Jennifer Creamer (left) and Le Ma (right) are both Ph.D. students at Missouri S&T. Creamer recently published her research into methods of improving the accuracy of 5-axis machine tools. (Image courtesy of Missouri S&T.)
Key to this new approach is the generation of compensation tables that can be fed into the machines to allow them to compensate for, and greatly reduce, errors. This is a significant departure from the piecemeal and less effective approaches of the past that forced operators to continuously recalibrate and cobble together several methods to reduce errors.
Although simultaneous 5-axis machining allows for the creation of complex contours or cures, the larger the part, the more difficult it is to hold tight tolerances. Fixing flaws in the machining of these large parts can be problematic, as 5-axis machine tools are known to have 41 basic geometric errors due in large part to their size.
This means that the way the machine is programmed to move can be very different from the way it actually moves during the machining process.
To compensate for these errors, operators must make adjustments in the calibrations of their machines. Unfortunately, there are several different methods, none of which completely solves all problems. As a result, operators will cobble together several approaches in order to avoid these errors. This approach can be very time consuming as well as costly if errors are not discovered before the parts are actually fabricated.
The new method seeks to eliminate this piecemeal approach to reducing errors by replacing it with a new model which would both identify the geometric errors and generate compensation tables, or maps of errors, that could then be inputted into the CNC machine.
To accomplish this, the Missouri S&T researchers used a laser tracker to measure over the entire workspace of the five-axis machine tools and then used the measurements to generate the compensation tables. Their goal was to hold errors to five thousandths of an inch over 120 feet. Additionally, the researchers reported that the compensation tables accurately captured kinematic errors regardless of whether they arose from expected or unexpected sources.
The study was led by Jennifer Creamer, a Ph.D. student in mechanical engineering at Missouri S&T and the study results were published in the Journal of Manufacturing Science and Engineering. Her fellow researchers included both professors at Missouri S&T and colleagues from Boeing where Creamer works as an engineer.