Prediction of Both E-Jet Printing Ejection Cycle Time and Droplet Diameter Based on Random Forest Regression.

This paper discusses a run-to-run iterative learning control (ILC) algorithm for Electrohydrodynamic jet (E-jet) printing. E-jet printing is a nano-manufacturing process that uses electric field induced fluid jet printing through nano-scale nozzles for achieving better control and resolution than traditional jet-printing processes. The printing process is controlled by changing the voltage potential between the nozzle and the substrate. However, it is difficult to maintain constant operating conditions such as stand-off height during a run of the printing process. The change in operating conditions results in fluctuating jet frequency. For stabilizing the jetting frequency across a single run, we propose a proportional ILC algorithm. We determine the jetting frequency by recording the electric current pulses when ink droplets are released from the nozzle. The frequency profile obtained from current measurements is then used to shape the voltage profile across a run to compensate for changing operating conditions. Experimental results are presented to validate the proposed control method.

High-resolution flexible electronic devices are widely used in the fields of soft robotics, smart human-machine interaction, and intelligent e-healthcare monitoring due to their mechanical flexibility, ductility, and compactness. The electrohydrodynamic jet printing (e-jet printing) technique is used for constructing high-resolution and cross-scale flexible electronic devices such as field-effect transistors (FETs), flexible sensors, and flexible displays. As a result, researchers are paying close attention to e-jet printing flexible electronic devices. In this review, we focused on the latest advancements in high-resolution flexible electronics made by e-jet printing technology, including various materials used in e-jet printing inks, the process control of e-jet printing, and their applications. First, we summarized various functional ink materials available for e-jet printing, including organic, inorganic, and hybrid materials. Then, the interface controlling the progress of e-jet printing was discussed in detail, including the physical and chemical properties of the functional ink, the interfacial wettability between the ink and substrate, and the micro-droplet injection behavior in a high-voltage field. Additionally, various applications of e-jet printing in the fields of flexible electrodes, FETs, flexible sensors, and flexible displays were demonstrated. Finally, the future problems and potential associated with the development of next generation e-jet printing technology for flexible electronic devices were also presented.

Electrohydrodynamic jet (E-jet) printing is a recent technique for high resolution additive micromanufacturing. With high resolution comes sensitivity to small disturbances, which has kept this technique from reaching its industrial potential. Closed loop control of E-jet printing can overcome these disturbances, but it requires an improved understanding of ink droplet spreading on the substrate and a physical model to predict printed feature locations and geometries from process inputs and disturbances. This manuscript examines a model of ink droplet spreading that uses assumptions that are important to the e-jet process. Our model leverages previous energy balance models that were derived for larger length scale droplets. At the smaller length scale, we find that viscous losses are a significant portion of the energy budget and must be accounted for; this is in contrast to models at length scales two orders of magnitude larger. Our model predicts the droplet height, base radius and contact angle in time from an initial volume and E-jet printing control parameters. The model is validated with published droplet spreading data and new measurements of E-jet printed droplets of diameter 8 μm. The viscous friction calculated in the new model is found to be significant compared to surface energy.

Globally, wind energy if properly harnessed, could serve as a source of energy generation in Africa. This study compared the performance of two Machine Learning (ML) algorithms (Linear regression and Random Forest) in predicting wind speed in five major cities in Africa (Yaoundé, Pretoria, Nairobi, Cairo and Abuja). Wind data were collected between January 1, 2000, and December 31, 2022, using the Solar Radiation Data Archive. The data preprocessing was carried out with 80% of the data used for training and 20% for validation. The performance of these ML algorithms was evaluated using Mean Square Error (MSE), Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE) and coefficient of determination (R2). The result shows that Nairobi (3.814795 m/s) closely followed by Cairo (3.606453 m/s) has the highest mean wind speed while Yaoundé (1.090512 m/s) has the lowest. Based on the performance metrics used, the two Machine Learning algorithms were competitive. Still, the Linear Regression (LR) algorithm outperformed the Random Forest Algorithm in predicting wind speed in all the selected major African cities. In Yaoundé (RMSE = 0.3892, MAE= 0.3001, MAPE =0.5030), Pretoria (RMSE=1.2339, MAE=0.9480, MAPE=0.7450) Nairobi (RMSE= 0.4223, MAE =0.6499, MAPE =0.1872), Nairobi (RMSE=0.6499, MAE=0.5171, MAPE =0.1872), Cairo (RMSE =1.0909, MAE =0.8544, MAPE =0.3541) and Abuja (RMSE = 0.70245, MAE =0.5441, MAPE= 0.4515) the Linear regression algorithms was found to outperformed Random Forest Regression. Therefore, the Linear regression algorithm is more reliable in predicting wind speed compared with the Random Forest regression.

Electrohydrodynamic jet printing of PZT thick film micro-scale structures

High-resolution additive direct writing of metal micro/nanostructures by electrohydrodynamic jet printing

Design, Development and Experimental Investigation of E-jet Based Additive Manufacturing Process

Electrohydrodynamic jet printing (e-jet printing) is a nascent additive manufacturing process most notable for extremely high resolution printing and having a vast portfolio of printable materials. These capabilities make e-jet printing promising for applications such as custom electronics and biotechnology fabrication. However, reliably fulfilling e-jet printing’s potential for high resolution requires delicate control of the volume deposited by each jet. Such control is made difficult by a lack of models that both capture the dynamics of volume deposition and are compatible with the control schemes relevant to e-jet printing. This work delivers such a model. Specifically, this work introduces a definition of “droplet volume” as a dynamically evolving variable rather than a static variable, and uses this definition along with analysis of high speed microscope videos to develop a hybrid dynamical system model of droplet volume evolution. This model is validated with experimental data, which involves the contribution of a novel technique for extracting consistent droplet volume measurements from videos.

This paper reports the use of advanced forms of electrohydrodynamic jet (e-jet) printing for creating micro- and nanoscale patterns of proteins on various surfaces ranging from flat silica substrates to structured plasmonic crystals, suitable for micro/nanoarray analysis and other applications in both fluorescent and plasmonic detection modes. The approaches function well with diverse classes of proteins, including streptavidin, IgG, fibrinogen, and γ-globulin. Detailed study reveals that the printing process does not adversely alter the protein structure or function, as demonstrated in the specific case of streptavidin through measurements of its binding specificity to biotin-modified DNA. Multinozzle printing systems enable several types of proteins (up to four currently) to be patterned on a single substrate, in rapid fashion and with excellent control over spatial dimensions and registration. High-speed, pulsed operational modes allow large-area printing, with narrow statistical distributions of drop size and spacing in patterns that include millions of droplets. The process is also compatible with the structured surfaces of plasmonic crystal substrates to enable detection without fluorescence. These collective characteristics suggest potential utility of e-jet techniques in wide-ranging areas of biotechnology, where its compatibility with various biomaterials and substrates with different topographies and surface chemistries, and ability to form deposits that range from thick films to submonolayer coatings, derive from the remote, noncontacting physical material transfer mode of operation.

Bioelectronic interfaces and biohybrid technologies have emerged as powerful solutions for sensing and manipulation applications. A unique method of underwater chemical sensing via a combination of engineered microbes and bioelectronic interfaces is discussed here with the goal of reporting to autonomous robotics through an electronic nervous system. The fabrication of these interfaces warrants a deposition method capable of precisely implementing the electronic, interfacial, and biological layers necessary for reliable devices. Electrohydrodynamic jet (e-jet) printing has been shown capable of doing so at a biologically relevant scale of 240 ± 50 nm. An e-jet printing system was fabricated and modified to use additive manufacturing to fabricate heterogeneous interfaces for the biological reporters Nitric Oxide and luminescence. These tasks included the fabrication of electronics that typically have multiple materials and thus multiple layers in their architecture; a potential point of failure in this liquid phase deposition method. The implementation of code based ejection modalities and a heated vacuum chuck allow for controlled volume deposition and rapid solvent evaporation, alleviating dissolution events in the sublayer. By eliminating these events, controlling film thickness to deter pinhole formation, and generating inks with solvents that were incompatible with sublayers, multi-layer electrochemical nitric oxide sensors and photosensors were produced. The fabricated multi-layer nitric oxide sensor has a lower limit of detection range of 188.5 nM - 314.5 nM; levels at which RAW 264.7 macrophage depolarization can be monitored in real time. An alternative bioelectronic interface is being studied wherein a modified version of this sensor is interfaced with Shewanella oneidensis to detect aqueous toxins. The photosensor architecture is comprised of a transparent PET substrate, an ITO cathode (pre-deposited), a bulk heterojunction P3HT:PC₇₁BM active layer, and a silver anode, with a final maximum spectral response of 7 mA/W and 90 mA/W/mm² at 514 nm, with a broad visible light spectrum from 400 nm to 620 nm. In interfacing these sensors with Vibrio fischeri, emitting light at 490 nm, the concentration of auto inducers in solution has been monitored in real time. These studies have created a clear path toward the fruition of olfactory guided, underwater autonomous robotics.

Numerical simulation of stable electrohydrodynamic cone-jet formation and printing on flexible substrate

Fabrication of piezoelectric thick-film stator using electrohydrodynamic jet printing for micro rotary ultrasonic motors

Estimating monthly average temperature by remote sensing in China

This research explores the optimization of mechanical properties and predictive modeling of polylactic acid (PLA) reinforced with boron nitride nanoplatelets (BNNPs) using data-driven machine learning (ML) models. PLA-BNNP composites were fabricated through injection molding, with a focus on how key processing parameters influence their mechanical performance. A Taguchi L27 orthogonal array was applied to assess the effects of BNNP composition (0.02 wt.% and 0.04 wt.%), injection temperature (135–155 °C), injection speed (50–70 mm/s), and pressure (30–50 bar) on properties such as tensile strength, Young’s modulus, and hardness. The results indicated that a 0.04 wt.% BNNP loading improved tensile strength, Young’s modulus, and hardness by 18.6%, 32.7%, and 20.5%, respectively, compared to pure PLA. Taguchi analysis highlighted that higher BNNP concentrations, along with optimal injection temperatures, improved all mechanical properties, although excessive temperatures compromised tensile strength and modulus, while enhancing hardness. Analysis of variance (ANOVA) revealed that injection temperature was the dominant factor for tensile strength (68.88%) and Young’s modulus (86.39%), while BNNP composition played a more significant role in influencing hardness (78.83%). Predictive models were built using machine learning (ML) models such as Random Forest Regression (RFR), Gradient Boosting Regression (GBR), and Extreme Gradient Boosting (XGBoost). Among the ML models, XGBoost demonstrated the highest predictive accuracy, achieving R2 values above 98% for tensile strength, 92–93% for Young’s modulus, and 96% for hardness, with low error metrics i.e., Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE). These findings underscore the potential of using BNNP reinforcement and machine learning-driven property prediction to enhance PLA nanocomposites’ mechanical performance, making them viable for applications in lightweight packaging, biomedical implants, consumer electronics, and automotive components, offering sustainable alternatives to petroleum-based plastics.

Accurate prediction of crop yield before harvest is critical to food security and importation. The calculated ten explanatory factors and autumn crop yield data were used as data sources in this research. Firstly, a Redundancy Analysis (RDA) was employed to carry out explanatory factors and feature selection. The simple effects of RDA were used to evaluate the interpretation rates of the explanatory factors. The conditional effects of RDA were adopted to select the features of the explanatory factors. Then, the autumn crop yield was divided into the training set and testing set with an 80/20 ratio, using Support Vector Regression (SVR), Random Forest Regression (RFR), and deep neural network (DNN) for the model, respectively. Finally, the coefficient of determination (R2), the root mean square error (RMSE), the mean absolute error (MAE), and the mean absolute percentage error (MAPE) were used to evaluate the performance of the model comprehensively. The results showed that the interpretation rates of the explanatory factors ranged from 54.3% to 85.0% (p = 0.002), which could reflect the autumn crop yields well. When a small number of sample training data (e.g., 80 samples) was used, the DNN model performed better than both SVR and RF models.