Thermal Localization Improves the Interlayer Adhesion and Structural Integrity of 3D printed PEEK Lumbar Spinal Cages

Abstract

Additive manufacturing (AM) is a potential application for polyetheretherketone (PEEK) spinal interbody fusion cages, which were introduced as an alternative to titanium cages because of their biocompatibility, radiolucency and strength. However, AM of PEEK is challenging due to high melting temperature and thermal gradient. Although fused filament fabrication (FFF) techniques have been shown to 3D print PEEK, layer delamination was identified in PEEK cages printed with a first generation FFF PEEK printer [1]. A standard cage design [2] was 3D printed with a second generation FFF PEEK printer. The effect of changing layer cooling time on FFF cages' mechanical strength was investigated by varying nozzle sizes (0.2 mm and 0.4 mm), print speeds (1500 and 2500 mm/min), and the number of cages printed in a single build (1, 4 and 8). To calculate the porosity percentage, FFF cages were micro-CT scanned prior to destructive testing. Mechanical tests were then conducted on FFF cages according to ASTM F2077 [2]. Although altering the cooling time of a layer was not able to change the failure mechanism of FFF cages, it was able to improve cages' mechanical strength. Printing a single cage per build caused a higher ultimate load than printing multiple cages per build. Regardless of the cage number printed per build, cages printed with bigger nozzle diameter achieved higher ultimate load compared to cages printed with smaller nozzle diameter. Printing with a bigger nozzle diameter resulted in less porosity, which might have an additional affect on the interlayer delamination failure mechanism.

Keywords: 3D printing; Fused Filament Fabrication; Polyetheretherketone; Spinal Cage.

Fig. 1.

Spinal cages were printed using a commercial FFF machine capable of reaching the high temperatures associated with printing PEEK (Apium P220)(a) and PEEK 450G™ filament (Invibio) (b). The brims and support structures (c-left) were removed prior to testing (c-right).

Fig. 2.

Control volumes (shown as blue boxes) were taken from the both ends of the micro-CT scanned cage (a), which were then used to measure the porosity of the printed cages (b).

Fig. 3.

Compression tests were conducted on the printed cages as per ASTM F2077 [31] (a) in the direction orthogonal to the build layers (b), ultimate load (N) and ultimate displacement (mm) values were calculated from the load/displacement curves (c).

Fig. 4.

Printing single cages achieved higher ultimate load than printing multiple cages per build with bigger nozzle diameter (a) (purple bar showing the machined PEEK cage values with its mean and the standard deviation [1]) and smaller nozzle diameter under higher speed (b). Furthermore, cages printed with the bigger nozzle diameter showed higher ultimate load than printing with the smaller nozzle diameter when printing single and four cages per build under slower speed (c), whereas for all printing conditions under higher speed (d).

Fig. 5.

Printing single cages with different layer thicknesses did not show a significant difference in cages’ ultimate load for both nozzle diameters when printed under higher speed with the second-generation printer (a-b). There was not a significant difference observed between printer generations when single and multiple printed cages’ ultimate load was compared (c-d). Purple bar in all graphs is showing the machined PEEK cage values with its mean and the standard deviation [1].

Fig. 6.

Printing with smaller nozzle resulted in higher porosity in cages when printed under both print speeds (a-b). 1^st generation cages had significantly higher porosity than the 2^nd generation printer (c). A representative pore when printed with slower speed using the 1^st generation printer (d).