The main goal of this doctoral study is to investigate and develop knowledge for the use of additive manufacturing (AM) in repair of large metal components.
When purchasing a component, price is one of the main selection criteria. However, when a repair is needed, this changes drastically as lead time plays a much larger role. This is especially true for critical components, e.g. valves and other large components in the maritime, oil and gas industry. In such industries, the loss of income due to downtime tends to be so substantial that repair cost is of little significance. Limited access to highly skilled welders has posed a problem for this industry for years, a problem that became even worse during the Covid pandemic with travel restrictions and closed borders. This backdrop makes it very interesting to develop prequalified technologies that make repair work, either in the field or in workshops, less sensitive to operator skill.
SINTEF Manufacturing have a long history of research on powder bed methods. Such methods can produce high resolution, intricate geometries and nice surfaces. The only problems are that build chamber size will put strong restrictions on component size and the process is relatively slow. Using such methods for repair is also problematic, since the powder bed methods are limited to starting its build form a flat, even surface without any points protruding beyond the starting level. To broaden both competence and capabilities, SINTEF Manufacturing have invested in a "directed energy deposition", DED, build head from Meltio, which by integration on a robotic arm is capable of producing large parts. The process is slower, but more precise than arc-based methods, and therefore, it is something in-between powder bed and arc-based methods. This equipment enables a good balance between process control, high productivity and generous geometrical freedom.
API 20S and DNV-ST-B203 are standards that defines quality- and qualification- requirements for metallic components manufactured by AM. However, both these standards are characterized by conservatism as many will interpret the requirements in a way that each component must first be destructively tested for qualification. This has severe impact on the lead time and cost. This doctoral work will attempt to demonstrate that parameter windows can be determined for such processes for any component. Publishing results from testing of different material properties and qualities should help build knowledge of the link between material properties and build parameters and strategies.
The first months of the doctoral work has been spent assembling the cell, followed by integration of the Meltio equipment and the robot. The next step will be to look into and establish a methodology for efficient development of process parameters for new materials. It is desirable to look at a relatively wide range of material qualities to see if it is possible to define an acceptable parameter window at a relatively elementary level so that one can move from component qualification to machine qualification within a certain parameter range.
The parameter development will start with building simple geometries to compare the effect from different parameter combinations on buildability, production speed, melt pool depth and material properties. Following this, more complex geometries will be built to see the effect on buildability and material properties.
Expected future work will be to look into hybrid technologies for combining grinding, machining, machine vision and 3D scanning, since all these operations are likely to be necessary activities during component repair.
Progress as off 2021-12-31:
Initially, the doctoral work has consisted of a literature study, as well as a trip to the Formnext trade fair in Frankfurt. The trip was recommended by the supervisors to get an overview of what equipment and technologies are available in AM for metal. In summary, there are few that offer similar technologies as the Meltio system offer. Thus, there exist great opportunities to publish innovative research results, which can contribute to increased understanding of and use of multi-laser DED/AM techniques.