*2.1. Materials*

When researching the system, di fferent investigations were performed with two di fferent materials, which were presented in Table 1. The samples were shafts with a diameter of 30 or 40 mm and a length of 120 mm. Mono-material components were selected to investigate tool breakage and tool wear. When studying the performance of material detection friction-welded specimens, two mono-material shafts were used.



#### *2.2. Machining and Data Acquisition*

Experimental tests were carried out on the turning center Gildemeister CTX420 linear (Germany). An industrial personal computer (IPC) was connected to an open platform communication server and fieldbus interface of the Siemens Powerline 840D machine control (München, Germany) as well as to the CAN bus of the feeling turret. For the longitudinal turning, an indexable insert DNMG150404-FP5 by Walter AG (Tübingen, Germany) was used. The machining operations were all carried out without the use of cooling lubricant. The process force was measured by the sensory abilities of the turret, which is equipped with strain gauges (HBM, Darmstadt, Germanycountry). The strain gauges were set up as a Wheatstone-bridge and glued into the notch ground. The notches were designed by finite element simulation to increase the measured strain. A total of six strain gauges were set up as Wheatstone-bridges are placed on the turret. The signals are digitized by electronic devices close to the measuring position and transmitted to the IPC via a CAN-BUS [21]. According to the occurring strains during machining, the forces were calculated. The determined force resolution of the feeling turret is 64 N in the x-direction, 43 N in the y-direction, and 44 N in the z-direction. A dynamometer (Kistler 9129A, Winterthur, Switzerland) was used as a reference for the force measurement. Both systems have a sampling rate of 1000 Hz. The low-pass frequency for the feeling turret is 25 Hz and 30 Hz for the dynamometer. The passive force *Fp* acts in the x-direction, the cutting force *Fc* acts in the y-direction, and the feed force *Ff* acts in the z-direction. The numerical control (NC) provided the axis positions with a frequency of 83 Hz. The measurement setup is shown in Figure 1.

Within the evaluation, the process force was also simulated to investigate the process parameter range in smaller steps. The force model was used according to Kienzle [22]. The Kienzle parameters applied for the modelling were determined experimentally on the presented test set-up for the material and tool combinations. These parameters are shown in Table 2.


**Table 2.** Summary of Kienzle parameters.

**Figure 1.** Presentation of the measurement setup.

#### *2.3. Simulation of Process Errors*

A tool breakage can result in a sudden increase in process force. This can be caused by the broken cutting-edge jamming between the tool and workpiece. Afterward, the process force decreases because the gap between tool and workpiece must first be overcome by the travel movement. Then the remaining part of the tool collides with the workpiece, which increases the process forces again [23]. The described drop in force was simulated in the experiments by a material defect depicted in Figure 2a. The missing material results in an identical force characteristic as in the case of a tool breakage. The advantage of these simulated errors is the reproducibility and the fact that the machine tool components are not critically stressed. To investigate, if tool wear can be identified with the systems, the resulting force for machining with a new and a worn tool was measured. The worn tool exhibits a cutting-edge failure with a length of 465 μm, which is depicted in Figure 2b.

**Figure 2.** (**a**) Prepared workpiece and (**b**) presentation of the new and worn tool.
