Search Results

There is an increasing demand in automated manufacturability analysis of metal castings at the initial stages of their design. This paper presents a system developed for virtual manufacturability analysis of casting components. The system can be used by a casting designer to evaluate manufacturability of a part designed for various manufacture processes including casting, heat treatment, and machining. The system uses computational geometrics and geometric reasoning to extract manufacturing features and geometry characteristics from a part CAD model. It uses an expert system and a design database consisting of metal casting, heat treatment and machining process knowledge and rules to present manufacturability analysis results and advice to the designer. Application of the system is demonstrated for the manufacturability assessment of automotive cast aluminum components.

Recent advancements in simulation software and computational hardware make it realizable to simulate a full vehicle system comprised of multiple sub-models developed in different modeling languages. The so-called, co-simulation allows one to develop a control strategy and evaluate various aspects of a vehicle system, such as fuel efficiency and vehicle drivability, in a cost-effective manner. In order to study the feasibility of the synchronized parallel processing in co-simulation this paper presents two co-simulation frameworks for a complete vehicle system with multiple heterogeneous subsystem models. In the first approach, subsystem models are co-simulated in a serial configuration, and the same sub-models are co-simulated in a parallel configuration in the second approach.

Model-Based Development processes in the automotive industry typically use high-level modeling languages to build the reference models of embedded controllers. One can use formal verification tools to exhaustively verify these design models against their requirements, ensuring high quality models and a reduction in the cost and effort of functional testing. However, there is a gap, in terms of processes and tools, between the informal requirements and the formal specifications required by the verification tools. In this paper, we propose an approach that tries to bridge this gap by (i) identifying the verifiable requirements through a categorization process, (ii) providing a set of templates to easily express the verifiable requirements, and (iii) generating monitors that can be used as specifications in design verification tools. We demonstrate our approach using the Simulink Design Verifier tool for design verification of Simulink/Stateflow models.

In 2006, the Insurance Institute for Highway Safety (IIHS) released a new Low Speed Bumper Test Protocol for passenger cars1. The new test protocol included the development of a deformable barrier that the vehicle would impact at low speeds. IIHS positioned the new barrier to improve correlation to low speed collisions in the field, and also to assess the ability of the bumper system to protect the vehicle from damage. The bumper system must stay engaged to the barrier to protect other vehicle components from damage. The challenge is to identify the bumper system design features that minimize additional cost and mass to keep engagement to the barrier. The results of the Design for Six Sigma analysis identified the design features that increase the stiffness of the bumper system enable it to stay engaged to the barrier and reduce the deflection.

Delivering the appropriate material data for CAE analysis of plastic components is not as straight forward as it would seem to be. While a few of the properties typically used by resin manufacturers and material engineers to describe a plastic are useful to the analysis community (density, CLTE), most are not (flexural modulus, notched izod). In addition some properties such as yield stress are defined differently by the analysis community than by the materials community. Lastly, secondary operations such as painting or chrome plating significantly change the behavior of components with plastic substrates. The materials engineering community and the CAE analysis community must work together closely to develop the material data necessary to increase the capability of the analysis. This paper will examine case studies where these issues have required modifications to the material property data to increase the fidelity of the CAE analysis.

Decoratively plated plastic parts continue to be in high demand. One of the essential and challenging features of these finished goods is the adhesion between the metal plating and the plastic. As is the case with any bond between metals and plastics, combating the force from dissimilar thermal growth is an ongoing concern. When a plated plastic part is frozen and the plastic contracts, the failure mode for the plating manifests as a blister or “worm track”. On the other hand, when high heat causes plating failures from growth of the plastic, the problem is one of cracking in the plating. In this study, two methods are discussed that provide insight into the strength of the bond between the metal plating and the ABS and ABS+PC plastics. Peel testing is one means to evaluate the strength of the plating to plastic bond. Peel testing methodology and results are reported for both ABS and ABS+PC samples. A second means to evaluate the bond strength is through thermal cycle testing.

Engineering has continuously strived to improve the vehicle development process to achieve high quality designs and quick to launch products. The design process has to have the tools and capabilities to help ensure both quick to the market product and a flawless launch. To achieve high fidelity and robust design, mistakes and other quality issues must be addressed early in the engineering process. One way to detect problems early is to use the math based modeling and simulation techniques of the analysis group. The correlation of the actual vehicle performance to the predictive model is crucial to obtain. Without high correlation, the change management process begins to get complicated and costs start to increase exponentially. It is critical to reduce and eliminate the risk in a design up front before tooling begins to kick off. The push to help achieve a high rate of correlation has been initiated by engineering management, seeing this as an asset to the business.

Hybrid high voltage battery pack is not only heavy mass but also large in dimension. It interacts with the vehicle through the battery tray. Thus the battery tray is a critical element of the battery pack that interfaces between the battery and the vehicle, including the performances of safety/crash, NVH (modal), and durability. The tray is the largest and strongest structure in the battery pack holding the battery sections and other components including the battery disconnect unit (BDU) and other units that are not negligible in mass. This paper describes the mass optimization work done on one of the hybrid batteries using CAE simulation. This was a multidisciplinary optimization project, in which modal performance and fatigue damage were accessed through CAE analysis at both the battery pack level, and at the vehicle level.

This paper presents ‘Virtual Shaker Table’: a CAE method that enables random frequency structural response and random vibration fatigue analyses of a battery system. The Virtual Shaker Table method is a practical and systematic procedure that effectively assesses battery system vibration performance prior to final design, build and testing. A random structural frequency response analysis identifies the critical frequencies and modes at which the battery system is excited by random inputs. Fatigue life may be predicted after PSD stresses have been ascertained. This method enables frequency response analysis techniques to be applied quickly and accurately, thereby allowing assessment of multiple design alternatives. Virtual Shaker Table facilitates an elegant solution to some of the significant challenges inherent to complex battery system design and integration.

The paper will discuss the results of a study to develop lightweight steel proof-of-concept front lower control arm (FLCA) designs that are less expensive and achieve equivalent structural performance relative to a baseline forged aluminum FLCA assembly. A current production forged aluminum OEM sedan FLCA assembly was selected as an aggressive mass target based on competitive benchmarking of vehicles of its size. CAE structural optimization methods were used to determine the initial candidate sheet steel and forged designs. Two (2) sheet steel FLCA designs and one (1) forged steel FLCA design were selected and developed to meet specified performance criteria. An iterative optimization strategy was used to minimize the mass of each design while meeting the specified stiffness, durability, extreme load, and longitudinal buckling strength requirements.

A displacement-based CAE analysis is applied to liftgate chucking noise problems. A CAE simulation model of a small-size sport utility vehicle (SUV) is simulated with a set of realistic road loads as a time transient simulation. The model contains a trimmed vehicle, a liftgate and structural body-liftgate interface components such as the latch-striker wire, contact wedges and slam bumpers. Simulation design of experiments (DOE) is carried out with the model. As performance measures, the relative displacements at the contact points of the interface components are selected, since they are considered the direct cause of liftgate chucking. As design variables, body structure stiffness, liftgate stiffness, liftgate opening stiffness, stiffness characteristics of the interface components and additional liftgate mass are selected. Results of the simulation DOE is post-processed, and response surface models (RSM) are fit for the performance measures.

This paper deals with the development of a Hood Slam Process Automator (PA) to automate the pre-processing tasks of the virtual slam assessment with non-linear Nastran Transient Sol. 129 on all types of hoods. The slam analysis generally consumes a lot of analyst's time for building the slam models, typically six hours and is very tedious and has the potential for errors. The Hood Slam PA will automatically create and perform slam analysis pre-processing tasks within HyperMesh software such as creating latch striker interface, creating seals and bumpers with CBUSH1D elements, assigning transient slam speed to the hood and will finally generate the Nastran non-linear transient (Sol.129) hood slam analysis input files. The ready to run analysis input files will be submitted to the Nastran solver and the analysis results will then be post processed using HyperView software.

For many years, the computer aided math model has been the foundation for lowering cost and reducing time to market for many manufacturing industries. The automotive industry applies a variety of tools and methods to evaluate the expected vehicle performance to a forever expanding set of requirements. These mathematical predictions of performance are then repeated for both a set of design cycles and a multitude of vehicles in the product portfolio. This paper presents a CAE perspective of the unique problems of the large scale virtual vehicle engineering factory and a set of solutions. Different strategies to create the various complex math models required are explored. These strategies include using COTS FEA pre-processers, producing FEA models internal to the CAD tools, as well as custom built tools, macros and process automation tools.

For a system which involves a fluid medium contained inside a deformable structure, such as a liquid fuel system, a simulation which couples the structure and fluid may be required depending on the operating conditions and system performance metric of interest. Simulation methods for fluid / structure interaction (FSI) have been gradually developed by CAE engineers with the advent of increased computer power. A robust fuel system design requires carefully designed components that can withstand all loadings it may experience. This paper presents both LS-Dyna's Arbitrary Lagrange-Euler (ALE) and Abaqus' Coupled Eulerian-Lagrange (CEL) methods for predicting the structural performance of a fuel tank system and demonstrates that a fuel tank systems and their components can be numerically evaluated before the products release.

This paper will present the Systems Engineering perspective of HTC (High Throughput Compute) architecture for rapid meshing and assembly that resembles HPC (High Performance Compute) cloud architecture. The architectural framework addresses the modeling requirements in CAE (Computer Aided Engineering) and CFD (Computational Fluid Dynamics) domains. The architecture for building a rapid modeling framework is discussed along with a brief description of the procedure to create one such application. Meshing and preparing the math model for CAE/CFD analysis is a time and computation intensive process. Reducing the meshing and model preparation time increases the possible number of analysis iterations before the final design intent model is finalized. The HTC architecture helps in relieving the effort of redevelopment for changing business requirements and provides the flexibility to meet the evolving needs of engineering.

In the automotive industry, CAE methods are now widely used to predict several functional characteristics and to develop designs that are first-time-capable to meet programs targets. The N&V area is one of the increasing key factors for a product differentiation; costumers expect not only more powerful and more fuel efficient but also less noisy engines. The oil pan is one of the bigger contributors to engine radiated noise and to diesel knocking, so that great attention is paid within GM to optimize oil pans of Diesel engines by following a CAE-based approach to achieve a “first-time-capable” design for this component. This allows focusing the subsequent N&V testing activities to pinpoint modifications mainly on those components with shorter lead time. This paper describes the key-steps that are executed to optimize the oil pan design by using CAE methods with the main intent of reducing its noise radiation.

The sensor mounted on the door impact beam plays a major role in side impact events. The accelerations of side impact sensors are processed by sensing algorithms to make a decision on the air bag deployment. The sensing signal criterion for the deployable condition is a well understood process. However, the non-deployment sensing signal for the immunity to abuse conditions is a function of sensor attachment stiffness to the base structure. The base structure can be a door inner panel or door impact beam. In one of the production program, the acceleration based sensor attached to the impact beam showed immunity issues in the abusive door slams/opening to objects. Hence, the computer Aided Engineering (CAE) analysis was used to develop the sensor attachment criterion.

Developing a robust model that can simulate all real world conditions a vehicle can experience can be extremely difficult to predict. When working through the engineering process, Computer Aided Engineers (CAE) traditionally set modeling parameters and conditions to a nominal setting. This is done to simplify the models so that it avoided inputting too much tedious details into the system and wasting so much engineering time preparing the work. It was soon realized that this strategy did not capture all the possible conditions a hood on a vehicle could experience. There was a need to develop a formal approach and method to correlate an analysis model to real world conditions. The Design for Six Sigma (DFSS) process was utilized to develop robustness in the techniques used to accurately understand the vehicle environment. The DFSS process is normally used to design and develop robustness into physical parts.

This paper investigates changes in fuel economy of a mid-size sedan at various engine cooling fan power levels and front grille opening areas. A full vehicle model was built using MATLAB Simulink to calculate the fuel economy (MPG). The model utilized inputs from aerodynamic wind tunnel testing as well as FTP and MVEG dynamometer tests results. Simulation and testing was carried out at three front opening areas and three engine cooling fan power levels. The results provide a guideline for optimizing the front grille opening vs. engine cooling fan power combination at various driving conditions.

The FlexRay serial data protocol is being considered in automotive vehicle architectures as an enabler for active safety, time critical systems due to the advantages it provides for time-determinism, increased data bandwidth, and multiple data channels to support fault tolerance strategies. To improve the robustness/availability of the electrical/physical layer when used for these critical applications, a FlexRay Active Star device is available. The Active Star is part of the physical layer and facilitates the creation of robust, fault tolerant network systems by partitioning the network into individual branches connected to one or more ECUs. This partitioning allows fault confinements and isolation activities to be performed at individual branches with minimal disruption to network communication. This paper describes the investigation of Active Star capabilities and some complexities related to their network integration.