Difference between revisions of "Projecthp-MSFEMs"

From Wiki
Jump to: navigation, search
Line 12: Line 12:
 
==== Project Members ====
 
==== Project Members ====
 
* [http://www.asc.tuwien.ac.at/~schoeberl  Univ.Prof. Dr. Joachim Schöberl]
 
* [http://www.asc.tuwien.ac.at/~schoeberl  Univ.Prof. Dr. Joachim Schöberl]
* [https://online.tugraz.at/tug_online/visitenkarte.show_vcard?pPersonenId=2B467FDE88603B6D&pPersonenGruppe=3  Univ.Prof. Dr. Oszkár Bíró]
 
 
* [http://www.asc.tuwien.ac.at/~schoeberl/wiki/index.php/Markus_Sch%C3%B6binger Dipl.-Ing. Markus Schöbinger (Ph.D. student)]
 
* [http://www.asc.tuwien.ac.at/~schoeberl/wiki/index.php/Markus_Sch%C3%B6binger Dipl.-Ing. Markus Schöbinger (Ph.D. student)]
 
* Valentin Hanser (Master student)
 
* Valentin Hanser (Master student)
 
==== Former Members ====
 
* Haik Jan Silm, Dipl.-Ing.
 
* Richard, Prüller BSc (Student)
 
 
==== Workshop: ====
 
 
* [http://www.asc.tuwien.ac.at/mshom2016 Advances in Multi-Scale Methods and Homogenization for Laminates and Windings in Magnetic Fields MSHOM 2016]
 
  
 
==== Open Positions ====
 
==== Open Positions ====
Line 29: Line 20:
 
==== Funding ====
 
==== Funding ====
 
Fonds zur Förderung der wissenschaftlichen Forschung [http://www.fwf.ac.at/en/ FWF] <br>   
 
Fonds zur Förderung der wissenschaftlichen Forschung [http://www.fwf.ac.at/en/ FWF] <br>   
Grant number: P 27028 <br>   
+
Grant number: P 31926 <br>   
Funding periode: from September 1, 2014 to April 30, 2018
+
Funding periode: from November 1, 2018 to October 31, 2021
  
 
==== Abstract ====
 
==== Abstract ====
  
The simulation of eddy currents in laminated iron cores by the finite element method (FEM) is of great interest in the design of electrical machines and transformers. The iron core is made of ferromagnetic grain oriented laminates. The material properties are anisotropic and exhibit a magnetic hysteresis. The scales vary from the meter range for the iron core to the thickness of single laminates (typically in the range of 0.2-0.3mm). Clearly, modeling each laminate individually is not a feasible solution. Many finite elements (FEs) have to be used in such a model leading to extremely large nonlinear systems of equations. That's why an accurate simulation of eddy currents and the iron losses in laminated ferromagnetic cores with reasonable computer resources is by far not solved satisfactorily. It is still one of the major challenges in computational electromagnetics.
+
The iron core of electrical devices is laminated to reduce the eddy current (EC) losses. The
 
+
geometric dimensions are extremely di�erent. The thickness of iron laminates is about 0.3mm
Laminated cores represent a periodic microstructure and therefore are well suited for FEM with homogenization. Simulations with FEM and homogenization show a boundary layer quite similar to that which occurs in corresponding brute force models of such cores with anisotropic material properties. An accurate approximation of the boundary layer is essential for an exact evaluation of the iron losses. However, many FE layers are required, which considerably increases the total number of FEs in the model. The periodic nature of the lamination is interrupted by step lap joints or ventilation ducts or disturbed  by skewing leading to complex geometries which are costly in the FE modeling on its own.
+
and separated by quite small air gaps. On the other hand, the overall dimensions of the core
 
+
are in the meter range consisting of up to several thousands of laminates. Finite element
An accurate approximation by the FEM with standard polynomials also in case of equations with rough coefficients, for instance materials with a microstructure, and problems with a boundary layer, requires extremely fine meshes.  
+
(FE) simulations are indispensable for the optimal design of the devices. However, modeling
Therefore, we will develop new multiscale finite element methods (MSFEM) to cope with the microstructure, where the standard polynomial basis is augmented by special functions incorporating a priori information into the ansatz space to avoid fine FE meshes. Then, the MSFEM will be combined with the harmonic balance method to reduce the computational costs furthermore. To provide a comprehensive solution for the present topic, approaches for the boundary layer and for the above geometrical difficulties will be designed and integrated into MSFEM. Hysteresis will be considered by a Preisach model. Fast adapted numerical integration methods, a very important issue for an efficient MSFEM, will be developed which do not affect the accuracy of the approximation. All approaches will be developed for the time and frequency domain and for both potential formulations, the magnetic and the current vector potential.
+
of each laminate by FEs leads to an extremely large nonlinear system of equations, well above
 
+
hundreds of millions, which cannot reasonably be solved with present computer capacities.
All new MSFEM approaches will be incorporated into the open source hp-FEM code Netgen/NGSolve. A benchmark to provide measured data and the supercomputer VSC to compute very expensive reference solutions will ensure an optimal development of the new MSFEM approaches. The aim is to create highly accurate numerical solutions consuming minimal computer resources to run on personal computers without any difficulty.
+
The development of multiscale �nite element methods (MSFEMs) for ECs in laminated iron
 +
has brought the simulation capabilities a major step forward. Nevertheless, MSFEMs su�er
 +
from various shortcomings and the computational costs are still too high. There is no error
 +
estimator for MSFEMs which is a big problem to trust in MSFEM solutions. MSFEMs
 +
in 3D are severely restricted to simple problems. These di�culties are major challenges in
 +
computational electromagnetics. Completely new methods will be developed to achieve the
 +
breakthrough of MSFEMs.
 +
Adaptive MSFEMs facilitating MSFEMs with di�erent potential formulations, higher order
 +
MSFEMs, harmonic balance MSFEMs etc. are absolutely unique and novel. E�cient and
 +
reliable equilibrated local error estimators with a computable constant and based on the theorem
 +
of Prager and Synge to allow both h- and p-re�nement will be developed.
 +
The simulation of one laminate often su�ces for electrical machines assuming common simpli
 +
�cations. To avoid expensive 3D FE simulations, completely new space splitting 2-D/1-D
 +
methods will be developed considering in particular the edge e�ect and Biot-Savart-�elds,
 +
leading to a major reduction of computational costs.
 +
The missing models for interfaces with large stray �elds are a very serious shortcoming of
 +
MSFEMs in 3D. Feasible solutions in 3D are of exceptional importance, because stray �elds
 +
are present in almost all electrical devices. Current MSFEM approaches vanish in air which
 +
represents a serious problem. Thus, radically new approaches have to be found.
 +
Novel nonlinear model order reduction (MOR) schemes have to be developed for di�erent
 +
MSFEMs to achieve tremendous savings in computational costs. MOR schemes exploiting
 +
the speci�c nature of MSFEMs of ECs will be provided to cope with large nonlinear systems
 +
of equations.
 +
The aims of the project are a strong reduction of the high computational costs of MSFEMs
 +
to run such simulations on personal computers without any di�culty. Unique adaptive MSFEMs
 +
will guarantee highly accurate and e�cient MSFEM solutions. Novel MSFEMs will
 +
solve the still existing shortcomings in 3D.
 +
The project will be carried out by the applicant and the doctoral candidate.
  
 
==== Software ====
 
==== Software ====
 
Finite element package [http://sourceforge.net/projects/ngsolve/ ngsolve] for electromagnetic problems.
 
Finite element package [http://sourceforge.net/projects/ngsolve/ ngsolve] for electromagnetic problems.

Revision as of 11:49, 12 November 2018

Project Manager

Institute for Analysis and Scientific Computing
Wiedner Hauptstrasse 8-10
A-1040 Vienna, Austria

Tel: +43 1 58801 10116
Email: karl.hollaus@tuwien.ac.at

Project Members

Open Positions

There are currently no vacant positions.

Funding

Fonds zur Förderung der wissenschaftlichen Forschung FWF
Grant number: P 31926
Funding periode: from November 1, 2018 to October 31, 2021

Abstract

The iron core of electrical devices is laminated to reduce the eddy current (EC) losses. The geometric dimensions are extremely di�erent. The thickness of iron laminates is about 0.3mm and separated by quite small air gaps. On the other hand, the overall dimensions of the core are in the meter range consisting of up to several thousands of laminates. Finite element (FE) simulations are indispensable for the optimal design of the devices. However, modeling of each laminate by FEs leads to an extremely large nonlinear system of equations, well above hundreds of millions, which cannot reasonably be solved with present computer capacities. The development of multiscale �nite element methods (MSFEMs) for ECs in laminated iron has brought the simulation capabilities a major step forward. Nevertheless, MSFEMs su�er from various shortcomings and the computational costs are still too high. There is no error estimator for MSFEMs which is a big problem to trust in MSFEM solutions. MSFEMs in 3D are severely restricted to simple problems. These di�culties are major challenges in computational electromagnetics. Completely new methods will be developed to achieve the breakthrough of MSFEMs. Adaptive MSFEMs facilitating MSFEMs with di�erent potential formulations, higher order MSFEMs, harmonic balance MSFEMs etc. are absolutely unique and novel. E�cient and reliable equilibrated local error estimators with a computable constant and based on the theorem of Prager and Synge to allow both h- and p-re�nement will be developed. The simulation of one laminate often su�ces for electrical machines assuming common simpli �cations. To avoid expensive 3D FE simulations, completely new space splitting 2-D/1-D methods will be developed considering in particular the edge e�ect and Biot-Savart-�elds, leading to a major reduction of computational costs. The missing models for interfaces with large stray �elds are a very serious shortcoming of MSFEMs in 3D. Feasible solutions in 3D are of exceptional importance, because stray �elds are present in almost all electrical devices. Current MSFEM approaches vanish in air which represents a serious problem. Thus, radically new approaches have to be found. Novel nonlinear model order reduction (MOR) schemes have to be developed for di�erent MSFEMs to achieve tremendous savings in computational costs. MOR schemes exploiting the speci�c nature of MSFEMs of ECs will be provided to cope with large nonlinear systems of equations. The aims of the project are a strong reduction of the high computational costs of MSFEMs to run such simulations on personal computers without any di�culty. Unique adaptive MSFEMs will guarantee highly accurate and e�cient MSFEM solutions. Novel MSFEMs will solve the still existing shortcomings in 3D. The project will be carried out by the applicant and the doctoral candidate.

Software

Finite element package ngsolve for electromagnetic problems.