Date of Award


Degree Name

Doctor of Philosophy


Mechanical Engineering


Ryan Wicker


The fields of modern energy conversion, e.g., fossil fuel, natural gas, and coal power plants, are in increased need of health monitoring and data collection (temperature, pressure, and flow rate) to improve process efficiency and satisfy the demand for clean energy. In situ process monitoring can lead to a robust automated system to achieve the goal. The monitoring equipment or sensors in energy systems components (e.g., pressure tube, turbine blade, and fuel injector) undergo high temperature, high pressure, and corrosive environments; for example, combustion inlet conditions can reach up to 810 K and 2760 kPa [1]. Therefore, a reliable, accurate, and durable monitoring system is necessary that can withstand the requirements for operating in harsh environments. Wired or wireless sensors used in energy systems are typically bonded to the components or placed inside the cavities at targeted areas (e.g., flanges, connectors, and injectors). Placement of the sensor using these methods may require alterations to the design of the component that can affect the sensor's efficiency and increase its design complexity. In some cases, it could be difficult to access the inward areas of the component at which process monitoring is required. Moreover, a sensor's life can be reduced through its exposure to corrosive environments of high temperature and high pressure for a prolonged time.

This research primarily focused on the design and fabrication of smart parts with embedded piezoceramic material using powder bed fusion (PBF) additive manufacturing (AM) technology, and on the characterization of the sensor functionality and the performance of the smart parts. The process development and part design focused on the non-intrusive placement of a sensor within the additively manufactured structure, which can increase the sensor's life by limiting the contact of the sensor to corrosive combustor environments. Additive manufacturing is well known for fabricating complex shaped part based on computer aided design (CAD) data. The layer-by-layer fabrication process provides the prospect of obtaining access to a predesigned sensor cavity at any desired height for the embedding process to take place.

To demonstrate the smart part concept, EBM AM was used to fabricate a cylindrical shaped part where a piezoceramic material was embedded. A "stop and go" process was developed that included embedding of the piezoceramic material while pausing the EBM system at any desired location during fabrication. After fabrication was complete, the functionality of the fabricated part was evaluated using a compression-compression test. As a qualitative experimentation of the sensor functionality, the sensor showed voltage responses with the applied force at four different frequencies (i.e., 10 Hz, 15 Hz, 20 Hz, and 25 Hz) and obtained a highest average voltage-force ratio of 0.53 V/kN at 10 Hz frequency level.

After finishing the proof of concept cylindrical part, a simple smart tube was fabricated and tested for force and temperature sensing functionality. The smart tube was tested using the compression-compression test setup and a hot air setup for demonstrating force and temperature sensing, respectively. The smart tube showed force sensing capabilities at different loading conditions of 1 Hz, 10 Hz, and 15 Hz using a compression-compression test and temperature testing showed the measurement capabilities of temperatures from 150º C to 200º C using the embedded piezoceramic sensor material. The highest currents generated include 0.28 pA, 0.24 pA, and 0.2 pA for the three inlet conditions of 7.6 cm, 15.2 cm, and 30.5 cm, respectively.

Previously, the "stop and go" process required the use of a masking plate (a stainless steel plate was used), to work as a planar surface to continue fabrication of the smart part after embedding the sensor material within the cavity. The start plate in regular EBM fabrication process or masking plate in "stop and go" fabrication process is necessary to avoid interaction of electron beam with the metal powder that can create powder smoking within the powder bed and result in failure of fabrication process. Machining of the appropriate masking plate to fit a component can become cumbersome when fabricating a complex shaped geometry. As a means of discarding the use of the masking plate, a build scan pattern was developed, that allowed to restart the fabrication process after a pause, therefore allowing increased flexibility for the geometry of the part to be fabricated. The tensile properties of the specimens fabricated using "stop and go" process was evaluated and compared with the default or control specimens. The tensile properties are indicative measures for obtaining the bonding strength of the interface created by the "stop and go" fabrication process that helps in determining the applicability of the process for other engineering applications.

While the EBM "stop and go" process is performed, there is a need for accurate part registration. In the first part of this work, an IR camera-based image analysis process was explored for part positioning. Typically, the embedding process requires the removal of the base part from the powder bed, which creates part alignment issues when restarting the fabrication process. A beam positioning method was developed that allowed centering of the electron beam corresponding to the part within the powder bed after a fabrication pause was performed. This method was based on the calculation of the centroid position of the area of the paused surface layer. Furthermore, a procedure was also developed to calculate changes in the angular displacement of a part, before and after resuming fabrication, which allowed to reorient the part in the CAD followed by the build file preparation software previous to resuming fabrication to minimize part deviation. To evaluate the efficacy of procedures for part alignment mentioned above, a part registration study was performed which consisted of cylindrical part and rectangular prism to demonstrate accurate positioning of the part after pausing the system. A maximum misalignment of 0.17 mm and 0.87 mm was obtained for cylindrical part and rectangular prism, respectively that underwent "stop and go" fabrication process.

Finally, a smart injector was fabricated using a laser powder bed fusion (LPBF) technology employing the "stop and go" fabrication process. The main reason for employing LPBF for fabrication of the smart injector was the easy removal of metal powder from the internal channels and cavities in compare to the EBM fabricated parts. The LPBF technology does not require powder preheating steps that can partially sinter the metal powder in the powder bed unlike EBM fabrication process. The smart injector was tested in a combustion chamber that reached up to ~120º C to show the operative capability of a prototype smart part in harsh environments traditionally experienced by energy systems. In conclusion, this work resulted in the development of the "stop and go" fabrication process using PBF AM technologies, demonstrating the viability of this technique for producing multi-functional metallic components with applications in the manufacturing, aerospace, energy, automotive, and biomedical industries. This research will serve as a stepping stone for the future development of larger and higher complexity smart systems.

The sensing responses from the smart parts can be used in obtaining the structural health monitoring while in operation. The future developments can include improvement of build parameters to obtain an improved interface joint in compare to traditional joining methods (e.g., welding and adhesive). A similar fabrication approach that of "stop and go" can also be employed in repair and multi-material fabrication processes that can be a pathway for future applications using metal based AM technologies. The part positioning method can be employed in other AM based technologies to advancing the embedding, multi-material and hybrid AM fabrication processes.




Received from ProQuest

File Size

142 pages

File Format


Rights Holder

Mohammad Shojib Hossain