ML20010D272

Evaluating Flaw Detectability Under Limited Coverage Conditions

ML20010D272
Person / Time
Issue date:	01/10/2020
From:	Diaz A, Harrison J, Holmes A, Hutchinson C, Jacob R, Carol Nove, Prowant M NRC/RES/DE/CIB, Pacific Northwest National Laboratory
To:
Nove C
References
PNNL-SA-150350
Download: ML20010D272 (18)

Text

PNNL-SA-150350 Evaluating Flaw Detectability Under Limited Coverage Conditions Joel Harrison, Matthew Prowant, Aimee Holmes, Chris Hutchinson, Richard Jacob, and Aaron Diaz Technical Information Exchange Carol Nove, NRC COR

Introduction:

Limited Inspection Coverage

• Incomplete examination coverage of welds is a very common issue in the nuclear power industry that exists in every plant.

• In cases where welds susceptible to degradation are not inspectable or partially inspectable, the condition must be addressed in order to determine the structural integrity of the component

• Assuming a flaw existed in an uninspectable region, to what extent would it have to propagate into the inspectable region before it would be detected?

2

Historical Perspective of Limited Examinations

• Conventional wisdom has driven industrys position on limited examination coverage:

Shear Waves will not effectively propagate through austenitic material Longitudinal waves produce varying flaw responses, although more effective than shear waves in propagating through austenitic material

• A variety of UT techniques have been applied to address limited coverage conditions A broad assessment of these techniques has not been conducted.

The Performance Demonstration process only offers a Yes/No assessment

• No formal study has previously been initiated to evaluate and document the extent by which a flaw must propagate outside a limited coverage area in order to be detected 3

Limited Inspection Coverage

• PNNL conducted a search of NRCs ADAMS database to locate relief requests and reports associated with limited UT examination coverage issues.

ML17318A120 (PNNL-26157)

• Several weld configurations that often result in limited coverage were identified and were used to prioritize PNNLs assessment of the impact of incomplete coverage.

• Conditions that limit UT coverage of a specified examination volume restrict probe movement and include excessive weld crown width and outside surface configuration.

4

Examples of Limiting Conditions 5

Examples of Limiting Conditions

• Configuration limited due to taper from tee and weld geometry

• Material limited due to no single sided qualification for austenitic SS welds

• Component supports block probe motion 6

Design of Experiments: Factors and Levels Number Factors List of levels Notes of levels WSS - WSS

• Factor and levels based on Materials 3 CASS - CS No scans to be performed from the CASS sides.

CASS - SS typical conditions in the field Wall Thickness 2 Thin, Thick Thin 1.6 in.; Thick > 1.6 in.

• Conditions limiting coverage Weld Root Condition 1 None Assuming best case scenario of no weld root, although some specimens may have existing weld root.

include taper (weld and/or Probe Aperture 2 Small, Large component), physical access Single Element, restrictions, weld geometry, and Probe Type 3 Phased Array, Dual-Element TRL material microstructure (CASS) PA - 30°-70° Refracted Angle 4 30° ,45° ,60° ,70° Conventional - 45°, 60°, 70° TRL - 45°, 60°

• Metrics for quantifying coverage Wave Mode 2 Shear, Longitudinal Shear is only applicable for conventional probes and near-side exams.

will be tabulated Conventional - 2.25 MHz, 2 MHz, Probe Frequency 3 Phased Array - 2 MHz

• How do these factors limit 2.25 MHz TRL - 2 MHz Probability of Detection Length/Depth Ratio 3

<3 3-5 Ranges have been adjusted due to lack of specimens with high aspect ratios

>5 Ongoing assessment with respect to size distributions, location, orientation, Flaw Parameters and tilt. Other factors may also be included as assessment progresses.

7

Design of Experiments Matrix

• Data acquisition matrix resulting from the Design of Experiments analysis.

• Specimen list to include austenitic or dissimilar metal welds only.

8

Data Partitioning Data has been acquired and partitioned to simulate the following conditions:

1. No obstructions Unrestricted probe movement across the weld and away from the weld

2. Weld Crown Obstruction Probe cannot move across the weld; however, probe movement away from the weld is unrestricted

3. Component or Support Obstruction Probe can move across the weld; however, an obstruction prevents movement back away from the weld 9

Data Recorded for Each Flaw & Each Probe

1. Flaw Detected or Not Detected

2. Presence of a Flaw Tip or Not Detected

3. Flaw Response Amplitude

4. Flaw Length at 6 dB Below Max Amplitude

5. Flaw Length at Noise Floor

6. Maximum and Mean Noise on Each Side of the Flaw

7. Signal to Noise Ratio 10

Determining Signal to Noise 11

Common Limitation is WSS-WSS Piping Missed Detection of Shallow Near Side Flaw and Determining the Full Extent of Deep Flaws Weld Crown Width Prohibits Coverage of Heat Affected Zone After Weld Crown Removal Shallow Flaw is Detected and Deep Flaw Tips are Identified 12

Limited Coverage Flaw Response Simulations Flaw response simulations with weld crown

• Probe coverage limited by weld limitations crown obstruction

• Coverage limitation caused:

Virtually no response at 45° Incomplete flaw response at 60° Misleading response at 70° (looks like noise or fabrication flaw)

Flaw response simulation, full coverage, 45° 13

Limited Coverage Actual Flaw Responses 45° Shear Wave Near Side 45° Shear Wave Far Side 45° Longitudinal Wave Far Side 60° Longitudinal Wave Far Side 70° Longitudinal Wave Far Side 14

Ongoing Analysis Probe Location - 20 mm segments DETECTED? Y/N Y/N Y/N

• Parse data into 20 mm wide sections with 10 mm overlap.

• Determine whether or not the flaw is detected (Yes/No) for each section as objectively as possible (i.e., would an analyst call a flaw from this data?)

• Calculate detectability as a function of probe position and flaw depth.

Current Activity

• Complete data collection and analysis activities for WSS-WSS welds

• Creation of Probability of Detection (POD) curves from tabulated stainless steel data

• Technical Letter Report (TLR) March 30, 2020 on Phase 1 wrought stainless to wrought stainless.

• Analysis of dissimilar metal weld data 16

Summary

• Conditions that limited examination coverage continues to be an industry issue

• The impact a limited examination condition has on the probability of flaw detection must be evaluated and documented.

• The ultrasound response characteristics from a portion of a flaw resulting from inadequate ensonification must be determined in order to enhance the probability of detection 17

Thank you 18