Search Results

The human-robot interaction (HRI) is involved in a lift assistant system of manufacturing assembly line. The admittance model is applied to control the end effector motion by sensing intention from force of applied by a human operator. The variable admittance including virtual damping and virtual mass can improve the performance of the systems. But the tuning process of variable admittance is un-convenient and challenging part during the real test for designers, while the offline simulation is lack of learning process and interaction with human operator. In this paper, the Iterative learning algorithm is proposed to emulate the human learning process and facilitate the variable admittance control design. The relationship between manipulate force and object moving speed is demonstrated from simulation data. The effectiveness of the approach is verified by comparing the simulation results between two admittance control strategies.

This study evaluated the interactions between sheet steel, lubricant and measurement system under typical sheet forming conditions using a fixed draw bead simulator (DBS). Deep drawing quality mild steel substrates with bare (CR), electrogalvanized (EG) and hot dip galvanized (HDG) coatings were tested using a fixed DBS. Various lubricant conditions were targeted to evaluate the coefficient of friction (COF) of the substrate and lubricant combinations, with only rust preventative mill oil (dry-0 g/m2 and 1 g/m2), only forming pre-lube (dry-0 g/m2, 1 g/m2, and >6 g/m2), and a combination of two, where mixed lubrication cases, with incremental amounts of a pre-lube applied (0.5, 1.0, 1.5 and 2.0 g/m2) over an existing base of 1 g/m2 mill oil, were analyzed. The results showed some similarities as well as distinctive differences in the friction behavior between the bare material and the coatings.

During sheet metal forming, the friction and surface roughness change as the sheet slides, bends and stretches against the tools. This study assessed evolution of friction and surface roughness changes on aluminum sheet with two surface finish conditions, mill finish (MF) and electron discharge texture (EDT), in both the longitudinal and the transverse rolling directions of the sheet. The sheets were tested using a three pin Draw Bead Simulator (DBS). Surface roughness of the sheet evolved as a result of bending at the first shoulder, reverse bending at the middle pin, bending at the second shoulder and unbending at the exit. Stretching conditions and sheet-pin contact were also varied to see the impact on surface roughness. In general, the largest surface roughness change for the transverse direction was observed at the convex side of the exit shoulder pin and on the convex side of the first shoulder for the longitudinal direction.

Aluminum sheet is commercially available in three surface finishes, mill finish (MF), electric discharge texture (EDT), and dull finish (DF). This surface finish impacts the friction behavior during sheet metal forming. A study was done to compare ten commercially available sheet samples from several suppliers. The friction behavior was characterized in the longitudinal and transverse directions using a Draw Bead Simulator (DBS) test, resulting in a coefficient of friction (COF) value for each material. Characterization of the friction behavior in each direction provides useful data for formability analysis. To quantitatively characterize the surface finish, three-dimensional MicroTexture measurements were done with a WYKO NT8000 instrument. In general, the MF samples have the smoothest surface, with Sa values of 0.20-0.30 μm and the lowest COF values. The EDT samples have the roughest surface, with Sa values of 0.60-1.00 μm, and the highest COF values.

An iterative learning control (ILC) algorithm has been developed for a test cell electro-hydraulic, fully flexible valve actuation system to track valve lift profile under steady-state and transient operation. A dynamic model of the plant was obtained from experimental data to design and verify the ILC algorithm. The ILC is implemented in a prototype controller. The learned control input for two different lift profiles can be used for engine transient tests. Simulation and bench test are conducted to verify the effectiveness and robustness of this approach. The simple structure of the ILC in implementation and low cost in computation are other crucial factors to recommend the ILC. It does not totally depend on the system model during the design procedure. Therefore, it has relatively higher robustness to perturbation and modeling errors than other control methods for repetitive tasks.

A Machine Learning (ML) approach is presented to correlate in-cylinder images of early flame kernel development within a spark-ignited (SI) gasoline engine to early-, mid-, and late-stage flame propagation. The objective of this study was to train machine learning models to analyze the relevance of flame surface features on subsequent burn rates. Ultimately, an approach of this nature can be generalized to flame images from a variety of sources. The prediction of combustion phasing was formulated as a regression problem to train predictive models to supplement observations of early flame kernel growth. High-speed images were captured from an optically accessible SI engine for 357 cycles under pre-mixed operation. A subset of these images was used to train three models: a linear regression model, a deep Convolutional Neural Network (CNN) based on the InceptionV3 architecture and a CNN built with assisted learning on the VGG19 architecture.

This report describes a new driving simulator capability at General Motors (GM) Research and Development's (R&D) Vehicle Development Research (VDR) Laboratory and its application in an iterative HMI development process. The paper also provides an overview of three recent simulator usability tests supporting HMI development.

An engineering approach was developed to extract the failure plastic strain, thinning failure strain, and major in plane failure strain for finite element simulation applications. This approach takes into account the failure strain dependency on the element size when element deletion scheme is invoked in the simulation of material fracture. Both localized necking fracture and tensile shear fracture can be predicted when appropriate elements and material models are used in LS-DYNA simulations. This leads to a more accurate prediction of fracture and tearing in the finite element simulation of vehicle structure and crash loading conditions.

This paper outlines a process to assess the aerodynamic performance of different vehicle exterior surfaces. The initial section of the paper summarizes the details of white-light scanning process that maps entire vehicle to points in Cartesian co-ordinate system which is followed by the conversion of scanned points to theme surface. The concept of point-cloud modeling is employed to generate a smooth theme surface from scanned points. Theme surfaces thus developed are stitched to under-body/under-hood (UB/UH) parts of the base vehicle and the numerical simulations were carried out to understand the aerodynamic efficiency of the surfaces generated. Specifics of surface/volume mesh generated, boundary conditions imposed and numerical scheme employed are discussed in detail. Flow field over vehicle exterior is thoroughly analyzed. A comparison study highlighting the effect of front grilles in unblocked condition along with air-dam on flow field has been provided.

For decades, the industry standard for laboratory durability simulations has been based on reproducing quantified vehicle responses. That is, build a running vehicle, measure its responses over a variety of durability road surfaces and reproduce those responses in the laboratory for durability evaluation. To bring a vehicle to market quickly, the time between tightening the last bolt on a prototype test vehicle and starting the durability evaluation test must be minimized. A method to derive 4-Post simulator displacements without measuring or predicting vehicle responses is presented.

The imperative to get to the market faster with new and better products, has determined all automotive OEM to rethink their product development cycle, and, as a result, many hardware based processes were replaced and/or augmented with virtual, software based ones. However, the virtualization itself does not guaranties better and faster products. In the area of vehicle dynamics, we concentrate on improving the multi-body model development process, facilitating comprehensive virtual testing, and verifying the robustness of the design. The authors present a highly flexible and efficient environment that encourages, enforces, and facilitates model sharing, reusing of components, and parallelization of vehicle dynamics simulations, developed on top of an existing commercial off-the-shelf engineering software application.

Accurate material property data in both the elastic and plastic ranges of deformation is essential for accurate material representation in finite element simulations of vehicle systems. Variation of post formed material properties across a part are often of interest in different types of analyses, such as metal forming or fatigue life, for example. Depending on a part's shape it is not always possible to cut standard size tensile test specimens from all areas of interest across the part. Smaller size specimens with curved or tapered gage section may have to be used to promote strain localization and fracture at or near the gage center. This paper presents comparison of quasi-static tensile properties determined using two specimen gage section geometries, straight and tapered. Specifically, the following questions are addressed. How do the engineering strains computed from two-dimensional strain fields obtained by DIC compare to strains measured during standard tensile tests?

Dynamic modal frequency structural analysis incorporating ADAMS/Flex dynamic load prediction and structural modal stress can provide accurate dynamic stress history for fatigue analysis and synthesis. The amount of data input to finite element analysis is reduced significantly compared to traditional modal & direct transient finite element analysis techniques. Compared to traditional dynamic loads prediction, no additional simulation effort is required except for incorporating flexible body models of structural components into the ADAMS model. This structural analysis technique seamlessly comprehends the correct geometry and force boundary conditions together for long duration dynamic stress calculations. This technique also provided the solution for the deficiency of traditional quasi-static inertia relief method, which is particularly significant for structural system with either significant deformation or articulation.

In today's fast pace world of innovation and technology, lifelong learning has become a necessity for anyone working in industry. This is especially true for those in alternative propulsion or other such rapidly evolving fields. Universities and other learning institutions are delivering ever more Distance Learning certificates, degrees, and programs in an effort to re-tool the technical workerforce. To ensure Distance Learning programs are effective, successful, and advantageous, personal interaction and direct connection in some form between the students, industry, and academia is instrumental. Communication and interaction are vital to learning, whether it is Distance Learning or other more traditional methods, and personal interaction should not be overlooked for any of these learning delivery methods.

As part of the Federal Motor Vehicle Safety Standards, requirements for roof strength need to be met for all vehicles. On the other hand, automobile manufactures need to minimize vehicle mass for fuel economy and other objectives. It is important, therefore, for manufacturers to have a good understanding of the sources of variation in measured roof strength. An accurate estimation of such variation is important to achieving these objectives. This paper presents a method of using CAE simulation and vehicle tests to effectively estimate the range of variability in the roof crush tests. A number of vehicle and test variables which could potentially affect the measured roof strength were chosen, and their sensitivity was evaluated through CAE simulation. This knowledge of the sensitivity was then used to design a small number of vehicle tests, producing an estimation of the variation range in roof strength.

Due to the multitude of external design constraints, such as increasing fuel economy standards, and the increasing number of global vehicle programs, developers of automotive transmission controls have had to cope with increasing levels of system complexity while at the same time being forced by the marketplace to improve system quality, reduce development costs, and improve time to market. General Motors Powertrain (GMPT) chose to meet these challenges through General Motors Company's Road-to-Lab-to-Math (RLM) strategy, particularly the Math-based method of a virtual vehicle simulation environment called System Simulation. The use of System Simulation to develop transmission control algorithms has enabled GMPT to improve product quality and reduce development times and costs associated with the dependence on physical prototypes. Additionally, System Simulation has facilitated the reuse of GMPT controls development assets, improving overall controls development efficiency.

Performance testing and evaluation always plays an important role in the developmental process of a vehicle, which also applies to autonomous vehicles. The complex nature of an autonomous vehicle from architecture to functionality demands even more quality-and-quantity controlled testing and evaluation than ever before. Most of the existing testing methodologies are task-or-scenario based and can only support single or partial functional testing. These approaches may be helpful at the initial stage of autonomous vehicle development. However, as the integrated autonomous system gets mature, these approaches fall short of supporting comprehensive performance evaluation. This paper proposes a novel hierarchical and systematic testing and evaluation approach to bridge the above-mentioned gap.

In today’s race for improved fuel economy and lower emissions from gasoline engines, precise metering of delivered fuel is essential. Gasoline Direct Injection fuel systems provide the means for improved combustion efficiency through mixture preparation and better atomization. These improvements can be achieved from both increasing fuel pressure and using multiple injection events, which significantly reduce the required energizing time per injection, and in a number of cases, force the injector to operate at less than full stroke. When the injector operates in this condition, the influence of variation in injector dynamics account for a large percentage of the delivered fuel and require compensation to ensure accurate fuel delivery. Injector dynamics such as opening delay and closing time are influenced by operating conditions such as fuel pressure, energizing time, and temperature.

This work is to propose a new Iterative Learning Observer (ILO)-based strategy for State Of Charge (SOC) estimation. The ILO is able to estimate the SOC in real time while identifying modeling errors and/or disturbances at the same time. An Electrical-Circuit Model (ECM) is adopted to characterize the Lithium-ion battery behavior. The ILO is designed based on this ECM and the stability is proved. Several experiments are conducted and the collected data is used to extract ECM parameters. The effectiveness of the estimated SOCs via ILO is verified by the experimental results. This implies that the ILO-based SOC determination scheme is effective to identify the SOC in real time.

Purpose - This research aimed to investigate the process of lean manufacturing implementation in automotive industry in China in order to identify the critical success factors. Design/methodology/approach - A review of relevant literature is used to identify potential critical success factors for lean manufacturing implementations. The research had targeted lean-manufacturing management, practitioners, process users, and consultants working in automotive industry in China. Data were collected with an electronic survey which included 20 close ended questions, each measured by using five-point scale, Out of total 200 questionnaire distributed, 80 useable responses were received resulting in 40 % response rate. A judgmental sampling technique had been selected. Both descriptive and inferential statistics had been used to analyze this data.