The Art of Intentionally Flawed Specimens and NDE Performance Demonstration

An inside glance at what it takes to make these intricate samples.

AWS Publications | October 11, 2021 | Inspection
Welding Digest ►  The Art of Intentionally Flawed Specimens and NDE Performance Demonstration

An inside glance at what it takes to make these intricate samples.

The concept of making flawed specimens seems simple at first thought. Anyone can make a bad weld or product, right? This might be true in concept, but making a flawed specimen for nondestructive examination (NDE) purposes is another thing altogether. Flawed specimens must have accurate dimensions and no distractors other than the intended flaws or discontinuities to gauge the performance and ability of NDE technicians or American Welding Society (AWS) Certified Welding Inspectors (CWIs) to properly locate, size, and dispatch said discontinuities to a specific code, procedure, or industry standard.


In the Beginning

Since the beginning of inspection, we have been looking for anomalies, discrepancies, inconsistencies, irregularities, discontinuities, and flaws that have or are approaching being characterized or categorized as a flaw or defect that affects the serviceability of a product or structure. Preferably, we find them before they go into service, or worse yet, an actual failure.

Blocks or product segments with notches, side-drilled holes, flat bottom holes, and other discontinuity simulations are typically used for calibration to establish a standard response or detectability size. Standardization (a more common term now) for establishing a common baseline for inspections and calibration is now relegated to the equipment’s interval verification of functionality. In most training schools, we typically use specimens with discontinuities similar to those found in new fabrication or manufacturing (e.g., incomplete fusion, slag, incomplete joint penetration, and maybe laminations or nonbonded areas). These are all very important because we don’t want any new fabrications or manufactured products entering their service life defective.

Training and qualification of NDE technicians for in-service or post-event environments are completely different and somewhat forensic in nature, requiring a different body of knowledge. It requires specialized training, additional experience, and awareness of failure mechanisms of the particular asset. It also requires different types of specimens with flaw sets representing those types of discontinuities — Fig. 1.

Fig 1-Oct-11-2021-07-44-05-70-PMFig. 1 — An NDE technician uses the magnetic particle testing dry method verification of flaws to inspect a gear.


No Replacement for Experience

My experience has contributed significantly to my current work in flaw manufacturing. I have learned there is no replacement for experience, and the only way to get that experience is to do the job and always continue to increase your body of knowledge. I have been very fortunate over my 33-year career to have worked on a wide variety of projects and different scopes of work, and I have been exposed to a wide array of discontinuities. I have been with inspection service companies for 30 years and three years with FlawTech America. I have inspected hundreds of ships and buildings and worked on projects for every major amusement park in Southern California and a couple in Orlando. I have worked in most of the refineries in Southern California and one in Colorado. I was heavily involved with the post-seismic event evaluation after the 1994 Northridge earthquake and landmark projects such as the San Francisco-Oakland Bay Bridge and the Las Vegas High Roller. I have worked in seven countries and many states throughout the United States.

It would be great if every person in inspection could have the opportunity to work and evaluate real-life specimens made during fabrication or removed from service, but there are several drawbacks to this. First, there is no guarantee there are flaws in those specimens. Second, the availability of those specimens is extremely limited. Third, it usually requires destructive testing to quantify the discontinuity size fully because every NDE method or technique has its principles and limitations in characterizing and sizing said discontinuities.


How Are They Made?

One of the most frequent questions we receive is, “How do you make your flaws?” Many of those techniques are proprietary and have been developed over our more than 36 years of history. Many of the welding flaws, such as porosity, slag, incomplete fusion, incomplete joint penetration, undercut, and tungsten inclusions, are made just like they are in normal welding operations. Nothing too fancy, except we place them incredibly accurately for both location and size. Cracks are a little different, and we have developed several different techniques and types to simulate specific types of cracks. We produce most of our cracking using mechanical fatigue, thermal fatigue, or hybrid-type methods.

For some of the industry-driven crack specimens, such as hydrogen-induced cracking or stress corrosion cracking (intergranular or transgranular), we use a technique we call craze cracking, because we can’t precisely model the environment in the effective, timely manner that is required to develop the two previously mentioned cracking modes.

We are very successful in developing flaws that model the NDE responses for different NDE methods used to evaluate assets for these flaws. We have also developed what we call accelerated corrosion, which produces more realistic pitting, corrosion, and erosion. We can even age the exterior or interior of a tube or pipe to simulate it having been in service. What is most remarkable is the accuracy with which we can replicate these flaws. Virtually all flaws are mechanically measured and verified, so regardless of the NDE method being used, we know the actual size of the flaw manufactured. We call this the “flaw truth.” Manufactured flaws over the years have been sliced and destructively tested, confirming the manufacturing tolerances are highly accurate. With the continued improvement in NDE equipment, flaw manufacturing must keep step with accuracy, flaw characteristics, and replication of those found in the real environment and required to be found using various NDE methods and techniques.

Similar to how we approach developing kits, custom flawed samples are an even more collaborative process — Fig. 2. We rely heavily on customer input and feedback, encouraging them to be very involved with the process and inviting them to our location to verify progress and final inspection before shipping.

Fig 2-2Fig. 2 — This example of a custom sample shows a nuclear vessel wall with a nozzle.


The real key to the flaw manufacturing is the personnel. Training welders and machinists to become flaw technicians is no small feat. It takes approximately two years of experience building flaws, depending on the individual welder’s experience and skill set, to become a trustworthy flaw technician in the art of intentionally flawed specimens. It is very important to maintain a key group of qualified individuals in this highly specialized form of manufacturing.

The next-to-last phase of the flaw manufacturing process is the quality assurance that the flaws are detectable and there are no unintentional flaws or distractors per the customer’s requirements — Fig. 3. This is also where we can do flaw fingerprinting, or the customer is encouraged to do their own validation on site before shipment.

Fig 3-3Fig. 3 — Phased array ultrasonic testing verification of flaws in a circumferential weld.


The final and very important stage is the as-built computer-aided design (CAD) drawings and the associated document package provided to the customer with a certificate of conformance that indicates all the intended flaws are present and accurate to the tolerances determined by our standards or the customer’s requirements. Also included, if required, can be NDE reports, test sheets, material test reports for base material and welding electrodes, and measurement test equipment with National Institute of Standards and Technology traceable calibrations.


Industry and Codes

The flaw-manufacturing industry has developed many stock specimens and code-specific kits over the years. We rely heavily on customer input and feedback when designing new specimens and kits. As many know, our industry is vast, and virtually anything that deals with metals has a critical element that must be evaluated with some method of NDE and done proficiently to avoid failures.

We have seen a significant increase in in-line inspection (ILI) specimens for the piping industry. We are also working on developing specimens for composites and additive manufacturing.

The development of American Petroleum Institute (API) kits resulted from a presentation at the 2000 American Society for Nondestructive Testing (ASNT) Research Symposium in Denver, Colo., where a representative from the API gave a presentation concerning the beginning of the Qualified Ultrasonic Testing Examiner program due to a research study and its findings.

When going into any type of timed performance-based exam such as this, a high probability of passing requires practice using similar weld configurations and flaw types. CWI exam preparation and the new ASNT Industry Sector Qualification program are also huge drivers for kit development.

The AWS D1.8 seismic kit was developed to comply with the AWS D1.8 Supplemental UT Technician Qualification — Fig. 4. The AWS D1.8, Structural Welding Code — Seismic Supplement, first published in 2005, resulted from several Federal Emergency Management Agency studies initiated after the 1994 Northridge earthquake. The results affected design, welding, and NDE. For ultrasonic examination (UT), it was the SAC/BD-00/06 (Ref. 1), Round Robin Testing of Ultrasonic Testing Technicians, which determined that a significant number of fabrication-type flaws were inconsistently identified or missed by UT inspectors. AWS D1.8-qualified UT inspectors are required to evaluate a minimum of 20 flaws per AWS D1.8/D1.8M:2016, Annex F (Ref. 2), in various common weld joint configurations found in structural steel fabrication and are graded on indication rating, flaw depth, and flaw length within specific tolerances on their submitted report.

Fig 4-2Fig. 4 — AWS D1.8 seismic kit.



Fabricated or manufactured flaw specimens are an effective, reasonable, and cost-saving alternative to real-world flaw specimens, which can be very expensive, limited in numbers, and difficult to ascertain the actual size of the flaws without destroying the sample. Waiting for demolition or deactivation of assets is not a timely option to train, qualify, and validate the qualification of NDE practitioners and CWIs alike required to service our industry.



1. SAC/BD-00/06, Round-Robin Testing of Ultrasonic Testing Technicians, was funded by the Federal Emergency Management Agency (FEMA) and the California Office of Emergency Services (OES) and produced by the SAC Joint Venture (published 2000, 26 pages).

2. AWS D1.8/D1.8M:2016, Structural Welding Code — Seismic Supplement, 3rd Edition Annex F (Normative) Pages 53 and 54.


This article was written by Ricky L. Morgan, vice president of FlawTech America, Concord, N.C., for the American Welding Society.