Creating Machine-Specific Tool Paths
Article from Modern Machine Shop
Some CAM systems offer toolpath strategies that ensure a consistent load on a tool by controlling its engagement angle with the workpiece. Using this approach, the tool isn’t driven into internal corners where its engagement angle (thus, the force exerted upon it) greatly increases.
Delcam has recently developed a roughing strategy designed for solid carbide tools it calls Vortex that combines consistent tool engagement with a step-cutting strategy already available in the company’s PowerMill CAM software. With this approach, the tool does not step down immediately after each subsequent pass. Instead, extra cutting moves are added, working from the bottom of each step upwards. As a result, the initial cutter can take a much deeper axial depth of cut and remove a greater amount of material, minimizing the amount of subsequent rest roughing that’s required. Depending on the geometry of the part, the overall machining time can be reduced by as much as 60 percent, the company says. Vortex will be included within the soon-to-be-released PowerMill 2013 as well as many of the company’s other CAD/CAM offerings.
On its own, this certainly seems to be a viable, high speed roughing strategy. However, as with other CAM-generated tool paths, it’s generic. That is to say, it can be used to drive pretty much any suitable CNC machine once the appropriate postprocessor is written. There is a point, though, at which the operational limits of a machine and its control come into play. For instance, a machine simply might not have the ability to maintain the programmed feed rate that would otherwise result in optimal performance. In addition, controls differ in terms of their processing speed, high speed machining parameters and method for handling arc/line transitions among other characteristics. This spurred Delcam to look more closely at how the differences between individual machines, even machines that are very similar, affect the overall speed and effectiveness of a CAM-generated tool path.
The result of these efforts is MachineDNA, a technology the company developed more or less concurrently with Vortex. (Patents are pending for both offerings.) Directly integrated into a CAM system such as PowerMill, MachineDNA enables the CAM system to gather data from a specific machine to establish a performance baseline, and use that information to create an individualized tool path shaped by the machine’s condition and capabilities. By learning and applying a machine’s own traits, different, yet effective tool paths can result even though the same overall strategy (i.e. constant engagement angle) is applied.
Simply put, MachineDNA aims to determine what type of tool path a machine "enjoys” so the CAM system can generate such a path. Through a series of simple axis motion tests in which no actual cutting is performed, a machine defines its own capabilities. These tests need not be performed every time a new tool path is created—only when configuring a new machine or after a machine has been serviced or upgraded.
To run a test, users enter basic information such as model of machine, type of control and maximum feed rate into a simple interface within the CAM system. MachineDNA automatically generates the G code that drives the machine through the exercises. During testing, data is continuously fed back and stored within a machine-specific file that the CAM software will access each time it creates a new tool path for that particular machine.
Testing reveals a handful of key performance characteristics. These include the optimal trochoid size for a given feed rate, the machine’s preferred point distribution along curves and the most effective manner for moving through arc/line transitions. These attributes would be difficult if not impossible for even an experienced programmer to accurately identify for a specific machine.
In one exercise, the machine is directed to make a series of circular movements to determine the minimal trochoid radius at which the machine can maintain the programmed feed rate. For a trochoid radius that’s smaller than optimal, the machine typically won’t be able to maintain the programmed feed rate. Conversely, the machine might be able to maintain the programmed feed rate using a larger trochoid radius, but a bigger radius equates to a longer tool path and longer cycle time.
Testing also determines the optimal point distribution to match a machine’s control capabilities. Envision a simple oval with two lines and two arcs. The CAM system might output this arrangement as two lines and two arcs; leave the lines as they are, but linearize the two arcs for a smooth, even point distribution; or redistribute points at appropriate intervals around the entire oval shape. Based on testing performed on a range of controls, MachineDNA takes into consideration the characteristics of the control as it determines the most appropriate point distribution along the tool path. It also accounts for the control’s processing speed so as not to feed a slower control too many points, which could potentially cause stuttering due to data starvation.
Finally, when a tool goes from a straight line into an arc, some controls will dwell the tool slightly at the point of the transition. These dwells must be eliminated to maintain the programmed feed rate throughout the entire tool path. In understanding the idiosyncrasies of various controls, MachineDNA provides the data that enables the CAM system to generate an optimized tool path that accounts for such control peculiarities.
Article From:5/10/2012 Modern Machine Shop, Derek Korn, Senior Editor