Have we lost the plot? Analysis of national submissions to Combustion Symposia from 1949 to 2010

The following article, written by Dr David B Smith, appeared in the Spring 2012 Newsletter of the British Section of the Combustion Institute. Because of its general interest, we reproduce it here.



Introduction

About 12 months ago, Paul Fennell wrote an interesting article for the Section Newsletter (June 2011). Somewhat provocatively entitled “Has the UK forgotten how to do combustion Science”, he compared the paper submission and acceptance rates for various countries for the 32nd and 33rd Symposia. Although not giving a direct answer to his leading question, he did reach some significant conclusions.

In Newsletter 96/3, I wrote an article along broadly similar, though more historical, lines. This presented a breakdown of papers presented at the Combustion Symposia, according to the authors’ countries. That article updated a previous one, at July 1987. Among train spotters, people with exaggerated interest in detail are called rivet counters. I hope that I don’t suffer from a similar syndrome but I have now extended my analysis up to the most recent (Beijing) symposium. The methodology is simple:

• All symposia are included, from the Third (held in 1949) to the Thirty Third. The first two symposia are not included. They were held long before the Combustion Institute was formed, in 1928 and 1937, under the auspices of the American Chemical Society, and so do not form part of the main series. 
• Authors of papers are categorised by country. Where authors come from more than one country, a split is made according to the number of institutions from each country. With the internationalisation of research, this is increasingly common and, to my mind, wholly welcome. 
• Contributions from the following countries are compared: US, UK, France, Germany and Japan. This choice is entirely pragmatic. It made more sense 30-40 years ago, when these countries provided the major input. More recently, again linked with internationalisation, other Institute Sections have played an increasingly dominant role (e.g. in Europe: Italy, Sweden, Netherlands, Denmark; in Asia: Korea, India, China, Taiwan (ROC); Canada,… I certainly don’t intend any disrespect for the very significant input they make. My justification is that including all would render the picture very confusing, it would not easily permit historical comparison back to the earliest Symposia, and so obscure my main aim of how UK combustion research is faring? 

As a footnote to this introduction, I appreciate that there are more subtle and comprehensive ways in which one might assess the health of combustion research in the UK. That would be vastly more complex and I do not have the necessary resource.
  
Increasing Size of the Symposia
To provide some general context, I thought it worth looking at how the Symposia have grown in size. Figure 1 demonstrates that the Symposia have grown significantly over the years, from about 100 papers initially up to 400 or more now.

The pressure to include more papers has been accommodated in a number of ways. First, the number of parallel sessions has increased. I believe the early symposia were single session. This has now grown to 4, 6 and now typically 7 parallel sessions. At the 2002 Board meeting, the suggestion was made to go even further, possibly up to 10, though this was rejected. Another suggestion (again not accepted by the Board) was to reduce paper presentation time from the current 20 minutes to 5-10. Another ploy which was used at the 32nd Symposium (Montreal) was to have some accepted papers presented as posters rather than orally, though I don’t think this was universally appreciated. This paper growth has had another consequence: since the 26th Symposium (Naples, 1996), Symposium Proceedings have been published as two-volumes.

 As noted in the Introduction, one reason for the increased symposium size is the greater input from what I have loosely and disingenuously collated as ‘other’ countries, i.e. those which played small part in the early years. This is illustrated in Figure 1, which shows a growth from essentially zero in the earliest symposia to more than 20% of all papers now.

How has UK fared over the years?
The full analysis shows a complex pattern and I have tried to present it in several ways. The first slice (Figure 2) shows the percentage of papers over the years for those countries highlighted in the Introduction, i.e. UK, US, France, Germany and Japan. Being constrained to black on white makes it a very complicated graph, but I think several points do emerge from the confusion.

The earliest symposia were essentially US/UK affairs, with an approximately 2:1 split, and little input from elsewhere. Since then the US percent has dropped from about 60% to less than 40% most recently, while the UK contribution has dropped more sharply from 33% to 6%, at Beijing. Inputs from France, Germany and Japan have all grown steadily and, ignoring inevitable fluctuations, over the last 10 symposia these three countries plus UK each lie within a 5 - 15% band.



This percentage drop in UK input is partly due to the ‘dilution’ effect of increased internationalisation. In order to try to eliminate this, I show in Figure 3 the actual number of papers (rather than %) that we have presented, again using the US as a yardstick.


 
This shows markedly different patterns for the two countries. The number of US papers presented has grown roughly threefold (60 to 180-ish). In contrast, we have stayed in a 20 - 40 band, though with a slight overall decrease.  

Role of China
As an aside from the main theme of the article, I was interested in the contribution from China. Until fairly recently, the Chinese combustion community has had essentially no role in the Symposia but, as Figure 4 shows, this is now changing dramatically. In the last two symposia, their number of papers has equalled, even exceeded, those from UK, France or Germany. My guess is that such growth will continue and that we will see an increasing impact.  



Academic vs. Industrial Input from UK

Some British Section members will know that I spent my working life in combustion research within British Gas (as it was when I joined). I have, therefore, been particularly interested to what extent industrial labs have contributed to the symposia. And so, in my previous analysis of symposium presentations, I examined the balance between academic and industrial papers from the UK.

I continue this here, by looking at just four symposia spread over 40 years: the 7th (1958), 12th (1968), 22nd (1988) and 32nd (2008). The pattern shows a remarkable, though perhaps not surprising, shift. Figure 5 clearly shows a steady drop in the proportion of papers from industry, from almost 31% to barely 4% now.

It is also instructive to look at those industries which contributed to the 7th Symposium: the Government labs; Rocket Propulsion Establishment ,Westcott, near Aylesbury; Explosive Research & Development, Waltham Abbey; Fire Research Establishment, Boreham Wood; National Gas Turbines Establishment, Farnborough; Coal Research Establishment, Stoke Orchard : and from private industry; Rolls Royce; Joseph Lucas; CAV ; United Steel Companies, Rotherham; Armstrong Siddeley. Many members will not remember the last of these (though I’m sure our esteemed, retiring editor will); they were luxury car makers, when cars really were cars. Anyone interested in their model range can check this link.

 (Comment by John Griffiths: Dave tantalised me with this comment. The contribution from the “Combustion Department” of A.S. to the 7th Symposium was, “A Theoretical Analysis of Multi-stage Reaction Rate Controlled Systems.” It was based on the landmark Longwell and Weiss study in an adiabatic stirred flow reactor. On a more telling note, given that the Symposium was held in London and Oxford, 21 UK companies, and the Institute of Fuel, made financial contributions to support it. I think only three of those industrial names survive today, certainly in closely related form.)

There are at least two (not exclusive) reasons for the drop in industrial input. First, symposia are increasingly dominated by fundamental research. This concern was raised at the Heidelberg Symposium (2006) by Roman Weber (previously of the International Flame Research Station in Ijmuiden, Netherlands and now at the Technical University of Clausthal (Germany). His full argument is presented in the Section Newsletter, Vol 2007-1, March 2007, but briefly he expressed two specific concerns:

1. “The gap between combustion technology and combustion fundamentals has never been as wide as now.” I am not sure I agree with him on that. I think recent developments in e.g. diagnostics and modelling mean that we should be in a better position to start to bridge the gap. 
2. “We are almost exclusively academic people” – meaning symposia attendees. A check of the attendance lists of recent symposia indicates this is indeed so.

If this diagnosis is correct, it inevitably follows that industry will see our symposia as increasingly irrelevant to them.

The second reason for lower industrial input from the UK is more parochial. In earlier symposia, significant input came from publicly owned energy industries: gas, electricity and coal. Their privatisation and subsequent break-up in the 1980-90s led to a substantial drop in their research activity. Members aware of my political views will know I opposed this at the time. One unfortunate but easily foreseen consequence has been the decimation of relevant combustion research. But private industry associated with combustion, with the dazzling exception of Shell Thornton up to the present, also seems to have backed away from research. And the result? In the 22nd Symposium, there was input from Shell (3 papers), British Gas and the Coal Research Establishment; in the 32nd, from Shell, BP Exploration and Johnson Matthey.

I do not see how the Section can have any impact on the second reason for industrial research decline. But I do think we could consider whether the Symposium content has become unbalanced.

Conclusion
I think that I’ve presented a mixed but, I hope, not too gloomy a picture. In many ways, combustion symposia continue to be huge successes. But I seriously think the Institute needs to look at what it is aimed at, i.e. more or less exclusively academically based research, or a better balance between fundamental and applied activity. And on the UK scene, perhaps the British Section should seriously consider whether it could engage more with applied/industrial combustion interests here.

Publishing Science, Profits and the Public Interest

Invited article written by Professor Simone Hochgreb, University of Cambridge, Secretary of the Committee of the Combustion Institute British Section.

News erupted over the past year, starting with an article by George Monbiot on the obscene profits of academic publishers in August 2011. This has been followed by a number of articles, and the announcement by Fields Medal Winner Timothy Gowers not to submit articles to Elsevier journals on 21 January 2012, leading to a movement which collected 9000 signatures of scholars who pledged to refrain from working for publishers which restrict free sharing as authors, reviewers or editors.

Philosophiae Naturalis Principia Mathematica (wikipedia)

In parallel, Sir Mark Walport, director of Wellcome Trust, the second largest non-governmental funder of medical research, found out from moving from his prior job as an academic that he could not download publications which he might have published himself at his new organization. As a result, the Wellcome Trust is now launching a high calibre scientific journal free on the web. The momentum behind open source publishing is clearly gaining ground by the day.But first, let's do the numbers that have led to the belated outrage directed in particular at Elsevier:

1) The market: the academic publishing market was around 19 million USD, 43% controlled by Elsevier, Springer and Wiley, whilst the remainder fragmented among over 2000 publishers [1]. These numbers are from 2002, so assuming steady growth of 3%, the absolute numbesr could be over 30% higher.

2) These big three publishers control all top journals (Nature, Science, and the top journals in other scientific realms). Given that most libraries need these subscriptions, and most authors need to be published in the top journals to be considered for academic promotion, this gives the journals stupendous bargaining powers. Interestingly, the numbers on what they charge different universities are not necessarily available, as prices paid by universities are covered by confidentiality clauses with publishers. According to David Prosser, executive director of Research Libraries UK, "British universities spend around £200m a year on subscriptions to electronic databases and journals, which is around 10% of the block grants the institutions receive from government", that means, a sizable fraction of the overhead.

3) As a result, profit margins for these publishers are mind-boggling: Elsevier has just boasted results for 2011, with a profits/revenue ratio of (27%) which are similar to 2010 (27%) [2], and Springer not far behind with 18.6% for 2011 [3].

Before we all run out and buy some stock, let us consider for a moment what allows this state of affairs:

1) Research is produced, written up, reviewed, corrected, formatted and uploaded by each one of us, with additional hours devoted by the many editors and assistant editors -- for no pay.

2) We give the publishers the rights to own the labour and the good name of the journals, which are associated with the eminence of the editorial boards, for historical reasons: the various societies outsourced the formerly tedious job of editing to companies. Interestingly, the companies have automated the publishing job, but we still do all the editing, proofreading and fact checking.

3) We pay them through our university overheads or our direct downloads for the privilege of maintaining the servers that automate the process, and for the pleasure of negotiating bulk deals to buy back what we produced for incredibly high prices.

4) We proudly point to our list of publications on these journals during promotion and tenure decisions. The more prestigious and expensive (which correlates with the number of citations, as these are disseminated further) the journal, the more points we get.

In summary, this is a strange state of affairs, which is maintained through our complacent inertia, as this seems too big a problem to attack. It is also not a new issue, but one which has proven resilient to change, which is why the profits and the charges to our libraries continue. Other societies in our field of fluid mechanics in combustion (SAE, AIAA, ASME) can be even worse, providing less service in readying papers for publishing than journals, for higher charges, but creating an audience and the appearance of peer review.
Whereas there continues to be significant movement in open source publishing, the evolving picture of open access may not lend itself to a grand ending, but rather a series of small steps.

1) Insist on maintaining the copyright, and deposit material in archival repository such as dspace: many journals, including Elsevier, allow this, but the fine print is still moving and uncertain.

2) Consider publishing open source journals (PLoSOne), or society journals (AIP, RSC) which charge libraries significantly less.

3) Work with the different societies in the field to understand the costs of breaking the oligopoly of publishing, given that online publishing tools are so widely available: keep the peers and their review, but understand whether the same or better service in delivering on-line archival material can be obtained.
Finally, there remains the issue of the indexing and how we are all subject to the vagaries of Thomson-Reuters and its ownership of the largest abstract indexing engine. But that will be the subject of another article.

References

1. G. S. McGuigan, R. D. Russell, The Business of Academic Publishing: A Strategic Analysis of the Academic Journal Publishing Industry and its Impact on the Future of Scholarly Publishing, Electronic Journal of Academic and Special Librarianship, v. 9, n. 3 (2008).
2. Reed Elsevier 2011 Results.
3. Springer press release 2011 Results.

Has the UK Forgotten how to do Combustion Science?

by Dr Paul Fennell, Imperial College London
(article published in the Newsletter of the British Section of the Combustion Institute, Volume 2011-1 June 2011)

Introduction
You have prepared your manuscript for a Combustion Symposium, discussed it with your co-authors, and performed the most taxing test of your mental faculties for the year. Of course, here I am talking about getting the word count done, not writing the paper. You then send it off, and the waiting begins... For the first couple of months, the paper nags at you: did I do this correctly, was it clear enough? Then, you forget about it. Months pass.

Then, one day, an e-mail arrives. Heart pounding, you open it.

pa = “WHAT THE DEVIL? Who chose these morons?”
pb = “Fantastic! What excellent reviewers, and what a good Symposium session chair”.

Those of you who attended the 33rd Symposium may have noted that pa was quite high (on a historical basis) for the UK, whereas pb was quite low. Of course, there are always grumbles about the standard of refereeing, though this year there has been a larger than usual number reaching your humble correspondent’s ears. For interest, the success rates for submitted papers are reproduced below, grouped in a number of different ways, and including some data from the 32nd Symposium.

Success Rates

Table 1. Success rates for the Symposium in general (32nd Symposium data from Jan 09 British Section Newsletter)

Table 2. Success rates by country for the last two Symposia

The massive increase in submissions for the 33rd Symposium, relative to the 32nd Symposium, reduced the potential for acceptance, but France and the USA maintained position as the front runners, with Australia also having strong representation, whereas the UK’s acceptance rate was somewhat below average (34.2 % in 2010) and much lower than at the previous Symposium (87.2 % in 2008).

Table 3. Overall success 33rd Symposium, by Colloquia

Conspicuous in the success rate, when presented in the context of the colloquia, is the high proportion of accepted papers for detonations, kinetics research and soot related studies. Investigators of heterogeneous combustion and stationary combustion systems seem to have a particularly hard time. Although there is no compelling reason for success rates to be the same for each colloquium, the number of submissions to each one is certainly a reflection of the trends in combustion research activity.

Discussion
Four main possibilities come to mind as to why the British section has such a low overall success rate, even if it continues to have a similar total number of papers accepted to e.g. France and Germany.

1. Britain no longer does world-leading research in combustion science
2. Britain does research in the wrong areas
3. There is bias in the system
4. France, the USA and Australia are better prepared for the Symposia

Does Britain no Longer do World-Leading Research in Combustion? If not, why not?
If we leave out the hideous possibility that we are not as clever as we think, we need to consider whether we have adequate facilities in the UK for world-leading combustion research. Perhaps the community, accustomed to fighting each other for acceptance for a Combustion Symposium, fights itself too vehemently when it comes to reviewing grant proposals? The evidence from funding councils is that engineering disciplines review each other more harshly than others – but that they do not take into account the relative harshness of reviews within a particular discipline when awarding grants. Perhaps as a community we should take heed of certain other areas of research... those where everyone, uniformly, is world leading. The community knows that combustion underpins modern society: where would power generation or transportation be without our work? Which is more important? Combustion or nanotechnology? Which trumpets its achievements better and secures guaranteed funding in the UK? Why?

The imminent demise of BCURA as a funding agency shows just how far the star of combustion has fallen. In years gone by, the UK had the CRE and other similar establishments for electricity generation, gas and nuclear energy. Now, we have next to no centralised research facilities, and industry complains that the shortage of skilled engineers in the UK forces them to recruit from outside the EU. The combustion community needs to work collectively to promote our achievements and to explain how we can contribute to the very real challenges facing us in the next decades. Combustion scientists need to show that we are part of the clean energy solution, not the problem.

Does Britain do research in the wrong area?
A brief survey of the papers accepted for the Symposium this year from Britain (i.e. those that receiving a travel grant from the British section) indicates that Britain has particular strength in modelling (particularly LES) and flames. Of course, there is immense survivor bias in these results, since they are accepted papers, not submitted ones. (It might be helpful for the British section to request that members send details not just of accepted papers to the British section office, but also of submitted papers). It is worrying that whilst the UK still retains great strength in the field of heterogeneous combustion (amongst other fields), papers are not being accepted for publication in this colloquium. It would be interesting to see if this was caused by non-submission or by non-acceptance. Certainly, discussion with a number of eminent scientists in this area leads to the suspicion that the quality of the peer review process in certain areas is seen as a significant barrier. Furthermore, some have indicated that they do not submit papers to a Symposium because it is seen as increasingly non-relevant to industry.

Finally, the large numbers of papers which are written on the basis of shock-tube measurements would indicate that there is a relevance of this research field. It was an area which had tremendous strength in the UK, in earlier decades, but the kinetics aspect has declined to negligible activity. Fortunately, outstanding research in detonation related studies is still active. Numerical simulations of the results from a number of different types of shock tubes, along with jet-stirred reactors, etc, are world-leading. Surely, we need to obtain the facilities to conduct these experiments ourselves, rather than relying upon our colleagues around the world?

Bias
Table 3 shows that the acceptance rates in certain areas were very high at the most recent Symposium. The Institute as a whole seems at the moment to have a great deal of positive reinforcement in these fields – an extremely eminent member of the Institute recently made the statement that “we should only be doing research in important fields”. He then outlined which fields were important... the areas into which the Combustion Institute is more and more moving. Some might consider that the success rates indicate a bias in favour of kinetics research, though others would argue that the standards of this research, historically high, have simply been maintained. It is also possible that the standards of reviewing are superior in this field. (Over very many Symposia, my recollection is one of a proportionately greater success with submissions to Kinetics than to other colloquia. Perhaps crossing the boundaries from my combustion kinetics background was never going to be easy – yet the Combustion Symposia have seemed to me to be a natural forum for this to happen. Perhaps better scientists than me have a different story to tell. Ed.)

Given that the UK retains or is building significant research capability in industrial areas (such as Doosan-Babcock’s oxyfuel demonstration plant in Renfrew) and also that the Combustion Institute as a whole is worried at the highest levels about the rapid decline in the submission of papers from industry into the Symposia, perhaps the British Section should work hard to encourage the submission of papers from our industrial colleagues. It might also make sense to encourage British industrialists to take a stronger role in the British section and in time the highest echelons of the Institute. This might help to bolster our position in, for example, stationary combustion.

Finally, it is interesting to note the opacity of the methods chosen to add new members to the worldwide committee of the Combustion Institute. The Catholic Church seems to have stolen a march on the combustion community by signalling the new pope using coloured smoke. Many other features of the selection process appear to be similar.

Preparation for the Symposia
Reports indicate that the US Sections in particular are far better prepared for the International Symposia. Interesting Symposium-worthy material is, often, first presented at American Section meetings and comments sought well ahead of time on the potential for improvement. Furthermore, it was clear from the comments of the co-chairs in the introduction to the Symposium that what they were looking for was a clear message from the paper. It is possible that in the UK we are not so accustomed to writing the short, punchy, papers which are now required for a Symposium; perhaps we need to begin preparation of our papers with this in mind, rather than initially writing long papers and agonising

Conclusions
The underperformance of the British section (at least on a historical basis, and based on success rate rather than total number of accepted papers) is noteworthy. Some possible reasons for this have been explored. It is clear that the Combustion Institute has been moving in one particular direction for a number of years. Whether the British Section can do anything about this is unclear (even if we should want to - which is a topic that should be debated). Securing funding for combustion research is vital, particularly in the current environment. The countries who are outperforming the UK invest more money in combustion research, so that working together to promote combustion research in the UK and secure this outcome may be a key factor.

by Paul Fennell