Date on Master's Thesis/Doctoral Dissertation
Biles, William E.
Committee Co-Chair (if applicable)
Evans, Gerald W.
Leach, James A.
Additive manufacturing; Powder-bed; Bin packing; Revenue maximization; Optimization
Manufacturing processes--Automation; Programmable controllers
The goal of this research is to develop a procedure for the placement of a priority part into a planned build in a powder-bed additive manufacturing machine. Toward that goal, a heuristic procedure was developed that seeks to maximize the revenue in a scheduled build, subject to due-date constraints; i.e., all parts with the closest due date will not only be included in the build, but will be placed near the bottom of the build. Likewise, any part in the scheduled build that does not have an immediate due date is a candidate for removal in order to accommodate a higher priority part, the order for which arrives after the build is underway. The build volume for this experiment with the SLS2500+ machine is 13-inch x 15-inch x 18-inch rectilinear container. To achieve this optimization of the build, the experiment for this research involved a penalty which is based on failure to meet the due date constraint. The m parts in the build are placed such that are placed the highest priority parts, including those with immediate due date constraints, are positioned near the bottom of the build volume. The focus of this study is to determine which “unbuilt” parts x(i) will be removed to make room for higher priority parts. Toward that end, each part is assigned a priority p(i), i = 1, …,m until the build is full. If an order for a part having a higher priority p(i)than any of the parts in the build, the procedure developed here seeks to replace at least one of the parts in the scheduled build. Hence, the dimensions of the lower priority parts are key, since the part that is removed must provide sufficient space for the new part. Importantly, the enactment of the procedure developed here takes place in real time, with little or no interruption of the progress of the build. The resulting Excel file will be sorted as to place the immediate items at the bottom on the build. The solution is valid for a single and also for a multiple build. Simulations of this heuristic procedure were shown to substantially increase the revenue desired from the planned build by adding higher priority parts.
Dieder, Sonja, "A heuristic algorithm for determining the part set in a powder-bed additive manufacturing machine." (2014). Electronic Theses and Dissertations. Paper 345.