Search

This paper proposes a novel physics-based equivalent circuit model of the lithium-ion battery for electric vehicle applications that has comprehensive electrochemical significance and an acceptable level of complexity. Initially, the physics-based extended single particle (ESP) model is improved by adding a correction term to mitigate its voltage bias. Then, the equivalent circuit model based on the improved extended single particle (ECMIESP) model is derived. In this model, the surface state of charge (SOC) of solid particles is approximated using a capacity and multi first-order resistance-capacity equivalent circuits with only two lumped parameters. The overpotential of electrolyte diffusion is approximated using a first-order resistance-capacitance equivalent circuit. The electrochemical reaction overpotential is characterized by a nonlinear resistance. The voltage accuracies of ECMIESP and conventional 2RC equivalent circuit model (ECM2RC) are compared across the entire SOC range under various load profiles. The results demonstrate that the ECMIESP model outperforms ECM2RC model, particularly at low SOC or when the electrochemical reaction overpotential exceeds 50 mV. For instance, the ECMIESP model shows an 820.4 mV reduction in voltage error compared to the ECM2RC model at the endpoint during a 2C constant current discharge test. Lastly, the ECMIESP model was used for SOC estimation with extended Kalman filter, resulting in significantly improved accuracy compared to the conventional ECM2RC model. Therefore, the ECMIESP model has great potential for real-time applications in enhancing voltage and SOC estimation precision. [ABSTRACT FROM AUTHOR]

The high-precision trajectory tracking and chatter-free performance of the manipulators are of great significance to the improvement of the automation level. Therefore, in view of the problems of chattering and tracking performance degradation of multi-joint manipulators caused by uncertain factors such as external disturbances and modeling errors, an adaptive fuzzy sliding mode control of the manipulator based on an improved super-twisting algorithm is proposed. First, fuzzy control is chosen to approximate the uncertainty of the manipulator model, which reduces the influence of unknown disturbances on the system, and then the adaptive law is utilized to automatically modify and improve the fuzzy rules to enhance the approximation capability; Second, an improved super-twisting algorithm is used on this basis to put the discontinuous term into the integral term to avoid chattering and eliminate the approximation error; Finally, a comparative simulation experiment is carried out with the 3-DOF manipulator as the research object. The experimental results show that the method in this paper improves the convergence speed by up to 69.74% and the tracking error accuracy by up to 85.11% over the fuzzy control based on adaptive reaching law with no chattering and little difference in control inputs, thus verifying the feasibility and superiority of the method. [ABSTRACT FROM AUTHOR]

Traumatic rupture of the aorta (TRA) is highly lethal in motor vehicle collisions (MVCs). Human body biomechanical models (HUBMs) show promise as useful tools for better comprehension of the mechanical response and injury mechanism of TRA. However, the prediction accuracy largely depends on its material modeling. This study investigated the performance of the isotropic elastic material constitutive model (Mat_001) and the orthotropic elastic material constitutive model (Mat_002) for predicting aortic dynamic response and injury in blunt side impacts, via simulations against vertical and oblique side impact cadaver tests with previously developed HUBM. The aortic strain response, peak intraluminal pressure, thoracic deflections and injury distributions were compared. It was found that both Mat_001 and Mat_002 could be used to model aortic tissue, but Mat_002 presented more sensitivity to blunt side impacts in terms of strain responses. The thoracic rib cage and internal organs sustained more severe injuries under the oblique left side impact (AIS 5+) than those under the vertical left side impact (AIS 4+), whereas the aortic peri-isthmus exhibited a higher TRA risk in the vertical side impact than that in the oblique side impact, with regard to the peak longitudinal strain. The TRA could result from the aorta's sudden stretch, and the peak longitudinal strain could be regarded as a main contributing factor to it. This study enhances our understanding of the aortic material modeling and injury mechanisms. [ABSTRACT FROM AUTHOR]

This paper presents a sliding mode control based on particle swarm optimization neural network and adaptive reaching law, and the proposed control method solves the problem of chattering and tracking performance degradation of a multi-joint manipulator caused by uncertainties such as external disturbances and modeling error. First, to address the problem that the precise dynamic system of the manipulator is difficult to establish, the radial basis function neural network (RBFNN) is used to approximate the uncertainty of the manipulator model, and the parameters of the neural network are optimized through the adaptive natural selection particle swarm optimization algorithm (ASelPSO) to improve the approximation ability and reduce the approximation error. Second, to eliminate chattering, adaptive reaching law is selected to improve the dynamic quality of approaching motion. Finally, a comparative simulation experiment is carried out with a 3-DOF manipulator as the research object. The results show that the control method has obvious improvements in eliminating chattering, improving tracking accuracy, and increasing convergence speed, which verifies the feasibility and superiority of the control scheme. [ABSTRACT FROM AUTHOR]

High-tilt reclined seats are one of the most popular configurations in highly automated vehicles; however, current restraint systems cannot protect out-of-position occupants in this type of seat. To reduce the risk of injury to reclined occupants, this study proposes a swiveling seat driven by occupant inertia and rotated in the sagittal plane during impact. The effectiveness of the swiveling seat was evaluated based on kinematics and injury to a human biomechanical model in a frontal sled test. A simulation matrix was constructed to design and optimize various safety devices, including the belt, pre-tensioner, knee constraint, and rotation stiffness for the swiveling seat. The results showed that (1) submarining occurred when the reclined occupant was on a fixed seat with a normal three-point belt during impact; (2) a fixed seat with a dynamic locking tongue and passenger lap pretension prevented the submarining, but produced a high lumbar force of 5359 N, which was higher than the spine injury criterion; and (3) the proposed swiveling seat with a matched restraint system could prevent submarining and produce lumbar force of 1787 N. The results demonstrated that the swiveling seat has high potential for occupant protection in intelligent driving scenarios. [ABSTRACT FROM AUTHOR]

This paper provides the consequences for trajectory tracking of uncertain n-DOF robotic manipulators in the presence of external disturbances and backlash. The control strategy consists of decoupling system based on inverse dynamics, sliding mode controller, and optimal controller. The tendency of robotic manipulators with gear backlash is to function in two unique phases: contact phase or backlash phase. The sliding mode controller works in the contact phase, after the decoupling of the manipulator, a adaptive second-order sliding mode controller is designed to guarantee system achieve desirable tracking precision and robust stability. The optimal controller works in the backlash phase, which makes the motor quickly traverse the backlash gap and re-contact the joints without collision. Finally, the proposal is verified in simulation using a model of a PUMA560 robot manipulator identified from real data. [ABSTRACT FROM AUTHOR]

Hepatic injury induced by blunt abdominal impact is a major cause of death in vehicle crashes. However, few works have been done effectively in cadaver experiments to clarify the liver's specific dynamic behavior and mechanical characteristics. This paper described the dynamic behavior and mechanical characteristics of the liver under blunt impacts to the upper abdomen with the finite element model (FEM) of the Chinese human body — the 50 percentile-sized male (CHUBM-M50). The simulation matrix, three directions (frontal, oblique, lateral), and four speeds included in each group were designed with a 23.4 kg, rigid cylindrical impactor aligning the T11 level. The liver deformation contours displayed compression against the spine and rotation in the horizontal plane, which were the two main features in liver motion. Pressure distribution in the liver capsule and parenchyma was discussed to elucidate the biomechanical characteristics related to impact direction. Generally, the stress distribution in the capsule was 10 times higher than that in the parenchyma. A discussion of the injury mechanism of the liver capsule and parenchyma observed in the simulations was given upon the pressure distribution. It demonstrated that the capsule could protect liver parenchyma at low-speed impacts and should not be neglected for understanding liver injury mechanisms. [ABSTRACT FROM AUTHOR]

Ensuring the reliability and stability of long‐distance power transmission requires monitoring the transmission lines' and grid environment's health. For this purpose, this study proposes a triboelectric nanogenerator (TENG) with anti‐self‐excited vibration. In case of aeolian vibration in the transmission line, the anti‐self‐excited oscillation triboelectric nanogenerator (AO‐TENG) generates electricity via cohesive work between the spring and the additional weight block that drives the TENG. As a result, this prevents excessive surge voltage from damaging the back‐end circuit, and the capacitor contributes significantly to voltage drop. This solution solves the issue that arises due to the TENG's high output voltage, making it difficult to power commercial sensors. This study presents a new design solution to facilitate smart grids that guarantee stable operation of distributed sensor nodes in the grids. [ABSTRACT FROM AUTHOR]

To investigate the aortic injury under different conditions, this study performed three impact simulations of the occupant to the interior components of an automobile. Blood in the aorta and heart was considered by the method of the surface-based fluid cavity capability in Abaqus, and the blood flow through the aortic valve was also simulated. The material of the aortic wall was modeled by HGO model to characterize the fiber property. The result showed that: (1) the injury mechanisms were different under different impact conditions. Compression of the aorta between the spine and sternum caused the rupture in the frontal impact. Twist and stretch of the aortic arch led to rupture in left and right impact, respectively. (2) blood pressure in the aorta did not significantly affect the maximum strain at the aorta, while the movement of the aorta and other tissues was the predominant cause of the blunt aortic rupture. [ABSTRACT FROM AUTHOR]

Picking robots have become an important development direction of smart agriculture, and the position detection of fruit is the key to realizing robot picking. However, the existing detection models have the shortcomings of missing detection and slow detection speed when detecting dense and occluded grape targets. Meanwhile, the parameters of the existing model are too large, which makes it difficult to deploy to the mobile terminal. In this paper, a lightweight GA-YOLO model is proposed. Firstly, a new backbone network SE-CSPGhostnet is designed, which greatly reduces the parameters of the model. Secondly, an adaptively spatial feature fusion mechanism is used to address the issues of difficult detection of dense and occluded grapes. Finally, a new loss function is constructed to improve detection efficiency. In 2022, a detection experiment was carried out on the image data collected in the Bagui rural area of Guangxi Zhuang Autonomous Region, the results demonstrate that the GA-YOLO model has an mAP of 96.87%, detection speed of 55.867 FPS and parameters of 11.003 M. In comparison to the model before improvement, the GA-YOLO model has improved mAP by 3.69% and detection speed by 20.245 FPS. Additionally, the GA-YOLO model has reduced parameters by 82.79%. GA-YOLO model not only improves the detection accuracy of dense and occluded targets but also lessens model parameters and accelerates detection speed. [ABSTRACT FROM AUTHOR]

As a renewable clean energy with good application prospects, wind energy has attracted more and more attention. Harvesting energy from the ambient environment with triboelectric nanogenerators (TENGs), for use in electronic devices has become a feasible method. Nevertheless, the alternating‐current signal generated by TENGs needs to be converted into direct‐current signal to be effective in applications. Therefore, this article presents a power management circuit with a self‐triggering triboelectric‐electromagnetic hybrid generator (ST‐TEHG). The switch module of the power management circuit is composed of an electromagnetic (EMG), 555 single timer, and an N‐type metal–oxide–semiconductor, the module converts the AC signal output by EMG into a square wave signal under monostable state. The self‐triggering switching function is realized when in monostable operation. Systematic experiments demonstrate that the 555 single timer function can output waveforms of monostable operation. The back‐end of the circuit can output DC signals stably, the ripple voltage is 0.03 V, and the ripple coefficient is 0.52%. ST‐TEHG provides a new way of thinking for the power supply of microelectronic components. [ABSTRACT FROM AUTHOR]

Thermal conditions are strongly changeable in a vehicle cabin, where passengers could suffer consecutive self‐thermoregulation to such dynamic changing thermal stresses, though its HVAC system works well. To observe human overall and local thermal sensations in dynamic thermal conditions, a series of experiments under various conditions were carried out in a cabin‐like climate chamber. The results showed that the head, chest, back, and hands during hot exposure are warmer leading to the overall thermal sensation being hot. The thermal sensation of the head was warmer than the overall thermal sensation. During cold exposure, arms, hands, legs, and feet were the main areas causing coldness. In a dynamic thermal environment, the previous skin temperature state and thermal sensation form a thermal sensation overshoot, causing a shift in the body's neutral temperature point. This study proposes a thermal sensation model for the prediction of human thermal sensation local and overall based on skin temperature changes in a dynamic environment. Considering the airflow characteristics in the cabin, the human body is set into seven local parts in the local thermal sensation model. To compensate for sensation overshoot from this, defining recovery points rp for local parts differentiate temperature setpoints according to the experienced thermal state so that the effect resulting from the dynamic condition is integrated into the model algorithm. The model provides a scientific basis for guiding design optimization and intelligent regulation in the dynamic environment of the vehicle cabin, so as to achieve efficient energy utilization. [ABSTRACT FROM AUTHOR]

Human-driven vehicles and different levels of autonomous vehicles are expected to coexist on roads in the future. However, autonomous systems behave differently from their human-driver counterparts, these two behaviors are incompatible with one another, thereby negatively impacting traffic efficiency and safety. Herein, we present the construction of human-like trajectories for use in autonomous vehicles as a possible solution to this issue. We present a trajectory planning method based on the spatiotemporal geometric transformation of driving scenarios to generate human-like trajectories. Speed and safety redundancy data were collected through driving tests to understand human driving behaviors. Self-driving scenarios were abstracted as a Lorentz coordinate system under a three-dimensional Minkowski space-time. A surrounding-manifold tensor equation was established using differential geometry theory to depict the relationship between the trajectory constraints and the geometric spatiotemporal background. A metric tensor field can be solved from the equation to construct the corresponding “volcano space-time,” which is a three-dimensional general Riemannian space for placing the subject vehicle and the surroundings. The geodesics of the volcano space-time are solved using the geodesic equation and are projected back into the three-dimensional Minkowski space-time. Geodesic trajectories were fitted as Bézier curves in this study and corrected according to the vehicle dynamics constraints for trackability. In simulation and real vehicle tests, trajectories generated using the proposed algorithm exhibited collision avoidance and trackability, and the algorithm offered behaviors that were similar to those of human drivers under the same scenarios. [ABSTRACT FROM AUTHOR]

An adaptive second-order sliding mode control method based on RBF neural network is proposed for n-DOF robotic manipulators in the presence of external disturbances. First, RBF neural network is used to approximate the model information. Second, by using adaptive technology to compensate the uncertainty, whose prior knowledge about upper bound is not required. In addition, since the proposed control scheme is continuous, the chattering phenomenon is almost completely eliminated. Finally, the stability and finite time convergence of the proposed method are proved by Lyapunov stability theory. Through the simulation of 2-DOF manipulator and 5-DOF manipulator, the effectiveness and superiority of the control scheme are verified. [ABSTRACT FROM AUTHOR]

Deep learning has been widely used in removing image noise, but the blind denoising problem in agriculture limits the accuracy of citrus recognition by picking robots, and most denoising methods do not preserve the texture details of the image well. To solve these problems, this study proposes a multi‐module image denoising method based on feature pyramid network. The algorithm is based on the pyramid network structure, adding modules such as attention mechanism, dilated convolution, and feature fusion to denoise images. The attention mechanism and dilated convolution enhance the extraction of noisy features. The feature fusion module fuses the feature maps of each layer to solve the problem of incomplete image detail texture preservation. The experimental results of this algorithm on different datasets show that the performance of our method on various datasets with different noise levels is superior, and it is solving the problem of real image denoising. It is effective to perform both on and save the image texture details, which proves the practicability of the method in this study. After using this method to remove the real noise in the image, the average score of citrus detection accuracy is 52.4% higher than that before denoising. Practical Applications: Citrus picking is a labor‐intensive work. Traditionally, citrus picking depends on manpower. This harvesting method has high cost and low efficiency, which seriously hinders the development of citrus industry. In this study, the pyramid network structure is used to denoise the citrus image by adding modules such as attention mechanism, dilated convolution, and feature fusion, so as to guide the robot to recognize the citrus and improve the production efficiency. It is of great value to the recognition technology of industrial citrus picking robot. [ABSTRACT FROM AUTHOR]

In view of the disadvantages of the existing pose estimation algorithm, which has low real-time performance and the positioning accuracy will be greatly reduced in dynamic scene, a compound deep learning and parallel computing algorithm (DP-PE) is proposed. The detection algorithm based on deep learning is used to detect dynamic objects in the environment, and the dynamic feature points are removed before the matching of feature points to reduce the impact of dynamic objects on the positioning accuracy; A method for distinguishing "pseudo-dynamic objects" is proposed to solve the problem that the stationary vehicles and pedestrians in the environment are regarded as dynamic objects. The parallel computing framework for feature point extraction and matching is established on CPU-GPU heterogeneous platform to speed up DP-PE; In the localization part of DP-PE, we propose a 3D interior point detection strategy to achieve parallel search of map points, and the saturated linear kernel function is used to act on reprojection error to realize the parallelization of pose optimization. We verify the algorithm on KITTI dataset, the experimental results show that average speedup ratio of feature point extraction and matching is 6.5 times, and the overall computational efficiency of DP-PE is about 7 times higher than that before acceleration, which can realize high precision and efficient pose estimation in dynamic scene. [ABSTRACT FROM AUTHOR]

The optimal initial pre-conditioning parameter is essential to properly adjust the temperature within the cabin in an effective and accurate way, especially while passengers' thermal comfort and energy-saving properties are both considered. Under the various environmental thermal loads, the pre-conditioning solutions resulting from those pre-fixed cooling parameters are unfeasible for achieving accurately passengers' comfort temperature. In addition, it is also difficult in such a narrow car space to identify a lot of local attributes due to the different material properties and sizes of a variety of structural parts that have various thermal responses to environmental conditions. This paper presents a data-driven decision model to numerically identify the degrees of the cabin thermal characteristic to determine satisfactory pre-conditioning parameter schemes. Initially, based on the thermal data within a vehicle recorded through the whole year at a selected hot climate region of the Middle East, the study levels multiple climate scenes corresponding to change in the cabin air temperature. Then three classification algorithms (Support Vector Machines, Decision Tree, and K-nearest neighbor model) are used to comparatively identify climate levels according to the input conditions. Based on the identified climate level, an appropriate parameters scheme for this level is applied. A comprehensive evaluation index (CEI) is proposed to characterize the passengers' satisfaction in numerical computation, on considering multi-satisfaction objectives including Predicted Mean Vote (PMV), local temperature, air quality, and energy efficiency; and it formulates the pre-conditioning parameter scheme for each climate scene with CEI. Several scene cases are carried out to verify the effectiveness of the proposed models. The result shows that the pre-conditioning schemes of the model can effectively satisfy passengers in multi-satisfaction objectives. [ABSTRACT FROM AUTHOR]

Maize is a crop that is widely cultivated all over the world. Thus, the classification of maize seeds quality is important, while the traditional methods based on the texture, shape, and color which require repeated work is not efficient. Recently, deep learning reached the goal in the field of image processing, and a deep convolutional neural network (DCNN) is often used to do the image classification task. Here, we explored another neural network called Vision Transformer (ViT), which originally was applied to the natural language processing. Based on the self‐attention mechanism, ViT discards the convolutional structure. But when trained from scratch on medium‐sized datasets, ViT performed poorly compared to CNN. Due to the lack of local structure within the input image, tokenization cannot be used to generate a valid training set in the original ViT model. As a result, we proposed an improved ViT model SeedViT. Compared with the original ViT which could only train large datasets, SeedViT can train small and medium datasets to achieve SOTA (State of the Art) in vision classification with only 2,500 images in our study. The feasibility of SeedViT to classify maize seeds' quality was studied in this article, and we compared it with DCNN and traditional machine learning algorithms. The accuracy, sensitivity, specificity, and precision were 97.6%, 94.1%, 98.9%, and 97%, respectively. In addition, we employed ViT and VGG (Visual Geometry Group, a convolutional neural network) to extract image features, and SVM (support vector machine) was used as the classifier to classify them, with the result that ViT‐SVM was stable around 96.6% on the test set and VGG‐SVM was stable around 94.6%. At last, a visual attention map was generated by visualization technology. It showed that SeedViT can be a new and novel way for maize seed manufacturing. Practical applications: An algorithm for classifying and sorting out high‐quality maize seeds.Use the Transformer algorithm from the field of natural language processing instead of the convolutional neural network algorithm.Use GeLU function for activation and soft split with images to improve Vision Transformer model so that it can achieve great performance with small and medium datasets. [ABSTRACT FROM AUTHOR]

With the development of deep learning theory, the application of Yolov3 in fruit detection has been widely studied. Aiming at the problem that Yolov3 loses information during network transmission and the semantic feature extraction of small targets is not rich, this article proposed an improved Yolov3 cherry tomato detection algorithm. Firstly, the proposed algorithm uses dual path network as a feature extraction network to extract richer small target semantic features. Second, four feature layers with different scales are established for multiscale prediction. Finally, the improved K‐means++ clustering algorithm is used to calculate the scale of anchor boxes. Experiments showed that the algorithm has a precision rate of 94.29%, a recall rate of 94.07%, and an F1 value of 94.18%. The F1 value is 1.54% higher than Faster R‐CNN and 3.45% higher than Yolov3. It takes 58 ms on average to recognize an image, which provides a theoretical basis for the fruit detection. Practical Applications: Fruit picking is a labor‐intensive task. Traditionally, fruit picking relies on manpower. This method of harvesting has high cost and low efficiency, which seriously hinders the development of the fruit industry. This research uses deep learning algorithms to detect and recognize cherry tomatoes, guide robots in picking, and improve production efficiency. It is of great value to the recognition technology of industrial‐scale fruit picking robots. [ABSTRACT FROM AUTHOR]

This paper focuses on the mobile robot path planning problem of optimizing the performance metrics of bidirectional A* algorithm in randomized two-dimensional map environments. An algorithm called direction constraints adaptive extended bidirectional A* (DCAE-BA*), which is an improvement of the traditional target dynamic bidirectional A* algorithm (TTD-BA*), is proposed to improve the performance metrics of the algorithm. Regarding the improvement, we propose the adaptive extension method and the direction-constrained optimal node extension method (DCONE). Simulation experiments were conducted for DCAE-BA*, TTD-BA* and traditional A* algorithm (A*) in a large number of random two-dimensional map environments. The simulation experimental scenarios consider four types of start and end point relative directions and three obstacle proportions to objectively and comprehensively evaluate the performance of the proposed algorithms. The results show that different scenarios have a significant impact on the algorithm performance metrics. Finally, the overall performance of the proposed algorithm is evaluated with a large number of experiments in "random" scenarios, and the results show that DCAE-BA* obtains significantly better search time for all three obstacle proportions, and better path length and number of expanded nodes for 10% and 25% obstacle proportions. The effectiveness of the proposed DCAE-BA* algorithm is demonstrated, which provides an essential reference for the path planning of mobile robots in a random 2D map environment. • DCAE-BA* is improved by adaptive extension and direction-constrained optimal node extension (DCONE). • The Start-End Dir and obstacle proportions have a significant impact on the algorithm metrics. • Overall, the DCAE-BA* has the highest computational efficiency for all three types of obstacle proportion. • DCAE-BA* optimizes the path length in the "diagonal" scenario most significantly. • As the proportion of obstacles increases, the metric optimization effect decreases. [ABSTRACT FROM AUTHOR]

The gas diffusion layer (GDL) is currently manufactured from randomly stacked carbon fibers, with each layer forming a different pore structure. This investigation attempts to know the effect of the random poral microstructures on the water transport capacity. The stochastic reconstruction approach to numerically establishes random fiber distributions in GDL and a three-dimensional two-phase volume of fluid model are developed and validated. By combining the two models the liquid water transporting inside the randomly different GDLs is numerically processed. Thus, those observed relations, such as the water coverage ratio (WCR) of the fiber layer (FL) plane and gas channel wall, saturation and pressure drop, are comparable in simulations. The results show that even if the porosity of each FL is the same, the uncertainty of liquid water movement due to the random structure can cause large state deviations with different GDLs, within the studied range. They can lead to significant differences in oxygen content within the same GDL volume and ultimately results in significantly different fuel cell performance. The mechanism of the effect of the randomness of GDL microstructure on water transport capacity is beneficial for further mining valid characteristic information of FLs and structure improvement. • The GDL pore structure was established using the stochastic reconstruction method. • A model that captures the behavior of liquid water through the pores was developed. • The effect of random stacking of fibers on liquid water transfer was investigated. • The effect of fiber layers' random arrangement on water transfer was investigated. [ABSTRACT FROM AUTHOR]

Liver injury is one of the most severe traffic trauma. The development of accurate liver finite element model is beneficial for improving the biofidelity and validity in crash simulations, aiming to analyze the injury in accidents. 12 human liver finite element models in the current research were constructed from high resolution CT data of a Chinese male 50th percentile human subject, including the main structures like left lobe, right lobe, capsule, parenchyma and falciform ligament. The simulations based on Nava et al. experiment were conducted to validate the models and make comparisons of modeling method accuracy. The results demonstrated that the larger deviation happened to the tetra models due to the stiffer algorithm compared to the hex models, which were more sensitive to element size. The existence of capsule had significant effects on the liver mechanical responses, reducing the liver tissue pressure. Shell elements were more suitable for modeling the capsule.