Engineering Failure Analysis: Warwick Payten, Damien Charman, Andrew Chapman, Kourosh Bashiri, Phillip Bendeich

Engineering Failure Analysis xxx (2011) xxx–xxx
Contents lists available at ScienceDirect
Engineering Failure Analysis

journal homepage: www.elsevier.com/locate/engfailanal
Life assessment and inspection of a hot reheat turbine bifurcation

Warwick Payten a,⇑, Damien Charman b, Andrew Chapman b, Kourosh Bashiri b, Phillip Bendeich a
a
Australian Nuclear Science and Technology Organisation (ANSTO), Locked Bag 2001, Kirrawee DC, NSW 2232, Australia
b
Austpower Engineering, PO Box 1184, Warners Bay, NSW 2282, Australia
a r t i c l e i n f o a b s t r a c t
Article history: A life assessment of a hot reheat bifurcation has been undertaken. Surface breaking crack-
Received 16 October 2010 like indications were detected in the bifurcation weld. Subsequent replication showed that
Received in revised form 18 April 2011 this was due to significant creep damage in a number of material zones within the weld.
Accepted 19 April 2011
This damage was removed by hand dressing until it had been reduced to an acceptable
Available online xxxx
level below which good material remained with minimal damage; this was confirmed by
additional replication. A finite element analysis was undertaken based on the elastic rup-
Keywords:
ture reference stress multiplied by weld redistribution factors which quantifies stress
Creep
Stress rupture
redistribution within the weldments. The results indicated that creep damage would be
Weld evident on both the external surface of the saddle region and on the inside surface at
Bifurcation the crotch position. Additional analysis based on a RIKS procedure and an elastic-perfectly
plastic analysis confirmed the results of the rupture reference method. A full re-inspection
of the bifurcation branch was recommended before another 10,000 h of service was com-
pleted as a consequence of this study. At the next inspection recurrence of the damage was
found, and internal inspection confirmed the analysis results with damage also present on
the internal surface at the crotch position. Based on these findings replacement of the
bifurcation has been recommended.
Crown Copyright Ó 2011 Published by Elsevier Ltd. All rights reserved.
1. Introduction
High energy piping systems are designed to operate at high temperatures and high pressures, typically 540–560 °C and
40–60 MPa hoop stress; hot reheat pipes are included in this category. The remaining life of large high energy welded com-
ponents subject to creep is often limited by the life of the welds. Experience and testing has shown that the creep strength of
welds, and in particular of the heat-affected zone (HAZ), is generally less than that of the parent material. Estimating the
remaining life of welded vessels then often becomes a matter of estimating the remaining creep life of the welds. Welds
are heterogeneous, and the various material zones; the weld metal and the different regions of HAZ all have different creep
properties. During stress relaxation caused by creep, the complex geometry and material properties leads to stress redistri-
bution as a result of off-loading stresses occurring in the various material zones.
Failure of these types of branch intersections generally occurs at welds in the inter-critical zone of the HAZ and is cate-
gorised as Type IV failure [1]. Cr–Mo–V weldments are particularly prone to this type of cracking mechanism. Type IV fail-
ures is a worldwide problem in power generation systems and is attributed to localised microstructral region that has low
creep strength [2,3]. In large megawatt (MW) coal fired steam generation stations, the intermediate pressure turbine is gen-
erally fed by the hot reheat (HRH) line. In this unit, the HRH line was made up of four legs A to D that link to two bifurcations
C and D legs and A and B legs. This paper discusses the life assessment carried out on the HRH turbine bifurcations and
⇑ Corresponding author.
E-mail address: wmp@ansto.gov.au (W. Payten).
1350-6307/$ - see front matter Crown Copyright Ó 2011 Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.engfailanal.2011.04.010
Please cite this article in press as: Payten W et al. Life assessment and inspection of a hot reheat turbine bifurcation. Eng Fail Anal (2011),
2 W. Payten et al. / Engineering Failure Analysis xxx (2011) xxx–xxx
D Leg Crotch Positions

6 and 12 o’clock
C-Leg Turbine Terminal

Saddle Position Point
3 and 9 o’clock
Fig. 1. Diagram showing the four legs and two bifurcations of the hot reheat line. Welds tested are indicated.
details the findings for the C–D leg which was found to be the most significantly damaged leg this is illustrated in Fig. 1. The
body of the branch piece has an outer radius of 435 mm and a wall thickness of 57.5 mm. The branch pipe had an outer ra-
dius of 319 mm, and a wall thickness of 52.3 mm it has been manufactured from a 0.5Cr–0.5Mo–0.25V steel. A pressure load-
ing of 4 MPa was applied on the internal surface. This equates to an axial end loading of 11.39 MPa on the main line, and
9.89 MPa on the branch line. The unit service hours were 187,974.
2. Metallurgical inspection
Inspections were carried out at a power station at various welds that formed part of the HRH turbine bifurcations. The
inspection techniques used incorporated magnetic particle (MT), ultrasonics (UT) and replication. Creep damage was found
in a number of the welds. Microstructures were examined by taking surface replicas at both HAZs of the nominated welds, at
four test areas that were approximately 90° apart around the circumference.
2.1. Magnetic particle
Surface breaking crack-like indications (CLIs) were detected at the ‘‘D’’ leg branch weld. Intermittent indications with a
circumferential orientation were seen around the 3 o’clock location within the ‘‘C’’ leg side of the HAZ in a band that was
approximately 250 mm long. The longest single indication inside the cracking band was approximately 50 mm in length,
as shown in Fig. 2.
All of the CLIs were removed by hand dressing with a grinder leaving a profile similar to that shown in Fig. 3. Prior to
dressing, the HAZ’s were defined by macro-etching and marked up to allow placement of a groove 2 mm deep at the required
areas. These were subsequently blended to obtain a smooth profile when MT showed no further cracking.
Fig. 2. The crack-like indications (CLIs) that were revealed by magnetic particle testing at the bifurcation branch welds are shown.
W. Payten et al. / Engineering Failure Analysis xxx (2011) xxx–xxx 3
Fig. 3. A typical excavation profile that was used to remove surface breaking CLIs.
2.2. Ultrasonics
The CLIs revealed by MT at the branch welds did not exceed 5 mm in depth. There were no recordable indications or
reflectors observed at all other parts of the branch welds. The nominal measured parent material thickness for the bifurca-
tion ‘‘D’’ leg was 58 mm.
2.3. Replication
The replication assessment and classification guide uses a system based on the Neubauer system and the current ECCC
recommendations [4], in that it incorporates and simplifies provisions in the SAQ modified Neubauer classification system
and the VGB Guidelines classifications [4]. Tables 1 and 2 detail the modified system that was used.
Cavitation orientation was represented by classes A to E and cavitation density as a function of orientation is shown in
Table 2. Significant creep damage was detected at the ‘‘D’’ leg branch weld (Fig. 4) and the CLIs revealed by MT were caused
by creep cracks in the Type IV region of the HAZ. Damage appeared at diametrically opposed locations, and these were the
areas where stresses should be highest. Figs. 5–7 illustrate the creep damage that developed in the parent material and HAZ
at the 3 o’clock location. The 9 o’clock position had marginally reduced damage. Clearly, Type IV HAZ cracking was pro-
nounced. Damage in the parent material and coarse grain HAZ regions was similar in extent and featured the occurrence
of large elongated voids. Hand-dressed excavations, however, reduced the cavity densities to a minor level, suggesting that
surface-prevalent stresses were responsible for the damage. The weld metal near the 3 o’clock location showed oriented cav-
ities at a moderate density which in itself is a strong indicator of the large service stresses that acted. Typical weld metal
creep is illustrated by Fig. 8. Inspection during 2005 reported cracking at the 3 o’clock location, but it was classified as dor-
mant due to its heavily oxidised state. Also, virtually no creep features were observed which is in stark contrast to the pres-
ent condition. A significant increase in damage had occurred during the last 28,058 h of service, suggesting that a major
Table 1
Cavitation orientation.
Classification Damage
A No damage
B Isolated cavitation
C Oriented cavitation
D Microcracking
E Macrocracking
Table 2
Cavitation density (voids/mm2).
Damage number Voids/mm2 Description

1 0–100 Minor
2 100–250 Mild
3 250–500 Moderate
4 500–1000 Significant
5 >1000 Severe
Fig. 4. Two locations where creep cracks were removed by hand dressing are identified. Remnants of the line that marked the HAZs can be seen.
Fig. 5. Parent material at the leg ‘‘C’’ HAZ 600.
Fig. 6. Type IV region of the leg ‘‘C’’ HAZ 600.
operating change or a possible acceleration of creep damage was responsible for the attendant loss of creep life. Details of the
replication assessment are shown in Table 3.
2.4. Twelve month follow-up inspection
A 12 month follow-up inspection was recommended based on the extent of damage found in the previous inspection and
as a result of the finite element modelling results described in the following sections. At the follow-up inspection, it was also
Fig. 7. Coarse grain region of the leg ‘‘C’’ HAZ 600.
Fig. 8. Typical oriented creep voids in the weld metal microstructure from the leg ‘‘D’’ branch weld at the 3 o’clock position. Magnification: 600.
Table 3
HRH turbine bifurcation ‘‘C–D’’ leg branch weld.
Location Parent Type IV HAZ Coarse grain HAZ Weld

3 o’clock, leg ‘‘C’’ HAZ C4 E5 D5 C3
3 o’clock leg ‘‘C’’ HAZ excavated 5 mm B2 D4 C5 B1
3 o’clock, leg ‘‘C’’ HAZ excavated 8 mm B1 B1 B1 A
9 o’clock, leg ‘‘C’’ HAZ B4 D5 D5 B1
9 o’clock leg ‘‘C’’ HAZ excavated 3 mm B1 B1 B2 A
recommended that as access by boson chair into the branch piece was feasible, additional replication be performed on the
internal surface at both the saddle and crotch positions as considerable concern were raised that creep damage would also be
progressing from the inside surface based on the FEA results.
The results of the follow-up inspection (which occurred after 10 months from the original inspection) observed creep
microcracking in the C/D branch weld at the 9 o’clock position (Fig. 9). This microcracking had initiated since the previous
outage and was not a remnant of the previous cracking as it had a different morphology. The previous creep cracking
appeared as numerous large creep voids, where the present microcracking had a fine intergranular appearance. Both
forms of cracking are creep cracking occurring at the parent metal/Type IV HAZ interface. Microcrcaking was also present
in the coarse grain HAZ (Fig. 10). A further 3 mm excavation was required to reduce the damage to a minor level. It was
also recommended that replicas be taken on the inside surface as a consequence of the finite element analysis discussed
on the following sections. Replicas were taken across the branch weld HAZ in the bottom crotch position; results are de-
tailed in Table 4. Moderate creep damage in the Type IV HAZ and creep microcracking was observed in the C/D bifurca-
tion at the 6 o’clock position (Figs. 11 and 12). Replicas on the internal surface at the saddle position found no creep
damage.
Fig. 9. Creep microcracking observed in the leg C/D branch weld, 9 o’clock, leg C side, Type IV HAZ (mag. 200).
Fig. 10. Creep microcracking observed in the leg C/D branch weld, 9 o’clock, leg C side, coarse grain HAZ (mag. 400).
Table 4
HRH turbine bifurcation ‘‘C–D’’ leg branch weld Internal.
Location Parent Type IV HAZ Coarse grain HAZ Weld

6 o’clock, leg C HAZ A B3 D4 A
Fig. 11. Creep microcracking observed in the leg C/D branch weld, internal, 6 o’clock, leg C side, coarse grain HAZ (mag. 400).
Fig. 12. Moderate creep damage observed in the leg C/D branch weld, internal, 6 o’clcok, leg C side, Type IV HAZ (mag. 400).
3. Finite element analysis
3.1. Modelling
The models were generated using 3D solid tetrahedral elements. Due to symmetry only one half of the model was re-
quired to be generated. The finite element mesh consisted of approximately 400,000 elements, (Fig. 13). Although the weld
zones are not explicitly included in the model, the external surface weld profile (blend) was included.
3.2. Rupture reference stress calculation
A previous analysis of a branch [5] used a more detailed steady-state finite element creep analysis, where the weld zones
were explicitly included, and the stresses relaxed using a Norton creep law. Rupture times were estimated following the
methodology presented in [6]. A similar branch was used by Yu et al. [7] and extended to include a continuum damage
mechanics (CDM) approach. Some examples of other papers where CDM has been used to model weld zones for CMV mate-
rials include Sun et al. [8], Hyde and Sun [6], Hyde et al. [9,10], Hayhurst et al. [11,12], and Perrin and Hayhurst [13]. As a
result of the relative thinness of the branch and the large angle of the branch to include the three zones, parent metal (PM),
weld metal (WM) and HAZ, would be a more difficult meshing excise then a standard right angle branch. The commercial
realities of the project necessitated a quick turn around time and hence, a creep relaxed model or a more complex explicit
CDM model was beyond the scope of this project. As a result, the rupture reference stress was used and calculated via a num-
ber of methods based on Patel [14] and Volume 7 of the R5 methodology [15]. The method as well as examples of large scale
testing are also described in Budden [16,17].
The rupture reference stress defined in R5 for homogenous component is:
P W ry
rref ¼ ð1Þ
PL
Fig. 13. General dimensions and Meshed ABAQUS model.
where PL is the limit load for yield stress ry. The reference stress is in proportion to the working load Pw. The ratio of ry/PL is
independent of yield stress for a homogenous component.
The above relationship can for structures where the loading may be represented as membrane and bending stress resul-
tants be described by the following relationship:
" #12
2
PB PB 2
rref ¼ þ þ PL ð2Þ
3 3
The rupture reference stress rRref for ductile materials with a secondary creep stress exponent less then 7 is then calcu-
lated from R5 as:
rRref ¼ f1 þ 0:13½v 1grref ð3Þ

r el;max
where v ¼ is an adjustment factor for stress concentrations, and r
rref ,
el;max is the maximum elastically calculated equiva-
lent stress range.
The homogenous rupture reference stress, rRref is multiplied by a weld redistribution factor, k [3], which quantifies stress
redistribution within the weldments.
A second direct method based on using Eq. (1) was also analysed to provide a more accurate determination of the rupture
reference stress. This was based on the use of an elastic perfectly plastic FE limit load solution and a RIKS algorithm within
ABAQUS [18]. Once the limit load is found by either technique it can be directly used with Eq. (1) to calculated the rupture
reference stress.
3.3. Material data
The elastic material data used is detailed as follows:
Young’s modulus ¼ 178; 000 MPa

Poisson’s ratio ¼ 0:3
The welds are characterised by four distinct zones:
(1) Parent ½ CMV steel.

(2) 21=4 Cr1Mo filler.
(3) Inter-critical (Type IV heat affected zone (HAZ) material).
(4) Coarse grained HAZ (CGHAZ).
The HAZ material zone is the weakest zone with the weld metal stronger, and the parent metal the strongest, with respect
to the minimum creep rate.
3.3.1. Parent ½CMV steel

In PD6525 [5], the creep rupture data for 0.5Cr–0.5Mo–0.25V steel are represented by:
log t log ta
PðrÞ ¼ a þ b log r þ cðlog rÞ2 þ dðlog rÞ3 þ eðlog rÞ4 ¼ ð4Þ
ðT T a Þr
where P(r) is the creep rupture parameter; T is the temperature in K; t is the rupture time in h; r is the stress in MPa; r is the
temperature exponent; and ta, Ta, and a to e are constants. The creep expression is limited to a relatively high stress range of
123.5–308.7 MPa, and it exhibits turn-back at stresses lower than 61 MPa. A Sinh extrapolation [19] was used for lower
stresses.
3.3.2. 2 1=4 Cr1Mo weld

Eq. (4) again applies to the weld material with the constants were taken from PD6525 [20] for 2 1/2Cr1Mo.
3.3.3. Type IV zone

The following relationship [14] was used to describe the rupture properties of the Type IV zone. The data is valid for stress
ranges of 35–154 MPa and is valid provided the calculated rupture life is less than 5 105 h.
logðt r Þ 20:01
PðrÞ ¼ ¼ 2:081817 102 3:82507799 105 r þ 6:7801245 108 r2 ½20 ð5Þ
ðT þ 104Þ
3.3.4. Coarse grained HAZ (CGHAZ)

As insufficient data is available for the CGHAZ, similar to [14], the mean parent rupture data is used to provide a lower
bound estimate.
3.3.5. Stress redistribution factor k

From Volume 7 of the R5 guidelines [15] the homogenous rupture reference stress is multiplied by a factor, k, which
quantifies the stress redistribution within the weldments. Similar to [14] it is judged that the maximum principal stress lies
approximately parallel to the fusion line and hence as stress redistribution occurs, k may differ from unity and vary between
zones. This analysis assumed for conservatism a value of k = 1 for the Type IV zone. The advice in R5 was taken for the other
material; k = 0.7 for the weld and k = 1.4 for the coarse-grained HAZ.
4. Results – FEA analysis
4.1. Elastic analysis rupture reference stress
The results for the elastic analysis are shown in Figs. 14 and 15. Rupture reference stresses were calculated for three posi-
tions; the saddle side where the maximum damage was found and the two crotch positions. The saddle position had a max-
imum bending stress on the outside surface whereas for the crotch positions maximum bending was on the inside surface.
Linearisation profiles for all positions were taken through the approximate position of the weld Type IV zone. Table 5 details
the results.
4.2. RIKS analysis-rupture reference stress
The results for the RIKS analysis are shown in Figs. 16 and 17. The results display the onset of plastic collapse with a load
proportionality factor of 3.465. The saddle shows through-wall plasticity just before the acute angle crotch side. The refer-
ence stress is calculated on the plastic collapse solution for the component as a whole and leads to a reference stress of
57.7 MPa and a rupture reference stress of 62.9 MPa.
Fig. 14. Elastic stress internal views.
Fig. 15. Elastic stress saddle side.
Table 5
Stress components and rupture reference stress calculation.
Stress (MPa) Saddle Open crotch side Closed crotch side

Membrane 25.2 50.17 42.4
Bending 26 26.7 41.4
Peak 73.5 73 98.7
Rupture ref. stress 40.28 61.57 63.59
Fig. 16. RIKS analysis saddle side.
Fig. 17. RIKS analysis internal view.
4.3. Elastic perfectly plastic collapse rupture reference stress
The results for the elastic perfectly plastic analysis are shown in Fig. 18. The results display the onset of through-wall
plasticity. This occurs at a pressure load of 16 MPa; again the saddle shows through-wall plasticity just before the acute an-
gle crotch side. The rupture reference stress is calculated on the plastic collapse solution for the component as a whole and
leads to a rupture reference stress of 50 MPa and a rupture reference stress of 56.2 MPa.
4.4. Rupture calculations
The rupture reference stresses for the elastic method (62.9 MPa) and the RIKS method (62.98) showed similar results
when taken from the highest reference stress location (acute angle crotch position). The plastic collapse solution was lower
with a rupture reference stress of 56.2 MPa. A surprising result was the low result for the calculated rupture reference stress
for the saddle position using the elastic method (40.2 MPa).
Fig. 18. Elastic plastic analyses cut view at through wall plasticity.
It is suggest that this may have occurred due to three factors. The first is a consequence of the steep angle that the branch
forms. The second factor is the relatively thinness of this branch-piece as a ratio of the diameter. The third factor is due to the
redistribution of the stresses from the crotch side to the saddle, that cause damage to accumulate at the saddle position fas-
ter then what would have been predicted by the elastic solution.
For a right angle branch [14], results indicate that the crotch region would fail first. The RIKS and elastic–plastic methods
suggest that the saddle and acute crotch failed at almost the same time indicating stress redistribution to the saddle area not
inferred by the elastic rupture reference stress solution. Although the bifurcation at the acute angle side of the crotch had a
high rupture reference stress, no damage was seen in the current inspections at this location. It is suggested that this may be
due to the maximum bending stress occurring on the inside face of the weld, whereas the saddle side showed the maximum
bending stress occurring on the external surface.
4.4.1. Elastic stress rupture lives

The elastic rupture lives are detailed in Tables 6 and 7 for mean and lower bound material properties. Those cases ital-
icised represent analysis where the rupture life of the material zone has expired.
Table 6
Mean elastic stress rupture lives (h).
Parent Weld Type IV CGHAZ

Saddle 2.7E6 1.5E6 4.8E5 1.3E6
Acute crotch 9.8E5 7E5 1.6E5 3.1E5
Open crotch 1.0E6 7.5E5 1.8E5 3.5E5
Table 7
Lower bound elastic stress rupture lives (h).

Saddle 1.7E6 1.1E6 3.0E5 7.2E5
Acute crotch 4.9E6 3.9E5 8.3E4 1.1E5
Open crotch 5.5E6 4.4E5 9.2E4 1.3E5
Table 8
Mean bound RIKS and plastic stress rupture lives (h).

RIKS 1.0E6 7.2E5 1.7E5 3.2E5
Plastic 1.3E6 8.9E5 5.1E5 7.3E5
Table 9
Lower bound RIKS and plastic stress rupture lives.

RIKS 5.1E5 4.2E5 1.1E5 1.2E5
Plastic 7.3E5 5.6E5 1.3E5 2.0E5
4.4.2. RIKS and elastic perfectly plastic rupture lives

Both the RIKS and elastic perfectly plastic method analysis the bifurcation as a whole component and hence arrive at a
single collapse value. The results (Tables 8 and 9) are indicative of both the saddle position and the acute side of the crotch as
collapse occurred at both positions almost simultaneously.
5. Discussion
The elastic stress rupture lives show that considerable through-wall life still remains in the saddle region where the
external damage was evident. In contrast, both the acute angle and open angle crotch side showed zero remaining live
for both the mean and lower bound material properties. The elastic rupture reference stress based on Eqs. (2) and (3)
may be a poor indicator of damage for this component if the saddle position is used. As suggested previously this may be
due to the steep angle that the branch forms, the relative thinness of this branch piece and the stress redistribution to
the saddle position. Generally this method should result in a conservative measure of the reference stress. To achieve a high-
er level of confidence in the results both a RIKS and elastic purely plastic analysis where undertaken. The RIKS and plastic
method are for this component should be a less conservative and more accurate method. However, both techniques indi-
cated that the Type IV zone material had almost no remaining life for mean and lower bound properties. Additionally the
CGHAZ for lower bound properties using the RIKS method indicated an advanced state of damage. A key feature of this anal-
ysis is the weld redistribution factor k for the Type IV zone. R5 recommends that k be taken as one. However, values of k < 1
may be used for the assessment of net-section rupture for certain reference stresses, and temperatures, based on hoop strain
compatibility arguments. Hence k could be as low as 0.85 [14]. This would lead to longer calculated remaining life, how-
ever, as insufficient detail is currently available within the R5 code for this argument to be used the conservative value of one
was applied.
As a number of methods, in particular the RIKS technique, for mean bound properties showed zero or minimal remaining
life for the Type IV zone, the bifurcation appears to have limited remaining life. Replication findings confirmed these results
which revealed damage in all zones of the saddle position; however, good material was evident after blending. As a conse-
quence, the unit was allowed to return to service for a limited amount of time and a full re-inspection of the bifurcation
branch were recommended before another 10,000 h of service is completed.
The FEA models of the bifurcations showed that it was possible that creep damage and cracking may also be present at the
bore in the bottom crotch position. The next inspection confirmed the damage at the bore internal crotch position. Addition-
ally reoccurrence of new creep damage was evident at the saddle location. The implication of the finite element analysis and
the new damage accumulated since the last outage is that the bifurcation has little useful remaining life. The bore damage
now verified from the internal inspection means that if a weld repair is conducted the entirety of the branch weld material
should be removed. If a partial repair is performed by an excavation of the damaged material, without the full thickness re-
moval of the weld, the subsequent post weld heat treatment would propagate the remaining creep damage. This would in all
likelihood significantly increase the size of the present microcracking. This would also have the effect of substantially reduc-
ing the life of the repair weld, and it would have to be considered as a temporary option only. Although a full repair option is
feasible by cutting out the entire weld, a replacement option has been considered to be the better strategy at the next major
outage.
6. Conclusions and recommendations
The reportable surface-breaking flaws detected at the branch welds were active and formed by creep cracking in the Type
IV region of the HAZ. They were removed by hand dressing but were treated as a temporary measure because it did not
address the underlying causes for the localised rapid loss of creep life.
Although localised surface damage was found, the NDT assessment indicated that sound material extended into the weld.
The FEA analysis indicated that the maximum damage would occur in the Type IV material with a resultant limited
remaining life based on the RIKS method.
The analysis showed that the internal acute crotch location also had substantial damage with limited useful remaining
life in the Type IV zone.
A full re-inspection of the bifurcation branch was recommended before another 10,000 h of service was completed.
Reoccurrence of damage at the saddle was found on the next inspection
Creep microcracking in the Type IV HAZ zone was observed in the C/D branch in the current inspection which had formed
in approximately 10 months of operation.
The use of simplified reference stress concepts, and Volume 7 of the R5 code results in reasonable predictions of the dam-
age, this was confirmed by the in-service inspection replication results.
References
[1] Viswanthan R. Damage mechanisms and life assessment of high temperature components. ASM Int 1989.
[2] Smith DJ, Walker NS, Kimmins ST. Type IV creep cavity accumulation and failure in steel welds. Int J Pres Ves Pip 2003;80(9):617–27.
[3] Kimmins ST, Coleman MC, Smith DJ. Kimmins ST, Coleman MC, Smith DJ. An overview of creep failure associated with heat affected zones of ferritic
weldments. In: Wilshire B, Evans RW, editors. Proceedings of the fifth international conference on creep and fatigue of engineering materials and
structure, Swansea; 1993.
[4] Residual life assessment and microstructure. ECCC recommendations. vol. 6. AC/MC/93 WG1.1; 2005 [issue 1].
[5] Payten W. Large scale multi-zone creep finite element modelling of a main steam line branch intersection. Int J Pres Ves Pip 2006;83(5):359–64.
[6] Hyde TH, Sun W. Effect of bending load on the creep failure behaviour of a pressurised thick walled CrMoV pipe weldment. Int J Pres Ves Pip
2002;79(5):331–9.
[7] Yu T, Yatomi M, Shi H-J. Numerical investigation on the creep damage induced by void growth in heat affected zone of weldments. Int J Pres Ves Pip
2009;86(9):578–84.
[8] Sun W, Hyde TH, Becker AA, Williams JA. Comparison of the creep and damage failure prediction of the new, service-aged and repaired thick-walled
circumferential CrMoV pipe welds using material properties at 640 °C. Int J Pres Ves Pip 2000;77(7):389–98.
[9] Hyde TH, Sun W, Becker AA. Effect of geometry change on the creep failure life of a thick-walled CrMoV pipe with a circumferential weldment. Int J Pres
Ves Pip 2004;81(4):363–71.
[10] Hyde TH, Sun W, Becker AA, Williams JA. Effect of weld angle and axial load on the creep failure behaviour of an internally pressurised thick walled
CrMoV pipe weld. Int J Pres Ves Pip 2001;78(5):365–72.
[11] Hayhurst RJ, Mustata R, Hayhurst DR. Creep constitutive equations for parent, Type IV, R–HAZ, CG–HAZ and weld material in the range 565–640 °C for
Cr–Mo–V weldments. Int J Pres Ves Pip 2005;82(2):137–44.
[12] Hayhurst RJ, Vakili-Tahami F, Hayhurst DR. Verification of 3-D parallel CDM software for the analysis of creep failure in the HAZ region of Cr–Mo–V
crosswelds. Int J Pres Ves Pip 2009;86(8):475–85.
[13] Perrin IJ, Hayhurst DR. Continuum damage mechanics analyses of Type IV creep failure in ferritic steel crossweld specimens. Int J Pres Ves Pip
1999;76(9):599–617.
[14] Patel RD. Creep life assessment of welded trunnion and branch components using the R5 procedure. Int J Pres Ves Pip 2003;80(10):695–704.
[15] R5. Assessment procedure for the high temperature response of structures. Gloucester, UK: British Energy; 2003 [issue 3].
[16] Budden PJ. Validation of the high-temperature structural integrity procedure R5 by component testing. Int J Pres Ves Pip 80(7–8):517–26.
[17] Budden PJ. Analysis of the Type IV creep failures of three welded ferritic pressure vessels. Int J Pres Ves Pip 1998;75(6):509–19.
[18] ABAQUS Version 8.7. DSS Inc; 2009.
[19] Leinster MG. A method of creep rupture data extrapolation based on physical processes. Int J Pres Ves Pip 2008;85(10):701–10.
[20] PD6625-Part 1: 1990. Elevated temperature properties for steels for pressure purposes. British Standards Institute; 1990.

Engineering Failure Analysis: Warwick Payten, Damien Charman, Andrew Chapman, Kourosh Bashiri, Phillip Bendeich

Uploaded by

Copyright:

Available Formats

Engineering Failure Analysis: Warwick Payten, Damien Charman, Andrew Chapman, Kourosh Bashiri, Phillip Bendeich

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Engineering Failure Analysis: Warwick Payten, Damien Charman, Andrew Chapman, Kourosh Bashiri, Phillip Bendeich

Uploaded by

Copyright:

Available Formats

Engineering Failure Analysis xxx (2011) xxx–xxx

Contents lists available at ScienceDirect

Engineering Failure Analysis

Life assessment and inspection of a hot reheat turbine bifurcation

D Leg Crotch Positions

C-Leg Turbine Terminal

2.1. Magnetic particle

Damage number Voids/mm2 Description

Fig. 5. Parent material at the leg ‘‘C’’ HAZ 600.

Fig. 6. Type IV region of the leg ‘‘C’’ HAZ 600.

2.4. Twelve month follow-up inspection

Fig. 7. Coarse grain region of the leg ‘‘C’’ HAZ 600.

Location Parent Type IV HAZ Coarse grain HAZ Weld

Location Parent Type IV HAZ Coarse grain HAZ Weld

3. Finite element analysis

3.2. Rupture reference stress calculation

Fig. 13. General dimensions and Meshed ABAQUS model.

rRref ¼ f1 þ 0:13½v 1grref ð3Þ

3.3. Material data

The elastic material data used is detailed as follows:

Young’s modulus ¼ 178; 000 MPa

(1) Parent ½ CMV steel.

3.3.1. Parent ½CMV steel

3.3.2. 2 1=4 Cr1Mo weld

3.3.3. Type IV zone

3.3.4. Coarse grained HAZ (CGHAZ)

3.3.5. Stress redistribution factor k

4. Results – FEA analysis

4.1. Elastic analysis rupture reference stress

4.2. RIKS analysis-rupture reference stress

Fig. 14. Elastic stress internal views.

Fig. 15. Elastic stress saddle side.

Stress (MPa) Saddle Open crotch side Closed crotch side

Fig. 16. RIKS analysis saddle side.

Fig. 17. RIKS analysis internal view.

4.3. Elastic perfectly plastic collapse rupture reference stress

4.4. Rupture calculations

4.4.1. Elastic stress rupture lives

Parent Weld Type IV CGHAZ

Parent Weld Type IV CGHAZ

Parent Weld Type IV CGHAZ

Parent Weld Type IV CGHAZ

4.4.2. RIKS and elastic perfectly plastic rupture lives

6. Conclusions and recommendations

You might also like

rRref ¼ f1 þ 0:13½v 1grref ð3Þ