Recording of real cutting forces along the milling of complex parts

In this paper, a data acquisition system that simultaneously allows the recording of cutting forces and cutting tool position (coordinates X, Y and Z) is presented. Thus, the geometry of the surface being machined and the values of the cutting forces generated during the surface milling call be correlated. In this manner, two usual problems in milling experimentation are solved. First, those derive from the continuous changes in the feed rate value due to the special look ahead functions (being on-line applied by the numerical control in all high-speed milling centres), which affects the correlation between part geometry and forces. Second, those originate by the presence of unexpected stocks coming from previous semi-finishing operations, which changes the value of cutting forces respect to those calculated taking into account the theoretical tool engagement into the part. The objective of this work is the development of a diagnostics tool for allowing the detection of potential milling problems by researchers, making more profitable the performance of machining test on real complex parts. This tool can be used in the optimization of the milling of test-parts, machining real geometries that incorporate problems different from those observed in the linear tests at constant linear feeds (in end-milling conditions). The final goal of the system is the generation of cutting forces maps as a function of the geometry of real workpieces, which is an advance respect the usual forces recording in function of time or tool rotation typical of linear machining tests. In order to do that, the system incorporates a dynamometric plate for the measurement of the three components of the cutting force, as well as an acquisition card connected to the analog output of the position control loops. After machining, the file containing position and force data is post-processed and chromatic and vector maps are generated. The force analysis utility has been applied to three cases, in order to assess its feasibility in machining research projects. The first example shows a reduction of the number of tests and therefore machining time (even by 18 times) in the validation of a mechanistic cutting force model. Second example is focused oil the detection of unexpected tool engagement conditions in complex parts. And the last one addresses to the milling of thin walls, for investigation of static and dynamic milling problems. (C) 2005 Elsevier Ltd. All rights reserved.