Computer-aided Engineering Model Research Articles

Fusion of Multiple Data Sources for Vehicle Crashworthiness Prediction Using CycleGAN and Temporal Convolutional Networks

Abstract Computer-aided engineering (CAE) models play a pivotal role in predicting crashworthiness of vehicle designs. While CAE models continue to advance in fidelity, an inherent discrepancy between CAE model predictions and the responses of physical tests remains inevitable. Machine learning (ML) models have shown promising potential in improving the prediction accuracy of CAE models. Nevertheless, the scarcity of vehicle crash data poses a significant challenge for the training of such models. This paper overcomes these challenges by fusing multiple data sources from two different types of vehicles. More specifically, the Cycle-Consistent Generative Adversarial Neural networks (CycleGAN) are first employed to translate features of time-series test data from one domain to another using cycle consistency loss. Such a translation allows for the generation of synthetic crash test data for the second vehicle type by leveraging existing tests from both vehicle types. In parallel, an initial temporal convolutional network (TCN) model is trained using CAE simulation data and physical test data of the first vehicle type. This pre-trained TCN model is then fine-tuned using three sources of data from the second vehicle type, namely the CAE data, test data, and the augmented virtual test data generated using CycleGAN. Through these data fusion, the crashworthiness prediction accuracy of the second vehicle type can be improved. The proposed method is demonstrated by a real-world case study with a small-size SUV and a medium-size SUV. Results show enhancements in the predictive performance of the medium-size SUV model.

Digital Twins to Predict Crack Propagation of Sustainable Engineering Materials under Different Loads

Computer-aided engineering (CAE) is an essential tool in a digital twin not only to verify and validate a virtual twin before it is transformed into a physical twin, but also to monitor the use of the physical twin for enhanced sustainability. This paper aims to develop a CAE model for a digital twin to predict the fatigue life of materials. Fatigue damage is represented by the size of a macro-crack that grows with a cluster of micro-cracks subjected to three different loads. The growth angle is related to the maximum circumferential tensile stress, and the growth rate is determined by the stress intensity factor (SIF) at the crack tip. The prediction model takes into consideration the main factors, including micro-cracks, crack closures, and initial configurations. Simulations are developed for the growth of macro-cracks with radially distributed micro-cracks and randomly distributed micro-cracks, and we find that (1) the macro-crack in the second case grows faster than that in the first case; (2) a pure shear load affects the macro-crack propagation more than a combined shear and tensile load or a tensional load; (3) the external stresses required to propagate are reduced when the inclination angle of the micro-crack is small and within (−25° < β < 25°); (4) micro-cracks affect the propagating path of the macro-crack and generally guide the direction of propagation. The developed model has been verified and validated experimentally for its effectiveness in predicting the fracture or fatigue damage of a structure.

Vehicle Crashworthiness Performance Prediction Through Fusion of Multiple Data Sources

Abstract This study aims to improve the prediction accuracy of the computer-aided engineering (CAE) model for crashworthiness performance evaluation at speeds beyond those defined by current regulations and public domain testing protocols. One way of achieving this is by integrating data from a few physical crash tests with the CAE data using machine learning models. In this study, two scenarios are investigated: (1) improving CAE model prediction accuracy using test data of a vehicle type that is the same as that of the CAE model; (2) improving CAE model prediction accuracy using test data from two different types of vehicles (e.g., two different sizes of SUVs). In the first scenario, a novel approach is proposed in the displacement domain (deceleration versus displacement) to enable data fusion to help recover the unmodeled physics in the CAE model. A nonlinear spring-mass model is used to simulate rigid-barrier vehicle frontal impact. A Gaussian process regression (GPR) model is then applied in conjunction with a Gaussian mixture model to capture the model bias of the nonlinear spring constant under a dynamic analysis scheme. In the second scenario, we propose a time-domain method (deceleration versus time) based on temporal convolutional network (TCN) and transfer learning. An initial TCN model is first trained by fusing CAE data with physical test data of the first vehicle type based on data augmentation. This data-augmented TCN model is then fine-tuned through transfer learning using CAE and test data of the second vehicle type. It leverages the domain-invariant representations of the two types of vehicles to enhance the CAE model prediction accuracy of the second vehicle type. Case studies are used to validate the proposed approaches and to demonstrate their efficacy in improving the prediction accuracy of the CAE models.

Expert Knowledge Transfer from CAE Models to CNN Models Using Enhanced Adversarial Domain Adaptation

It is a common belief that convolutional neural networks (CNN) are incapable of acquiring knowledge from domain experts for fault detection and diagnosis. To address the challenge, this paper proposes a knowledge-transfer scheme from computer-aided engineering (CAE) models to CNN models. Domain experts build the CAE models that emulate the faulty behavior of rotating machines by incorporating fault symptom and controlling the degree of fault severity. Fault data are hardly acquired from rotating machines in the field, while a sufficient number of fault data can be generated using the CAE models. Then, a domain adaption model is trained using synthetic data (i.e., normal and fault data) from the CAE models and real data (i.e., normal data only) from rotating machines. To evaluate the validity of the proposed method, a small-scale testbed is regarded as the target system that does not have any fault data. This study contributes to resolve the dearth of fault data from most safety-related engineering assets such as power plant steam turbines, wind turbines, and urban air mobility.

(Invited) Simulation of Battery Chemo-Mechanical Physics Incorporating Degradation

Computer-aided engineering models have accelerated the design cycle for large-format battery cells, packs, and systems. Building on that success, researchers are developing physics-based models to accelerate advancements in battery manufacturing, degradation/lifetime prediction, and materials research. This talk highlights achievements and challenges. The first simulates the reaction sequences of solid/electrolyte interfacial growth, coupling atomistic- to continuum-scale models to identify pathways to extend the calendar life of the silicon anode. The second describes electrochemical-mechanical degradation of 3D polycrystalline cathode particles to quantify degradation processes and identify optimal material architectures and fast-charge protocols. Third, artificial intelligence algorithms are coupled with electrochemical and degradation physics to identify electrochemical performance parameters, diagnose degradation, and predict lifetime.

On the Indispensability of Isogeometric Analysis in Topology Optimization for Smooth or Binary Designs

Recently, isogeometric analysis (IGA), which unifies the computer-aided design (CAD) model and the computer-aided engineering (CAE) model, has been adopted to develop the isogeometric topology optimization (ITO) framework. However, a critical study on the indispensability of IGA in topology optimization to take the place of the conventional finite element method (FEM) is still lacking. In the current work, two important problems are extensively discussed: (1) The lower numerical precision of the FEM resulting from the disunification between the CAD and CAE models damages the effectiveness of the topology optimization, which suggests the indispensability of IGA in the replacement of the FEM in optimization; (2) a material penalization model is required to ensure the generation of a full loading-transmission path during optimization in classic density-based methods, which causes a greater overestimation of structural stiffness and also suggests the necessity of an ersatz material model. The current paper describes a promising ITO method with point-wise design to gain smooth or binary symmetrical topologies, for which an extended density distribution function (DDF) was constructed to describe the structural topology. Two benchmarks of the stiffness-maximization problem and compliant mechanism are studied in the context of the above issues. Finally, several topologically optimized designs with symmetry are obtained using the ITO method.

Development of a CAE Modeling Technique for Heavy Duty Cargo Weight using a DFSS Methodology

<div class="section abstract"><div class="htmlview paragraph">Cargo box is one of the indispensable structures of a pickup truck which makes it capable of transporting heavy cargo weights. This heavy cargo weight plays an important role in durability performance of the box structure when subjected to road load inputs. Finite element representation for huge cargo weight is always challenging, especially in a linear model under dynamic proving ground road load durability analysis using a superposition approach. Any gap in virtual modeling technique can lead to absurd cargo box modes and hence durability results. With the existing computer aided engineering (CAE) approach, durability results could not correlate much with physical testing results. It was crucial to have the right and robust CAE modeling technique to represent the heavy cargo weight to provide the right torsional and cargo modes of the box structure and in turn good durability results.</div><div class="htmlview paragraph">As there are multiple variables owing to heavy cargo weight, Design for Six Sigma (DFSS) methodology was used to develop a robust CAE modeling technique. This project was defined with multiple key control factors considering geometrical and material properties of cargo along with the multiple ways to connect cargo ballast with the box floor bed. The manufacturing variability of box panels was considered as a noise factor. L18 Orthogonal Array matrix was chosen to optimize the variables based on factorial effects of signal to noise (S/N) ratio. Finally, the validation was done for the achieved optimum solutions using numerical tests.</div><div class="htmlview paragraph">This paper describes the DFSS project in detail, and emphasizes on how the DFSS approach could be very effective to predict the influential control factors and how the combination of these factors was utilized to find out the optimum solution. This paper also highlights how the optimized/proposed CAE modeling technique improves the accuracy of the cargo box modal and durability results.</div></div>

A methodology to improve simulation of multibody systems using estimation techniques

ABSTRACT This paper presents a methodology for state estimation and accuracy improvement of computer simulations of computer aided engineering (CAE) models based on prediction and correction state estimation techniques and sensing. The aim is to simulate the dynamic behaviour of a real system, which can be sensed, and obtain values of states that are not measurable due to economic or technical limitations. This methodology can be applied to both optimization of design processes and on-line control of complex systems. State estimation techniques are currently used only on mathematical models, where the relationships among system variables are expressed by means of mathematical language, making state observer implementation possible but leading to limitations in system modelling and knowledge. Favoured over mathematical models, multibody CAE models (created by means of computer-aided engineering software) have become the essential tool for complex system development, simulation, analysis, optimization and control, such as multibody systems; one of their main advantages is the ease and flexibility in creating and modifying them, allowing the faithful modelling of complex systems.

MOOSAS – A systematic solution for multiple objective building performance optimization in the early design stage

There is great potential for building performance optimization (BPO) in the early design stage, but there is still a lack of methods, algorithms, and tools to support the BPO process in this stage. Through a comprehensive review, this study identified three critical issues that affect the implementation of the BPO process in the early design stage: model integration, real-time performance analysis, and interactive optimization design. This study provides a systematic solution to these three critical issues. In terms of model integration, a feature-based and graph-based 3D building space recognition algorithm is proposed to automatically convert the computer-aided design (CAD model) into a computer-aided engineering model (CAE model). In terms of real-time performance analysis, a simplified physical method, an HPC-accelerated method, and an AI-based method are explored, and a real-time energy modeling module and a real-time daylighting analysis module are developed. In terms of the interactive optimization design, a preference-based multi-objective BPO design algorithm that can consider user preferences is proposed to make full use of the decision-making ability of humans and the computing power of machines and significantly improve the optimization efficiency and result satisfaction. Based on the systematic solution, a multi-objective BPO design software, MOOSAS, is developed. MOOSAS allows real-time performance feedback, dynamic parameter analysis, and interactive optimization, supporting the BPO process in the early design stage. The innovations of this study are as follows: first, this study proposes a systematic solution to the three critical issues of the BPO process, i.e., model integration, real-time performance analysis, and interactive optimization design; second, this study develops a multi-objective BPO design software (MOOSAS) for the early design stage.

α-Stirling hydrogen engines for concentrated solar power

α-Stirling engines are receiving more and more attention for applications of concentrated solar power in small power installations (15–30 kW). The design of these engines has not experienced in recent years the breakthrough needed to deliver close to the Carnot Cycle energy conversion efficiencies. The delivered efficiencies are limited to mid-to-high 20% in the typical installations “Dish Stirling”. Here we review the latest studies made on α-Stirling engines, unfortunately mostly based on theoretical models of limited reliability, but also including very few examples of Computer-Aided Engineering (CAE) modeling followed by prototyping and testing. Finally, we present in detail one CAE model of an α-Stirling engine delivering energy conversion efficiencies of 42% with hydrogen as working fluid and adopting one hot cylinder, one cold cylinder, and one regenerator, with the hot fluid temperature of 800 °C. This efficiency is much higher than current air microturbines, which may deliver efficiencies of only about 20% working at much lower temperatures.

A framework to automate the design of digitally-fabricated timber plate structures

The current study uses knowledge from digital architecture, computer science, engineering informatics, and structural engineering to formulate an algorithmic framework for integrated Computer-Aided Design (CAD) and Computer-Aided Engineering (CAE) of Integrally-Attached Timber Plate (IATP) structures. The algorithm is designed to take the CAD 3D geometry of an IATP structure as input and automates the construction and analysis of the corresponding CAE model using a macroscopic element, which is an alternative to continuum Finite Element (FE) models. Each component of the macro model is assigned a unique tag that is linked to the relevant geometric and structural parameters. The CAE model integrity is maintained through the use of the common data model (CDM) concept and object-oriented programming. The relevant algorithms are implemented in Rhinoceros 3D using RhinoCommon, a .NET software development kit. Once the CAE macro model is generated, it is introduced to the OpenSees computational platform for structural analysis. The algorithmic framework is demonstrated using two case structures: a prefabricated timber beam with standard geometry and a free-form timber plate arch. The results are verified with measurements from physical experiments and FE models, where the time needed to convert thousands of CAD assemblies to the corresponding CAE models for response simulation is considerably reduced.

Data Augmentation-Based Prediction of System Level Performance under Model and Parameter Uncertainties: Role of Designable Generative Adversarial Networks (DGAN)

Owing to uncertainty factors present in the system, computer-aided engineering (CAE) models suffer from limitations in terms of accuracy of test model representation. This paper proposes a new predictive model, termed designable generative adversarial network (DGAN), which applies the Inverse generator neural network to GAN, one of the methods employed for data augmentation. Statistical model-based validation and calibration technology, employed for improving the accuracy of a predictive model, is used to compare the prediction accuracy of the DGAN. Statistical model-based technology can construct a predictive model through calibration between actual test data and CAE data by considering uncertainty factors. However, the achievable improvement in prediction accuracy is limited, depending on the degree of approximation of the CAE model. DGAN can construct a predictive model through machine learning using only actual test data, improve the prediction accuracy of an actual test model, and present design variables that affect the response data, which is the output of the predictive model. The performance of the proposed prediction model was evaluated and verified, as a case study, through a numerical example and system level vehicle crash test model including parameter uncertainties.

Occupant’s lower extremities’ impact-based model calibration for vehicle under-belly blast

Computer-aided engineering (CAE) technology is an important tool in the research of vehicle under-belly blast. A high-quality CAE model is required to guarantee accurate prediction. This requires an efficient model calibration process with physical testing, but the traditional model calibration method based on trial and error and operators’ experience is usually very inefficient and arbitrary. In this paper, an automatic model calibration process based on the Error Assessment of Response Time Histories (EARTH) metric is applied. The EARTH metric evaluates the discrepancies between test and CAE by quantifying the phase error, magnitude error, and slope error between their time histories. Then, an automatic calibration process can be achieved by solving a multi-objective optimization problem for which the optimization object is minimizing discrepancies between testing and CAE and results in automatic calibration of key parameters of the blast simulation CAE model. A case of occupant’s lower extremities’ responses under vehicle under-belly blast load is studied as an example to demonstrate the feasibility of the applied method.

Isogeometric independent coefficients method for fast reanalysis of structural modifications

PurposeThis paper aims to provide designers/engineers, in engineering structural design and analysis, approaches to freely and accurately modify structures (geometric and/or material), and then quickly provide real-time capability to obtain the numerical solutions of the modified structures (designs).Design/methodology/approachThe authors propose an isogeometric independent coefficients (IGA-IC) method for a fast reanalysis of structures with geometric and material modifications. Firstly, the authors seamlessly integrate computer-aided design (CAD) and computer-aided engineering (CAE) by capitalizing upon isogeometric analysis (IGA). Hence, the authors can easily modify the structural geometry only by changing the control point positions without tedious transformations between CAE and CAD models; and modify material characters simply based on knots vectors. Besides, more accurate solutions can be obtained because of the high order degree of the spline functions that are used as interpolation functions. Secondly, the authors advance the proposed independent coefficients method within IGA for fast numerical simulation of the modified designs, thereby significantly reducing the enormous time spent in repeatedly numerical evaluations.FindingsThis proposed scheme is efficient and accurate for modifying the structural geometry by simply changing the control point positions, and material characters by knots vectors. The enormous time spent in repeated full numerical simulations for reanalysis is significantly reduced. Hence, enabling quickly modifying structural geometry and material, and analyzing the modified model for practicality in design stages.Originality/valueThe authors herein advance and propose the IGA-IC scheme. Where, it provides designers to fasten and simple designs and modify structures (both geometric and material). It then can quickly in real-time obtain numerical solutions of the modified structures. It is a powerful tool in practical engineering design and analysis process for local modification. While this method is an approximation method designed for local modifications, it generally cannot provide an exact numerical solution and its effectiveness for large modification deserves further study.

A Digital Shadow cloud-based application to enhance quality control in manufacturing

In Industry 4.0 era, rapid changes to the global landscape of manufacturing are transforming industrial plants in increasingly more complex digital systems. One of the most impactful innovations generated in this context is the “Digital Twin”, a digital copy of a physical asset, which is used to perform simulations, health predictions and life cycle management through the use of a synchronized data flow in the manufacturing plant. In this paper, an innovative approach is proposed in order to contribute to the current collection of applications of Digital Twin in manufacturing: a Digital Shadow cloud-based application to enhance quality control in the manufacturing process. In particular, the proposal comprises a Digital Shadow updated on high performance computing cloud infrastructure in order to recompute the performance prediction adopting a variation of the computer-aided engineering model shaped like the actual manufactured part. Thus, this methodology could make possible the qualification of even not compliant parts, and so shift the focus from the compliance to tolerance requirements to the compliance to usage requirements. The process is demonstrated adopting two examples: the structural assessment of the geometry of a shaft and the one of a simplified turbine blade. Moreover, the paper presents a discussion about the implications of the use of such a technology in the manufacturing context in terms of real-time implementation in a manufacturing line and lifecycle management.

A new calibration metric that considers statistical correlation: Marginal Probability and Correlation Residuals

Computer-aided engineering (CAE) models have been indispensable to virtual testing for designing and evaluating engineered systems to satisfy reliability requirements. However, it is not easy to fully characterize the variability in the model input variables due to limited resources. Statistical model calibration is thus of great importance as a strategy to improve the predictive capability of a CAE model. Optimization-based statistical model calibration is formulated as an unconstrained optimization problem that infers the unknown statistical parameters of input variables associated with a CAE model by maximizing statistical similarity between predicted and observed output responses. A calibration metric is defined as the objective function to be maximized that quantifies statistical similarity. One important challenge in formulating a calibration metric is how to properly consider the statistical correlation in output responses. Thus, this study proposes a new calibration metric: The Marginal Probability and Correlation Residual (MPCR). The foundational idea of the MPCR is to decompose a multivariate joint probability distribution into multiple marginal probability distributions, while considering the statistical correlation between output responses. The MPCR has favorable properties, such as normalization, boundedness, and marginalization. Two mathematical and two engineering examples are presented to demonstrate the effectiveness and potential benefits of the MPCR.

Review of statistical model calibration and validation—from the perspective of uncertainty structures

Computer-aided engineering (CAE) is now an essential instrument that aids in engineering decision-making. Statistical model calibration and validation has recently drawn great attention in the engineering community for its applications in practical CAE models. The objective of this paper is to review the state-of-the-art and trends in statistical model calibration and validation, based on the available extensive literature, from the perspective of uncertainty structures. After a brief discussion about uncertainties, this paper examines three problem categories—the forward problem, the inverse problem, and the validation problem—in the context of techniques and applications for statistical model calibration and validation.

Fast high fidelity CFD/CSM fluid structure interaction using RBF mesh morphing and modal superposition method

PurposeThis paper aims to present a fast and effective approach to tackle complex fluid structure interaction problems that are relevant for the aeronautical design.Design/methodology/approachHigh fidelity computer-aided engineering models (computational fluid dynamics [CFD] and computational structural mechanics) are coupled by embedding modal shapes into the CFD solver using RBF mesh morphing.FindingsThe theoretical framework is first explained and its use is then demonstrated with a review of applications including both steady and unsteady cases. Different flow and structural solvers are considered to showcase the portability of the concept.Practical implicationsThe method is flexible and can be used for the simulation of complex scenarios, including components vibrations induced by external devices, as in the case of flapping wings.Originality/valueThe computation mesh of the CFD model becomes parametric with respect to the modal shape and, so, capable to self-adapt to the loads exerted by the surrounding fluid both for steady and transient numerical studies.

Investigation and optimization for the dynamical behaviour of the vehicle structure

Conceptual modelling is known as a well suited alternative approach for Computer-Aided Engineering (CAE) analysis in automotive industry. In this paper, an improved conceptual modelling method in which beams and panels of the structure are modelled as simplified beam elements has been proposed. To explore the advantage of conceptual modeling in determining the resonant frequencies/mode-shapes, a case study for wheelhouse was performed. Firstly, an experimental test and advanced CAE analysis were carried out to measure the wheelhouse dynamic characteristics. The advanced CAE model was then validated by means of Modal Assurance Criterion and natural frequencies by associated experimental measurements. The results of wheelhouse concept model compared to the advanced CAE and experimental model in low frequency range, showed that the error percent of natural frequency is lower than 10% and the Modal Assurance Criterion is above 0.75 for the first four mode shapes of wheelhouse structure. Finally, the conceptual model is used as a baseline for optimization. The genetic algorithm was implemented to maximize the first natural frequency to 41.74 Hz. So The genetic algorithm successfully provided new possibility for optimization by attempting to influence the first mode shape by means of the cross section characteristics. Due to the accuracy and reliability of developed conceptual model, this modelling approach can be a crucial tool in CAE and vibration analysis of vehicle in the early design phase. The proposed method allows the designer to give the results of design changes very quickly by neglecting details. Therefore, for the analysis of the vehicle performance in NVH domain, the proposed method could be considered in the conceptual design phase.

A bi-directional interface for improved interaction with engineering models in virtual reality design reviews

Leveraging virtual reality (VR) technology to enhance engineering design reviews has been an area of significant interest for researchers since the advent of modern VR. The ability to interact meaningfully with 3D computer-aided engineering models in these VR design reviews is an important, though often neglected, capability due to the difficulty of performing data translation between native computer-aided design (CAD) data and VR compatible file formats. A bi-directional interface was developed between a VR design review environment and a commercial CAD package that streamlines the data translation process. Transmitting both geometric data and selected metadata from the CAD system enabled the development of enhanced model interaction tools in a VR design review application. User experiments were performed that compared the enhanced tools developed to a baseline toolset. Participants success using these toolsets was measured as they performed tasks related to design understanding and decision making, such as counting the number of gears in a gearbox or evaluating the feasibility of a proposed design change in a four-cylinder engine. The analysis of the data from these experiments showed a statistically significant improvement in participants ability to understand the geometry of the model correctly, confidently, and quickly, as well as in participants ability to correctly and confidently understand the implications of a proposed design change when using the Enhanced Toolset. We conclude that the bi-directional interface concept developed in this work can be extended to enable advanced interaction with a diversity of engineering data in VR.