Bone-mimicking gradient porous NiTi shape memory alloys(SMAs)are promising for orthopedic im-plants due to their distinctive superelastic functional properties.However,premature plastic deformation in weak areas such ...Bone-mimicking gradient porous NiTi shape memory alloys(SMAs)are promising for orthopedic im-plants due to their distinctive superelastic functional properties.However,premature plastic deformation in weak areas such as thinner struts,nodes,and sharp corners severely deteriorates the superelasticity of gradient porous NiTi SMAs.In this work,we prepared gradient porous NiTi SMAs with a porosity of 50%by additive manufacturing(AM)and achieved a remarkable improvement of superelasticity by a simple solution treatment regime.After solution treatment,phase transformation temperatures dropped signif-icantly,the dislocation density decreased,and partial intergranular Ti-rich precipitates were transferred into the grain.Compared to as-built samples,the strain recovery rate of solution-treated samples was nearly doubled at a pre-strain of 6%(up to 90%),and all obtained a stable recoverable strain of more than 4%.The remarkable superelasticity improvement was attributed to lower phase transformation tem-peratures,fewer dislocations,and the synergistic strengthening effect of intragranular multi-scale Ti-Ni precipitates.Notably,the gradient porous structure played a non-negligible role in both superelasticity deterioration and improvement.The microstructure evolution of the solution-treated central strut after constant 10 cycles and the origin of the stable superelastic response of gradient porous NiTi SMAs were revealed.This work provides an accessible strategy for improving the superelastic performance of gra-dient porous NiTi SMAs and proposes a key strategy for achieving such high-performance architectured materials.展开更多
Texture is inevitably introduced during the manufacturing of most NiTi shape memory alloys(SMAs),and the textured nanocrystalline NiTi has been extensively employed in engineering.However,the effect of texture,and the...Texture is inevitably introduced during the manufacturing of most NiTi shape memory alloys(SMAs),and the textured nanocrystalline NiTi has been extensively employed in engineering.However,the effect of texture,and the joint effect of grain size(GS)and texture on the functional properties of NiTi SMAs and the corresponding microscopic mechanisms have not been clearly understood yet.In this work,based on the phase field method,the effect of texture on the GS-dependent functional properties of NiTi SMAs,including super-elasticity(SE),one-way shape memory effect(OWSME),and stress-assisted two-way shape memory effect(SATWSME),is investigated,and the corresponding microscopic mechanisms are revealed.Moreover,the samples with discrete geometrical gradients and/or texture gradients are designed to achieve graded functional properties.The simulation results indicate that the dependence of functional properties on texture is due to the effect of crystallographic orientation on martensite transformation and reorientation,which can lead to different inelastic strains.In the designed samples with texture gradients,the stress–strain responses of sheets with various textures are different,allowing for the coordination of overall deformation of the sample by combining such sheets,with varying inelastic deformation degrees.Thus,the overall response of the sample differs from that without texture gradient,leading to the achievement of graded functional properties.The simulation results and new findings in this work contribute to a deeper understanding of the effects of texture,GS,and their interaction on the functional properties of SMAs,and provide valuable reference for the design and development of SMA-based devices with desired functional properties.展开更多
Gradient nanostructured(GNS)metallic materials are commonly achieved by gradient severe plastic de-formation with a gradient of nano-to micro-sized structural units from the surface/boundaries to the center.Certainly,...Gradient nanostructured(GNS)metallic materials are commonly achieved by gradient severe plastic de-formation with a gradient of nano-to micro-sized structural units from the surface/boundaries to the center.Certainly,such GNS can be inversely positioned,which however has not yet been reported.The present work reports a facile method in deformation gradient control to attain inverse gradient nanostructured(iGNS),i.e.,tailoring the cross-section shape,successfully demonstrated in Ti-50.3Ni shape memory alloy(SMA)wire through cold rolling.The microstructure of the rolled wire is characterized by a macroscopic inverse gradient from boundaries to the center—the average sizes of grain and martensite domain evolve from micrometer to nanometer scale.The iGNS leads to a gradient martensitic transforma-tion upon stress,which has been proved to be effectively reversible via in-situ bending scanning electron microscopy(SEM)observations.The iGNS Ti-50.3Ni SMA exhibits quasi-linear superelasticity(SE)in a wide temperature range from 173 to 423 K.Compared to uniform cold rolling,the gradient cold rolling with less overall plasticity further improves SE strain(up to 4.8%)and SE efficiency.In-situ tensiling synchrotron X-ray diffraction(SXRD)analysis reveals the underlying mechanisms of the unique SE in the iGNS SMAs.It provides a new design strategy to realize excellent SE in SMAs and sheds light on the advanced GNS metallic materials.展开更多
This study investigates the application of Learnable Memory Vision Transformers(LMViT)for detecting metal surface flaws,comparing their performance with traditional CNNs,specifically ResNet18 and ResNet50,as well as o...This study investigates the application of Learnable Memory Vision Transformers(LMViT)for detecting metal surface flaws,comparing their performance with traditional CNNs,specifically ResNet18 and ResNet50,as well as other transformer-based models including Token to Token ViT,ViT withoutmemory,and Parallel ViT.Leveraging awidely-used steel surface defect dataset,the research applies data augmentation and t-distributed stochastic neighbor embedding(t-SNE)to enhance feature extraction and understanding.These techniques mitigated overfitting,stabilized training,and improved generalization capabilities.The LMViT model achieved a test accuracy of 97.22%,significantly outperforming ResNet18(88.89%)and ResNet50(88.90%),aswell as the Token to TokenViT(88.46%),ViT without memory(87.18),and Parallel ViT(91.03%).Furthermore,LMViT exhibited superior training and validation performance,attaining a validation accuracy of 98.2%compared to 91.0%for ResNet 18,96.0%for ResNet50,and 89.12%,87.51%,and 91.21%for Token to Token ViT,ViT without memory,and Parallel ViT,respectively.The findings highlight the LMViT’s ability to capture long-range dependencies in images,an areawhere CNNs struggle due to their reliance on local receptive fields and hierarchical feature extraction.The additional transformer-based models also demonstrate improved performance in capturing complex features over CNNs,with LMViT excelling particularly at detecting subtle and complex defects,which is critical for maintaining product quality and operational efficiency in industrial applications.For instance,the LMViT model successfully identified fine scratches and minor surface irregularities that CNNs often misclassify.This study not only demonstrates LMViT’s potential for real-world defect detection but also underscores the promise of other transformer-based architectures like Token to Token ViT,ViT without memory,and Parallel ViT in industrial scenarios where complex spatial relationships are key.Future research may focus on enhancing LMViT’s computational efficiency for deployment in real-time quality control systems.展开更多
In this paper, a new class of three term memory gradient method with non-monotone line search technique for unconstrained optimization is presented. Global convergence properties of the new methods are discussed. Comb...In this paper, a new class of three term memory gradient method with non-monotone line search technique for unconstrained optimization is presented. Global convergence properties of the new methods are discussed. Combining the quasi-Newton method with the new method, the former is modified to have global convergence property. Numerical results show that the new algorithm is efficient.展开更多
Laser additive manufacturing (AM) of lattice structures with light weight, excellent impact resistance, and energy absorption performance is receiving considerable attention in aerospace, transportation, and mechanica...Laser additive manufacturing (AM) of lattice structures with light weight, excellent impact resistance, and energy absorption performance is receiving considerable attention in aerospace, transportation, and mechanical equipment application fields. In this study, we designed four gradient lattice structures (GLSs) using the topology optimization method, including the unidirectional GLS, the bi-directional increasing GLS, the bi-directional decreasing GLS and the none-GLS. All GLSs were manufactureed by laser powder bed fusion (LPBF). The uniaxial compression tests and finite element analysis were conducted to investigate the influence of gradient distribution features on deformation modes and energy absorption performance of GLSs. The results showed that, compared with the 45° shear fracture characteristic of the none-GLS, the unidirectional GLS, the bi-directional increasing GLS and the bi-directional decreasing GLS had the characteristics of the layer-by-layer fracture, showing considerably improved energy absorption capacity. The bi-directional increasing GLS showed a unique combination of shear fracture and layer-by-layer fracture, having the optimal energy absorption performance with energy absorption and specific energy absorption of 235.6 J and 9.5 J g-1 at 0.5 strain, respectively. Combined with the shape memory effect of NiTi alloy, multiple compression-heat recovery experiments were carried out to verify the shape memory function of LPBF-processed NiTi GLSs. These findings have potential value for the future design of GLSs and the realization of shape memory function of NiTi components through laser AM.展开更多
The inelastic deformations of shape memory alloys(SMAs)always show poor controllability due to the avalanche-like martensite transformation,and the effective control for the deformation of precision de-vices has been ...The inelastic deformations of shape memory alloys(SMAs)always show poor controllability due to the avalanche-like martensite transformation,and the effective control for the deformation of precision de-vices has been not yet mature.In this work,the phase field method was used to investigate the shape memory effects(SMEs)of NiTi SMAs undergoing grain size(GS)engineering,to obtain tunable one-way and stress-assisted two-way SMEs(OWSME and SATWSME).The OWSME and SATWSME of the systems with various gradient-nanograin structures and bimodal grain structure,as well as that with geometric gradients were simulated.The simulated results indicate that due to the GS dependences of martensite transformation and reorientation,the occurrence and expansion of martensite reorientation,martensite transformation and its reverse can be efficaciously controlled via the GS engineering.When combining the GS engineering and geometric gradient design,since the effects of GS and stress gradient can be su-perimposed or competing,and the responses of martensite reorientation,martensite transformation and its reverse to this are different,the OWSME and SATWSME of the geometrically graded systems with various nanograin structures can exhibit different improvements in controllability.In short,the reorienta-tion hardening modulus during OWSME is increased and the transformation temperature window during SATWSME is widened by GS engineering,indicating the improved controllability of SMEs.The optimal GS engineering schemes revealed in this work provide the basic reference and guidance for designing tun-able SMEs and producing NiTi-based driving devices catering to desired functional performance in various engineering fields.展开更多
针对现有基于深度确定性策略梯度(deep deterministic policy gradient,DDPG)算法的再入制导方法计算精度较差,对强扰动条件适应性不足等问题,在DDPG算法训练框架的基础上,提出一种基于长短期记忆-DDPG(long short term memory-DDPG,LST...针对现有基于深度确定性策略梯度(deep deterministic policy gradient,DDPG)算法的再入制导方法计算精度较差,对强扰动条件适应性不足等问题,在DDPG算法训练框架的基础上,提出一种基于长短期记忆-DDPG(long short term memory-DDPG,LSTM-DDPG)的再入制导方法。该方法采用纵、侧向制导解耦设计思想,在纵向制导方面,首先针对再入制导问题构建强化学习所需的状态、动作空间;其次,确定决策点和制导周期内的指令计算策略,并设计考虑综合性能的奖励函数;然后,引入LSTM网络构建强化学习训练网络,进而通过在线更新策略提升算法的多任务适用性;侧向制导则采用基于横程误差的动态倾侧反转方法,获得倾侧角符号。以美国超音速通用飞行器(common aero vehicle-hypersonic,CAV-H)再入滑翔为例进行仿真,结果表明:与传统数值预测-校正方法相比,所提制导方法具有相当的终端精度和更高的计算效率优势;与现有基于DDPG算法的再入制导方法相比,所提制导方法具有相当的计算效率以及更高的终端精度和鲁棒性。展开更多
基金the financial support of the National Natural Science Foundation under Grant No.52274387project support by the Shanghai Science and Technology Com-mission(Grant No.20S31900100).
文摘Bone-mimicking gradient porous NiTi shape memory alloys(SMAs)are promising for orthopedic im-plants due to their distinctive superelastic functional properties.However,premature plastic deformation in weak areas such as thinner struts,nodes,and sharp corners severely deteriorates the superelasticity of gradient porous NiTi SMAs.In this work,we prepared gradient porous NiTi SMAs with a porosity of 50%by additive manufacturing(AM)and achieved a remarkable improvement of superelasticity by a simple solution treatment regime.After solution treatment,phase transformation temperatures dropped signif-icantly,the dislocation density decreased,and partial intergranular Ti-rich precipitates were transferred into the grain.Compared to as-built samples,the strain recovery rate of solution-treated samples was nearly doubled at a pre-strain of 6%(up to 90%),and all obtained a stable recoverable strain of more than 4%.The remarkable superelasticity improvement was attributed to lower phase transformation tem-peratures,fewer dislocations,and the synergistic strengthening effect of intragranular multi-scale Ti-Ni precipitates.Notably,the gradient porous structure played a non-negligible role in both superelasticity deterioration and improvement.The microstructure evolution of the solution-treated central strut after constant 10 cycles and the origin of the stable superelastic response of gradient porous NiTi SMAs were revealed.This work provides an accessible strategy for improving the superelastic performance of gra-dient porous NiTi SMAs and proposes a key strategy for achieving such high-performance architectured materials.
基金The National Natural Science Foundation of China(12202294 and 12022208)the Project funded by China Postdoctoral Science Foundation(2022M712243)the Fundamental Research Funds for the Central Universities(2023SCU12098)are acknowledged.
文摘Texture is inevitably introduced during the manufacturing of most NiTi shape memory alloys(SMAs),and the textured nanocrystalline NiTi has been extensively employed in engineering.However,the effect of texture,and the joint effect of grain size(GS)and texture on the functional properties of NiTi SMAs and the corresponding microscopic mechanisms have not been clearly understood yet.In this work,based on the phase field method,the effect of texture on the GS-dependent functional properties of NiTi SMAs,including super-elasticity(SE),one-way shape memory effect(OWSME),and stress-assisted two-way shape memory effect(SATWSME),is investigated,and the corresponding microscopic mechanisms are revealed.Moreover,the samples with discrete geometrical gradients and/or texture gradients are designed to achieve graded functional properties.The simulation results indicate that the dependence of functional properties on texture is due to the effect of crystallographic orientation on martensite transformation and reorientation,which can lead to different inelastic strains.In the designed samples with texture gradients,the stress–strain responses of sheets with various textures are different,allowing for the coordination of overall deformation of the sample by combining such sheets,with varying inelastic deformation degrees.Thus,the overall response of the sample differs from that without texture gradient,leading to the achievement of graded functional properties.The simulation results and new findings in this work contribute to a deeper understanding of the effects of texture,GS,and their interaction on the functional properties of SMAs,and provide valuable reference for the design and development of SMA-based devices with desired functional properties.
基金supported by the National Natural Science Foundation of China(Grant Nos.52171007,52101166,51931004)the 111 Projects 2.0(Grant No.BP0618008).
文摘Gradient nanostructured(GNS)metallic materials are commonly achieved by gradient severe plastic de-formation with a gradient of nano-to micro-sized structural units from the surface/boundaries to the center.Certainly,such GNS can be inversely positioned,which however has not yet been reported.The present work reports a facile method in deformation gradient control to attain inverse gradient nanostructured(iGNS),i.e.,tailoring the cross-section shape,successfully demonstrated in Ti-50.3Ni shape memory alloy(SMA)wire through cold rolling.The microstructure of the rolled wire is characterized by a macroscopic inverse gradient from boundaries to the center—the average sizes of grain and martensite domain evolve from micrometer to nanometer scale.The iGNS leads to a gradient martensitic transforma-tion upon stress,which has been proved to be effectively reversible via in-situ bending scanning electron microscopy(SEM)observations.The iGNS Ti-50.3Ni SMA exhibits quasi-linear superelasticity(SE)in a wide temperature range from 173 to 423 K.Compared to uniform cold rolling,the gradient cold rolling with less overall plasticity further improves SE strain(up to 4.8%)and SE efficiency.In-situ tensiling synchrotron X-ray diffraction(SXRD)analysis reveals the underlying mechanisms of the unique SE in the iGNS SMAs.It provides a new design strategy to realize excellent SE in SMAs and sheds light on the advanced GNS metallic materials.
基金funded by Woosong University Academic Research 2024.
文摘This study investigates the application of Learnable Memory Vision Transformers(LMViT)for detecting metal surface flaws,comparing their performance with traditional CNNs,specifically ResNet18 and ResNet50,as well as other transformer-based models including Token to Token ViT,ViT withoutmemory,and Parallel ViT.Leveraging awidely-used steel surface defect dataset,the research applies data augmentation and t-distributed stochastic neighbor embedding(t-SNE)to enhance feature extraction and understanding.These techniques mitigated overfitting,stabilized training,and improved generalization capabilities.The LMViT model achieved a test accuracy of 97.22%,significantly outperforming ResNet18(88.89%)and ResNet50(88.90%),aswell as the Token to TokenViT(88.46%),ViT without memory(87.18),and Parallel ViT(91.03%).Furthermore,LMViT exhibited superior training and validation performance,attaining a validation accuracy of 98.2%compared to 91.0%for ResNet 18,96.0%for ResNet50,and 89.12%,87.51%,and 91.21%for Token to Token ViT,ViT without memory,and Parallel ViT,respectively.The findings highlight the LMViT’s ability to capture long-range dependencies in images,an areawhere CNNs struggle due to their reliance on local receptive fields and hierarchical feature extraction.The additional transformer-based models also demonstrate improved performance in capturing complex features over CNNs,with LMViT excelling particularly at detecting subtle and complex defects,which is critical for maintaining product quality and operational efficiency in industrial applications.For instance,the LMViT model successfully identified fine scratches and minor surface irregularities that CNNs often misclassify.This study not only demonstrates LMViT’s potential for real-world defect detection but also underscores the promise of other transformer-based architectures like Token to Token ViT,ViT without memory,and Parallel ViT in industrial scenarios where complex spatial relationships are key.Future research may focus on enhancing LMViT’s computational efficiency for deployment in real-time quality control systems.
文摘In this paper, a new class of three term memory gradient method with non-monotone line search technique for unconstrained optimization is presented. Global convergence properties of the new methods are discussed. Combining the quasi-Newton method with the new method, the former is modified to have global convergence property. Numerical results show that the new algorithm is efficient.
基金supported by the financial support from the National Natural Science Foundation of China(Nos.51735005 and U1930207)the Basic Strengthening Program(No.2019-JCJQ-JJ-331)+1 种基金National Natural Science Founda-tion of China for Creative Research Groups(No.51921003)the 15th Batch of‘Six Talents Peaks’Innovative Talents Team Program(No.TD-GDZB-001).
文摘Laser additive manufacturing (AM) of lattice structures with light weight, excellent impact resistance, and energy absorption performance is receiving considerable attention in aerospace, transportation, and mechanical equipment application fields. In this study, we designed four gradient lattice structures (GLSs) using the topology optimization method, including the unidirectional GLS, the bi-directional increasing GLS, the bi-directional decreasing GLS and the none-GLS. All GLSs were manufactureed by laser powder bed fusion (LPBF). The uniaxial compression tests and finite element analysis were conducted to investigate the influence of gradient distribution features on deformation modes and energy absorption performance of GLSs. The results showed that, compared with the 45° shear fracture characteristic of the none-GLS, the unidirectional GLS, the bi-directional increasing GLS and the bi-directional decreasing GLS had the characteristics of the layer-by-layer fracture, showing considerably improved energy absorption capacity. The bi-directional increasing GLS showed a unique combination of shear fracture and layer-by-layer fracture, having the optimal energy absorption performance with energy absorption and specific energy absorption of 235.6 J and 9.5 J g-1 at 0.5 strain, respectively. Combined with the shape memory effect of NiTi alloy, multiple compression-heat recovery experiments were carried out to verify the shape memory function of LPBF-processed NiTi GLSs. These findings have potential value for the future design of GLSs and the realization of shape memory function of NiTi components through laser AM.
基金The National Natural Science Foundation of China(12022208)the Project funded by China Postdoctoral Science Foundation(2022M712243)the Fundamental Research Funds for the Cen-tral Universities are acknowledged.
文摘The inelastic deformations of shape memory alloys(SMAs)always show poor controllability due to the avalanche-like martensite transformation,and the effective control for the deformation of precision de-vices has been not yet mature.In this work,the phase field method was used to investigate the shape memory effects(SMEs)of NiTi SMAs undergoing grain size(GS)engineering,to obtain tunable one-way and stress-assisted two-way SMEs(OWSME and SATWSME).The OWSME and SATWSME of the systems with various gradient-nanograin structures and bimodal grain structure,as well as that with geometric gradients were simulated.The simulated results indicate that due to the GS dependences of martensite transformation and reorientation,the occurrence and expansion of martensite reorientation,martensite transformation and its reverse can be efficaciously controlled via the GS engineering.When combining the GS engineering and geometric gradient design,since the effects of GS and stress gradient can be su-perimposed or competing,and the responses of martensite reorientation,martensite transformation and its reverse to this are different,the OWSME and SATWSME of the geometrically graded systems with various nanograin structures can exhibit different improvements in controllability.In short,the reorienta-tion hardening modulus during OWSME is increased and the transformation temperature window during SATWSME is widened by GS engineering,indicating the improved controllability of SMEs.The optimal GS engineering schemes revealed in this work provide the basic reference and guidance for designing tun-able SMEs and producing NiTi-based driving devices catering to desired functional performance in various engineering fields.
文摘针对现有基于深度确定性策略梯度(deep deterministic policy gradient,DDPG)算法的再入制导方法计算精度较差,对强扰动条件适应性不足等问题,在DDPG算法训练框架的基础上,提出一种基于长短期记忆-DDPG(long short term memory-DDPG,LSTM-DDPG)的再入制导方法。该方法采用纵、侧向制导解耦设计思想,在纵向制导方面,首先针对再入制导问题构建强化学习所需的状态、动作空间;其次,确定决策点和制导周期内的指令计算策略,并设计考虑综合性能的奖励函数;然后,引入LSTM网络构建强化学习训练网络,进而通过在线更新策略提升算法的多任务适用性;侧向制导则采用基于横程误差的动态倾侧反转方法,获得倾侧角符号。以美国超音速通用飞行器(common aero vehicle-hypersonic,CAV-H)再入滑翔为例进行仿真,结果表明:与传统数值预测-校正方法相比,所提制导方法具有相当的终端精度和更高的计算效率优势;与现有基于DDPG算法的再入制导方法相比,所提制导方法具有相当的计算效率以及更高的终端精度和鲁棒性。
