Design and heat treatment of 3D printed spinal cages

Abstract

Back pain and back instability affect close to 80% of the population at some point during their life. One solution is to insert a spinal cage, which results in spinal fusion, meaning two vertebrae fuse together to create one big vertebra. This study focuses on the investigation of both the processing aspects as the design aspects of producing a functional spinal cage. The spinal cage investigated is made of 3D printed Ti6Al4V using selective laser melting (SLM). The original design showed some undesirable qualities, which are a higher apparent Young's modulus compared to the surrounding bone, causing stress shielding; and the creation of high stress concentrations due to the implant design, which reduces the fatigue life. A new design is created and tested using a combination of FEM software and mechanical testing. Comparing the new spinal cage to the original spinal cage it was found that the Young's modulus is 4 times lower; the yield strength 27% lower, but does not yield when a force of 4 kN is applied as required; and the fatigue life 2.2 times higher. Currently, implants are post treated using hot isostatic pressing (HIP), which is an expensive procedure. One of the objectives of this research is to investigate the effectiveness of a vacuum oven as an alternative heat treatment method. This is done with SLM 3D printed Ti6Al4V samples with a porosity of 59%, which are heat treated at 850 and 1050 degrees Celsius for 2 hours. When comparing the vacuum heat treatment performed at 1050 with 850 degrees Celsius, the alpha grain thickness is increased 2.8 times, the Young's modulus with 13% and the yield strength with 7%. The vacuum heat treatments increase the fatigue life with 20% due to the removal of residual stresses and the transformation of the 'alpha-martensite to an alpha + beta microstructure as is required. Between the two heat treatments no significant difference in fatigue life is found. Comparing the fatigue life of the vacuum heat treatments with HIP at 920 degrees Celsius for 2 hours, it was observed that at 10,000 cycles the fatigue life of HIP is 20% better and at 1,000,000 cycles 1% worse. The main difference between the heat treatments is the reduction of internal pore size after HIP. The internal porosity affects the fatigue life. However, the effect of internal porosity will reduce faster compared to the effect of surface quality resulting in a similar fatigue life at high cycle fatigue. For spinal cages high cycle fatigue is more interesting to look at. Finally, commercial spinal implants are universally designed based on the requirements and life-style of a 30 year old male. In order to improve the success-rate of spinal fusion, more patient-specific implants can be designed and produced using additive manufacturing. In this research five different patient specific groups are defined based on expected load and bone density. Using these requirements five different spinal cages are designed based on different porosities. Using FEM software the minimum apparent Young's modulus which can be achieved for every patient group is found. The apparent Young's modulus found for every group ranges between 1.40 and 3.17 GPa. Finally, the fatigue graph and apparent Young's modulus obtained from the mechanical tests are used to validate the simulation. The simulated fatigue life was around three times lower compared to the actual fatigue life and the apparent Young's modulus was 6.3% higher compared to the actual apparent Young's modulus. These findings are used to correct the simulations and the new apparent Young's modulus for every patient group lies between 1.32 and 2.98 GPa which is even closer to the Young's modulus of the surrounding bone.