Milestone-Proposal talk:Theoretical Foundation of Finite-Element-Method for Electromagnetics

From IEEE Milestones Wiki

Advocates and reviewers will post their comments below. In addition, any IEEE member can sign in with their ETHW login (different from IEEE Single Sign On) and comment on the milestone proposal's accuracy or completeness as a form of public review.

Advocates’ Checklist

  1. Is proposal for an achievement rather than for a person?
  2. Was proposed achievement a significant advance rather than an incremental improvement to an existing technology?
  3. Were there prior or contemporary achievements of a similar nature?
  4. Has the achievement truly led to a functioning, useful, or marketable technology?
  5. Is proposal adequately supported by significant references (minimum of five) such as patents, contemporary newspaper articles, journal articles, or citations to pages in scholarly books? At least one of the references from a peer-reviewed scholarly book or journal article. The full text of the material, not just the references, shall be present. If the supporting texts are copyright-encumbered and cannot be posted on the ETHW for intellectual property reasons, the proposers shall email a copy to the History Center so that it can be forwarded to the advocate. If the advocate does not consider the supporting references sufficient, the advocate may ask the proposer(s) for additional ones.
  6. Are the scholarly references sufficiently recent?
  7. Is proposed citation readable and understandable by the general public?
  8. Does the proposed plaque site fulfill the requirements?
  9. Is the proposal quality comparable to that of IEEE publications?
  10. Scientific and technical units correct? (e.g. km, mm, hertz, etc.) Are acronyms correct and properly upperercased or lowercased?
  11. Date formats correct as specified in Section 6 of Milestones Program Guidelines? https://ieeemilestones.ethw.org/Helpful_Hints_on_Citations,_Plaque_Locations

Reviewers’ Checklist

  1. Is suggested wording of the Plaque Citation accurate?
  2. Is evidence presented in the proposal of sufficient substance and accuracy to support the Plaque Citation?
  3. Does proposed milestone represent a significant technical achievement?
  4. Were there similar or competing achievements? If so, have the proposers adequately described these and their relationship to the achievement being proposed?

Original Citation Title and Text -- Administrator4 (talk) 16:06, 29 July 2022 (UTC)

Theoretical Foundation of Finite-Element-Method for Electromagnetics, 1979-1991

From 1979 to 1991, an electromagnetic-wave research group in the department of electrical engineering, National Taiwan University, laid the theoretical foundation for the Finite Element Method for general linear Electromagnetic field problems. They also found a straightforward derivation as a generalized Galerkin's method. It has been widely applied in academic researches and commercial software development, like the early versions of HFSS.


Review by Expert 1 (inserted by the Advocate) -- JaninA (talk) 10:34, 8 July 2022 (UTC)

EXPERT REVIEW No. 1

Proposal for an IEEE Milestone “Theoretical Foundation for Finite Element Method for Electromagnetics (No. 2021-12)”

Thank you for sending me this proposal, I have read it with interest. My immediate reaction is that I would support very strongly the idea of recognising the introduction of the finite element method modelling in electromagnetics as an IEEE Milestone. As someone who has been active in this field since early 70s, I have witnessed, and contributed to, many significant developments and fully agree that this was an important advancement in the area of electrical and electronic engineering.

Having expressed my general support, I need to share, however, my significant concerns regarding the way this proposal has been prepared. These relate to the following three aspects.

First, I am not sure why only the 'theoretical foundation' is proposed to be recognised. In fact the mathematical formulations already existed in the early 70s and it was the implementation, associated numerical schemes, actual algorithms and overall 'a package of developments' which made the finite element method in electromagnetics so successful and revolutionised the simulation and design methodologies. Thus I suggest that the title should be changed to reflect this, perhaps something along the lines of 'Application of finite element method to modelling and design methodologies in electromagnetics'

Secondly, the proposal appears to suggest that 'electromagnetics' is synonymous with solving a wave equation; the text itself gives the following areas: 'from radiation, scattering, waveguide, to high-frequency circuits'. But this completely ignores the significant advances in the low frequency electromagnetics, where the 'displacement current' can be neglected. It is in fact in the low frequency applications where the finite element modelling has made particular advances and almost completely dominated the methodologies, whereas in high frequency (wave equations) other techniques (such as FDTD) seem to be preferred nowadays. Thus in my opinion it is crucial that the milestone refers to finite element modelling across the full frequency spectrum, not just high frequency.

Finally - and rather importantly - the current proposal provides a list of publications in support of the application, almost all of which come from the authors in Asia. While there have indeed been many papers published around the world on these topics, the most important and original contributions originated in North America, the UK and other European countries. The current list completely ignores such publications and the names of the most seminal authors such as Trowbridge, Lowther, Webb, Carpenter, Coulomb, Rubinacci, Weiland - and many others - are never mentioned. Thus the historical picture is completely distorted.

To summarise, I would be very supportive of the idea of creating an IEEE Milestone related to the introduction of the finite element techniques into electromagnetics as I believe this was indeed a significant milestone. However, the proposal, as is currently written, does not correctly reflect the real achievements in the field, is too narrow (just high frequency applications) and fails to give credit where the credit is due.

Re: Review by Expert 1 (inserted by the Advocate) -- Ruey-Beei Wu (talk) 13:48, 2 August 2022 (UTC)

We thank the expert comments and revised accordingly to clarify our proposal. Our responses to the comments are given below:

Comment 1. I would support very strongly the idea of recognising the introduction of the finite element method modelling in electromagnetics as an IEEE Milestone. As someone who has been active in this field since early 70s, I have witnessed, and contributed to, many significant developments and fully agree that this was an important advancement in the area of electrical and electronic engineering.

Response: We thank the support of expert 1.


Comment 2. First, I am not sure why only the 'theoretical foundation' is proposed to be recognised. In fact the mathematical formulations already existed in the early 70s and it was the implementation, associated numerical schemes, actual algorithms and overall 'a package of developments' which made the finite element method in electromagnetics so successful and revolutionised the simulation and design methodologies. Thus I suggest that the title should be changed to reflect this, perhaps something along the lines of 'Application of finite element method to modelling and design methodologies in electromagnetics.'

Response: We agree that the mathematical formulations of FEM already existed in the early 70s, and the implementation details contribute very much to the success of the FEM in computational electromagnetics. However, just like the low-frequency EM field problems, researchers in every engineering domain need to apply their domain knowledge to derive and interpret equations and use the computed results for different purposes. Different disciplines have various concerns in their formulation in addition to the general mathematical equations given in, say, structural mechanics. For example, the nonlinear effect of magnetization for quasistatic magnetic field problems is often needed. However, such a nonlinear consideration is unnecessary, while the radiation condition must be considered when studying antenna radiation or computing the RCS of an F-35B fighter. We believe that there is much to explore theoretically to apply FEM successfully for the linear high-frequency EM field problems.

With this understanding, our previous title was not very precise, as indicated by the Expert Reviewer. Hence the title for the proposal is changed to "Theoretical Foundation of Applying Finite-Element Method to Linear High-Frequency Electromagnetic Wave Problems." To clarify the confusion, we have added some paragraphs to the section, "What is the historical significance of the work (its technological, scientific, or social importance)?".


Comment 3. Secondly, the proposal appears to suggest that 'electromagnetics' is synonymous with solving a wave equation; the text itself gives the following areas: 'from radiation, scattering, waveguide, to high-frequency circuits'. But this completely ignores the significant advances in the low frequency electromagnetics, where the 'displacement current' can be neglected. It is in fact in the low frequency applications where the finite element modelling has made particular advances and almost completely dominated the methodologies, whereas in high frequency (wave equations) other techniques (such as FDTD) seem to be preferred nowadays. Thus in my opinion it is crucial that the milestone refers to finite element modelling across the full frequency spectrum, not just high frequency.

Response: We agree that 'electromagnetics' should include low-frequency cases and is not correctly described in our work. We changed it to 'high-frequency electromagnetic waves' in this revised version. We claim only our contributions to applying FEM to high-frequency problems. The reason is that an IEEE milestone plaque is to be mounted at a site and recognizes only the contributions of people related to that place, according to the IEEE template in preparing the proposal. It is not designed to commemorate the achievements of all researchers worldwide on that topic. We also agree that the FEM for low-frequency problems is a huge success, which dominates many complicated applications. We thus believe that the development and implementation of the application of FEM for quasistatic field problems also deserves one or more IEEE milestones. Finally, we agree that there exist compelling alternatives nowadays for simulating high-frequency electromagnetic waves, such as the method of moments (MoM), the finite-difference time-domain (FDTD) method, and the finite integral technique (FIT). Each method has its niches and has been implemented as commercial/non-commercial software. However, we have to mention that, based on the studies on applying the FEM to high-frequency problems, the software HFSS (High-Frequency Structure Simulator) was the first commercially available software for solving electromagnetic radiation and scattering problems. With HFSS, researchers and designers can conduct "virtual prototyping" for the first time by running accurate S-parameter simulations of fully 3D microwave devices and high-frequency circuits. In the new version of our proposal, paragraphs to reflect the above points have also been added to the section, "What is the historical significance of the work (its technological, scientific, or social importance)?"


Comment 4. Finally - and rather importantly - the current proposal provides a list of publications in support of the application, almost all of which come from the authors in Asia. While there have indeed been many papers published around the world on these topics, the most important and original contributions originated in North America, the UK and other European countries. The current list completely ignores such publications and the names of the most seminal authors such as Trowbridge, Lowther, Webb, Carpenter, Coulomb, Rubinacci, Weiland - and many others – are never mentioned. Thus the historical picture is completely distorted.

Response: As explained in responses to Comment 3, we claim only our contributions to the application of FEM to high-frequency problems. Thus the list of supportive publications contains papers mainly by the FEM@NTU team. However, we agree that the pioneers mentioned in this Comment should be noted. We thus included a list of such publications in this new proposal. They are mentioned in the section, "What is the historical significance of the work (its technological, scientific, or social importance)?" The publications on quasistatic fields are collected for those from 1971 to 1997. Those pioneers' work on high-frequency fields earlier than 1984 is also included since our major theoretical studies were published in 1980 and 1984, respectively. We apologize for not being able to cover all the excellent work. We will be more than happy to include any important literature suggested by the Expert's comments on this revised proposal.


Comment 5. To summarise, I would be very supportive of the idea of creating an IEEE Milestone related to the introduction of the finite element techniques into electromagnetics as I believe this was indeed a significant milestone. However, the proposal, as is currently written, does not correctly reflect the real achievements in the field, is too narrow (just high frequency applications) and fails to give credit where the credit is due.

Response: As explained in responses to Comment 3, we do not intend to apply for an IEEE milestone for all researchers studying the FEM for all EM field problems. It is thus natural that our proposal cannot reflect all achievements in the field. However, in this new revision, we have added words to clarify major concerns pointed out by the experts and recognize the success of the FEM for quasistatic fields. A list of literature by authors suggested by the Expert Reviewer and published about the time of our work is compiled, though perhaps not complete enough.

Re: Review by Expert 1 (inserted by the Advocate) -- Ruey-Beei Wu (talk) 05:59, 15 August 2022 (UTC)

Replace this text with your reply

Review Expert 2 by Dr Eric W. Lucas (uploaded by Advocate) -- JaninA (talk) 04:34, 9 July 2022 (UTC)

Dr. Eric W. Lucas Alpha Omega Electromagnetics, LLC. Tel: 410-852-8996 www.alphaomegaem.com elucas@alphaomegaem.com

To: Janina Mazierska Re: IEEE Milestone proposal (No. 2021-12)

It is an honor to offer my full recommendation in affirmation of the IEEE Milestone proposal entitled, “Theoretical Foundation for Finite Element Method for Electromagnetics”. The proposed Plaque Citation is an accurate and concise statement of this important technical contribution.

The Finite Element Method (FEM) has historically been a primary simulation methodology for several areas of engineering science. However, it’s full impact in the area of electromagnetic (EM) simulation somewhat lagged other engineering disciplines. Part of the difficulty was due to the formulation complexity of coupling finite volumetric domains to regions with wave-based field solutions of otherwise infinite extent. Further formulation difficulties were encountered when the electromagnetic systems under consideration were non-self-adjoint, for example when the simulation domain involved certain types of complex tensorial media operators. The attempted generation of variational principles for such problems using strictly mathematical techniques was extremely complicated and far out of the reach of most engineering practitioners.

At the heart of this Milestone contribution is a systematic variational formulation procedure that builds upon a physically observable quantity referred to as “electromagnetic reaction”. The research group at the National Taiwan University was able to leverage and generalize the well- known concept of reaction reciprocity in order to build a straightforward and methodical variational formulation framework that enabled the application of the FEM to a host of important EM simulation applications. The “Theory of Variational Electromagnetics (VEM)” presented this systematic procedure and in great detail explained the necessary considerations for formulating a fundamental variational principle (FVP) that inherently included non-self-adjoint systems. I personally feel that this is one of the most underrated yet broadly significant theoretical contributions to the field of computational EM to date. The general theory was further consolidated and subsequently explained in the presentation of the “Variational Reaction Theory” (VRT). A further practical insight highlighted the concept of “partial variations” and showed how the FVP could be further reduced and related to the VRT via a “Partial Variational Principle” (PVP). The PVP facilitated a somewhat more direct discretization process that could be understood as a generalization of the well-known “Galerkin method.”

These important variational reaction concepts laid the foundation for a generalization of the FEM known as the “Transfinite Element Method.” This advanced formulation enabled the direct coupling of the volumetric FEM with RF transmission lines and waveguide ports. This breakthrough technology, augmented by adaptive mesh refinement methodologies, enabled for the first-time accurate S-Parameter simulations of fully 3D, industrially significant microwave devices and high-frequency circuits. This gave birth to the original “HFSS” commercial FEM software package, which became the forerunner to several industrially indispensable EM simulation tools that are now used by electrical engineers worldwide for the “virtual prototyping” of complex RF devices and related technologies. The consolidated contributions referenced in “Theoretical Foundation for Finite Element Method for Electromagnetics” have significantly contributed to advances in the field of computational EM. The EM simulation software technologies that followed have revolutionized an entire industry and enabled the development of products that one way or another affect every life. I wholeheartedly support the recognition of this historical milestone and congratulate all of the associated contributing researchers at the National Taiwan University.

Dr. Eric W. Lucas June 28, 2022

Re: Review Expert 2 by Dr Eric W. Lucas (uploaded by Advocate) -- Ruey-Beei Wu (talk) 13:29, 2 August 2022 (UTC)

We would like to express our deepest appreciation to the reviewer for the kind words and strong support. Thanks so much. Regards, Ruey-Beei

An updated proposal, 15 October 2022 -- JaninA (talk) 04:31, 5 November 2022 (UTC)

An updated proposal with a new title with a new title "Variational Reaction Theory for Finite Element Formulation, 1979-1991" was submitted on 15 October 2022 (to replace a first update done after the review by Expert 1).


SCHOOL OF ELECTRICAL AND COMPUTER ENGINEERING 29 October 2022 To: Dr. (Ms.) Janina Mazierska Re: IEEE Milestone proposal (No. 2021-12): Theoretical Foundation of Finite-Element-Method for Electromagnetics

Dear Dr. Mazierska:

It is my great pleasure to review and support the proposal. I would say the work of Prof. CHEN has profoundly influenced my work. I documented part of Prof. CHEN’s contribution in my book “Waves and Fields in Inhomogeneous Media”, especially in Chapter 5 where I made liberal references to the work coming from Prof. CHEN’s group. My book is widely referenced. I just checked on the stats on my book, the 1995 version (I first published my book in 1990). It has received almost 6,000 citations. My work, overall, has received about 45,000 citations, and my H index is now over 90.

I am not writing to brag about my achievements, but we live in an interconnected world where “butterfly effect” prevails. Even though I was not deeply involved in finite element method, but Prof. CHEN’s work inspired me as a young man even when I was studying at MIT. He was from a developing country in Asia, (at that time Taiwan), and I grew up in Malaysia until 1970’s. It was an inspiring experience to read about Prof. CHEN’s works. He was ahead of his time to think about a rigorous finite element formulation for electromagnetics problems, which are not self-adjoint for many applications. He correctly identified that one should define an auxiliary problem for this class of problems. He was also the first one to propose an extended operator to remove the need for an essential boundary condition for solving finite element method for electromagnetics. You can read more about this in the book I wrote on “Waves and Fields in Inhomogeneous Media”.

When I was young, I read that Taiwan aspired to be the electronics center of the world. That has actually happened in a number of fronts. TSMC (Taiwan Semiconductor Manufacturing Company) has emerged to be a leader in chip manufacturing, with a revenue of 57B USD. NVIDIA, founded by a Taiwanese engineer boasts a revenue of 27B. That’s what I mean by “butterfly effect”. Even though not directly responsible, the education system in Taiwan deeply affects many people around the world. Prof. CHEN was at the heart of this education system in National Taiwan University. He has trained many students in electromagnetics/computational electromagnetics who later become leaders in the Taiwanese education system, including RB Wu who was the head of ECE at National Taiwan University. Dr. Jin-Fa LEE is a by product of this educational system. LEE was influenced by Prof. CHEN and his teaching at National Taiwan University. Later, LEE was the prime mover and the force behind HFSS, the most popular software in electromagnetics simulation. I know both LEE and Zol Cendes very well. Cendes was the teacher of LEE for LEE’s PhD dissertation, and Cendes had the lofty vision to found ANSOFT, later acquired by ANSYS to become a leader in EM simulation software. ANSYS and ANSOFT greatly benefited from the high-frequency code that Jin-Fa LEE had written. It was the first robust code that solves electromagnetics problem broadband, and for complex geometries. It has become immensely popular among electrical engineers, including those in the EDA, electronics, and antenna design industry and many more. Electromagnetics is a fundamental component of electrical engineering, and its education has far reaching impact in this world. Now ANSYS thrives on the repertoire of codes that reply on capturing wave physics accurately including those for mechanical and aerospace engineering.

It is true that Zienkiewics brought finite element method to the engineering community. His work has been mainly impactful in the civil engineering community. But the civil engineering community is not so tightly coupled into wave physics as the electromagnetics community. Hence, the issue of non-self-adjoint-ness is not as important in the civil engineering as the electrical/electromagnetics engineering community. I would say that electromagnetics education in Taiwan has a deep influence in electrical/electronics engineering education there. It would be correct to say that the standard of electromagnetics education in Taiwan is high mainly due to the leadership of Prof. CHEN. And thus, they have raised a new generation of electrical engineers who serve well the electronics industry in Taiwan, making Taiwan into indeed an electronics powerhouse in the modern era. This indirectly contributes to the success of TSMC, the leading chip maker in the world. I am also familiar with the work of Thomas Weiland. Weiland’s claim to fame is his finite- integration technique in solving Maxwell’s equations, and his ability to commercialize his idea via his company, CST studio. Indeed, Weiland has an elegant way of solving Maxwell’s equations that is worth lauding. But Weiland hardly publishes and hence keeps his knowledge very much as a company secret. He has a smaller footprint as he has trained fewer students in the area of computational electromagnetics.

Also, I have colleagues in the U of Illinois who came from Taiwan, Prof. Shung-Wu LEE and Prof. Shun-Lien CHUANG. Prof. LEE was the chief architect of XPATCH, the leading software in defense electromagnetics. Prof. CHUANG, who graduated from National Taiwan University, had taken courses from Prof. CHEN. He later foray into microelectronics making enormous contributions to many optoelectronics works. His works also need foundational knowledge in electromagnetics.

All in all, Prof. CHEN has influenced many researchers, including those far from the shore of Taiwan. I support strongly this milestone proposal on Theoretical Foundation of Finite- Element-Method for Electromagnetics.

If you have any questions, please do not hesitate to contact me.

Sincerely, Prof. Weng Cho Chew Distinguished Professor of Electrical and Computer Engineering Purdue University Elmore Family School of Electrical and Computer Engineering Electrical Engineering Building 465 Northwestern Ave. West Lafayette, Indiana 47907-2035

Ps: I would suggest that you capitalize Prof. CHEN’s family name. The placement of the family name among the Chinese community has been a source of confusion for many. In written and spoken Chinese (or Japanese, Korean, or Vietnamese), the family name is usually put at the beginning. In many places, such as Singapore and Malaysia, the family name is still placed at the beginning. But when communicating with the West, many people would put their family name in the end. For the Japanese, this has been automated. All Japanese automatically put their family name after their given names, such as Akira KUROSAWA (黑泽明) in Western publications, but not people from China and overseas Chinese. For example, we often hear of LEE Kuan Yew where LEE is his family name, so do we when we hear MAO Ze Dong, XI Jin Ping. Important people will have their family names placed first before their given names. Moreover, the Japanese capitalize their family names to avoid this confusion, but not the Chinese. Hence, to avoid the confusion for historical record, it is better to refer to Professor CHEN Chun Hsiung with a capitalized family name. Better still, many international journals allow Chinese scientists to publish their names with Chinese characters. Then the identity is unambiguous. In this case, his name in Chinese would be 陳俊雄。Each Chinese character in Mandarin can have up to five tones. So when Chinese characters are transliterated into Western alphabets, there is a collapse of potentially 125 sounds to 1 sound. Hence, IEEE should record the names of Asian authors in Chinese characters, and likewise, for the names of other nationalities.

https://engineering.purdue.edu/wcchew/ https://en.wikipedia.org/wiki/Weng_Cho_Chew https://ece.illinois.edu/directory/profile/w-chew Phone: 765-494-5402 Email: wcchew@purdue.edu Office: Wang 3053

Review 2 by Expert 1 for the first modification of the proposal August 2022 -- JaninA (talk) 05:05, 5 November 2022 (UTC)

REVIEVER NO 1 21.8.2022

Dear Janina,

Revised Proposal for an IEEE Milestone “Theoretical Foundation of Applying Finite-Element Method to Linear High-Frequency Electromagnetic Wave Problems (No. 2021-12)”

Thank you for sending me the revised proposal. Regrettably, I am somewhat disappointed with the revisions. I was hoping for an enhancement of the original idea to recognise the introduction of the finite element techniques into electromagnetics as a significant milestone, covering the whole frequency spectrum. Instead, the prospers narrowed the scope to high frequency electromagnetic waves.

At the same time, I am wondering if my original understanding of the ‘milestone’ initiative was correct. I assumed this was to do with a particular technique (methodology) making significant impact in the area of electrical and electronic engineering, whereas perhaps this is more with recognising a contribution of an individual (or a local group). If the latter is the case, then I suggest that I may not be the right person to give a fair judgement about the proposal as I was not following at the time – and I am not very familiar with – the publications of Professor Chun Hsiung Chen and his group.

I would like to make the following observations. First, the proposers still insist on emphasising 'theoretical foundation', even if they acknowledge ‘that the mathematical formulations of FEM already existed in the early 70s’. Thus, to repeat my original comment: ‘it was the implementation, associated numerical schemes, actual algorithms and overall a package of developments which made the finite element method in electromagnetics so successful’. If the mathematics was already there, then what ‘theoretical foundations’ are to be recognised?

The proposers argue that ‘the software HFSS (High-Frequency Structure Simulator) was the first commercially available software for solving electromagnetic radiation and scattering problems’. This is arguably true, but according to available evidence the software was originally developed by Zoltan Cendes and his students. If the group of Professor Chun Hsiung Chen made a significant contribution to this development, this would need to be documented through appropriate publications and/or other ‘evidence’.

Finally, HFSS is one of the current software packages, although it is commonly accepted that CST (Computer Simulation Technology, now part of SIMULIA CST Studio Suite) is currently the leading software in the design and analysis of electromagnetic components at high frequency. CST was originally developed by Thomas Weiland and his contribution in the field of high frequency electromagnetic modelling and simulation cannot be overestimated. The proposal makes no reference at all to these developments, many of which took place between the 70s and the 90s.

To conclude, I regret but I am unable to support this proposal in its current form. I appreciate, however, that the contribution of Professor Chun Hsiung Chen and his group might deserve a recognition, but I suggest that a different reviewer should be approached who is more familiar with the work of that group.

Kind regards,

Comment of Expert 1 about the updated proposal dated 29.10.2022 -- JaninA (talk) 05:09, 5 November 2022 (UTC)

Dear Janina,

As I said from the beginning, I would be very supportive of the idea of creating an IEEE Milestone related to the introduction of the finite element techniques into electromagnetics. However, the proposal - although improved from its first draft - is very narrow and concerns a marginal development in this otherwise broad and important field. I really have nothing more to add and suggest that you approach a different reviewer, perhaps someone more familiar with the local group.

Best regards,

needs rephrasing? -- Amy Bix (talk) 15:59, 30 November 2022 (UTC)

I am not confident that the History Committee is currently ready to proceed with approving this, given these repeated concerns expressed by expert 1. Obviously, a concise plaque cannot and does not need to detail every step of global advances in a field. But if I am reading JaninA's comments correctly, the concern is that the current citation inaccurately portrays this one lab's contributions to the field by exaggerating its role in relationship to work elsewhere. I am not enough of an expert in this area to suggest ways of rephrasing to modify this citation to pay tribute to work done by this one research group while not overstating its role. But for me, there is enough of a concern raised here that I think this needs to be resolved by careful rewording before we can definitively move ahead.


New title is much better, yet this implies 12 years of development which saw the contributions of dozens of researchers. setting the level above which a contribution deserves to be considered will be very difficult and we will probably end with all sections having contributed. Where such a plaque will be placed then?

Stefano

No Apparent Changes Since December 2022 Discussion -- Bberg (talk) 13:51, 6 February 2023 (UTC)

This proposal was discussed during the December 2022 History Committee meeting, and it was sent back to the proposer since the citation was completely void of any explanation of the benefits of what was being recognized, including any benefit to humanity. However, if the citation has changed since that time, it does not yet appear to have been improved in that regard, and there was very broad agreement during that Decemer meeting that the citation was unsatisfactory. As such, without any real improvement made to the citation, I am surprised that this is coming up for discussion again.

Re: No Apparent Changes Since December 2022 Discussion -- Dmichelson (talk) 16:19, 9 February 2023 (UTC)

I agree with Brian's concerns. Milestones capture our imaginations by focussing on a time and place when our perceptions of what's possible change due to some technical advance or breakthrough. This proposal needs to identify that inflection point and build the proposal around that.

Response to Review 2 of Expert 1 in Nov. 2022: -- Ruey-Beei Wu (talk) 05:55, 27 February 2023 (UTC)

  • Reply to paragraphs 1 and 2

We agree with the comment: I assumed this was to do with a particular technique (methodology) making significant impact in the area of electrical and electronic engineering, whereas perhaps this is more with recognising a contribution of an individual (or a local group).

However, with the rapid advancement of science and technology, the fast accumulation of knowledge and invention of tools, and the fast growth of researchers in many nations, an important technique (methodology) making a significant impact often involves essential contributions from many researchers and research teams worldwide. The finite element method (FEM) may be just the case: its applications to high- and low-frequency electromagnetic (EM) problems were developed by different groups. Many have substantial impacts, and their contributions deserve recognition of IEEE milestones. Based on this thinking, we narrowed the scope to the general variational formulation for using the FEM to solve general linear electromagnetic field problems, especially for radiation, scattering, and propagation of electromagnetic waves in high frequency.

Meanwhile, the IEEE template for preparing the proposal for milestone requires the proposers to:

Describe briefly the intended site(s) of the milestone plaque(s). The intended site(s) must have a direct connection with the achievement (e.g. where developed, invented, tested, demonstrated, installed, or operated, etc.).

This statement means the milestone plaque(s) can be placed in different sites. Thus, a plaque among them at a specific place will inevitably recognize the contributions of a related individual or a local group. In our proposal, the plaque is installed at National Taiwan University (NTU) and is to acknowledge the work of the FEM@NTU group led by Professor Chun Hsiung Chen.

On the other hand, it is still possible to recognize high-impact techniques (methodology) developed by more than one researcher or group.

From the requirement mentioned above,

Describe briefly the intended site(s) of the milestone plaque(s)

we wonder if the IEEE can issue multiple milestone plaques for a common achievement to more than one researcher or group with reviewed essential contributions. If acceptable, every plaque citation can state that the cited contribution is part of the major milestone. Then, the citation in each plaque can continue with descriptions of the specific contribution of the local group. In our case, introducing the finite element method (FEM) into electromagnetics is a major milestone in electrical engineering, and our group’s work is an essential part of it. We thus propose to revise our plaque citation as:

Being part of introducing the finite element method (FEM) into electromagnetics, from 1979 to 1991, an NTUEE research group proposed and refined Variational Reaction Theory for non-self-adjoint linear boundary value problems. As the first formulation to simulate high frequency electromagnetic wave scattering and propagation phenomena, its variational nature laid the foundation for modern FEM’s to solve full-wave Maxwell equations. Subsequently, some widely adopted commercial software's reliability and accuracy are firmly established.

作為將有限元素法引入電波工程的一部分,1979至1991 台大電機系的電波研究團隊完成變分互作用系列理論,應用於線性非自伴邊界值問題,成為第一個可模擬高頻電磁波散射與傳播的變分原理。此法的變分性質為現代有限元素法求解 Maxwell 方程式奠定基礎,一些廣泛使用的商用有限元素軟體也因此確立了可靠性與準確性。

  • Reply to paragraph 3

Yes, in our early version of the proposal, we indeed wrote (in the section "What is the historical significance of the work?" ): that the mathematical formulations of FEM already existed in the early 70s

Since it confuses and to be more concise, we have completely rewritten this section and replaced related sentences by:

By the 1980s, the applications of finite element methods for static and quasistatic fields had been pretty mature, and the variational formulation was widely available for the Computational Electromagnetics community. For electrostatics/magnetostatics problems, the formulation can be simplified as the minimization of stored energy in the electric/magnetic field. The involved operator is Hermitian and the solution region is usually confined with Dirichlet and/or Neumann boundary conditions. However, the underlying physics for high-frequency electromagnetic wave radiations and scattering problems is non-Hermitian and with radiation conditions. The corresponding variational principle was lacking in the 1980s. Not until the VRT series of research publications did the corresponding variational formulation for 3D full-wave Maxwell equations become primarily elusive.

We believe this clarification, along with other new contents in the same section of the updated version, explained that what made the FEM successful in linear high-frequency electromagnetics is not only the improvement of implementation, numerical schemes, and algorithms, as Expert 1 said.

  • Reply to paragaph 4

Yes, the software HFSS was developed by Cendes and his students, especially Dr. Jin-Fa Lee, a major author of the computer codes and one of the proposers of this proposal. According to Lee, the software HFSS was the direct consequence of his Ph.D thesis work under the supervision of Cendes.

To elaborate, for HFSS, the theory and integration of 2D waveguiding structure and three-dimensional vector finite elements, the transfinite element method, were documented fully in Jin-Fa Lee's Ph.D. thesis (Jin-Fa Lee, Finite Element Methods for Modeling Passive Microwave Devices, Ph.D. thesis, Carnegie Mellon University, May 1989). Based on Lee's thesis, the paper authored by Jin-Fa Lee, “Analysis of passive microwave devices by using three-dimensional tangential vector finite elements,” International Journal of Numerical Modelling, Electronic Networks, Devices and Fields, Dec. 1990, pp. 235-246., was the only academic publication related to HFSS.

Although Dr. Jin-Fa Lee pursued his graduate studies at Carnegie Mellon University, his first encounter with finite element methods was through his undergraduate studies at NTU under Prof. Chun Hsiung Chen's instructions. In particular, he studied the VRT (Variational Reaction Theory) in depth during his senior year, which laid down a solid theoretical foundation for him and influenced the breakthrough of the transfinite element method.

  • Reply to paragraph 5

We mentioned HFSS because it is the first one for high-frequency EM simulation. We also believe that all EM simulation software has pros and cons from different aspects. The CST may be well accepted, but we cannot find objective data to show that it is a leading software and is more popular than HFSS.

The CST was developed by Thomas Weiland and his group. I think its early version is based on Weiland's famous and influential work, Finite Integration Technique (FIT), which is more like the Finite-Difference Time-Domain (FDTD) than the finite element method. We also agree that FIT was well-established for high-frequency modelling and simulation. We did mention FIT as a powerful tool in our proposal (in subsection Later development):

Of course, there exist compelling alternatives for simulating high-frequency electromagnetic waves, such as the method of moments (MoM), the finite-difference time-domain (FDTD) method, and the finite integral technique (FIT). Each method has its niche and has been implemented as commercial/non-commercial software

We did not refer to FIT-related references in the above excerpt because we think it is a well-known method. The reviewers can easily find more detailed introduction if necessary. We did the same for MoM and FDTD; both are also well-known, powerful, and well-accepted.

Since our proposal is focused on the finite element method, we also searched google scholar with the keywords "T. Weiland, finite element" on February 23, 2023, to understand Weiland's contributions to the finite element method, and obtained about 3,550 items. After checking the first ten pages of the results, we found that except for seven articles related to the FIT, everything we saw was published after the year 2000. Since the proposed milestone for the work of FEM@NTU covers only 1979-1991, it seems reasonable that Weiland's work need not be included in our proposal. The seven FIT articles, which did not deal with the finite element method, are given below. Of course, more articles on FIT will come out if the search keyword is "Thomas Weiland."

“Lossy waveguides with arbitrary boundary contour and material distribution,” by T Weiland - Archiv Elektronik und Uebertragungstechnik, 1979 - ui.adsabs.harvard.edu

[PDF] “On the numerical solution of Maxwell's equations and applications in the field of accelerator physics,” by T Weiland - Part. Accel., 1984 - cds.cern.ch

[PDF] “On the unique numerical solution of Maxwellian eigenvalue problems in three dimensions,” by T Weiland - Part. Accel., 1984 - cds.cern.ch

“Iterative methods for the solution of very large complex symmetric linear systems of equations in electrodynamics,” by M Clemens, T Weiland - 1996 - osti.gov

“Time domain electromagnetic field computation with finite difference methods,” by T Weiland - … Modelling: Electronic Networks, Devices and Fields, 1996 - Wiley Online Library

“Generation of 3D isosurfaces by means of the marching cube algorithm,” by M Bartsch, T Weiland, M Witting - IEEE transactions on …, 1996 - ieeexplore.ieee.org (Postprocessing for the FIT code MAFIA)

“Anderson localization in a string of microwave cavities,” by …, HD Gräf, R Hofferbert, H Rehfeld, A Richter, T Weiland - Physical Review E, 1999 – (using MAFIA FIT code for solving cavity problems)

Response to the Comment by Expert 1 in Nov. 2022: -- Ruey-Beei Wu (talk) 06:02, 27 February 2023 (UTC)

Again, we are sorry that Expert 1 could not support our updated proposal last year and hope Expert 1 can support our proposal with a new plaque citation this year.

Response to the Comment by Mrs. Amy Bix on Nov. 30, 2022: -- Ruey-Beei Wu (talk) 06:04, 27 February 2023 (UTC)

We thank the reviewer's comments and have replied to the comments of Expert 1 more appropriately this time. We also revised the plaque citation to describe the work of FEM@NTU more accurately. Specifically, we point out that their contributions are for the variational formulation of non-self-adjoint linear boundary value problems, which leads to the first FEM simulation of high-frequency electromagnetic wave scattering and propagation. We believe our clarification should have avoided leaving readers the impression of "exaggerating its role in relationship to work elsewhere."

Response to the Comment by Mr. Brian Berg on Feb. 6, 2023: -- Ruey-Beei Wu (talk) 06:07, 27 February 2023 (UTC)

We are sorry we overlooked the new information on the Discussion web page and did not respond in time. We have revised the plaque citation to make the benefits expressed more concretely, which read:

Being part of introducing the finite element method (FEM) into electromagnetics, from 1979 to 1991, an NTUEE research group proposed and refined Variational Reaction Theory for non-self-adjoint linear boundary value problems. As the first formulation to simulate high frequency electromagnetic wave scattering and propagation phenomena, its variational nature laid the foundation for modern FEM’s to solve full-wave Maxwell equations. Subsequently, some widely adopted commercial software's reliability and accuracy are firmly established.

and in Chinese

作為在電機工程引入有限元素法的部分貢獻,1979至1991台大電機系的電波研究團隊完成變分互作用系列理論,應用於線性非自伴邊界值問題,成為第一個可模擬高頻電磁波散射與傳播的變分原理。此法的變分性質為現代有限元素法求解 Maxwell 方程式奠定基礎,一些廣泛使用的商用有限元素軟體也因此確立了可靠性與準確性。

Re: Response to the Comment by Mr. Brian Berg on Feb. 6, 2023: -- Amy Bix (talk) 22:51, 21 March 2023 (UTC)

A milestone citation has a top limit of 70 words (with hyphenated words counting as two), with 60-65 more desirable. Accordingly, I have taken the proposed citation here and tried to create a more concise and clear phrasing, with the following recommendation:

National Taiwan University electrical engineering researchers proposed and refined Variational Reaction Theory for non-self-adjoint linear boundary value problems, creating the first formulation to simulate high frequency electromagnetic wave scattering and propagation phenomena. Introducing the finite element method (FEM) into electromagnetics helped lay the foundation for modern FEM’s to solve full-wave Maxwell equations. This discovery improved reliability and accuracy in some widely adopted commercial software.

A few notes: 1) I have not included a date in this citation, assuming that the dates will be listed in the plaque title; 2) I have spelled out NTUEE, since the meaning of this abbreviation may not be immediately evident to the general public reading this plaque; 3) I have done my best to suggest an improved phrasing of this citation, but given that the technical questions here are not my field, it will be important for others to check whether the statements and phrasing here are accurate.

Re: Re: Response to the Comment by Mr. Brian Berg on Feb. 6, 2023: -- Ruey-Beei Wu (talk) 07:09, 22 March 2023 (UTC)

We appreciate and fully agree with the revised wording. It's more concise and descriptive. We also note that the first phrase in our original citation, "Being part of introducing the finite element method (FEM) into electromagnetics," is now omitted. This paragraph is mainly in response to the review comments of expert 1, emphasizing that the application of FEM to electromagnetics spans many application areas, and the proposed milestone is a significant part of it. May the committee decide to keep or remove such a phrase. Thanks.

Re: Re: Re: Response to the Comment by Mr. Brian Berg on Feb. 6, 2023: -- Bberg (talk) 19:10, 26 March 2023 (UTC)

Amy did an excellent job in rewriting this citation, and I hope that this will be incorporated into the main page before this week's History Committee meeting. My only comment is that FEM's should be FEMs since a plural noun does not require an apostrophe, as described at https://www.hamilton.edu/academics/centers/writing/seven-sins-of-writing/4

Response to the Comment by Mr. Dave Michelson on Feb. 9, 2023: -- Ruey-Beei Wu (talk) 06:08, 27 February 2023 (UTC)

We agree that a milestone should include information about the time and space of technical advances or breakthroughs. We believe that the work of the FEM team at National Taiwan University from 1979 to 1991 has indeed advanced in formulating non-self-adjoint boundary problems and applying the finite element method to accurately simulate high-frequency electromagnetic waves' scattering, propagation, and radiation. Such formulation and applications in FEM did not exist before the series of research on VRT.

== Supplement: Did VRT work have a regional/community/industry/service impact? -- Ruey-Beei Wu (talk) 07:57, 30 March 2023 (UTC)

Thanks for the query. We are sure that the proposed work has academic impact. As for the impact on the industry, we know the following facts:

1) Comments from expert 2, who has his own microwave consulting company in the USA, said, “These important variational reaction concepts laid the foundation for a generalization of the FEM known as the “Transfinite Element Method.” This advanced formulation enabled the direct coupling of the volumetric FEM with RF transmission lines and waveguide ports. This breakthrough technology, augmented by adaptive mesh refinement methodologies, enabled for the first-time accurate S-Parameter simulations of fully 3D, industrially significant microwave devices and high-frequency circuits. This gave birth to the original “HFSS” commercial FEM software package, which became the forerunner to several industrially indispensable EM simulation tools that are now used by electrical engineers worldwide for the “virtual prototyping” of complex RF devices and related technologies. The consolidated contributions referenced in “Theoretical Foundation for Finite Element Method for Electromagnetics” have significantly contributed to advances in the field of computational EM. The EM simulation software technologies that followed have revolutionized an entire industry and enabled the development of products that one way or another affect every life.” Note that “Theoretical Foundation for Finite Element Method for Electromagnetics” is the title of the first version of this proposal.

2) As mentioned in our reply to comment 3 of Expert 1, one of the proposers, Dr. Jin-Fa Lee, is a major developer of HFSS, the first and very successful FEM commercial software for FEM. Dr. Lee took the FEM course taught by Professor Chun Hsiung CHEN at National Taiwan University. Especially in his senior year, he studied VRT (Variational Reaction Theory) in depth, which laid a solid theoretical foundation for him and influenced the breakthrough of the transfinite element method he proposed. To our knowledge, the HFSS, along with all other commercial FEM software, appears at least several years after the publication of VRT. Mathematically, the scattering, radiation, and propagation problems become non-self-adjoint. Though often not cited explicitly, the VRT or its equivalent form for those non-self-adjoint problems is believed to have been applied to derive desired matrix equations, at least for early versions of many commercial FEM software. The VRT also ensures that such formulation is valid and accurate. Here, by “equivalent form,” we mean the hybrid formulation using the Galerkin method or the method of weighted residue.

3) We only mention the milestone development span of 1979-1991 in our proposal. Yes, we later applied this method to some real industry-oriented 3D problems. We set different boundary conditions for the exterior problem, while VRT is used for the interior region. Due to different industry considerations, not all of these works have been published. Some of them were allowed to be published and will be presented here. For example, we applied the method to calculate the admittance of an inclined slot in a narrow wall of a rectangular waveguide. VRT is necessary to handle the inclined slot which is of irregular shape. The external problem uses a 90-degree wedge Green's function to model scattering off the waveguide corners. By comparing the calculated results with the literature results, it was found that the analysis is accurate, reliable and efficient [1]. The VRT method has also been used to deal with discontinuities in three-dimensional (3-D) waveguides of arbitrary shape. The approach is to use an edge-based tetrahedral finite element method for the junction region. External problems use analytical mode expansion technique for the waveguides. In addition to the variational formulation of the scattering coefficients, this study also presents several unique features that are also critical for successful analysis, including improved Delaunay triangulation for mesh generation and a frontal solution technique for the sparse matrix solutions. This approach was used to design a rectangular to dielectric-filled circular waveguide transition with less than -20 dB return loss over 40% bandwidth by using a suitable modified dielectric rod transformer [2]. Another interesting effort is our pioneering work combining FDTD and FEM to solve transient electromagnetic problems associated with structures of curved surfaces. A tetrahedral edge-based finite element scheme is only used to model the region near the curved surface, while the FDTD method is employed for most of the regular region. Without any interpolation of the field on the curved surface, nor any additional stability constraints due to finer divisions near the curved surface, the scheme was found to have second-order accuracy, unconditional stability, programming simplicity, and computational efficiency. The hybrid approach was applied to solve electromagnetic scattering from three-dimensional (3-D) arbitrarily shaped dielectric objects to demonstrate its superior performance [3]. [1] C. G. Jan, R. B. Wu, and P. Hsu, "Variational analysis of inclined slots in the narrow wall of a rectangular waveguide," IEEE Transactions on Antennas and Propagation, vol. 42, pp. 1455-1458, October 1994, [2] R. B. Wu, "A wideband waveguide transition design with modified dielectric transformer using edge-based tetrahedral finite element analysis," IEEE Transactions on Microwave Theory and Techniques, vol. 44, pp.1024-1031, July 1996. [3] R. B. Wu and T. Itoh, "Hybrid finite-difference time-domain modeling of curved surfaces using tetrahedral edge elements," IEEE Transactions on Antennas and Propagation, Vol. 45, pp. 1302-1309, August 1997.

Re: == Endorsement: Did VRT work have a regional/community/industry/service impact? by Prof. D. C. Chang -- Ruey-Beei Wu (talk) 07:59, 30 March 2023 (UTC)

I, Dau-Chyrh Chang, IEEE Life Fellow, the director (1982-1998) of the antenna division in the National Chung Shan Institute of Science and Technology (former Chung Shan Institute of Science and Technology, CSIST), and one of the co-authors of this paper:

C.G. Jan, R.B. Wu, P. Hsu, D.C. Chang, “Analysis of Edge Slots in Rectangular Waveguide with Finite Wall Thickness,” IEEE Transactions on Antenna and Propagation, Vol. 4, No. 8, August 1996.

This paper results from the many-year cooperation of CSIST and National Taiwan University. By using the FEM techniques, which Prof. Chen and his research group developed, the antenna division has developed many slot antennas for military applications. One of the slot array antennas was used for S-band PODAR (point defense array radar), and another was used for X-band PODAR. In addition to the ground base radar applications, many fuze slot antennas on various kinds of missiles were also developed by using the FEM technique.

Proposed Citation Rewording; No Documents Are Available -- GeoffT (talk) 01:17, 25 August 2023 (UTC)

I propose that the first 11 words of the proposed citation ("Researchers at electrical engineering department of the National Taiwan University proposed") be changed to "Researchers at National Taiwan University’s Electrical Engineering Department developed". This would put the university's name closer to the start, and properly capitalize the department's name. Most importantly, changing "proposed" to "developed" would much more accurately characterize the entire work honored by this proposal.

When I tried to look at some of the 60 references included in this proposal, only 2 hyperlinks were included, and both allowed access to only the Abstract of a paper. Are any papers or articles or stories available for free reading?

Re: Proposed Citation Rewording; No Documents Are Available -- Ruey-Beei Wu (talk) 03:21, 24 November 2023 (UTC)

Replace this text with your reply

I fully agree with the suggested change in the citation. It would be better to put the university name closer to the start. Also, I have inserted the hyperlinks for the 60 references except two regarding to dissertation. It can ease the readers to surf the references. Thanks for the comments.

Citation Improvement Suggestion -- Bberg (talk) 15:20, 20 November 2023 (UTC)

The first portion of the citation could be nicely improved by changing "Researchers at electrical engineering department of the National Taiwan University" to "Researchers at National Taiwan University's Electrical Engineering Department" as this would reduce the word count by 2, but more importantly would name the university first. As the department's name would normally be capitalized, I have made that change as well.

Re: Citation Improvement Suggestion -- Administrator5 (talk) 19:24, 20 November 2023 (UTC)

I agree with Brian. Ooverall, a good proposal...a lot of work to get it into this form

Re: Re: Citation Improvement Suggestion -- Bberg (talk) 13:08, 1 March 2024 (UTC)

I hope that the new wording that I proposed on November 20 will be used to update the citation on the main page.

Re: Re: Re: Citation Improvement Suggestion -- Ruey-Beei Wu (talk) 03:51, 12 March 2024 (UTC)
Thanks so much. Sorry for delay since I do not frequently surf the website. I have used the new wording in the proposal.