ADVANCED MATERIALS & PROCESSES | OCTOBER 2025 24 adapters, the number of exposed (i.e., unengaged) male threads on the exemplar mixing valve varied from two to five. This can be attributed to dimensional variations between the two mating components and may be further affected by the amount and type of sealant used. Examination of the fitting surface revealed no evidence of any physical condition, gross damage, or other anomalies that would suggest overtightening contributed to the failure. Thus, the number of threads engaged was not a reliable indicator of the stress applied to the adapter during installation. INSTALLATION STRESS FACTORS These points distill the assessment of installation stresses to the following. There are numerous factors that would contribute to the magnitude of sustained longitudinal tensile stresses for an installed adapter. These factors include the dimensions of each component of the threaded connection, thread sealant within the connection, and the number of turns past hand tight conducted during installation. These factors are interrelated and could not be comprehensively assessed for the subject fractured adapter after the failure occurred. In addition to the stresses imparted during installation, residual stress from manufacturing can also contribute to SCC of brass plumbing components[3,4]. A thin layer of deformed grains at the thread surfaces of the subject adapter suggested a high degree of material deformation during the machining process that created the adapter threads (Fig. 6). Abusive machining practices and the associated material surface deformation is a common source of residual stress in brass plumbing components and, in the experience of this author, is a common feature identified during investigation of SCC failures that occur in domestic water service. Thus, residual stress may have been a contributing factor to this failure. Finally, in addition to stresses from installation and residual stresses from manufacturing, a sustained hoop stress would be imparted to all components within the plumbing system due to normal operating water pressure. There were no indications of abnormal operating conditions in this system that would have subjected the failed com- ponents to a higher than typical water pressure. Thus, unusual hoop stresses due to system water pressure were most likely not a factor in this failure. RESULTS AND DISCUSSION After the investigation and filing of expert reports, the manufacturer, general contractor, and installer could all agree on two things: The fitting was cracked and the crack was the source of the leak, and the homeowner was not at fault. The homeowner ultimately agreed to release their claims against the other three parties in exchange for $85,000. Prior to releasing their claims in this matter, the expert witness for the homeowner (this author) issued the following opinions based on their examination and testing: The leak at the shower mixing valve was caused by a crack in the threaded adapter at the hot-side valve inlet; the mechanism for the crack was stress corrosion cracking (SCC); the material comprising the failed adapter was highly susceptible to SCC; the water supplied to this residence was not unusually corrosive; the potable water service environment at the residence was advertised as a suitable environment for the subject adapter; the crack was caused by sustained circumferential stresses that originated from both installation and manufacturing of the adapter, in addition to the plumbing system’s operating pressure; how the installation practices contributed to the failure cannot be reliably assessed; whether or not the adapter was installed according to manu- facturer instructions cannot be determined in this case; and the adapter likely would not have failed if it had been fabricated from copper or a brass alloy with lower zinc concentration. The remaining three parties disputed the underlying contributing factors to the crack that was agreed to be the source of the leak. The manu- facturer contended that both the general contractor and installer are at fault due to improper product selection and faulty installation. However, the general contractor and installer contended that: SCC of the brass fitting was a product defect caused by high zinc content in the brass alloy that the manufacturer selected and used; elevated levels of zinc reflected a design defect for which the manufacturer failed to warn of a known risk; and the fitting, which was purchased from a plumbing wholesale supply shop, was used for its intended purpose and properly installed. Fig. 6 — Typical metal microstructure near center of fitting wall (top) and near a typical thread root (bottom).
RkJQdWJsaXNoZXIy MTYyMzk3NQ==