Machining of dental restorations
The project addresses the fully automatic manufacturing of dental prosthetical implants and restorations. The area of application is characterized by a very high degree of individualization and autonomy, and differs from conventional milling especially in respect of these two aspects. The individualization is manifested by the diversity of dental restorations, which causes no crown, bridge or implant to equal another one. In conjunction with the high diversity of raw material, individualized machining strategies become essential to guarantee an economical production.
The autonomy is manifested by the typical operational area that consists of dental offices and laboratories in close proximity to the patient, which requires the minimization of the manual effort of the dental surgeon or technician to plan, perform and post-process the production in order to allow for a small latency and high average utilization. Dental offices and laboratories primarily employ budget-friendly miniature milling machines that exhibit an optimal trade-off between costs and benefits. This fact contrasts to the use of conventional heavy and large milling machines within external milling centres that require a significantly higher throughput in order to operate economically. With respect to everyday practice and apart from economic aspects, an outsourcing of implant machining to such an external milling centre would particularly result in increased latencies and reduced capabilities of individualization. As a result, the acceptance of an immediate and easy to handle in-house production close to the patient increases steadily. However, miniature milling machines exhibit significant mechanical and thermodynamic restrictions, which have to be compensated for algorithmically, in order to allow processing the diversity of materials and models that are common to the dental branch.
Image-based extraction of the available blank area and waste-minimizing object arrangement
The high diversity of compositions and colours of raw materials that are used for dental restorations demand a frequent change of blanks. On the other hand, economic manufacturing requires a parsimonious and efficient exploitation of the expensive raw material. In order to allow for big blanks to be reused or multiply occupied, respectively, a database-based logging of blank utilization, as well as an image-based extraction of the free blank area is employed. Based on the free blank area, target objects are subsequently positioned automatically by minimizing the resulting cutting waste. Algorithmically, this denotes an irregular two-dimensional nesting problem.
Minimally invasive fixation of objects
The manufacturing process requires the target object to be fixated within the remaining blank by means of pins.
These pins in turn distort the object's surface and require manual post-processing. By optimizing the position, radius and number of fixation pins with respect to object distortion and stability, manual efforts are minimized to plan a suitable fixation, as well as to post-process the pin/object-contact surfaces.
Tool path generation
Tool paths which employ 3/4/5 tool axes are generated fully automatically for individual manufacturing steps like roughing, finishing, grinding, polishing.
The adaptation of the tool path planning strategy to characteristics of the object surface leads to an optimization of the trade-off between machining time and necessary surface quality. For instance this includes a coarser and therefore faster machining of surface regions that are faced afterwards, in order to reduce machining time.
Simulation and feed rate optimization
Budget-friendly miniature milling machines, which are primarily employed in dental offices and laboratories, exhibit kinematical, dynamical and thermodynamical restrictions that hamper or hinder a conventional uniform machining of metallic or glass ceramic raw material. An algorithmic compensation for these restrictions demand an adaptation of tool paths to the machine's capabilities. One approach of adaptation includes the feed rate and cutting speed, whose optimization takes place predictively by means of an offline tool path simulation, as well reactively by means of a control loop.
The tool path simulation bases on implicit representations of all relevant shapes by means of level-sets (material surface) as well as analytically specified distance functions (tool sweep). Non-uniform level-set data structures, which enable a position- and time-dependent adaption of resolution to the simulation progress, ensure a compressed representation.
The reconstruction of the material surface from the tool paths enables a quantitative evaluation of the target object with respect to approximation quality and smoothness, and constitute the origin of further machining steps.