When radiographing an austenitic stainless steel weld with an appreciable weld deposit size, selecting a low radiographic kiloelectronvoltage (keV) can contribute to producing a radiographic indication that is not an imperfection. The contributors to this mottled condition are both radiographical and metallurgical. Electrons from low keV can diffract or absorb when penetrating through the dendritic grain structure of a weld. The increase in keV, or using gamma ray–equivalent isotopes, produces a marked change in electron output and penetration in material.
Certain combinations of the welding process and types of material being welded can result in dendritic grain formation. A larger weld deposition rate such as that created with the submerged arc welding (SAW) process affects the dendritic grain size. The elemental properties of an austenitic stainless steel contribute to microalloy segregation because of the quantity of elements that austenitic (300 series) stainless steel is composed of (chromium, nickel, iron, and carbon). It can have as many as 13 other minor alloys (elements). It is best known for its corrosion resistance and is used for applications from eating utensils to nuclear power plant fuel storage containers.
The combination of these factors can create a mottled radiographic indication that radiographers may incorrectly identify as incomplete fusion or incomplete penetration. Mottling is a radiographic indication that can appear as an indistinct area of more or less dense images. Mottling is caused by interaction of the object’s grain boundary material with low-energy X-rays (300 keV or lower).
Introduction
The image shown in Figure 1 illustrates how X-rays are generated by directing a stream of high-speed electrons at a target material such as tungsten, which has a high atomic number. When the electrons are slowed or stopped by the interaction with the atomic particles of the target, X-radiation is produced.
Nondestructive testing (NDT) textbooks address the appearance of mottling and its relationship to X-ray keV, which is a primary means to correct this condition. After reading these texts, it is commonplace to conclude that low keV causes mottling (Kodak 1980).
What textbooks do not address is that mottling can occur in quality radiographs displaying the proper penetrameter holes or image quality indicator (IQI) wires. Therefore, if quality objectives are met, X-ray keV is a contributor, but not the single cause, of a mottled image.
Under certain conditions, the combination of metallurgical and radiographic physics jointly create the cause of a mottled image.
Figure 2 shows a digital radiograph of a longitudinal austenitic stainless steel weld, approximately 0.5 in. (12.7 mm) thick. The surfaces of the weld were mechanically worked to be flush. Understandably, radiographers may disposition this image as having either an incomplete (lack of) weld penetration or an incomplete (lack of) fusion.
Incomplete (or inadequate) penetration (IP), also known as lack of penetration, occurs when the weld metal fails to penetrate the joint. It is one of the most deleterious types of weld discontinuities. Lack of penetration creates a natural stress riser from which a crack may propagate. The appearance on a radiograph is a dark area with well-defined straight edges that follow the land or root face down the center of the weldment (Figure 3).
(Right) Figure 4. Incomplete weld fusion: (a) diagram; (b) radiograph.
Incomplete fusion (IF), also known as lack of fusion (LOF), is a condition in which the weld filler material does not properly fuse with the base metal. It usually appears on a radiograph as an intermittent dark line or lines oriented in the direction of the weld seam along the weld prep or between weld stringers (NDT Resource Center 2021).
Radiographs are dispositioned based upon acceptance criteria in industry codes and standards. For example, the acceptance criteria in subsection 5320 of Section III of the ASME Boiler and Pressure Vessel Code (hereby referred to as the Code) establishes the following radiographic acceptance standards for a Class 1 weld (ASME 2007a): “Indications shown on the radiographs of welds and characterized as imperfections are unacceptable under the following conditions.”
Section V, Mandatory Appendix I of the Code provides definitions of these terms (ASME 2007b):
- Indications: “the response or evidence from a nondestructive examination that requires interpretation to determine relevance.”
- Interpretation: “the determination of whether indications are relevant or nonrelevant.”
- Imperfection: “a departure of a quality characteristic from its intended
- condition.”
- Relevant indication: “an NDT indication that is caused by a condition or type of discontinuity that requires evaluation.”
To implement these acceptance standards, radiographers use an interpretation process to assess relevant indications and imperfections.
Ways of Challenging Indication Relevancy
There are three ways to determine if an indication in a weld is an imperfection (weld defect).
Re-radiographing the weld by re-angulating: Figure 5 shows how the weld shown in Figure 2 was originally exposed. If a radiograph of a weld taken from normal (0°) to the surface produces questionable indications, the weld can be re-radiographed by shooting the weld at a different angle (as shown in Figure 6). Aside from tight imperfections with a depth that is normal to the surface, incomplete fusion or incomplete penetration imperfections observed at 0° will still be observed at 10° (for example) with slight dimensional changes. However, nonrelevant indications observed at 0° will abruptly change or partially (or totally) disappear (ASM 1976).
Re-radiographing the weld with a different radiographic source: Another technique to determine indication relevancy is to change the radiographic testing (RT) source from X-ray (machine generated) to a radioisotope source. Figure 7 shows the weld shown in Figure 2 reshot, obtaining the same overall film density as Figure 2 using iridium 192 (Ir-192) with no change in source-to-film distance or film type. In both illustrations, the 1B ASTM 11 IQI
wire required by the Code is observed as it substantiates the quality of the radiographic image. However, the indication observed in Figure 1 no longer exists in Figure 6.
Interpreting indications by using another NDT method: Another solution used in determining radiographic indication relevancy is to use a different volumetric NDT examination. The weld shown in Figure 2 was re-examined from both sides, using straight beam and shear wave ultrasonic testing (UT) techniques per Code requirements. The straight beam examination produced no response equal to or greater than the established calibration reference level (CRL) amplitude. Shear wave produced responses near the distance amplitude curve (DAC) half leg and full leg. The same results were obtained from the front and back side of the plate. This response was expected due to the dendritic grain structure, and was not indicative of any weld indication/imperfection. There were no indications observed in the weld equal to or greater than the CRL or DAC (EPRI 2003). Another NDT technique that can be used to determine relevancy of imperfections is computed tomography/radiography.
To reiterate, relevant indications (defects) do not disappear when re-angulating a shot, reshooting using a different radiographic source, or using a different volumetric NDT method. While these NDT techniques assist in determining indication relevancy, the radiographer needs to consider other factors when it is questionable whether an indication is an imperfection.
Background Questions to Consider
What is the welding process?
The weld shown in Figure 2 is a two-pass SAW. Figure 8 provides a cross-sectional view of a typical SAW weld. The weld process was selected due to its favorable weld quality and deposition rate.
Considering this, we can rule out an incomplete (lack of) fusion imperfection, as the RT indication appears to be running in the center of the weld, and the weld process is not multipass on each side. Also, joint geometry and the weld metal deposition makes LOF highly unlikely. An incomplete (lack of) penetration imperfection could exist; however, there are no sharp edges to the Figure 2 imperfection.
When the weld shown in Figure 2 was reshot using Ir-192, the weld indication disappeared. If either incomplete (lack of) penetration or incomplete (lack of) fusion existed, it would most likely not disappear.
When performing a shear wave UT examination to determine relevancy, the expectation of seeing a defect located approximately halfway between the half-leg and full-leg DAC points, which would support possible incomplete (lack of) penetration, did not occur.
Why is there a difference between Figures 2 and 7 when changing the energy levels (keV)?
In RT, the combination of keV and milliamperage determines the energy and intensity of the X-rays traveling through the weld. The plane(s) of the dendritic columnar and equiaxed grain structure diffracts and absorbs a portion of the X-rays (see Figure 8).
All materials not only absorb and transmit X-rays to varying degrees, but also diffract them by grain boundaries, as radiation of longer wavelength in all directions (Watanabe et al. 1988).
The lower the keV, the weaker the radiation, and the greater possibility of X-ray diffraction and X-ray absorption. Figure 2 was shot using 230 keV. This energy is much lower than Ir-192, which was used to produce the radiograph in Figure 7 (Ir-192 has an energy equivalency of 316 to 613 keV) (Mintern and Chaston 1959).
The application of keV is a contributor, but not the singular cause of the mottled images, as both RT techniques in Figures 2 and 7 met IQI quality requirements as specified in the Code.
The questionable indications observed disappeared when using Ir-192.
Therefore, the cause of the images starts with the structure of the weld. It is important to understand the weld geometry and chemistry.
Does the shape of the weld, or chemistry of the weld affect the radiograph?
Metals are alloys consisting of various elements. Alloying elements have different material densities and different melting/freezing points. Austenitic stainless steels may contain up to 18 elements. During alloy solidification, microsegregation occurs as a result of the differences in the melting/freezing temperatures of the elements. Certain elements with relatively high concentrations (such as chromium in austenitic stainless steels) may microscopically coagulate. This coagulation creates what radiographers call a “grainy” radiograph, which can contribute to the creation of a mottled radiographic.
Figure 9 shows a dendritic grain pattern in a weld and its heat-affected zone (HAZ). Dendrites in metals are the crystals that form in the liquid during freezing, which generally follow a pattern consisting of a main branch with many appendages. A crystal with this morphology slightly resembles a pine tree and is called a dendrite, which is a branching, tree-like structure. Dendrites are formed when crystals grow in defined planes due to the crystal lattice they create, with freezing occurring from the base metal into the weld center. Secondary dendrite arms branch off the primary arm, and tertiary arms off the secondary arms, and so on. This is the main contributor to causing a mottled radiographic image (ASM 1983).
The combination of relatively low keV during the digital X-ray process, along with the dendritic grain (columnar and equiaxed) structure, creates X-ray diffraction and X-ray absorption, which contributes to a mottled radiograph.
Mottled Image Results
Mottling results in a light spot on the radiograph corresponding to the position of the crystal. It may also produce a dark spot in another location if the diffracted or reflected beam strikes the film or detectors. Should this beam strike the film or detectors beneath the thickness of the weld, the dark spot may be mistaken for a void (ASNT 1985).
The X-ray film or detectors receive the X-rays and create a mottled pattern. When the weld grain size is large enough to be an appreciable fraction of the part thickness, the chance of a mottled image increases (Kodak 1980).
Welding processes and weld geometry influence the size of the grain structure. Figure 8 provides a cross-sectional view of a 0.5 in. (12.7 mm) austenitic stainless steel SAW weld (welded from both sides). The SAW grain structure is significantly different than the crystalline structure of the base material or the structure of a multipass shielded metal arc welding (SMAW) weld.
For a comparison, Figure 10 shows a conventional 0.5 in. (12.7 mm) multistringer austenitic stainless steel weld using the SMAW process. RT mottling will not be as pronounced because of the overlap of the weld stringers and dendritic grain structures occurring at varying angles.
Figures 8 and 10 show the differences in grain size and structure of the weld and base material. The grain pattern of the welds is dendritic columnar and equiaxed shaped, while the base material has a more homogenous, randomly oriented grain size.
ASME Code Interpretation III-1-83-88 addresses radiographic mottling due to X-ray diffraction. For convenience, the interpretation is included in Figure 11.
As shown in Figure 11, Interpretation Reply (1) states that for ASME III applications, mottling is not an imperfection or a weld defect.
Conclusion
Mottled images in welds are nonrelevant indications and not imperfections. Mottling is not a reason to reject a weld. When radiographing austenitic stainless steel welds, the chance of having a mottled image is influenced by the welding process used and possible occurrences of microalloy segregation. Mottling is the result of X-ray diffraction and X-ray absorption occurring because of a relatively low keV used while attempting to penetrate through a dendritic grain structure. The welding process selection influences the grain size. The larger the grain size as an appreciable fraction of the part thickness, the greater the possibility of mottling occurring.
References
ASM, 1976, Metals Handbook, 8th edition, Vol. 11: Nondestructive Inspection and Quality Control, American Society of Metals, Materials Park, OH, p. 144
ASM, 1983, Metals Handbook, 9th edition, Vol. 6: Principles of Joining Metallurgy, American Society of Metals, Materials Park, OH, pp. 22–49
ASME, 2007a, Boiler and Pressure Vessel Code; Section III, Construction of Nuclear Facility Components, Subsection NB Class 1 Components, American Society of Mechanical Engineers, New York, NY
ASME, 2007b, Boiler and Pressure Vessel Code, Section V: Nondestructive Examination, Mandatory Appendix I, Terms and Definitions, American Society of Mechanical Engineers, New York, NY
ASNT, 1985, Nondestructive Testing Handbook, Vol. 3: Radiography and Radiation Testing, second edition, American Society for Nondestructive Testing, Columbus, OH, pp. 211–212
EPRI, 2003, Ultrasonic Examination Technology – Level II Practical Specific Module 2PS, Electric Power Research Institute, pp. 9–13
Kodak, 1980, Radiography in Modern Industry, fourth edition, Eastman Kodak Co., Rochester, NY, pp. 31, 44, 58
Mintern, R.A., and Chaston, J.C., 1959, “Gamma Radiography with Iridium 192: Advantages in the Nondestructive Testing of Castings and Welded Structures,” Platinum Metals Review, Vol. 3, No. 1, pp. 12–16
NDT Resource Center, 2021, “Radiograph Interpretation – Welds,” https://www.nde-ed.org/educationresources/communitycollege/radiography/techcalibrations/radiographinterp.htm, accessed 23 November 2021
Watanabe, T., T. Sofue, Y. Saeki, and H. Mizukoshi, 1988, “Effects of Segregation on Mottling Seen on Radiographs of Stainless Steel Weldments,” Welding International, Vol. 2, No. 3, pp. 224–228, https://doi.org/10.1080/09507118809446529
_______
Author
Albert M. Wenzig, Jr., ASNT NDT Level III (PT, MT, RT, UT, VT, LT), ASNT Fellow, and AWS Certified Welding Inspector (CWI)
Citation
Materials Evaluation 80 (1): 22–26 https://doi.org/10.32548/2022.me-800122_2
©2022 American Society for Nondestructive Testing
Find additional papers published in Materials Evaluation on the ASNT NDT Library https://ndtlibrary.asnt.org/. Information for submitting a paper to Materials Evaluation can be found on the ASNT website.
Additionally, you may verify that the “Mottling” indications do not mask any unacceptable indications by the use of gamma ray energy, and fine grain film, as all indications observed in an area of interest should be clearly explained by the examiner in the radiographic interpretation report.
Very well written. Clearly demonstrates how mottling occurs and how it should be dispositioned.
Excellent article.