Search Results

Verification and validation (V&V) of autonomous vehicles (AVs) is a challenging task. AVs must be thoroughly tested, to ensure their safe functionality in complex traffic situations including rare but safety-relevant events. Furthermore, AVs must mitigate risks and hazards that result from functional insufficiencies, as described in the Safety of the Intended Functionality (SOTIF) standard. SOTIF analysis includes iterative identification of driving scenarios that are not only unsafe, but also unknown. However, identifying SOTIF’s unknown-unsafe scenarios is an open challenge. In this paper we proposed a systematic optimization-based approach for identification of unknown-unsafe scenarios. The proposed approach consists of three main steps including data collection, feature extraction and optimization towards unknown unsafe scenarios.

The performance of autonomous vehicle (AV) sensors, such as lidars or cameras, is often hindered during rain. Rain droplets on the AV sensors can cause beam attenuation and backscattering, which in turn causes inaccurate sensor readings and misjudgment by AV algorithms. Most AV systems are equipped with cleaning systems to remove contaminants, such as rain, from AV sensors. One such mechanism is to blow high-speed air over the AV sensors. However, the cleaning air can be hindered by incoming headwind, especially at higher vehicle speeds. An innovative idea proposed here is to use a visor to improve the cleaning performance of AV cleaning systems at higher vehicle speeds. The effectiveness of a baseline visor design was studied using computational fluid dynamics (CFD) air flow analysis and Lagrangian rain droplet tracking. The baseline visor improved the AV sensor cleaning performance in two ways. First, the visor protects the cleaning air flow from being disturbed by headwind.

As the mobility being developed becomes more complex and numerous, it is becoming difficult and inefficient to apply current vehicle-test-based development. To overcome this, research on combining test and simulation models has been actively conducted to perform objective and subjective evaluations more accurately and efficiently in the advance stage without a vehicle over the years. At first, test models for various systems such as tire, suspension and body were made compatible with simulation models by using various methodologies such as blocked forces, FBS decoupling, and Virtual Point Transformation (VPT). The second step was to objectively estimate road noise by using FBS coupling with system models and to deeply analyze transfer paths and system’s sensitivity. The results were verified by comparing with what was measured and analyzed on vehicle.

Simulation tools play a significant role in the automotive industry due to their cost-reducing capabilities in new model development. Computational Fluid Dynamics (CFD) is extensively utilized in various applications, such as vehicle aerodynamics and engine thermal management. However, its application in manufacturing engineering is not yet widespread. One crucial process in automotive manufacturing is the application of the base coat, which provides protection for the final paint layer. This process involves three key steps: bodywork immersion, electrophoretic deposition (E-coat), and bodywork removal from the bath. Each of these steps can be evaluated using appropriate CFD models. During the immersion step, the primary objective is to minimize the presence of trapped air. In the E-coat step, the focus is on controlling the paint layer thickness on the Body-in-white (BIW).

Automotive Heating Ventilation and Air Conditioning (HVAC) system is essential in providing the thermal comfort to the cabin occupants. The HVAC noise which is typically not the main noise source in IC engine vehicles, is considered to be one of the dominant sources inside the electric vehicle cabin. As air is delivered through ducts and registers into the cabin, it will create an air-rush/broadband noise and in addition to that, any sharp edges or gaps in flow path can generate monotone/tonal noise. Noise emanating from the HVAC system can be reduced by optimizing the airflow path using virtual tools during the development stage. This paper mainly focuses on predicting the noise from the HVAC ducts and registers. In this study, noise simulations were carried-out with ducts and registers. A Finite Volume Method (FVM) based 3-dimensional (3D) Computational Fluid Dynamics (CFD) solver was used for flow as well as acoustic simulations.

Increasing fuel and electricity prices create high pressure to develop efficient external aerodynamics of road cars. At the same time, development cycles are getting shorter to meet changing customer preferences while physical testing capacities remain limited, creating a pressing need for fast and accurate turbulence models to predict aerodynamic performance. This paper introduces and discusses different turbulence modelling approaches beyond the well-known and established models used today in the industry. The RANS Lag Elliptic Blending (Lag EB) k − ϵ model, which enables highly accurate steady-state RANS, was chosen as the baseline approach. As a medium fidelity approach Scale-Resolving Hybrid (SRH) model was utilized, which modifies a RANS base model to produce a smooth transition between URANS and LES behavior. The Wall-Modelled LES (WMLES) method was chosen for high fidelity simulations.

The vehicle industry being in the middle of transformation, the development of electric drives has come into engineers’ focus. The parameter evaluation of dynamic systems can be cumbersome when having nonlinearity in the structure, for example nonlinear stiffness characteristics. In such case, the standard linear approach, including EMA (Experimental Modal Analysis), modal superposition, FRF measurement (Frequency Response Function) and modal synthesis can not be applied. However, one of the main challenges in addressing nonlinearities is the lack of general tools to approach them. In this paper, a general framework to study nonlinearities in a structural dynamic context is presented. The method relies on standard random and sine sweep testing approaches to detect and localize nonlinearities, and on dedicated processing techniques to analyze the data and extract information on the nature of the analyzed nonlinearity.

The automotive industry is facing new emission regulations, changing customer preferences and technology disruptions. All have in common, that external aerodynamics plays a crucial role to achieve emission limits, reduce fuel consumption and extend electric driving range. Probably the most challenging components in terms of numerical aerodynamic drag prediction are the wheels. Their contribution to the overall pressure distribution is significant, and the flow topology around the wheels is extremely complicated. Furthermore, deltas between different rim designs can be very small, normally in the range of only a few drag counts. Therefore, highly accurate numerical methods are needed to predict rim rankings and deltas. This paper presents experimental results of four different production rim designs, mounted to a modified production car. An accurate representation of the loaded, deformed tire geometry is used in all calculations for comparable conditions between wind tunnel and CFD.

A model-based control approach is proposed to give proper reference for the feed-forward combustion control of Partially Pre-mixed Combustion (PPC) engines. The current study presents a simplified first principal model, which has been developed to provide a base estimation of the ignition properties. This model is used to describe the behavior of a single-cylinder heavy-duty diesel engine fueled with a mix of bio-butanol and n-heptane (80vol% bio-butanol and 20 vol% n-heptane). The model has been validated at 8 bar gross Indicated Mean Effective Pressure (gIMEP) in PPC mode. Inlet temperature and pressure have been varied to test the model capabilities. First the experiments were conducted to generate reference points with BH80 under PPC conditions. And then CFD simulations were conducted to give initial parameter set up, e.g. fuel distribution, zone dividing, for the multi-zone model.

Gasoline engine control strategies ensure a combustion control around stoichiometry. That is because the three-way catalytic converter allows CO and HC oxidation under lean operating conditions while ensuring NOx reduction for rich mixtures. In case of engine malfunction, the controller must adapt to compensate for potential torque loss and other critical attributes, potentially leading to significant deviation of the fuel-air mixture richness from stoichiometry and higher emission levels. Therefore, during development of the engine fault diagnostics, the impact on the pollutant emissions must be considered. In this paper, a model-based development process is proposed. It is based on system simulation modelling techniques, where a complete exhaust line is represented in order to predict tail-pipe emissions under stoichiometric, lean and rich conditions, for engine control design purposes. Two different modelling approaches are applied and evaluated in this paper.

The new trends in industry with for instance the electrification of systems leads to new challenges in the domains of acoustics and vibrations. The tonal noise of electric machines and their gearboxes, masked by the internal combustion engine in traditional vehicles, can now be significantly perceived by passengers, resulting in acoustic discomfort. Industry engineers now need to spend substantial effort in understanding the noise and vibration behavior of such components, with the main objectives of decreasing the weight for functional performance and reducing unpleasant sounds. Efficient vibro-acoustic simulation methods are then necessary to achieve these goals, starting from the modelling techniques up to the actual computation and solution post-processing. A complete workflow fully integrated within a single software platform is shown in this paper to achieve the simulation of a gearbox housing exterior radiation.

This paper presents a comprehensive decision-making algorithm for highway overtaking maneuvers, one of the highest risk maneuvers. For such, an overtaking scenario is divided into four phases: approach phase in which the host or overtaking vehicle (HV) detects a slow-moving lead vehicle (LV) in the same lane; left lane change and passing phase in which the HV performs a left lane change and passes the LV; right lane change phase in which the HV comes back to its original lane and free-flowing phase in which the HV maintains its lane and the initially set velocity. Depending on the phase-wise safety zones, the decision-making algorithm makes two decisions: change lane (1 = left lane change, 0 = maintain the same lane, -1 = right lane change) and adjust speed (1 = accelerate, 0 = maintain the current speed, -1 = decelerate). The proposed decision-making algorithm complements the human driver’s decision-making process and can be easily adapted for individual users.

For transmission engineers, numerical simulations open new possibilities to find the best design out of multiple design variations. For efficient usage of such methods, each design variation must be calculated in a timely manner. The overall performance of a transmission is evaluated throughout several rotations calculated on a 3D moving part. Such calculations are computationally expensive and faster modelling approaches are needed. Volume of fluid is one of the most common numerical approaches used for these configurations. In order to achieve numerical stability and accuracy, the mesh size must be small enough to capture the small scales at the interface. In this work a new mixture multiphase - large scale interface (MMP-LSI) approach was applied for which a large-scale interface individuation criterion is introduced along with the mixture multiphase approach.

Thermal management in electric vehicles (EVs) has attracted more attention due to its increasing significance, and computer aided engineering (CAE) plays an important role in its development. A 1D-3D online coupling approach is proposed to completely characterize transient thermal performance of an electric vehicle on a high performance computer (HPC) platform. The 1D thermal management model, consisting of air conditioning, motor cooling and battery cooling systems, is integrated with the 1D control strategy model and powertrain model consisting of motor, battery, driver and vehicle models. The 3D model is established for the air flow around the full vehicle and through its underhood. The 3D model gives boundaries such as heat exchanger air flowrates and heat flows on some component surfaces to the 1D model, while 1D gives back boundaries such as heat exchanger heat loads, component surface temperatures and fan speed simultaneously.

The use of sensors in advanced driver-assistance systems (ADAS) and autonomous vehicles has been accelerating over the past few years largely driven by regulatory and consumer interest in safety applications. These sensors help to prevent accidents and protect drivers by assisting with the monitoring, warning, braking, and steering tasks. As several unfortunate examples have highlighted these valuable systems can reduce safety if the sensors are not operating un-impaired. Planning for harsh weather environments is critical to the success of these systems. This study presents a fully automated workflow for an industrial side mirror geometry optimization for improved sensor performance under soiling conditions. The methodology includes CAD parametrization, multiphase simulation setup, intelligent design optimization and a detailed result analysis. All relevant aspects like external flow, geometrical fidelity and multiphase interaction are considered.

Water transport at high current densities is of main concern for polymer electrolyte membrane (PEM) fuel cells. The water content of the membrane must be high enough to provide maximum electrical conductivity and thus optimal stack performance. Dry-out may also lead to membrane degradation. However, a too high level of humidity leads to cell flooding, blocking the air and fuel flows to the catalyst sites and thus the reactions, resulting in a drop-in efficiency. Fuel cells water transport physics requires further investigation due to its complexity [1,2] and numerical modelling can improve the fundamental understanding of the phenomena. In this work, a 3D comprehensive model for fuel cells is presented. The PEM fuel cell is modelled in Siemens Simcenter STAR-CCM+ [3]. Anode and cathode GDL are modelled as porous media, with electrochemical reactions calculated in an infinitely thin catalyst layer.

Diesel spray mixture formation is investigated at target conditions using multiple diagnostics and laboratories. High-speed Particle Image Velocimetry (PIV) is used to measure the velocity field inside and outside the jet simultaneously with a new frame straddling synchronization scheme. The PIV measurements are carried out in the Engine Combustion Network Spray A target conditions, enabling direct comparisons with mixture fraction measurements previously performed in the same conditions, and forming a unique database at diesel conditions. A 1D spray model, based upon mass and momentum exchange between axial control volumes and near-Gaussian velocity and mixture fraction profiles is evaluated against the data.

Green zones are challenging problems for the thermal management systems of hybrid vehicles. This is because within the green zone the engine is turned off, and the only way to keep the aftertreatment system warm is lost. This means that there is a risk of leaving the green zone with a cold and ineffective aftertreatment system, resulting in high emissions. A thermal management strategy that heats the aftertreatment system prior to turning off the engine, in an optimal way, to reduce the NOx emissions when the engine is restarted, is developed. The strategy is also used to evaluate under what conditions pre-heating is a suitable strategy, by evaluating the performance in simulations using a model of a heavy-duty diesel powertrain and scenario designed for this purpose.

Multiple injection strategies are commonly used in conventional Diesel engines due to the flexibility for optimizing heat-release timing with a consequent improvement in fuel economy and engine-out emissions. This is also desirable in low-temperature combustion (LTC) engines since it offers the potential to reduce unburned hydrocarbon and CO emissions. To better utilize these benefits and find optimal calibrations of split injection strategies, it is imperative that the fundamental processes of multiple injection combustion are understood and computational fluid dynamics models accurately describe the flow dynamics and combustion characteristics between different injection events. To this end, this work is dedicated to the identification of suitable methodologies to predict the multiple injection combustion process.

Rubber is a non-linear viscoelastic material which properties depend upon several factors. In a tire two of these factors, namely the temperature and excitation frequency, are significantly influenced by the vehicle operating conditions. In the past years, applied research studied how rubber viscoelastic characteristics affect structural and frictional tire properties. The present study focuses on how these effects interact with the vehicle handling response. Based on state of the art theory of friction, structural properties of rubber and on experimental evidence, the dependency of key tire parameters on temperature and rolling speed is established. These results are then used in combination with a single-track vehicle model to assess their impact on key vehicle parameters; as an example, the understeer coefficient, yaw resonance peak / damping and maximum acceleration are studied.