Feb/March_AMP_Digital

A D V A N C E D M A T E R I A L S & P R O C E S S E S | F E B R U A R Y / M A R C H 2 0 2 1 2 1 The elemental distribution maps show that, due to dilution of the Ni- rich C-276 weld with Fe-rich base plate, there is an increase in the amount of Fe and a corresponding decrease in the Ni and Cr content in the first weld overlay layer. There is also evidence of dilution, albeit to a lesser extent, in the second overlay layer of deposited C-276. The µ-XRF results further show that dilution is not uniform and there could be elemental variation even with- in an individual weld bead. Elemen- tal analysis using both SEM-EDS and XRF in the first layer of weld provided Fe values ranging from 15 to 25 wt%, the second layer of weld ~8 wt%, and the fillet weld region ~1.25 wt%. Al- though it would be possible to repli- cate the XRF analysis in the SEM using SEM-EDS mapping, it would be much slower to cover the area shown and less accurate. Hence, µ-XRF provides a rap- id way of performing single field chemi- cal distribution mapping of large areas. It should also be emphasized that bulk analysis using EDS as a screening meth- od will not provide perspective on lo- cal maximum values in elements of interest within an individual weld bead. This could be an important criterion for long-term, in-service performance (depending on the weld metal expo- sure to the environment). It is thus the case that future methods of quality as- surance or enhanced weld procedure development may need to consid- er reporting of maximum, mean, and minimum percent of dilution. The val- ue of mapping the distribution of el- ements for such assessments using µ-XRF is that it can take such complex- ities into account that would other- wise not be possible by means of a bulk EDS measurement. CASE STUDY 2: LONG-SEAM WELD FAILURE IN A HOT REHEAT SYSTEM In power plants, welds play a cru- cial role inmaintaining the structural in- tegrity of pressure-retaining parts that operate at high temperatures. Welds can also represent locations of inherent increased susceptibility to damage de- velopment due to a range of concerns linked to design, fabrication, operation, construction scope changes, and other complexities. For the power generation industry, a particular concern at elevat- ed temperature is the introduction of time-dependent damage (e.g., creep). In one recent example, a long seam welded Grade 22 hot reheat sys- tem operating at ~1000°F (540°C), expe- rienced a leak during normal operation after ~270,000 hours [2] . Figure 3 (left) shows an etched cross-sectional im- age of the seamweld along with a crack that initiated at the pipe inner diameter (ID) near the weld and traveled over half the thickness of the wall. Also shown is a micro-hardness map (right) display- ing the distribution of hardness across the double V-groove seam weld. Prom- inent in this map is a region of signifi- cantly lower hardness (105 HV0.5 to 125 HV0.5, in blue) within the weld re- gion, immediately adjacent to the crack at the tube ID. Note that this same re- gion exhibits a marked difference in etching response in the macrograph. To inform the failure analysis in- vestigation and provide contributing factors for the root cause evaluation, a large-area µ-XRF scan was performed on the weld cross section. Distribution maps of key elements are shown in Fig. 4. The µ-XRF scan revealed that the zone with lower hardness has low- er Cr and Ni (and correspondingly higher Fe) content compared to the adjacent parent material and other regions of the weld near the outer diam- eter. This finding suggests that the long seam weld had been fabricated and/or repaired during the fabrication stage with a filler material that was consistent with a carbon steel composition rather than a matching AWS-type B3 consum- able. In this example, µ-XRF mapping was key in establishing that off-specifi- cation filler material (so-called “rogue filler metal”) was used to make the long seamweld and this likely contributed to the accelerated nature of the observed failure. CASE STUDY 3: ELEMENTAL SEGREGATION IN FINAL PRODUCT FORMS Segregation of alloying elements is commonly observed in engineer- ing alloys [3] . Heterogeneity in such al- loys is a consequence of the many steps required to make the final product, Fig. 3 — Cross section of a long-seamweld in a high energy pipe, showing a crack near the weld (left) and a hardness distribution map of this weld showing region of lower hardness near the ID. Fig. 4 — Micro-XRF elemental distribution maps of Fe, Cr, and Ni from an uncracked section of weld in Fig. 3.