Incoloy 800H is an austenitic stainless steel alloy with nominal composition Fe-32Ni-21Cr developed by the Special Metals Corporation in the 1950s. The alloy was developed for the purpose of creating a corrosion- and creep-resistant metal which was cheaper than its nickel-based superalloy counterparts. It has a solid substitution strengthened single-phase Fe-Ni-Cr matrix, with additional strengthening from Ti(C,N) and M23C6 precipitates. As a result of this, it has since been used in various applications. In particular, it has been used in the petrochemical industry for methane reformer exit tubes. These tubes are created from 800H billets by a pilgering process, which involves several multiaxial, pseudo-periodic deformation steps in order to reduce the wall thickness to the desired level. Following the pilgering process, the tubes are solution annealed to recrystallize the material and increase the average grain size to the American Society for Testing and Materials (ASTM) grain size number 5 or coarser. In the field, the pigtail tubes can operate at temperatures in excess of 900 °C and stresses up to 10 MPa. They have a target service life of 20 years.

To quantify the creep performance of a material, researchers devised creep deformation mechanism maps (CDMMs). These maps plot isominimum strain-rate lines as a function of normalised stress and homologous temperature. During creep deformation, there are several competing creep mechanisms, such as the diffusion-controlled climb of edge dislocations, or the diffusion of vacancies through grain boundaries. CDMMs present regions on the map where each mechanism is the dominant mechanism, which is defined as the mechanism which contributes the largest amount to the total minimum strain-rate. Despite being used extensively in creep-based applications, there is a significant lack of knowledge about the creep performance capabilities of 800H. To date, a CDMM for 800H has not been created. This means that the industry uses the alloy without any knowledge of the true failure mechanism during service, or the limits to which they can maximise the temperature and stress during service. There is also significant lack of knowledge about how the microstructure, in particular the average grain size, affects the creep performance. Filling these voids of knowledge will be the primary objective of this research.

For this research, 800H was received in an as-pilgered condition. This was useful as it meant that there was more control over the microstructure, as each batch of material could be heat-treated for different average grain sizes by solution annealing at different temperatures. In this research, separate batches of creep test samples with mean grain diameters of 109 μm, 180 μm and 248 μm were achieved by solution annealing for 2 hours at 1100 °C, 1150 °C and 1200 °C respectively. With sixteen samples from each heat-treatment, and an addition ten samples of unknown heat-treatment, a total of fifty-eight samples were creep tested. These tests were performed at temperatures ranging from 750 - 1020 °C and stresses ranging from 15 - 105 MPa. Of the fifty-eight tests, twenty-four were tested until rupture, and the remainder were terminated after the minimum strain-rate had been passed in order to save time. Test times ranged between 10 - 2000 hours. By fitting data to a hyperbolic-sine power-law model, minimum strain-rates were reliably extracted from the raw creep test data. Minimum strain-rates varied from approximately 1.6 x 10-6 - 2.0 x 10-10 s-1. These testing conditions covered a comprehensive region of the CDMM.

The primary creep mechanisms of interest are power-law creep, which can be divided into high-temperature and low-temperature submechanisms, Coble creep, and Nabarro-Herring creep. By summing the constitutive equations for these models, the total minimum strain-rate as a function of temperature, stress and average grain size can be modelled. Using a computational optimisation algorithm called the genetic algorithm (GA), the material constants for the model were numerically fitted to the experimental data, a novelty for this application. The solutions were then assessed using a training and testing technique. Once the material constants were optimised, the CDMM was created for 800H. This map was then expanded to three-dimensions by showing a series of maps with different average grain sizes. Several novel CDMM concepts were explored, such as adding colour to visually represent the contribution of each mechanism to the total minimum strain-rate, converting the axes to absolute units, and truncating the axes of the figure to show more detail in the more important region of the map. Other creep models, such as a hyperbolic-sine model, and a model based on the Larson-Miller and Monkman-Grant equations were also explored and compared to the traditional model.

After finalising the CDMM for 800H, it was determined that these maps alone are not sufficient for practical applications. The maps show the mean minimum strain-rate based on the fit of the model to the experimental data. Although error for the fitting process was minimised by using a GA, the error still exists and is not represented on any maps that exist today. Therefore, it was decided to create a map which showed the level of certainty for the minimum strain-rate as a function of the position on the CDMM. Using several statistical analyses of the data to quantify aspects such as the variability of data tested at the same temperature and stress, and the extrapolability of the data, an error contour map was created. This novel concept is a first approximation and will require further development before it can be implemented.

The final section of work focused on microstructural aspects during creep deformation of 800H. Using several different microscopy techniques, in particular, energy-dispersive X-ray spectroscopy (EDS) and electron backscatter diffraction (EBSD), the microstructure of creep deformed 800H was explored. This work focused on the accumulation of damage as a function of creep strain, and the determination of the creep mechanism involved during the deformation. Key microstructural aspects that were assessed include the formation of subgrain boundaries, precipitation and void formation. The gathered information was compared to a 20-year service 800H pigtail tube, which was used to determine the level of strain in the observed microstructure.

Overall, this work provides a significant contribution towards the understanding of the creep performance characteristics of 800H. This includes novel contributions to the fields of creep model optimisation, statistical analysis of creep data, and microstructural characterisation, all of which can be potentially extrapolated to related materials. Specifically, this work acts as a foundation for the understanding of creep deformation of 800H, which others can build upon. With this work, industries should be able to more reliably predict the creep deformation performance of 800H.