Failures Related to Welding

This article describes some of the welding discontinuities and flaws characterized by nondestructive examinations. It focuses on nondestructive inspection methods used in the welding industry. The sources of weld discontinuities and defects as they relate to service failures or rejection in new construction inspection are also discussed. The article discusses the types of base metal cracks and metallurgical weld cracking. The article discusses the processes involved in the analysis of in-service weld failures. It briefly reviews the general types of process-related discontinuities of arc welds. Mechanical and environmental failure origins related to other types of welding processes are also described. The article explains the cause and effects of process-related discontinuities including weld porosity, inclusions, incomplete fusion, and incomplete penetration. Different fitness-for-service assessment methodologies for calculating allowable or critical flaw sizes are also discussed.

Extraction of Construction Quality Requirements from Textual Specifications via Natural Language Processing

Construction inspection is an essential component of the quality assurance programs of state transportation agencies (STAs), and the guidelines for this process reside in lengthy textual specifications. In the current practice, engineers and inspectors must manually go through these documents to plan, conduct, and document their inspections, which is time-consuming, very subjective, inconsistent, and prone to error. A promising alternative to this manual process is the application of natural language processing (NLP) techniques (e.g., text parsing, sentence classification, and syntactic analysis) to automatically extract construction inspection requirements from textual documents and present them as straightforward check questions. This paper introduces an NLP-based method that: 1) extracts individual sentences from the construction specification; 2) preprocesses the resulting sentences; 3) applies Word2Vec and GloVe algorithms to extract vector features; 4) uses a convolutional neural network (CNN) and recurrent neural network to classify sentences; and 5) converts the requirement sentences into check questions via syntactic analysis. The overall methodology was assessed using the Indiana Department of Transportation (DOT) specification as a test case. Our results revealed that the CNN + GloVe combination led to the highest accuracy, at 91.9%, and the lowest loss, at 11.7%. To further validate its use across STAs nationwide, we applied it to the construction specification of the South Carolina DOT as a test case, and our average accuracy was 92.6%.

Automatic Generation of Customized Checklists for Digital Construction Inspection

Construction inspection plays a critical role to ensure the quality and long-term performance of infrastructure. The current construction inspection practice at state transportation agencies (STAs) in the United States, which requires inspectors to manually gather and personally interpret the construction requirements from standard specifications, is subjective, error-prone, and time-consuming. This paper presents an intelligent database approach to automatically generate customized checklists of construction requirements at the pay item level. The proposed approach consists of three components: (1) identification of the functional requirements by consulting with the end users, (2) development of a construction inspection knowledge model via ontology to guide the database design, and (3) devising mechanisms to automate the generation of customized construction checklists for the work under inspection with all the necessary details in relation to what, when, and how to check, as well as the risks and actions when noncompliance is encountered. Specifically, the following functions now can be performed within the new system: (1) automatic generation of a customized checklist at the pay item level; (2) access to a checklist display that aligns with the repetitive/cyclical nature of construction workflows; (3) navigation between cross-referenced check items; (4) subgroupings based on responsibility, risk level, and inspection frequency; and (5) real-time links to training materials such as photos, videos, textual documents, and websites. This newly developed tool is currently being implemented and is expected to greatly reduce the workload for inspectors and enhance the effectiveness of the construction inspection process.