| γ’ strengthened cobalt-base superalloys have the potential to become a new generation of high-temperature structural materials,and were praised as one of the seven development tendencies of superalloys in the keynote report of Eurosuperalloys 2014.However,there are still many challenges in the development and application of this class of alloys,such as low γ’ solvus temperature,narrow γ/γ’two-phase region,poor oxidation resistance and high alloy density.The development of multi-component CoNi-base superalloys is an effective way to solve the above problems.With the increase of alloying components,the alloying principle of complex alloy system becomes one of the key scientific problems which needs to be solved urgently,especially the effects of alloying elements on the microstructure stability above 1000℃.However,the researches on this topic are still limited at present.On the other hand,it takes a long period and a high cost to solve the above scientific problem through conventional research methods.Therefore,it is necessary to develop a research approach with high efficiency to accelerate the study of alloying principles in multi-component CoNi-base superalloys.Based on the previous research works and the concept of MGE(Material Genome Engineering),the following studies were carried out for multi-component CoNi-based superalloys in this thesis:1)A high-throughput experimental approach based on multi-component diffusion multiple was developed.Based on this approach,the effects of y and y’ forming elements on the microstructure stability of CoNi-base superalloys at 1000℃ were studied.At the same time,an experimental database for CoNi-base superalloys was established,which contains 1700 data of the quantitative relationship between the composition and microstructure.2)Based on the above database,a model for microstructural prediction of CoNi-base superalloys was established by machine learning,in order to develop a data-driven alloy design method.3)The influence of Ni,Cr and Al/W ratio on the micro structure stability at 1000-1150℃,nano hardness of γ and γ’ phase at room temperature,as well as oxidation property at 1000℃ of CoNi-base superalloys were studied by combining various characterization methods,in order to further understand the alloying principles of these three key parameters and guide the alloy design.In this study,a multi-component diffusion multiple consisted of 15 diffusion couples and 7 diffusion triples was designed.It contains a base alloy Co-20Ni-7Al-8W-1Ta-4Ti and other 8 senary or septenary CoNi-base superalloys.The objectives of the diffusion multiple are to study the alloying effects of Ni,Cr,Al,W,Ti,Ta,Mo and Nb on the microstructure stability of base alloys after being aging at 1000℃ for 1000 h,and accumulate experimental data for machine learning.The results of the diffusion multiple indicate that the γ’ volume fraction is increase by increasing Ni content.However,a very high content of Ni promotes the precipitation of harmful x phase.Increasing the Cr/W ratio can significantly increase the maximum addition amount of Cr,devoid of the precipitation of secondary phase.Furthermore,the addition of Cr increases the γ’ volume fraction,which compensates the reduction of γ’ volume fraction caused by decreasing W content to some extent.With substituting Al for W,i.e.,increasing Al/W ratio,theγ’ volume fraction increases first and then decreases,and the alloy density reduces significantly.However,the β phase precipitates when the value of Al/W ratio is too high.The substitution of W with Ta,Mo and Nb promote the precipitation of x phase and further decrease the γ’ volume fraction.However,substituting Ti for W by 6%does not promote the precipitation of any secondary phase,and only slightly reduces the γ’ volume fraction.The potency of γ’ forming elements for promoting the precipitation of detrimental phases is Nb>Ta≈Al>Mo>Ti.A large amount of experimental data on the quantitative relationship between the composition and microstructure of CoNi-based superalloys were accumulated by the diffusion multiple,including composition-phase constituent and composition-γ’ volume fraction at 1000℃.Based on these data,a machine learning model was established to predict the quantitative relationship between the alloy composition and microstructure parameters(phase constituent and γ’ volume fraction at 1000℃)of CoNi-base superalloys.By comparing the prediction results of machine learning and CALPHAD(CALculation of PHase Diagrams),it is indicated that the prediction accuracy of machine learning model for multi-component CoNi-base superalloys is much higher,which compensates the disadvantage of CALPHAD in this aspect.However,CALPHAD can compensate the weak extrapolation capability of machine learning.Therefore,the combination of machine learning and CALPHAD is more beneficial to the alloy development of multi-component alloy systems.The results regarding the effects of Ni,Cr and Al/W ratio on the microstructure stability at 1000-1150℃ and properties of multi-component CoNi-base superalloys indicate that simultaneously increasing the Ni content and Al/W ratio significantly improve the microstructure stability at high temperature.Alloy 30Ni-B with a high Ni content and Al/W ratio still maintain a high γ’ volume fraction devoid of secondary phase after being aging at 1150℃ for 1000 h.The increase of Al/W ratio results in the change of partitioning behavior of γ/γ’ phases,and further changes the morphology of γ’ phase significantly.In addition,increasing the Al/W ratio also promotes the formation of a dense continuous Al2O3 oxide layer at 1000℃,which significantly improves the oxidation resistance.However,increasing the Al/W ratio obviously reduces the room temperature hardness of y and γ’phases.Cr addition promotes the formation of continuous(but not dense)Cr2O3 oxide layer at 1000℃,and improve the oxidation resistance.Cr addition also significantly increases the hardness of γ’ phase at room temperature.However,increasing Cr content promotes W distributing from γ’ phase to γ phase,as well as decreases the solid solubility of W in γ phase,resulting in the precipitation of μphase.In summary,this study developed an alloying design approach for multi-component CoNi-base superalloys based on multi-component diffusion multiple and machine learning.This approach can not only accelerate the development of this class of alloys,but also provide a demonstration for the application of MGE technology in the research field of superalloys.This approach can also be applied in the development of other metal materials.On the other hand,the effects of alloying elements on the microstructure stability of multi-component CoNi-base superalloys at 1000-1150℃ are systematically studied to further develop the alloying principle of γ’ strengthened cobalt-base superalloys,which can provide a physical metallurgical basis for the compositional design of this class of alloys. |
PDF Full Text Request