Development and Validation of a Machine Learned Turbulence Model

A stand-alone machine learned turbulence model is developed and applied for the solution of steady and unsteady boundary layer equations, and issues and constraints associated with the model are investigated. The results demonstrate that an accurately trained machine learned model can provide grid convergent, smooth solutions, work in extrapolation mode, and converge to a correct solution from ill-posed flow conditions. The accuracy of the machine learned response surface depends on the choice of flow variables, and training approach to minimize the overlap in the datasets. For the former, grouping flow variables into a problem relevant parameter for input features is desirable. For the latter, incorporation of physics-based constraints during training is helpful. Data clustering is also identified to be a useful tool as it avoids skewness of the model towards a dominant flow feature.

Download Full-text

Machine Learning for Turbulence Modeling and Predictions

Volume 3: Computational Fluid Dynamics; Micro and Nano Fluid Dynamics ◽

10.1115/fedsm2020-20038 ◽

2020 ◽

Author(s):

S. Bhushan ◽

Greg W. Burgreen ◽

D. Martinez ◽

Wes Brewer

Keyword(s):

Machine Learning ◽

Turbulence Model ◽

Turbulence Modeling ◽

Training Data ◽

Relevant Parameter ◽

Arbitrary Initial Condition ◽

Integral Boundary ◽

Boundary Layer Equations ◽

Viable Approach ◽

Flow Variables

Abstract A stand-alone machine learned turbulence model is applied for the solution of integral boundary layer equations, and issues and constraints associated with the model are discussed. The results demonstrate that grouping flow variables into a problem relevant parameter for input during machine learning is desirable to improve accuracy of the model. Further, the accuracy of the model can be improved significantly by incorporation of physics-based constraints during training. Data driven machine learning training requires trial-and-error approach, shows oscillations in a posteriori predictions, and shows unphysical results when used with arbitrary initial condition, as the query is essentially extrapolations. Physics informed machine learning addresses the above limitations, and is identified to be a viable approach for development of machine learned turbulence model.

Download Full-text

Development and validation of the Global Competency Scale(GCS)

The Korean Journal of Industrial and Organizational Psychology ◽

10.24230/kjiop.v25i4.801-831 ◽

2012 ◽

Vol 25 (4) ◽

pp. 801-831

Author(s):

Sangchoong Roh ◽

Hongsik Jung ◽

Youngwon Suh

Keyword(s):

Training System ◽

Global Competency ◽

Human Resource Department ◽

Survey Results ◽

Systematic Selection ◽

Gcs Score ◽

The Right ◽

Development And Validation ◽

Criterion Variables ◽

And Training

As the world economy is becoming globalized, more domestic businesses are branching to overseas. Thereupon the number of expatriate workers who are getting assigned to overseas are increasing, and needs for systematic selection and training system for overseas expatriate workers are in dire needs. Nevertheless researches in this area are not enough and still inadequate level domestically. Therefore we developed the Global Competency Scale (GCS) with the purpose of the local businesses to use it to predict the possibility of successful overseas job performance and to select and train the right overseas expatriate workers. To develop the scale we conducted researches on documentations and interviews with former overseas expatriate workers and expatriate program managers in human resource department(HRD). Based on these results we developed 14 initial factors with 138 items. Using theses items we conducted both on & offline survey to people who work at global and multinational companies in Korea. With the 381 people's survey results, we implemented the cross validity. After cross validating we generated final 6 factors with 24 items. The GCS score we developed in this research shows that the degree of their goal achievement during past overseas experience and level of their satisfaction was significantly high in those criterion variables proving the criterion-related validity. Especially the GCS we developed in this research shows that after controlling the effect of English skills, still appear to have significant effect on criterion variables. Finally based on research results we discussed academical and operational implication and limitations for the further researches.

Download Full-text

Modeling of mass transfer in biofilms in oscillatory flow conditions using k-ɛ turbulence model

Water Science & Technology Water Supply ◽

10.2166/ws.2003.0104 ◽

2003 ◽

Vol 3 (1-2) ◽

pp. 201-207

Author(s):

H. Nagaoka ◽

T. Nakano ◽

D. Akimoto

Keyword(s):

Numerical Simulation ◽

Mass Transfer ◽

Turbulence Model ◽

Uptake Rate ◽

Oscillatory Flow ◽

Substrate Uptake ◽

Flow Conditions ◽

Mass Transfer Mechanism ◽

Substrate Uptake Rate ◽

Good Agreement

The objective of this research is to investigate mass transfer mechanism in biofilms under oscillatory flow conditions. Numerical simulation of turbulence near a biofilm was conducted using the low Reynold’s number k-ɛ turbulence model. Substrate transfer in biofilms under oscillatory flow conditions was assumed to be carried out by turbulent diffusion caused by fluid movement and substrate concentration profile in biofilm was calculated. An experiment was carried out to measure velocity profile near a biofilm under oscillatory flow conditions and the influence of the turbulence on substrate uptake rate by the biofilm was also measured. Measured turbulence was in good agreement with the calculated one and the influence of the turbulence on the substrate uptake rate was well explained by the simulation.

Download Full-text

Beyond Network Pruning: a Joint Search-and-Training Approach

Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence ◽

10.24963/ijcai.2020/358 ◽

2020 ◽

Author(s):

Xiaotong Lu ◽

Han Huang ◽

Weisheng Dong ◽

Xin Li ◽

Guangming Shi

Keyword(s):

Random Perturbation ◽

Search Space ◽

Fine Tuning ◽

Superior Performance ◽

Network Pruning ◽

Training Approach ◽

Target Network ◽

Tuning Strategy ◽

Coarse To Fine ◽

And Training

Network pruning has been proposed as a remedy for alleviating the over-parameterization problem of deep neural networks. However, its value has been recently challenged especially from the perspective of neural architecture search (NAS). We challenge the conventional wisdom of pruning-after-training by proposing a joint search-and-training approach that directly learns a compact network from the scratch. By treating pruning as a search strategy, we present two new insights in this paper: 1) it is possible to expand the search space of networking pruning by associating each filter with a learnable weight; 2) joint search-and-training can be conducted iteratively to maximize the learning efficiency. More specifically, we propose a coarse-to-fine tuning strategy to iteratively sample and update compact sub-network to approximate the target network. The weights associated with network filters will be accordingly updated by joint search-and-training to reflect learned knowledge in NAS space. Moreover, we introduce strategies of random perturbation (inspired by Monte Carlo) and flexible thresholding (inspired by Reinforcement Learning) to adjust the weight and size of each layer. Extensive experiments on ResNet and VGGNet demonstrate the superior performance of our proposed method on popular datasets including CIFAR10, CIFAR100 and ImageNet.

Download Full-text

Evidence-based training approach for higher education faculty: brief model of inclusion and training of the disabled

International Journal of Educational Management ◽

10.1108/ijem-04-2021-0150 ◽

2021 ◽

Vol ahead-of-print (ahead-of-print) ◽

Author(s):

Gulnaz Zahid

Keyword(s):

Higher Education ◽

Students With Disabilities ◽

Social Action ◽

Faculty Members ◽

Content Type ◽

Training Approach ◽

The Social ◽

Post Test ◽

Action Model ◽

And Training

PurposeThis interventional study aims to test the effectiveness of thek training approach for higher education faculty members to facilitate students with disabilities (SwD) to promote inclusion in higher education by operationalising approaches on the basis of the social action model. It presents an evidence-based training model created on recognised theories and strategies in the field of disability.Design/methodology/approachThe study follows a single-case pre/post-test intervention design in which data were analysed quantitatively, followed by a thematic analysis of participants' feedback and trainer's reflections. Training sessions were aligned to the social action model, the perspective of reasonable accommodations and introduction to technological support for teaching-learning and policy issues. Eighty faculty members from different schools of a multi-disciplinary Pakistani university participated in these sessions. Data from only 63 faculty members were available for analyses.FindingsTeacher Perceptions of Facilitating Students with Disabilities (TP-FSD) scale served as a pre- and post-test measure. The quantitative assessment revealed knowledge and attitudinal gains after brief trainings. However, when findings were interpreted considering effect sizes and supported by qualitative findings, moderate effectiveness level was evident. Effectiveness can be interpreted by the internal and external validity checks and findings of multiple assessments.Practical implicationsThis study can be replicated by adapting the training approach and by considering its strengths and shortcomings mentioned in detail in the discussion section.Originality/valueThe study tested the effectiveness of brief faculty training to support SwD in a multi-disciplinary university having faculty with varied education and training experience.

Download Full-text

Keuring en opleiding van rekenaarstelselontleders

SA Journal of Industrial Psychology ◽

10.4102/sajip.v0i0.317 ◽

1980 ◽

Author(s):

L. J. Fick ◽

I. Van W Raubenheimer

Keyword(s):

Computer Science ◽

Computer Systems ◽

Job Analysis ◽

Extensive Study ◽

Job Description ◽

Training Requirements ◽

Systems Analysts ◽

Development And Validation ◽

And Training ◽

Selection Of

Some of the results of an extensive study on selection and training of computer systems analysts are reported. Special attention is devoted to a job description and job analysis as a basis for identifying the critical attributes and training requirements involved. The development and validation of a battery for the selection of computer systems analysts and students of computer science are discussed.

Download Full-text

Integrated Research and Training in Enterprise Information Systems

Global Implications of Modern Enterprise Information Systems ◽

10.4018/978-1-60566-146-9.ch011 ◽

2009 ◽

pp. 195-208

Author(s):

Chen-Yang Cheng

Keyword(s):

Information Systems ◽

Integrated Approach ◽

Enterprise Information ◽

Enterprise Information Systems ◽

Integrated Research ◽

Training Approach ◽

Implementation Approach ◽

Training And Research ◽

And Control ◽

And Training

The success of implementing Enterprise Information System (EIS) depends on exploring and improving the EIS software, and EIS software training. However, the synthesis of the EIS implementation approach has not been investigated. In this chapter, the authors propose an integrated research and training approach for students and employees about enterprise information systems (EIS) that are encountered in an organization. Our integrated approach follows the different stages of a typical EIS project from inception to completion. These stages, as identified, are modeling, planning, simulation, transaction, integration, and control. This ensures that an employee who is trained by this plan has an acquaintance with the typical information systems in an organization. Further, for training and research purposes the authors developed prototype information systems that emulate the ones usually found in organizations. This ensures the EIS software logic is consistent with the business logic. This chapter also discuss some of the case studies conducted with the prototype systems.

Download Full-text

Unsteady flow in a viscous oil transporting centrifugal pump

Open Engineering ◽

10.2478/s13531-011-0013-x ◽

2011 ◽

Vol 1 (4) ◽

Cited By ~ 1

Author(s):

Wen-Guang Li

Keyword(s):

Unsteady Flow ◽

Centrifugal Pump ◽

Acoustic Resonance ◽

High Viscosity ◽

Broadband Noise ◽

Centrifugal Pumps ◽

Viscous Oil ◽

Flow Feature ◽

Flow Variables ◽

Ldv Measurement

AbstractAcoustic resonances are frequently fatal problems in centrifugal pump operations. Low pressure pulsation of fluid in the blade pass frequency is helpful to prevent from such problems. In addition, for a high quality centrifugal pump, a lower broadband noise level is also on demand. The acoustic resonance and broadband noise are associated with unsteadiness of flow in the pump. Even there exist extensive analyses of unsteady flow in centrifugal pumps by means of CFD so far, the effect of high viscosity of fluid pumped on the unsteadiness of flow feature remains unclear. Thus, the unsteady flow in an experimental centrifugal pump was exploited numerically when it transported the liquids with different viscosities. The velocity profiles at the impeller discharge were validated with the results of LDV measurement for water. The viscosity effect on the fluctuation of flow in the volute was clarified quantitatively. It was shown the increasing viscosity of fluid makes the fluctuation in flow variables less substantial and results into a less noticed tendency of separation of flow from the blade pressure side.

Download Full-text