properties at 'h thk): Curves representing
minimum values.
In order to illustrate this trend with some industrial data,
figures 13 and 14 hereafter provide some results obtained on
A542D at different stages of the thermal history, that is to say
after quenching and tempering (Q&T), after minimum P\VHT
and after maximum PWHT as well. Moreover, a large range of
thickness is considered, starting from 127mm up to 230mm.
tnflwoc& of temp-erlng parameter on UTS
G""127mm thkk "'14Srnm thM ; i9Jffim thkk. ' 230mm J!i@

""'"
21:200
Figure 13: Influence of LMP and thickness on UTS for the
grade 21.4Cr-1Mo-1.4V.
Figure 13 shows that for usual maximum PWHT (70SOC
during 33h), the UTS requirement is met. It also appears that
the next step of (710C/33h) will be acceptable. This need for
slight increase in PWHT temperature is currently under
discussion in order to provide the weld metal with a larger
margin in terms of toughness. One can also underline that the
effect of thickness is more present for low LMP than for the
largest ones. It is interesting to notice the decrease as a function
of the tempering parameter. In particular, this implies that the
initial target in the Q&T state must be close to the upper limit of
UTS requirement.
7 Copyright 2010 by ASME
lnfltnnce of tempering parameter on YS
L.!J.fu!if!l_ thick o 149/nm thick 4 193mm thick "
""
W>+----
202>:{1
'"'' '""
21000
hmperlng paramHu
Figure 14: Influence of the thickness and the tempering
parameter on Yield Strength for grade 2JACr-1Mo-%V.
Figure 14 above describes the evolution of the yield
strength as a function of the tempering parameter. Even if the
trend is obviously the same as for UTS, it appears that the
margin is larger regarding the minimum requirement to be met.
Anyway these two last figures clearly show the influence
of multiple P\VHT on the mechanical properties of the base
metal. The chemical analysis optimization has given some extra
margin but the material is close to its metallurgical limit.
Figure 15 shows the average values of Charpy-V impact
testing obtained at -l8C (for Q&T) and -29C after some
P\VHT. Results concern plate thickness between 127 and
230mm with quarter thickness and half thickness sampling. It
can be noticed that the impact properties remain completely
satisfactory for all the considered LMP range, whatever the
thennal history of the plate. Thus toughness does not appear to
be a limiting factor for the time being in tem1s of acceptable
PWHT.
Averagl) CharpyVva!ue (18'C Q&T; 29'C PWHT)
r..-mmm thick. o 14Srnm thick _. tnmm thick <
m 1n::
i' E
"' i . -t, ''t=-----1,,
3""t----! i -'' .
p ' ,, i !
g IS'J --t
w [J
toll , t.
----- ... M --== ..
moo 21200
Figure 15: lnfluencc of the thickness and the tempering
parameter on impact properties for grade 2%Cr-1 Mo-%V.
To summarize, the optimization carried out on 2'.4Cr-
1Mo-'4V steel has allowed to reach the level of mechanical
properties required for 250 mm thick plates while respecting a
LMP at 21200.
CONCLUSIONS
Modern steels have to be considered with care when
addressing the way pressure vessels fabrication will be
performed, in particular the associated heat treatments such as
PWHT. The need to make service conditions safer and safer has
lead to stringent specifications for steel supply. The best
compromise must then be found between materials properties
and heat treatments needed for welds. Too high post weld heat
treatment will deteriorate the base metal properties and
consequently decrease the service performance of the vessels.
Some improvements can be made but always targeting the
whole behaviour of the equipment. This can be done through
adaptation of chemical composition, especially for CrMo V
steels, or manufacturing route as in the case of C-Mn where
quench and temper can bring solutions for very thick products.
To save some PWHT margin, it is also possible to perform the
tempering treatment at a lower temperature than the PWHT and
this solution is more and more used today in the industry to
match both properties and manufacturing constraint at the same
time, with no adverse effect on the final properties of the parent
metal nor the weld.
REFERENCES
[I) Jaffe L.D., Buffum D.C., "Upper-Nose Embrittlement in
Ni-Cr Steel", Trans. AIME, 1957.
[2] Libsch, J.F., Powers A.E., Bhat G, IVfemper Embrittlement
in Plain Carbon Steels", Trans. ASM, Vo1.44, pp.1058-
1075, 1952.
[3] Pellini \V.S., Queneau B.R., "Development of Temper
Ernbrittlement in Alloy Steel", Trans. ASM, Vol.39,
pp.l36-161, 1947.
8 Copyright 2010 by ASME

S-ar putea să vă placă și