Quality Control of Cadweld (Mechanical) Splices

Test data for cadweld splicing of reinforcing steel collected during a study of quality assurance practices on nine nuclear power plant construction projects are presented and evaluated. These data lead to an important hypothesis that the visual inspection identifies procedural deficiencies, and the tensile test identifies material defects. It is also suggested that a material testing program and the visual inspection will detect essentially all substandard cadwell splices. This would permit the deletion of the expensive tensile testing program. Accordingly, most quality control programs require overtesting and overdocumentation of cadweld splices; and furthermore, these programs fail to recognize material defects. The project specifications and quality control requirements for the nine projects are compared. Where possible, these are evaluated against the industry standards and Federal regulations. It is shown that there are a number of deficiencies in these standards, and that in most cases, the testing requirements are not commensurate with the quality that is being achieved in the field.