Mineral admixture concrete carbon capture technology has emerged as a research focus in negative emission technologies due to its substantial carbon sequestration potential and high utilization rate of industrial soli...Mineral admixture concrete carbon capture technology has emerged as a research focus in negative emission technologies due to its substantial carbon sequestration potential and high utilization rate of industrial solid waste.However,early-age carbonation significantly increases the risk of steel reinforcement corrosion.Existing carbonation models,primarily based on 28-day standard-cured specimens,exhibit considerable conservative bias when predicting the carbonation behavior of concrete containing high-volume mineral admixtures and subjected to early-age exposure.To address this limitation,fifteen groups of fly ash(FA)and ground granulated blast furnace slag(GGBS)concrete specimens were prepared and subjected to accelerated carbonation tests after 14 days of standard curing.Building upon the Huang Shiyuan model framework,a Mineral Admixture Carbonation Rate Influence Coefficient(Km)was introduced for the first time to characterize the early-age effects of admixture type and dosage,and the Particle Swarm Optimization(PSO)algorithm,with global search capability in multi-parameter nonlinear systems,was used to optimize the parameters.The newly introduced Kmcoefficient quantitatively captures the differential effects of FA and GGBS,supported by microstructural reasoning.Independent validation confirmed its general applicability even for concrete incorporating FA,GGBS,and a wider range of water-to-binder ratios(W/B)(measured-to-predicted ratio:0.93±0.15).This model provides a theoretical tool for managing the early-age steel reinforcement corrosion risk in carbon capture concrete.展开更多
Recent advances in topology optimization have highlighted the importance of integrating design and manufacturing considerations.While much of the existing work has focused on additive manufacturing,there is growing in...Recent advances in topology optimization have highlighted the importance of integrating design and manufacturing considerations.While much of the existing work has focused on additive manufacturing,there is growing interest in exploiting the capabilities of multi-axis machining as part of design for manufacturing.In particular,modern 5-axis CNC milling machines offer substantial geometric flexibility by enabling tool access from multiple orientations,making them well suited for fabricating topology-optimized components with complex features.However,most existing methods rely on a limited set of predefined milling directions,and extending them to arbitrary tool orientations often incurs significant computational cost.This paper presents a practical adaptive framework,based on the level-set method,for handling discretized arbitrary milling access directions in topology optimization.At each iteration,the boundary velocity field is first computed without manufacturing constraints.Tool accessibility is then evaluated,and the velocity field is adaptively modified to enforce manufacturability while minimizing performance degradation.Rather than embedding milling constraints directly into the optimization formulation,the proposed approach applies incremental,localized corrections.A key contribution is the incorporation of a manufacturing-feasibility update loop into the approximate problem,which enhances computational efficiency.The proposed method has been implemented in the commercial code OptiStruct.Numerical examples demonstrate that the resulting 5-axis milling designs closely match unconstrained topology optimization solutions in both geometry and performance.The results indicate that the design freedom offered by 5-axis milling can,in most cases,be effectively utilized to obtain manufacturable designs with negligible performance penalties.展开更多
目的分析急性缺血性脑卒中患者步态异常的早期风险标志物,并构建预警模型。方法纳入2023年1月至2025年1月于解放军总医院接受治疗的急性缺血性脑卒中患者300例,依据患者发病后3个月是否表现出步态异常情况分为异常组105例和正常组195例...目的分析急性缺血性脑卒中患者步态异常的早期风险标志物,并构建预警模型。方法纳入2023年1月至2025年1月于解放军总医院接受治疗的急性缺血性脑卒中患者300例,依据患者发病后3个月是否表现出步态异常情况分为异常组105例和正常组195例,收集患者一般资料,使用均方根曲线构建列线图预测模型,并评估其准确度与一致性;采用ROC曲线评估列线图的临床预测价值。结果步态异常组年龄、美国国立卫生研究院卒中量表(National Institutes of Health Stroke scale,NIHSS)(下肢)评分、弥散加权成像(diffusion-weighted imaging,DWI)早期病灶体积、皮质脊髓束(corticospinal tract,CST)受累程度1级、神经元特异性烯醇化酶(neuron-specific enolase,NSE)、超敏C反应蛋白(high-sensitivity C-reactive protein,hs-CRP)水平明显高于正常组,功能步态分级(functional ambulation classification,FAC)明显低于正常组(P<0.05,P<0.01)。基于年龄、NIHSS(下肢)评分、简易精神状态检查表(mini-mental state examination,MMSE)评分、FAC、DWI早期病灶体积、CST受累程度、NSE和hs-CRP构建预测模型,列线图内部验证一致性指数为0.979(95%CI:0.975~0.983),ROC曲线下面积为0.985(95%CI:0.975~0.994)。结论急性缺血性脑卒中患者预后步态异常的早期风险因素包括年龄、NIHSS评分、DWI早期病灶体积、CST受累程度、NSE以及hs-CRP,保护因素包括MMSE、FAC,基于这些因素构建的列线图模型具有较好的预测性能。展开更多
摘要Mineral admixture concrete carbon capture technology has emerged as a research focus in negative emission technologies due to its substantial carbon sequestration potential and high utilization rate of industrial solid waste.However,early-age carbonation significantly increases the risk of steel reinforcement corrosion.Existing carbonation models,primarily based on 28-day standard-cured specimens,exhibit considerable conservative bias when predicting the carbonation behavior of concrete containing high-volume mineral admixtures and subjected to early-age exposure.To address this limitation,fifteen groups of fly ash(FA)and ground granulated blast furnace slag(GGBS)concrete specimens were prepared and subjected to accelerated carbonation tests after 14 days of standard curing.Building upon the Huang Shiyuan model framework,a Mineral Admixture Carbonation Rate Influence Coefficient(Km)was introduced for the first time to characterize the early-age effects of admixture type and dosage,and the Particle Swarm Optimization(PSO)algorithm,with global search capability in multi-parameter nonlinear systems,was used to optimize the parameters.The newly introduced Kmcoefficient quantitatively captures the differential effects of FA and GGBS,supported by microstructural reasoning.Independent validation confirmed its general applicability even for concrete incorporating FA,GGBS,and a wider range of water-to-binder ratios(W/B)(measured-to-predicted ratio:0.93±0.15).This model provides a theoretical tool for managing the early-age steel reinforcement corrosion risk in carbon capture concrete.
摘要Recent advances in topology optimization have highlighted the importance of integrating design and manufacturing considerations.While much of the existing work has focused on additive manufacturing,there is growing interest in exploiting the capabilities of multi-axis machining as part of design for manufacturing.In particular,modern 5-axis CNC milling machines offer substantial geometric flexibility by enabling tool access from multiple orientations,making them well suited for fabricating topology-optimized components with complex features.However,most existing methods rely on a limited set of predefined milling directions,and extending them to arbitrary tool orientations often incurs significant computational cost.This paper presents a practical adaptive framework,based on the level-set method,for handling discretized arbitrary milling access directions in topology optimization.At each iteration,the boundary velocity field is first computed without manufacturing constraints.Tool accessibility is then evaluated,and the velocity field is adaptively modified to enforce manufacturability while minimizing performance degradation.Rather than embedding milling constraints directly into the optimization formulation,the proposed approach applies incremental,localized corrections.A key contribution is the incorporation of a manufacturing-feasibility update loop into the approximate problem,which enhances computational efficiency.The proposed method has been implemented in the commercial code OptiStruct.Numerical examples demonstrate that the resulting 5-axis milling designs closely match unconstrained topology optimization solutions in both geometry and performance.The results indicate that the design freedom offered by 5-axis milling can,in most cases,be effectively utilized to obtain manufacturable designs with negligible performance penalties.
摘要目的分析急性缺血性脑卒中患者步态异常的早期风险标志物,并构建预警模型。方法纳入2023年1月至2025年1月于解放军总医院接受治疗的急性缺血性脑卒中患者300例,依据患者发病后3个月是否表现出步态异常情况分为异常组105例和正常组195例,收集患者一般资料,使用均方根曲线构建列线图预测模型,并评估其准确度与一致性;采用ROC曲线评估列线图的临床预测价值。结果步态异常组年龄、美国国立卫生研究院卒中量表(National Institutes of Health Stroke scale,NIHSS)(下肢)评分、弥散加权成像(diffusion-weighted imaging,DWI)早期病灶体积、皮质脊髓束(corticospinal tract,CST)受累程度1级、神经元特异性烯醇化酶(neuron-specific enolase,NSE)、超敏C反应蛋白(high-sensitivity C-reactive protein,hs-CRP)水平明显高于正常组,功能步态分级(functional ambulation classification,FAC)明显低于正常组(P<0.05,P<0.01)。基于年龄、NIHSS(下肢)评分、简易精神状态检查表(mini-mental state examination,MMSE)评分、FAC、DWI早期病灶体积、CST受累程度、NSE和hs-CRP构建预测模型,列线图内部验证一致性指数为0.979(95%CI:0.975~0.983),ROC曲线下面积为0.985(95%CI:0.975~0.994)。结论急性缺血性脑卒中患者预后步态异常的早期风险因素包括年龄、NIHSS评分、DWI早期病灶体积、CST受累程度、NSE以及hs-CRP,保护因素包括MMSE、FAC,基于这些因素构建的列线图模型具有较好的预测性能。