AbstractsBiology & Animal Science

The influence of stacked generators on source impedance of a piezoelectric spinal fusion implant

by Nathan Goetzinger

Institution: University of Kansas
Department: Bioengineering
Degree: M.E.
Year: 2014
Keywords: Biomedical engineering
Record ID: 2025582
Full text PDF: http://hdl.handle.net/1808/15925


The failure rate for difficult-to-fuse patients undergoing spinal fusion surgeries can be as high as 29-46%. The primary adjunct treatments which currently address this problem are bone morphogenetic proteins (BMPs) and DC electrical stimulation. BMPs have had clinical problems with off-label use and ectopic bone growth. DC electrical stimulation has had much clinical success, but it requires longer procedures (compared to surgeries with only spinal fusion cages and instrumentation) and it can require revision surgeries to remove the battery pack. To overcome these problems with current adjunct therapies, a piezoelectric spinal fusion implant has been developed which produces electrical stimulation without an external battery pack. One of the primary obstacles with creating a piezoelectric spinal fusion implant was the extremely high source impedance due to the capacitive nature of piezoelectric materials. Stacked actuators have been used with monolithic piezoelectric elements to lower source impedance, so this approach was utilized with a 3-1 composite piezoelectric material made from PZT 5A1 fibers with EPO-TEK 301 epoxy used as the matrix material. Specimens were made with 1 layer, 3 layers, 6 layers, or 9 layers and the effects of the number of layers on average maximum power and the load resistance at which average maximum power occurred (optimal load resistance) were studied. The effects of mechanical preload, mechanical load frequency, mechanical load amplitude, and poling electric field strength were also studied to further characterize the composite material. As the number of layers was varied from 1 to 9 with mechanical load amplitude of 1000 N and mechanical load frequency of 2 Hz, average maximum power was not significantly changed (p<0.05) and the optimal load resistance was lowered from 1 GOhms to 17 MOhms. When mechanical load frequency was increased from 1 to 5 Hz, the average maximum power for 9 layer implants with mechanical load amplitude of 1000 N increased from 551 µW to 2848 µW while shifting the optimal load resistance from 30 MOhms to 6 MOhms. Mechanical load amplitude was varied from 100 to 1000 N for 9 layer implants with mechanical load frequency of 2 Hz and resulted in average maximum power increasing from 8 µW to 1190 µW, with no significant effect on the optimal load resistance. Mechanical preload and poling electric field strength did not have a significant effect on either average maximum power or optimal load resistance. The optimal load resistance was successfully lowered by approximately 3 orders of magnitude without significantly affecting the average maximum power generated by the specimen. Future work will develop circuitry needed for rectification and conditioning of the signal to deliver current densities which are delivered by devices currently in clinical use. Work will also continue to attempt to lower the optimal load resistance of the piezoelectric spinal fusion implants to improve performance with the developed circuitry.