Upload
others
View
6
Download
0
Embed Size (px)
Citation preview
Experimental Testing and Failure
Analysis for High Temperature Plant
Environments
Department of Mechanical Engineering
June 13th, 2017
Dr Catrin Mair Davies
Dr Keith Tarnowski, Dr Joe Corcoran, Mr Mustafa Nasser,
Dr Priyesh Kapadia, Dr Fred Cegla, Dr Marc Chevalier
Outline
Structural Integrity Challenges for Plant Operation
High Temperature Operation
Flexible Operation
Welding and Residual Stress Effects
Environmental Effects
Research Challenges for Current and Future Plant Operation
2
High Temperature Plant Operation
High operating temperatures is key to exploit power plant efficiency
Failure of high temperature plant components need needs to be accurately
predicted for existing and future plant
Failure principally due to:
Creep and fatigue processes
Enhanced by Residual Stresses
Principally in Weldments
Significant research into creep deformation and crack growth mechanisms
3
In-service failure of a branched
connection
Grain Boundary Creep Cavitation
Discrete micro cracks
Macroscopic crack growth
Reheat crack in 316H header component
Flexible Plant Operation
4
Flexibility is the ability to adapt to dynamic and changing conditions
balancing supply and demand by the hour or minute
Wind energy is a dominant renewable energy source
Renewables are intermittent – not easy to forecast availability
Need alternative supply to rapidly respond to fluctuations from renewables
20000
25000
30000
35000
40000
45000
50000
55000
20
17
-01
-25
15
:25
:01
20
17
-01
-26
07
:40
:03
20
17
-01
-26
23
:55
:01
20
17
-01
-27
16
:10
:04
20
17
-01
-28
08
:25
:01
20
17
-01
-29
00
:40
:01
20
17
-01
-29
16
:55
:04
20
17
-01
-30
09
:10
:01
20
17
-01
-31
01
:25
:05
20
17
-01
-31
17
:40
:04
20
17
-02
-01
09
:55
:04
20
17
-02
-02
02
:10
:01
20
17
-02
-02
18
:25
:04
20
17
-02
-03
10
:40
:05
20
17
-02
-04
02
:55
:02
20
17
-02
-04
19
:10
:02
20
17
-02
-05
11
:25
:01
20
17
-02
-06
03
:40
:04
20
17
-02
-06
19
:55
:02
20
17
-02
-07
12
:10
:04
20
17
-02
-08
04
:25
:03
20
17
-02
-08
20
:40
:43
20
17
-02
-09
12
:55
:04
20
17
-02
-10
05
:10
:01
20
17
-02
-10
21
:25
:04
20
17
-02
-11
13
:40
:04
20
17
-02
-12
05
:55
:01
20
17
-02
-12
22
:10
:04
20
17
-02
-13
14
:25
:04
20
17
-02
-14
06
:40
:02
20
17
-02
-14
22
:55
:04
20
17
-02
-15
15
:10
:01
20
17
-02
-16
07
:25
:01
20
17
-02
-16
23
:40
:09
20
17
-02
-17
15
:55
:05
20
17
-02
-18
08
:10
:02
20
17
-02
-19
00
:25
:01
20
17
-02
-19
16
:40
:02
20
17
-02
-20
08
:55
:04
20
17
-02
-21
01
:10
:04
20
17
-02
-21
17
:25
:04
20
17
-02
-22
09
:40
:05
20
17
-02
-23
01
:55
:04
20
17
-02
-23
18
:10
:04
20
17
-02
-24
10
:25
:02
20
17
-02
-25
02
:40
:04
Po
we
r (G
W)
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
20
17
-01
-25
15:2
5:0
1 2
017
-01
-26
07:4
0:0
3 2
017
-01
-26
23:5
5:0
1 2
017
-01
-27
16:1
0:0
4 2
017
-01
-28
08:2
5:0
1 2
017
-01-2
9 0
0:4
0:0
1 2
017
-01
-29
16:5
5:0
4 2
017
-01
-30
09:1
0:0
1 2
017
-01
-31
01:2
5:0
5 2
017
-01
-31
17:4
0:0
4 2
017
-02-0
1 0
9:5
5:0
4 2
017
-02
-02
02:1
0:0
1 2
017
-02
-02
18:2
5:0
4 2
017
-02
-03
10:4
0:0
5 2
017
-02
-04
02:5
5:0
2 2
017
-02
-04
19:1
0:0
2 2
017
-02
-05
11:2
5:0
1 2
017
-02
-06
03:4
0:0
4 2
017
-02
-06
19:5
5:0
2 2
017
-02
-07
12:1
0:0
4 2
017
-02
-08
04:2
5:0
3 2
017
-02
-08
20:4
0:4
3 2
017
-02
-09
12:5
5:0
4 2
017
-02
-10
05:1
0:0
1 2
017
-02
-10
21:2
5:0
4 2
017
-02
-11
13:4
0:0
4 2
017
-02
-12
05:5
5:0
1 2
017
-02
-12
22:1
0:0
4 2
017
-02
-13
14:2
5:0
4 2
017
-02
-14
06:4
0:0
2 2
017
-02
-14
22:5
5:0
4 2
017
-02-1
5 1
5:1
0:0
1 2
017
-02
-16
07:2
5:0
1 2
017
-02
-16
23:4
0:0
9 2
017
-02
-17
15:5
5:0
5 2
017
-02
-18
08:1
0:0
2 2
017
-02
-19
00:2
5:0
1 2
017
-02
-19
16:4
0:0
2 2
017
-02
-20
08:5
5:0
4 2
017
-02
-21
01:1
0:0
4 2
017
-02
-21
17:2
5:0
4 2
017
-02
-22
09:4
0:0
5 2
017
-02
-23
01:5
5:0
4 2
017
-02
-23
18:1
0:0
4 2
017
-02
-24
10:2
5:0
2 2
017
-02
-25
02:4
0:0
4
Po
we
r (G
W)
Wind availability Jan/Feb 2017 UK Energy demands Jan/Feb 2017
Power Demand and Response
5
0
10000
20000
30000
40000
50000
60000 2
017
-01
-25
15:2
5:0
1 2
017
-01
-26
09:3
5:1
2 2
017
-01
-27
03:4
5:0
1 2
017
-01
-27
21:5
5:0
2 2
017
-01-2
8 1
6:0
5:0
1 2
017
-01
-29
10:1
5:0
1 2
017
-01
-30
04:2
5:0
4 2
017
-01
-30
22:3
5:0
5 2
017
-01
-31
16:4
5:0
1 2
017
-02
-01
10:5
5:0
1 2
017
-02
-02
05:0
5:0
4 2
017
-02-0
2 2
3:1
5:0
1 2
017
-02
-03
17:2
5:0
3 2
017
-02
-04
11:3
5:0
2 2
017
-02
-05
05:4
5:0
1 2
017
-02
-05
23:5
5:0
4 2
017
-02
-06
18:0
5:0
4 2
017
-02
-07
12:1
5:0
4 2
017
-02
-08
06:2
5:0
3 2
017
-02
-09
00:3
5:0
2 2
017
-02
-09
18:4
5:0
1 2
017
-02
-10
12:5
5:0
4 2
017
-02
-11
07:0
5:0
1 2
017
-02
-12
01:1
5:0
2 2
017
-02
-12
19:2
5:0
1 2
017
-02
-13
13:3
5:0
5 2
017
-02
-14
07:4
5:0
2 2
017
-02
-15
01:5
5:0
4 2
017
-02
-15
20:0
5:0
2 2
017
-02
-16
14:1
5:0
3 2
017
-02
-17
08:2
5:0
2 2
017
-02
-18
02:3
5:0
4 2
017
-02
-18
20:4
5:0
1 2
017
-02
-19
14:5
5:0
1 2
017
-02
-20
09:0
5:0
4 2
017
-02
-21
03:1
5:0
4 2
017
-02
-21
21:2
5:0
1 2
017
-02
-22
15:3
5:0
4 2
017
-02
-23
09:4
5:0
2 2
017
-02
-24
03:5
5:0
4 2
017
-02
-24
22:0
5:0
3
Demand
CCGT
Coal
Wind
Power
Generation
for Month
P
ow
er
(GW
)
The Issue with Flexing
Most existing power plant designed for sustained operation at full load to
Maximize efficiency
Reliability
The need for flexibility causing
more frequent shutdowns when market or grid conditions warrant
more aggressive ramp rates (rate of output change)
6
The Issue with Flexing
Plant flexing leads to thermal fatigue failures
Fatigue damage interacts with creep damage to cause
Interactive creep-fatigue failures
Next generation plant need to accommodate flexing
7
Cracking at pipe penetration due to cyclic operation
(Courtesy of E.ON)
Cold feedwater introduced to a hot header caused the crack in this economizer header. The cold water created a large through-wall temperature gradient change in temperature during startup and during off-line top-off opportunities. Courtesy of EPRI
Environmental Effects : Carburisation
Advanced Gas Cooled Reactors (AGR)
Primary coolant gas:
CO2, CO, CH4, H, H2O
Stainless Steel 316H components
exposed at high temperatures (550°C)
Creep
Carburisation corrosion forms a
hardened outer surface layer
Component cracking.
Creep-Carburisation interaction not
considered in defect assessment
procedures.
Carburisation
Outer Surface
Inner Bulk
550C
CO2
Coolant Carburisation
H2O
316H
Oxide
Oxide
8
Research Challenges
Experimental Techniques for Creep Strain Characterisation
Parent material and weldments
Understand and Predict:
The role of residual stresses on creep crack initiation and
growth
Interactive creep-fatigue effects on failure
Influence of long-term operation and environment on the
integrity of plant components
• Carburisation for AGR components
Develop and Implement:
Plant monitoring
Lifetime prediction techniques
9
Weld Characterisation
Components mainly joined by welds
Complex inhomogeneous microstructures
Gradient of material properties
Parent material
Fusion zone
Heat effected zone (HAZ)
Deformation properties characterised
Digital Image Correlation
10
Laser Beam Weld in
Aluminium
0
50
100
150
200
250
300
350
400
0.00 0.02 0.04 0.06 0.08 0.10
Tru
e s
tre
ss
(M
Pa
)
True Strain
HAZ 2
HAZ 1
Weld
Weld Characterisation
Inhomogeneous elastic-plastic material properties determined
Incorporated in finite element analyses
11
High Temperature DIC
Custom designed furnace
2D or 3D DIC
Tensile or Fracture Sample
Time dependent creep strain distribution mapping
CMOD measurement
Creep Strain Evolution at 760 °C
13
Start 100h 300h 500h
630h 760h 835h Failure @ 843h
Creep Strain
Low Frequency ACPD Strain Measurement Technique
Technique is based on a square array, directional electrode configuration
Measures sub-surface
Small currents (~100 mA) sufficient even in quasi-DC regime
Phase detection in ACPD provides superior measurements to DC
Creep can be monitored through variation in the resistance ratio
Non-directional thermal effects rejected by calculating the resistance ratio
Permanently attached probe behaves like strain gauge
Geometrical variations have larger effect than microstructural changes
14
I V V I
11
1
VR
I
22
2
VR
I
1 2R R
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0 500 1000 1500 2000 2500
PD
Me
as
ure
d S
tra
in
Time (hours)
Low Frequency ACPD Strain Measurement
Local creep strain measurements by PD method
PD measurements verified by LVDT
Average array strain equals the global strain measurement
Can measure micro-crack formation prior to sample failure 15
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0 500 1000 1500 2000 2500S
tra
in
Time (hours)
LVDT
NRR Inversion
PD Measurement on Welded Sample
Rate of change in resistance clearly indicates onset of failure
Failure in HAZ region 16
0.000
0.002
0.004
0.006
0.008
0.010
0 200 400 600
ΔN
RR
/Δt
Time (hours)
CMV/2.25Cr 80 MPa 625 °C
Measuring Crack Initiation and Growth
Creep crack initiation (CCI) occupies large fraction of a components
lifetime
Verifying CCI models requires accurate experimental measurements
DCPD often used to monitor crack growth
Noisy signals
Requires assumptions for CCI point
17
Issues with current PD measurement technique and standards:
Any change in PD after load-up is attributed to crack growth
Method does not differentiate between creep strain and crack growth
Plasticity can also effect measurements
Challenges:
Reduce the noise in the PD response
Identify the point of CCI on the PD response
0.000
0.002
0.004
0.006
0.008
0.010
0.012
0.014
0 200 400 600 800
ΔV
/V0
Time [h]
Measuring Crack Initiation and Growth
18
CCI measurement
Recently demonstrated a point of inflection occurs on a plot of PD vs. CMOD
CCI times can be measured more accurately
19
0.00
0.01
0.02
0.03
0.04
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
ΔV
/V0
CMOD [mm]
Exp: Load-Up Exp: Creep FE: Load-up FE: Creep
Onset of Crack Growth in FE Model
Experimental Observation of CCI
• 4 nominally identical tests on C(T) samples of 316H SS:
CCG316_CT01: Significant crack growth
CCG316_CT02: Interrupted after 0.2 mm CCG
CCG316_CT03: Interrupted at point of inflection
CCG316_CT04: Interrupted before point of inflection
20
-0.02
0.00
0.02
0.04
0.06
0.08
0.0 0.1 0.2 0.3 0.4
ΔV
/V0
CMOD [mm]
550 °C
Interrupted Tests: CCG316_CT03
2.5mm from Mid-plane
5.0mm from Mid-plane 7.5mm from Mid-plane
Mid-plane
• Δa = 0.01 mm
• ΔaASTM = 0.02 mm
• ΔaNEW = 0.01 mm
Interrupted at point of inflection
Interrupted Tests: CCG316_CT04
2.5mm from Mid-plane
7.5mm from Mid-plane
Mid-plane
5.0mm from Mid-plane
• Δa = 0.00 mm
• ΔaASTM = 0.02 mm
• ΔaNEW = 0.00 mm
Interrupted before point of
inflection
Role of Residual Stresses on Creep Crack Initiation and Growth
Range of techniques to introduce residual stress (RS) into test samples
Difficult to distinguish role of RS and plasticity
23 Data for pre-compressed and side punched specimens from: Hossain, S. et al. Fatigue & Fracture of Engineering Materials & Structures, 2011,
34(9), pp. 654-666 Turski, M. et al. Acta Materialia, 2008, 56(14), pp. 3598-3612
Pre-Compressed
Side Punched Wedge-Loaded
EB Welded
EB and Wedge Loaded Sample
EB weld – high stress triaxiality
Combined loading possible
24
1 mm 0.5
0.4
0.3
0.2
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
FE DIC
Wedge doesn’t require test machine
Can easily be tested in environment
Metallography shows intergranular crack growth
Crack tunnelling – plane strain conditions and larger K at mid-thickness
WC(T)10 at 550°C for 1,004 hours:
Crack Growth Metallography
25 3.1 mm
4.5 mm
5.0 mm
5.8 mm
6.0 mm
CCG
Post test
fatigue
EDM
Notch
Crack Growth Measurements
26
WC(T)9
Specimen
ID
WC(T)8 44
WC(T)9 47
WC(T)10 44
Crack growth measured using low-frequency ACPD system
Crack Growth Prediction of Wedge Samples in FE
Creep deformation modelled using RCC-MR
model which includes primary and secondary
creep regimes
Implemented in ABAQUS using a CREEP
subroutine with time hardening
Damaged defined by:
Once elements are damaged (D = 1), elements
are effectively removed by reducing the load
bearing capacity using the USDFLD subroutine
27
Crack growth estimates at mid-thickness of the samples compared to
measurements in EB welded and wedge loaded C(T) specimens
Crack Growth Measurements vs Predictions
28
0
2
4
6
8
10
0 2 4 6 8 10
Ob
se
rve
d Δ
a (
mm
)
Predicted Δa (mm)
Non-conservative
Conservative
WC(T)8
FE Prediction
Characterising Creep-Fatigue Deformation
Perform low-cycle-fatigue test of prior crept sample
29
P91 Steel
600 °C
Characterising Creep-Fatigue Crack Growth Properties
Creep crack growth accelerated by fatigue loading
30
da
/dt (m
m/h
)
1E+01
1E+00
1E-01
1E-02
1E-03
1E-04
1E-05
C* (MPam/h) 1E-06 1E-05 1E-04 1E-03 1E-02 1E-01 1E+00
Plant Monitoring Tools
Techniques developed to monitor creep strain and crack growth
31
FurnaceUltrasonic
array
Specimen
DCPD
High Temperature
Ultrasonic Crack
Growth Imaging
ACPD Creep Strain
Monitoring of Plant
Components
Pipeline Life Assessment Tool (PLATO)
Developing a modelling tool that can be used to predict failure of power
plant pipelines
Estimate pipeline stresses during normal operation and fault conditions
Automate generation of the FE model, analysis and post-processing to
assess accumulation of creep-fatigue damage during service
32
Damage
Parameter Model
whole
pipeline
Sub model
of bend
region
Sub-Model of Pipe Bend Region
33
Global Model
Deadweight Analysis
Sub-Model
Deadweight Analysis
Thermal Model
Thermal Analysis
Global Model
Stress Analysis
Sub-Model
Stress Analysis
Fatigue Assessment
Sub-Model
Creep Analysis
Pipeline displacement determined from global model:
Through thickness temperature
variation from thermal model:
Temperature fields
transferred to sub-model
using pre-defined fields
Sub-model to determine pipe
stresses:
Global model
used to define
boundary
conditions of
sub-model
Pipe Stresses (1)
34
Global Model
Deadweight Analysis
Sub-Model
Deadweight Analysis
Thermal Model
Thermal Analysis
Global Model
Stress Analysis
Sub-Model
Stress Analysis
Fatigue Assessment
Sub-Model
Creep Analysis
Normal operation Hanger S28 locked up
S28
Bend intrados
Pipe Stresses (2)
35
Creep and Fatigue Damage Modelling
36
Uniaxial Creep Testing – Rupture Life
Carburised
As-Received
Carburised
As-Received
37
Carburisation Effects : Creep Strain Rate and Rupture
38
100
1000
100 1000 10000
Tru
e S
tres
s (M
Pa)
Rupture Time (h)
As-Received
Carburised
Carburised Uniaxial Creep - Magnification x 15
x 350 A
A
Carburisation – Fatigue Interaction
39
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
100 1000 10000 100000
Tota
l S
trai
n R
ang
e (%
)
Nf30
Carburised
As-Received
R66 Best Estimate
Carburised LCF - Magnification x 20
x 200
Widespread Surface
Cracking
Major Fatigue Cracks
Oxide Spallation
TOF SIMS Analysis
conducted by the
University of Manchester
Summary
Significant experience in high temperature plant operation
New challenges due to flexible operation
Creep-fatigue interaction effects can be significant and require detailed
understanding
DIC and Low Freq. ACPD Technique provide important information on
creep strain development in weldments
Low Freq. ACPD technique provides more accurate measurement of CCI
Wedge loaded samples are effective in determining influence of residual
stress on CCI and CCG
CCG can be accurately predicted using FE models
Plant monitoring tools being developed to monitor creep strain and image
crack growth
FE based Pipeline Life Assessment Tool being developed
Significant challenges remain in understanding the role of environment
effects on creep and creep-fatigue
40