THE MACHINE TOOL GENOME PROJECT DESCRIPTION
The objective of the Machine Tool Genome Project (MTGP) is to enable pre-process milling parameter selection for “first part correct” production. This will replace the current practice of trial-and-error part path validation and will be achieved by predicting the tool point frequency response function (i.e., the vibration response of the tool-holder-spindle-machine assembly as reflected at the free end of the cutting tool) using the Receptance Coupling Substructure Analysis (RCSA) algorithm. Given the tool point response, frequency-domain algorithms for stability and surface location error (caused by forced vibrations) will be applied to separate feasible and infeasible zones within the spindle speed-axial depth of cut domain. Once acceptable spindle speed-axial depth of cut combinations are determined, they will be presented in a new user-friendly format, the Tool Dashboard, similar to an automotive dashboard display.
Discrete part production by subtractive machining remains a key technology in the US manufacturing sector. Limitations to machining productivity include tool wear, positioning errors of the tool relative to the part, spindle error motions, fixturing concerns, programming challenges, and the machining process dynamics. Tool wear limits the maximum available surface speed and is addressed commercially through the continual improvement of coatings and edge geometries by tool manufacturers. Machine limitations, such as parametric error motions and thermal errors, are addressed by machine tool manufacturers using compensation schemes embedded in the controller. Computer-aided manufacturing software is constantly being improved to increase contour accuracy and avoid collisions. The process dynamics, however, remain an industry obstacle to high performance machining because a turn-key solution is difficult to realize.
Machine tool manufacturers have developed design expertise and computer-aided modeling tools to produce machining centers that meet customer requirements for accuracy, throughput, reliability, and safety. However, the selection of the cutting tool and holder is generally left to the customer and is based on the particular application. While this paradigm is reasonable given the broad range of potential part geometry-material combinations that may be implemented on a particular computer numerically-controlled (CNC) machine, it places the user in a difficult position because the cutting tool is often the most flexible element in the machine tool structure. Its compliance, therefore, often limits productivity due to forced and self-excited (chatter) vibrations during material removal. It is the large number of tools and holders that may be used on a given machine that poses the most significant challenge to realizing a turn-key solution to milling parameter selection.
In the proposed MTGP solution, the RCSA approach will be used to couple the measured spindle machine dynamics to models of the tool and holder and predict the assembly’s dynamic response. Given this information, frequency-domain algorithms may be applied to select feasible spindle speed-depth of cut pairs that enable first part correct production. These feasible cutting conditions will then be displayed in the user-friendly Tool Dashboard interface that does not require knowledge of machining dynamics or interpretation of model-specific diagrams. This physics-based, pre-process selection of operating parameters, coupled with an effective information display, will reduce the time to production and improve machining reliability by enabling operating parameter selections that minimize the occurrence of chatter and large vibrations which can damage the tool, part, and spindle.
This approach is analogous to the Human Genome Project (HGP), an international scientific research effort with the primary goal of determining the sequence of chemical base pairs which make up DNA and identifying and mapping the 20,000 to 25,000 genes of the human genome. HGP was launched in 1990 and completed in 2003. In the Machine Tool Genome Project (MTGP) proposed here, the “genes” are the tool, holder, and machine and the “mapping” is performed using RCSA to predict the tool point frequency response, i.e., the “body characteristics”. In the MTGP, the spindle-machine “genes” will be measured once by BlueSwarf personnel (using impact testing) and archived. The customer can then define the desired tool and holder “genes” using the BlueSwarf software application, select the spindle-machine from the BlueSwarf database, and receive the corresponding Tool Dashboard (via the Internet) to enable selection of preferred operating parameters which respect the limitations imposed by the process dynamics. The “mapping” steps of predicting the tool point frequency response, calculating the corresponding stability lobe diagram, and representing this information using the Tool Dashboard will be transparent to the user.
The objective of the MGTP is to establish a new paradigm for selecting optimized milling process parameters. Cutting tool information, for example, has many components (diameter, number of teeth, helix angle, teeth spacing, rake angle, flute length, shank length, coating, etc.) and is currently made available to the machinist electronically and in printed catalogs. While the process performance certainly depends on the tool geometry, the dynamic behavior is strongly dependent on the combination of machine tool, spindle, adaptor (holder), and cutting tool selected by the user. The current means of parameter selection using prior experience, combined with cutting test validation to verify the part program performance, is both costly and time consuming and does not guarantee optimal cutting conditions.
The new paradigm proposed here is to build a database of component receptances (dynamic responses) that will be joined analytically to predict the assembly response for the many tool-holder-spindle-machine combinations available to the modern machinist. The database will include receptances for spindle-machines, holders, and tools. Based on this goal, specific technical objectives have been identified and will be explored in this effort.
Dr. Tony Schmitz