Generative artificial intelligence (AI) is increasingly used in landscape architecture to accelerate design development and enhance stakeholder engagement. This study presents a new landscape concept design tool to support collaborative greenspace problem-solving by integrating rapid scenario generation with a Pollinator Habitat Indicator (PHI). We define this as the PHI-AI Landscape Design Tool. The model combines text-to-image generative workflows with a simplified pollinator-oriented reference that compares AI-generated scenarios based on broad landscape cover composition. The PHI-AI Landscape Design Tool generated varied concept layouts with planting and pathway configurations responsive to descriptive prompts and user-defined parameters. It generates layout variations that reflect visually coherent planting configurations, and circulation connections to existing infrastructure for user functionality. This enables the rapid production of multiple design alternatives that remain legible, adaptable, and flexible to human-controlled editing prompts for stakeholder engagement. This study offers a framework for integrating a simplified ecological reference into the planning phase of landscape design.
Nudiviruses are double stranded DNA viruses known to infect arthropod hosts. Around 20 different nudiviruses have been identified from crustacean hosts alone. In crustaceans, the hepatopancreas is the primary target of infection, where nuclei appear hypertrophic under histological preparation due to a growing intranuclear viroplasm. The Jonah crab, Cancer borealis, is the target species of an emerging fishery located in the New England (USA) region. Histological and metagenomic tools were used to characterize a new nudivirus, including pathological characterization and sequencing of the 97,677 bp genome, which encoded 93 open reading frames. The genome housed a genetic synteny that was similar to other gammanudiviruses. The virus also phylogenetically clusters with other members of the Gammanudivirus genus based on a 17-protein concatenated phylogenetic tree. Comparative genomics revealed up to 66 conserved genes across all Gammanudivirus species, including a putative deneddylase that appears not to be present in all other nudivirus lineages. Discovering and identifying this nudivirus provides greater detail on a new pathogen that poses a potential risk to the emerging fishery.
This study investigates machine-learning-based prediction of bio-oil yield and quality indicators during the hydrothermal liquefaction (HTL) of lignocellulosic biomass. A curated literature-derived database was used to develop multilayer perceptron, support vector regression, decision tree, random forest (RF), and extreme gradient boosting (XGB) models for predicting bio-oil yield, elemental ratios, carbon retention, oxygen removal efficiency, and higher heating value. Predictor selection combined process knowledge with correlation analysis, while model hyperparameters were optimized using particle swarm optimization. XGB and RF generally provided the strongest performance; however, the models developed for most secondary outputs exhibited limited generalization. Consequently, detailed post-prediction analysis was restricted to bio-oil yield, for which the XGB model showed the most reliable predictive performance. Inclusion of the effective water fill ratio (Vw/Vr) substantially improved yield prediction by representing variations in solvent availability, reactor filling, and associated physicochemical interactions. SHapley Additive exPlanations ranked Vw/Vr as the most influential predictor, whereas Spearman analysis identified reaction time as the strongest monotonic factor (rho approximate to -0.43), followed by Vw/Vr (rho approximate to +0.34). Segmented regression and localized partial-dependence analyses reconciled these rankings by showing that the influence of reaction time was concentrated mainly below approximately 80 min, while Vw/Vr produced stronger nonlinear and regime-dependent changes, particularly below a modelderived breakpoint near 13.9. Joint-bootstrap Monte Carlo and residual-conformal analyses were further used to quantify predictive uncertainty and establish a model-credibility envelope based on interval width and applicability-domain support. Overall, the proposed framework provides an interpretable and uncertainty-aware approach for modeling lignocellulosic HTL and prioritizing conditions for subsequent process optimization.
Impinging jet flow and crossflow play different roles in drying. Jets enhance heat and mass transfer by disrupting the boundary layer at the material interface, while crossflow removes moisture accumulated in the drying volume. Excessive crossflow, however, can adversely affect the impinging jets, reducing their effectiveness. This study investigates how the relative strength of crossflow and impinging jets affects vapor transport and mass transfer. To understand the competing effects of impinging jets and crossflow on the drying performance, the local relative humidity, surface temperature, and drying rate were measured under different crossflow-to-jet velocity ratios (R). Experiments were conducted during the constant-rate drying period using a multiple-jet impingement system with superimposed crossflow. The results show that increasing R from 0 to 0.5 enhanced the local Nusselt and Sherwood numbers. Further increasing R from 0.5 to 1.0 reduced local relative humidity but progressively decreased the local heat transfer and drying rate due to deterioration of jet impingement regions. When crossflow became dominant (R > 1), the mass transfer efficiency of the jet impingement declined substantially. These findings demonstrate that mass transfer is governed by the balance between boundary-layer disruption, vapor transport, and heat transfer, rather than by vapor removal alone. Real-time measurements of local temperature and humidity fields provided insights into the mass transfer boundary conditions, enabling optimization of jet impingement drying systems.
This study investigates the electrochemical corrosion behavior of phosphor bronze (Cu-Sn-P) alloy fabricated by wire-arc additive manufacturing (WAAM) compared to conventionally as-cast material. The study examined three conditions: as-cast, as-built WAAM, and heat-treated (HT) WAAM alloys. Microstructural characterization showed that the as-cast alloy exhibited heterogeneous dendritic structures with interdendritic regions, while the as-built WAAM sample displayed predominantly columnar grain structure. Heat treatment produced a more equiaxed, homogeneous microstructure and notably altered grain-boundary characteristics. Electrochemical corrosion tests in 3.5 wt.% NaCl solution using electrochemical impedance spectroscopy (EIS), potentiodynamic polarization (PDP), and cyclic potentiodynamic polarization (CPP) techniques. Results show that the HT alloy exhibits the lowest corrosion current density and significantly higher impedance during long-term immersion, while the as-cast condition showed more severe degradation. X-ray photoelectron spectroscopy (XPS) showed that the HT condition has a higher presence of Cu+ and Sn oxidized species, suggesting that corrosion behavior is influenced by alloy chemistry, microstructural heterogeneity, and the composition and stability of the surface corrosion-product layer. These findings underscore the potential of post-printing heat treatment to significantly enhance the corrosion resistance of WAAM-fabricated Cu-Sn-P alloys for marine applications.
