At present, artificial joint prosthesis is all prepared by injection molding process, but limited by complex shape structure, expensive mold development and long preparation cycle. With the improvement of patients' demand for therapeutic effect, personalized joint prosthesis must be the future direction. 3D printing can directly build personalized appearance structure in a short time through stacking materials layer by layer, which can well meet the needs of individuation and economy.
Compared with other molding processes, the orientation of filaments is unique and crucial for tribological properties of 3D printing PEEK. The friction coefficient and wear rate of 3D printing PEEK and UHMWPE friction pair were 0.14 and 0.777×10− 6 mm3/N∙m (Figs. 2 and 3), slightly larger than the friction pair of injection molded PEEK and UHMWPE [21]. The reason could be ascribed to the following: (Ⅰ) the 3D printing PEEK and UHMWPE friction pair performed 10,000 cycles and just entered the stable period of wear. While the study of Cowie et al. executed 1 million cycles and friction pair had been completely in the stable period, the wear rate calculated from the run-in period was obviously greater than the stable period [22]. The purpose of this study is to obtain the optimal printing direction of tribological property, and finally achieve the reasonable distribution of the printing path on the prosthesis based on overall wear. In fact, the tribometer in Fig. 1c can only execute one specimen at a time during friction. Therefore, considering the research purpose and time cost, the test was performed 10,000 cycles. Unsworth et al. removed the wear volume in the run-in period and severe wear period, and only calculated the wear rate in the stable period, so its wear rate was far less than that of other studies [19]. (Ⅲ) The 3D-printed surface is formed by regular arrangement of filaments, which will generate groove structure (Fig. 5) and roughness, this is one of the reasons for the high wear rate.
The mechanism of UHMWPE forming transfer film on the surface of 3D printing PEEK was discussed. In the initial stage of wear, the micro bulges on the surface of PEEK were embedded in the interior of UHMWPE due to higher hardness, and cut off the surface material of UHMWPE like tool to form wear debris during sliding. Afterwards, these wear debris would be mechanically captured by the groove texture of PEEK with hills and valleys, temporarily stored in the valleys. Finally, the UHMWPE wear debris gradually adhered to the surface of PEEK to form transfer film under the reduplicative action of contact stress. The hills stress of groove texture was too high for wear debris to enter, and the valleys stress was too low for wear debris to adhere on the surface of PEEK, so the transfer film mainly followed the transition area between the hills and valleys, namely the edge distribution of groove texture, which could be proved from Figs. 4 and 5.
Meanwhile, the effect of 3D printing orientation on the formation of transfer film was also analyzed. The suffered contact stress of wear debris on the hills of groove texture was much greater than that in the valleys during sliding wear. When the printing orientation of PEEK filaments was parallel to the sliding direction, the wear debris could continuously slide along the hills and valleys of groove texture (Fig. 10). As a result, the transfer film formed at 0° orientation possessed larger number and size (Fig. 4a1-a3). However, when the printing orientation of PEEK filaments was perpendicular to the sliding direction, the wear debris would alternately experience hills and valleys between groove textures, corresponding contact stress would change from the minimum of valleys to the maximum of hills and then repeated intermittently, which wasn’t conducive to form firm transfer film on the surface of PEEK. Consequently, the number and size of transfer film on the surface of PEEK was less at 90° orientation, which could be evident from Fig. 4e1-e3. Therefore, it can be concluded that steady stress was an important factor affecting the formation of transfer film.
Current some researches had reported the tribological behavior of PEEK, but the surface didn’t appear transfer film [19, 21, 22]. The reason could be attributed to the unique surface morphology formed by 3D printing process. These PEEK specimens were all fabricated by traditional injection molding process, and the surface roughness was less than Ra 0.1 \({\mu }\text{m}\), this smooth surface wasn’t conducive to capturing and storing wear debris. Therefore, UHMWPE wear debris couldn’t be adhered on the surface of injection molding PEEK to form transfer film. The 3D printing surface of PEEK was formed by regular arrangement of filaments, which possessed groove texture (Fig. 5a1-a4). They could mechanically capture wear debris, and then form transfer film under the action of contact stress. The friction coefficient and wear rate were all decreased with increasing formation of transfer film (Figs. 2 and 3). In conclusion, 3D printing PEEK showed unique advantage in the construction of groove morphology to form transfer film.
Furthermore, the wear mechanism of PEEK against with UHMWPE was also analyzed. The surface of UHMWPE appeared lots of furrows, and they changed from light and thin to deep and wide with the increasing orientation (Fig. 7). These furrows could be attributed to the micro-cutting effect of PEEK groove texture to UHMWPE. At 0° orientation, the sliding contact area between PEEK and UHMWPE was smaller compared with other orientations, the micro-cutting effect was slight, so the corresponding tribological properties were the best. Based on the above analysis, it could be known that the wear mechanism of 3D printing PEEK against with UHMWPE was abrasive wear.
The wear debris flowing into the serum is harmful and can cause osteolysis [38–41]. In this study, part of the wear debris flowed into the serum, while the other part adhered to the surface of PEEK to form transfer film. It can not only cover the matrix material of friction pair to improve tribological properties, but also reduce the flow of debris into the serum resulting in osteolysis. In addition, although the PEEK pins were cleaned by ultrasonic after wear test, it can be seen from the microstructure that the UHMWPE transfer film still adhered on the surface of PEEK and didn’t fall off with ultrasonic treatment (Figs. 4 and 5), which indicated that the UHMWPE transfer film has been firmly adhered to PEEK. A strong bond between transfer film and matrix would enhance the wear life of transfer film and avoid shedding into the serum causing biological damage.