Inside Machines: Practical Feedrate Limits for 5-Axis Machining of Thin Parts
When looking at cutting feedrate, a typical specification for 5-axis machines, consider “worst-case” part geometry.
When purchasing a 5-axis machine, one of the typical specifications is the maximum feedrate for cutting. The machine’s ability to meet this specification is commonly verified during buy-off with the Circle-Diamond-Square and Cone-Frustum tests as defined in National Aerospace Standards 979 (NAS 979).
For traditional aerospace parts, where a large percentage of aluminum is removed from a billet, this is a very reasonable test when run at the planned production feedrates. As more aircraft parts are being made from composites and titanium stampings, where a part is trimmed out of a contoured sheet, NAS 979 may need to be run at much higher speeds to represent the real machine usage. These concepts frequently apply to milling machines, waterjets, lasers, and fiber placement machines.
A simple example of this is a 0.1-in.-thick carbon fiber sheet with a “C” cross-section, intended for use as a beam, where edges need to be trimmed to the final size with an end mill. For this example, the intended path feedrate is 50-in. per minute (IPM) at the tooltip, the tooltip is 0.05-in. through the skin, and a 5-axis milling machine with a “C carries A” forked head is being used. Also assume a pivot length (tool tip to A axis) of 18 in. and that the machine performed perfectly on the NAS 979 tests at 300 IPM. During the flat sections of this contour, the machine should have no problem; the radius is the concern. If the radius (R) is 2 in. (2.15 in. at the outside of the skin), then the length (L) along the path is 3.1 in. (pi*R/2). At 50 IPM, this means the machine should take about 3.8 sec to make this radius. During this move, “A” moves 90 degrees to stay perpendicular to the surface. When the pivot length is considered, and excluding the distance for the programmed radius, X and Z each need to move 18 in. just to keep the tool center point (TCP) stationary. At a 2-in. radius, X would need to average 285 IPM to make this move. This is still well within the 300 IPM from the NAS 979 test. If the radius is decreased, then X needs to move even faster to maintain the 50 IPM at the TCP as the diagram and table show.
As the radius is decreased, axis speeds start exceeding what was tested with the NAS 979 for this machine, even though the path feedrate is not increasing. For this reason, when considering a machine for purchase, the geometry of the production parts needs to be considered. When considering a new machine for purchase, the worst-case part geometry should be determined early in the bidding process, as a machine that performs well at 285 IPM is in a very different price class from one at 1,140 IPM.
- Matt Goska, mechatronics engineer, Siemens, Elk Grove Village, Ill., is among Control Engineering Leaders Under 40, Class of 2011. Posted by Chris Vavra, Control Engineering, www.controleng.com