martes, 27 de agosto de 2019

Maxwell and The Nerds My favorite article of carl sagan















Maxwell and The Nerds

Why should we subsidize intellectual curiosity?
– Ronald Reagan, campaign speech, 1980
There is nothing which can better deserve our patronage than the promotion of science and literature. Knowledge is in every country the surest basis of public happiness.
– George Washington, address to Congress, 8 January 1790
Stereotypes abound. Ethnic groups are stereotyped, the citizens of other nations and religions are stereotyped, the genders and sexual preferences are stereotyped, people born in various times of the year are stereotyped (Sun-sign astrology), and occupations are stereotyped. The most generous interpretation ascribes it to a kind of intellectual laziness: instead of judging people on their individual merits and deficits, we concentrate on one or two bits of information about them, and then place them in a small number of previously constructed pigeonholes.
This saves the trouble of thinking, at the price in many cases of committing a profound injustice. It also shields the stereotyper from contact with the enormous variety of people, the multiplicity of ways of being human. Even if stereotyping were valid on average, it is bound to fail in many individual cases: human variation runs to bell-type curves. There's an average value of any quality, and smaller numbers of people running off in both extremes.
Some stereotyping is the result of not controlling the variables, of forgetting what other factors might be in play. For example, it used to be that there were almost no women in science. Many male scientists were vehement: this proved that women lacked the ability to do science. Temperamentally, it didn't fit them, it was too difficult, it required a kind of intelligence that women don't have, they're too emotional to be objective, can you think of any great women theoretical physicists?... and so on. Since then the barriers have come tumbling down. Today women populate most of the subdisciplines of science. In my own fields of astronomy and planetary studies, women have recently burst upon the scene, making discovery after discovery, and providing a desperately needed breath of fresh air.
So what data were they missing, all those famous male scientists of the 1950s and 1960s and earlier who had pronounced so authoritatively on the intellectual deficiencies of women? Plainly, society was preventing women from entering science, and then criticizing them for it, confusing cause and effect:
You want to be an astronomer, young woman? Sorry.
Why can't you? Because you're unsuited.
How do we know you're unsuited? Because women have never been astronomers.
Put so baldly, the case sounds absurd. But the contrivances of bias can be subtle. The despised group is rejected by spurious arguments, sometimes done with such confidence and contempt that many of us, including some of the victims themselves, fail to recognize it as self-serving sleight of hand.
Casual observers of meetings of sceptics, and those who glance at the list of CSICOP Fellows, have noted a great preponderance of men. Others claim disproportionate numbers of women among believers in astrology (horoscopes in most 'women's' but few 'men's' magazines), crystals, ESP and the like. Some commentators suggest that there is something peculiarly male about scepticism. It's hard-driving, competitive, confrontational, tough-minded - whereas women, they say, are more accepting, consensus-building, and uninterested in challenging conventional wisdom. But in my experience women scientists have just as finely honed sceptical senses as their male counterparts; that's just part of being a scientist. This criticism, if that's what it is, is presented to the world in the usual ragged disguise: if you discourage women from being sceptical and don't train them in scepticism, then sure enough you may find that many women aren't sceptical. Open the doors and let them in, and they're as sceptical as anybody else.
One of the stereotyped occupations is science. Scientists are nerds, socially inept, working on incomprehensible subjects that no normal person would find in any way interesting - even if he were willing to invest the time required, which, again, no sensible person would. 'Get a life,' you might want to tell them.
I asked for a fleshed-out contemporary characterization of science-nerds from an expert on eleven-year-olds of my acquaintance. I should stress that she is merely reporting, not necessarily endorsing, the conventional prejudices:
Nerds wear their belts just under their rib cages. Their short-sleeve shirts are equipped with pocket protectors in which is displayed a formidable array of multicoloured pens and pencils. A programmable calculator is carried in a special belt holster. They all wear thick glasses with broken nose-pieces that have been repaired with Band-Aids. They are bereft of social skills, and oblivious or indifferent to the lack. When they laugh, what comes out is a snort. They jabber at each other in an incomprehensible language. They'll jump at the opportunity to work for extra credit in all classes except gym. They look down on normal people, who in turn laugh at them. Most nerds have names like Norman. (The Norman Conquest involved a horde of high-belted, pocket-protected, calculator-carrying nerds with broken glasses invading England.) There are more boy nerds than girl nerds, but there are plenty of both. Nerds don't date. If you're a nerd you can't be cool. Also vice versa.
This of course is a stereotype. There are scientists who dress elegantly, who are devastatingly cool, who many people long to date, who do not carry concealed calculators to social events. Some you'd never guess were scientists if you invited them to your home.
But other scientists do match the stereotype, more or less. They're pretty socially inept. There may be, proportionately, many more nerds among scientists than among backhoe operators or fashion designers or traffic wardens. Perhaps scientists are more nerdish than bartenders or surgeons or short-order cooks. Why should this be? Maybe people untalented in getting along with others find a refuge in impersonal pursuits, particularly mathematics and the physical sciences. Maybe the serious study of difficult subjects requires so much time and dedication that very little is left over for learning more than the barest social niceties. Maybe it's a combination of both.
Like the mad-scientist image to which it's closely related, the nerd-scientist stereotype is pervasive in our society. What's wrong with a little good-natured fun at the expense of scientists? If, for whatever reason, people dislike the stereotypical scientist, they are less likely to support science. Why subsidize geeks to pursue their absurd and incomprehensible little projects? Well, we know the answer to that: science is supported because it provides spectacular benefits at all levels in society, as I have argued earlier in this book. So those who find nerds distasteful, but at the same time crave the products of science, face a kind of dilemma. A tempting resolution is to direct the activities of the scientists. Don't give them money to go off in weird directions; instead tell them what we need - this invention, or that process. Subsidize not the curiosity of the nerds, but what will benefit society. It seems simple enough.
The trouble is that ordering someone to go out and make a specific invention, even if price is no object, hardly guarantees that it gets done. There may be an underpinning of knowledge that's unavailable, without which no one will ever build the contrivance you have in mind. And the history of science shows that often you can't go after the underpinnings in a directed way, either. They may emerge out of the idle musings of some lonely young person off in the boondocks. They're ignored or rejected even by other scientists, sometimes until a new generation of scientists comes along. Urging major practical inventions while discouraging curiosity-driven research would be spectacularly counterproductive. Suppose you are, by the Grace of God, Victoria, Queen of the United Kingdom of Great Britain and Ireland, and Defender of the Faith in the most prosperous and triumphant age of the British Empire. Your dominions stretch across the planet. Maps of the world are abundantly splashed with British pink. You preside over the world's leading technological power. The steam engine is perfected in Great Britain, largely by Scottish engineers, who provide technical expertise on the railways and steamships that bind up the Empire.
Suppose in the year 1860 you have a visionary idea, so daring it would have been rejected by Jules Verne's publisher. You want a machine that will carry your voice, as well as moving pictures of the glory of the Empire, into every home in the kingdom. What's more, the sounds and pictures must come not through conduits or wires, but somehow out of the air, so people at work and in the field can receive instantaneous inspirational offerings designed to insure loyalty and the work ethic. The Word of God could also be conveyed by the same contrivance. Other socially desirable applications would doubtless be found.
So with the Prime Minister's support, you convene the Cabinet, the Imperial General Staff, and the leading scientists and engineers of the Empire. You will allocate a million pounds, you tell them - big money in 1860. If they need more, just ask. You don't care how they do it; just get it done. Oh, yes, it's to be called the Westminster Project.
Probably there would be some useful inventions emerging out of such an endeavour - 'spin-off. There always are when you spend huge amounts of money on technology. But the Westminster Project would almost certainly fail. Why? Because the underlying science hadn't been done. By 1860 the telegraph was in existence. You could imagine at great expense telegraphy sets in every home, with people ditting and dahing messages out in Morse code. But that's not what the Queen asked for. She had radio and television in mind but they were far out of reach.
In the real world, the physics necessary to invent radio and television would come from a direction that no one could have predicted.
James Clerk Maxwell was born in Edinburgh, Scotland, in 1831. At age two he found that he could use a tin plate to bounce an image of the Sun off the furniture and make it dance against the walls. As his parents came running he cried out, 'It's the Sun! I got it with the tin plate!' In his boyhood, he was fascinated by bugs, grubs, rocks, flowers, lenses, machines. 'It was humiliating,' later recalled his Aunt Jane, 'to be asked so many questions one couldn't answer by a child like that.'
Naturally, by the time he got to school he was called 'Dafty' -not quite right in the head. He was an exceptionally handsome young man, but he dressed carelessly, for comfort rather than style, and his Scottish provincialisms in speech and conduct were a cause for derision, especially by the time he reached college. And he had peculiar interests.
Maxwell was a nerd. He fared little better with his teachers than with his fellow students. Here's a poignant couplet he wrote at the time:
Ye years roll on, and haste the expected time
When flogging boys shall be accounted crime.
Many years later, in 1872, in his inaugural lecture as professor of experimental physics at Cambridge University, he alluded to the nerdish stereotype: It is not so long ago since any man who devoted himself to geometry, or to any science requiring continued application, was looked upon as necessarily a misanthrope, who must have abandoned all human interests, and betaken himself to abstractions so far removed from all the world of life and action that he has become insensible alike to the attractions of pleasure and to the claims of duty.
I suspect that 'not so long ago' was Maxwell's way of recalling the experiences of his youth. He then went on to say,
In the present day, men of science are not looked upon with the same awe or with the same suspicion. They are supposed to be in league with the material spirit of the age, and to form a kind of advanced Radical party among men of learning. We no longer live in a time of untrammelled optimism about the benefits of science and technology. We understand that there is a downside. Circumstances today are much closer to what Maxwell remembered from his childhood.
He made enormous contributions to astronomy and physics -from the conclusive demonstration that the rings of Saturn are composed of small particles, to the elastic properties of solids, to the disciplines now called the kinetic theory of gases and statistical mechanics. It was he who first showed that an enormous number of tiny molecules, moving on their own and incessantly colliding with each other and bouncing elastically, leads not to confusion, but to precise statistical laws. The properties of such a gas can be predicted and understood. (The bell-shaped curve that describes the speeds of molecules in a gas is now called the Maxwell-Boltzmann distribution.) He invented a mythical being, now 'Maxwell's demon', whose actions generated a paradox that took modern information theory and quantum mechanics to resolve.
The nature of light had been a mystery since antiquity. There were acrimonious learned debates on whether it was a particle or a wave. Popular definitions ran to the style, 'Light is darkness - lit up'. Maxwell's greatest contribution was his discovery that electricity and magnetism, of all things, join together to become light. The now conventional understanding of the electromagnetic spectrum - running in wavelength from gamma rays to X-rays to ultraviolet light to visible light to infrared light to radio waves - is due to Maxwell. So is radio, television and radar.
But Maxwell wasn't after any of this. He was interested in how electricity makes magnetism and vice versa. I want to describe what Maxwell did, but his historic accomplishment is highly mathematical. In a few pages, I can at best give you only a flavour. If you do not fully understand what I'm about to say, please bear with me. There's no way we can get a feeling for what Maxwell did without looking at a little mathematics.
Mesmer, the inventor of 'mesmerism', believed he had discovered a magnetic fluid, 'almost the same thing as the electric fluid', that permeated all things. On this matter as well, he was mistaken. We now know that there is no special magnetic fluid, and that all magnetism - including the power that resides in a bar or horseshoe magnet - is due to moving electricity. The Danish physicist Hans Christian Oersted had performed a little experiment in which electricity was made to flow down a wire and induce a nearby compass needle to waver and tremble. The wire and the compass were not in physical contact. The great English physicist Michael Faraday had done the complementary experiment: he made a magnetic force turn on and off and thereby generated a current of electricity in a nearby wire. Time-varying electricity had somehow reached out and generated magnetism, and time-varying magnetism had somehow reached out and generated electricity. This was called 'induction' and was deeply mysterious, close to magic.
Faraday proposed that the magnet had an invisible 'field' of force that extended into surrounding space, stronger close to the magnet, weaker farther away. You could track the form of the field by placing tiny iron filings on a piece of paper and waving a magnet underneath. Likewise, your hair after a good combing on a low-humidity day generates an electric field which invisibly extends out from your head, and which can even make small pieces of paper move by themselves.
The electricity in a wire, we now know, is caused by submicroscopic electrical particles, called electrons, which respond to an electric field and move. The wires are made of materials like copper which have lots of free electrons -electrons not bound within atoms, but able to move. Unlike copper, though, most materials, say, wood, are not good conductors; they are instead insulators or 'dielectrics'. In them, comparatively few electrons are available to move in response to the impressed electric or magnetic field. Not much of a current is produced. Of course there's some movement or 'displacement' of electrons, and the bigger the electric field, the more displacement occurs.
Maxwell devised a way of writing what was known about electricity and magnetism in his time, a method of summarizing precisely all those experiments with wires and currents and magnets. Here they are, the four Maxwell equations for the behaviour of electricity and magnetism in matter:
∇ * E = ρ/ε0
∇ * B = 0
∇ x E = -Ḃ
∇ x B = μ0j + μ0ε0Ė
It takes a few years of university-level physics to understand these equations. They are written using a branch of mathematics called vector calculus. A vector, written in bold-face type, is any quantity with both a magnitude and a direction. Sixty miles an hour isn't a vector, but sixty miles an hour due north on Highway 1 is. E and B represent the electric and magnetic fields. The triangle, called a nabla (because of its resemblance to a certain ancient Middle Eastern harp), expresses how the electric or magnetic fields vary in three-dimensional space. The 'dot product' and the 'cross product' after the nablas are statements of two different kinds of spatial variation.
Ė and Ḃ represent the time variation, the rate of change of the electric and magnetic fields, J stands for the electrical current. The lower-case Greek letter ρ (rho) represents the density of electrical charges, while ε0 (pronounced 'epsilon zero') and μ (pronounced 'mu zero') are not variables, but properties of the substance E and B are measured in, and determined by experiment. In a vacuum, ε0 and μ are constants of nature.
Considering how many different quantities are being brought together in these equations, it's striking how simple they are. They could have gone on for pages, but they don't.
The first of the four Maxwell equations tells how an electric field due to electrical charges (electrons, for example) varies with distance (it gets weaker the farther away we go). But the greater the charge density (the more electrons, say, in a given space), the stronger the field.
The second equation tells us that there's no comparable statement in magnetism, because Mesmer's magnetic 'charges' (or magnetic 'monopoles') do not exist: saw a magnet in half and you won't be holding an isolated 'north' pole and an isolated 'south' pole; each piece now has its own 'north' and 'south' pole.
The third equation tells us how a changing magnetic field induces an electric field.
The fourth describes the converse - how a changing electric field (or an electrical current) induces a magnetic field.
The four equations are essentially distillations of generations of laboratory experiments, mainly by French and British scientists. What I've described here vaguely and qualitatively, the equations describe exactly and quantitatively.
Maxwell then asked himself a strange question: what would these equations look like in empty space, in a vacuum, in a place where there were no electrical charges and no electrical currents? We might very well anticipate no electric and no magnetic fields in a vacuum. Instead, he suggested that the right form of the Maxwell equations for the behaviour of electricity and magnetism in empty space is this:
∇ ⋅ E = 0
∇ ⋅ B = 0
∇ x E = -Ḃ
∇ x B = μ0ε0Ė
He set ρ equal to zero, indicating that there are no electrical charges. He also set j equal to zero, indicating that there are no electrical currents. But he didn't discard the last term in the fourth equation, με0Ė feeble displacement current in insulators.
Why not? As you can see from the equations, Maxwell's intuition preserved the symmetry between the magnetic and electric fields. Even in a vacuum, in the total absence of electricity, or even matter, a changing magnetic field, he proposed, elicits an electric field and vice versa. The equations were to represent Nature, and Nature is, Maxwell believed, beautiful and elegant. (There was also another, more technical reason for preserving the displacement current in a vacuum, which we pass over here.) This essentially aesthetic judgement by a nerdish physicist, entirely unknown except to a few other academic scientists, has done more to shape our civilization than any ten recent presidents and prime ministers.
Briefly, the four Maxwell equations for a vacuum say (1) there are no electrical charges in a vacuum; (2) there are no magnetic monopoles in a vacuum; (3) a changing magnetic field generates an electrical field; and (4) vice versa.
When the equations were written down like this, Maxwell was readily able to show that E and B propagated through empty space as if they were waves. What's more, he could calculate the speed of the wave. It was just 1 divided by the square root of ε0 times μ. But ε0 and μ had been measured in the laboratory. When you plugged in the numbers you found that the electric and magnetic fields in a vacuum ought to propagate, astonishingly, at the same speed as had already been measured for light. The agreement was too close to be accidental. Suddenly, disconcertingly, electricity and magnetism were deeply implicated in the nature of light.
Since light now appeared to behave as waves and to derive from electric and magnetic fields, Maxwell called it electromagnetic. Those obscure experiments with batteries and wires had something to do with the brightness of the Sun, with how we see, with what light is. Ruminating on Maxwell's discovery many years later, Albert Einstein wrote, To few men in the world has such an experience been vouchsafed.'
Maxwell himself was baffled by the results. The vacuum seemed to act like a dielectric. He said that it can be 'electrically polarized'. Living in a mechanical age, Maxwell felt obliged to offer some kind of mechanical model for the propagation of an electromagnetic wave through a perfect vacuum. So he imagined space filled with a mysterious substance he called the aether, which supported and contained the time-varying electric and magnetic fields - something like a throbbing but invisible Jell-O permeating the Universe. The quivering of the aether was the reason that light travelled through it - just as water waves propagate through water and sound waves through air.
But it had to be very odd stuff, this ether, very thin, ghostly, almost incorporeal. The Sun and the Moon, the planets and the stars had to pass through it without being slowed down, without noticing. And yet it had to be stiff enough to support all these waves propagating at prodigious speed.
The word 'aether' is still, in a desultory fashion, in use - in English mainly in the adjective ethereal, residing in the aether. It has some of the same connotations as the more modern 'spacy' or 'spaced out'. When, in the early days of radio, they would say 'On the air', the aether is what they had in mind. (The Russian phrase is quite literally 'on the aether', v efir.) But of course radio readily travels through a vacuum, one of Maxwell's main results. It doesn't need air to propagate. The presence of air is, if anything, an impediment.
The whole idea of light and matter moving through the aether was to lead in another forty years to Einstein's Special Theory of Relativity, E = mc2, and a great deal else. Relativity, and experiments leading up to it, showed conclusively that there is no aether supporting the propagation of electromagnetic waves, as Einstein writes in the extract from his famous paper that I reproduced in Chapter 2. The wave goes by itself. The changing electric field generates a magnetic field; the changing magnetic field generates an electric field. They hold each other up, by their bootstraps.
Many physicists were deeply troubled by the demise of the 'luminiferous' ether. They had needed some mechanical model to make the whole notion of the propagation of light in a vacuum reasonable, plausible, understandable. But this is a crutch, a symptom of our difficulties in reconnoitring realms in which common sense no longer serves. The physicist Richard Feynman described it this way:
Today, we understand better that what counts are the equations themselves and not the model used to get them. We may only question whether the equations are true or false. This is answered by doing experiments, and untold numbers of experiments have confirmed Maxwell's equations. If we take away the scaffolding he used to build it, we find that Maxwell's beautiful edifice stands on its own.
But what are these time-varying electric and magnetic fields permeating all of space? What do Ė and Ḃ mean? We feel so much more comfortable with the idea of things touching and jiggling, pushing and pulling, rather than 'fields' magically moving objects at a distance, or mere mathematical abstractions. But, as Feynman pointed out, our sense that at least in everyday life we can rely on solid, sensible physical contact to explain, say, why the butter knife comes to you when you pick it up, is a misconception. What does it mean to have physical contact? What exactly is happening when you pick up a knife, or push a swing, or make a wave in a waterbed by pressing down on it periodically? When we investigate deeply, we find that there is no physical contact. Instead, the electrical charges on your hand are influencing the electrical charges on the knife or swing or waterbed, and vice versa. Despite everyday experience and common sense, even here, there is only the interaction of electric fields. Nothing is touching anything.
No physicist started out impatient with common-sense notions, eager to replace them with some mathematical abstraction that could be understood only by rarified theoretical physics. Instead, they began, as we all do, with comfortable, standard, common-sense notions. The trouble is that Nature does not comply. If we no longer insist on our notions of how Nature ought to behave, but instead stand before Nature with an open and receptive mind, we find that common sense often doesn't work. Why not? Because our notions, both hereditary and learned, of how Nature works were forged in the millions of years our ancestors were hunters and gatherers. In this case common sense is a faithless guide because no hunter-gatherer's life ever depended on understanding time-variable electric and magnetic fields. There were no evolutionary penalties for ignorance of Maxwell's equations. In our time it's different.
Maxwell's equations show that a rapidly varying electric field (making Ė large) ought to generate electromagnetic waves. In 1888 the German physicist Heinrich Hertz did the experiment and found that he had generated a new kind of radiation, radio waves. Seven years later, British scientists in Cambridge transmitted radio signals over a distance of a kilometre. By 1901, Guglielmo Marconi of Italy was using radio waves to communicate across the Atlantic Ocean.
The linking-up of the modern world economically, culturally and politically by broadcast towers, microwave relays and communication satellites traces directly back to Maxwell's judgement to include the displacement current in his vacuum equations. So does television, which imperfectly instructs and entertains us; radar, which may have been the decisive element in the Battle of Britain and in the Nazi defeat in World War Two (which I like to think of as 'Dafty', the boy who didn't fit in, reaching into the future and saving the descendants of his tormentors); the control and navigation of airplanes, ships and spacecraft; radio astronomy and the search for extraterrestrial intelligence; and significant aspects of the electrical power and microelectronics industries.
What's more, Faraday's and Maxwell's notion of fields has been enormously influential in understanding the atomic nucleus, quantum mechanics, and the fine structure of matter. His unification of electricity, magnetism and light into one coherent mathematical whole is the inspiration for subsequent attempts - some successful, some still in their rudimentary stages - to unify all aspects of the physical world, including gravity and nuclear forces, into one grand theory. Maxwell may fairly be said to have ushered in the age of modern physics.
Our current view of the silent world of Maxwell's varying electric and magnetic vectors is described by Richard Feynman in these words: Try to imagine what the electric and magnetic fields look like at present in the space of this lecture room. First of all, there is a steady magnetic field; it comes from the currents in the interior of the earth - that is, the earth's steady magnetic field. Then there are some irregular, nearly static electric fields produced perhaps by electric charges generated by friction as various people move about in their chairs and rub their coat sleeves against the chair arms. Then there are other magnetic fields produced by oscillating currents in the electrical wiring - fields which vary at a frequency of 60 cycles per second, in synchronism with the generator at Boulder Dam. But more interesting are the electric and magnetic fields varying at much higher frequencies. For instance, as light travels from window to floor and wall to wall, there are little wiggles of the electric and magnetic fields moving along at 186,000 miles per second. Then there are also infrared waves travelling from the warm foreheads to the cold blackboard. And we have forgotten the ultraviolet light, the X-rays, and the radiowaves travelling through the room.
Flying across the room are electromagnetic waves which carry music of a jazz band. There are waves modulated by a series of impulses representing pictures of events going on in other parts of the world, or of imaginary aspirins dissolving in imaginary stomachs. To demonstrate the reality of these waves it is only necessary to turn on electronic equipment that converts these waves into pictures and sounds.
If we go into further detail to analyze even the smallest wiggles, there are tiny electromagnetic waves that have come into the room from enormous distances. There are now tiny oscillations of the electric field, whose crests are separated by a distance of one foot, that have come from millions of miles away, transmitted to the earth from the Mariner [2] space craft which has just passed Venus. Its signals carry summaries of information it has picked up about the planets (information obtained from electromagnetic waves that travelled from the planet to the space craft).
There are very tiny wiggles of the electric and magnetic fields that are waves which originated billions of light years away - from galaxies in the remotest corners of the universe. That this is true has been found by 'filling the room with wires' - by building antennas as large as this room. Such radiowaves have been detected from places in space beyond the range of the greatest optical telescopes. Even they, the optical telescopes, are simply gatherers of electromagnetic waves. What we call the stars are only inferences, inferences drawn from the only physical reality we have yet gotten from them - from a careful study of the unendingly complex undulations of the electric and magnetic fields reaching us on earth.
There is, of course, more: the fields produced by lightning miles away, the fields of the charged cosmic ray particles as they zip through the room, and more, and more. What a complicated thing is the electric field in the space around you!
If Queen Victoria had ever called an urgent meeting of her counsellors, and ordered them to invent the equivalent of radio and television, it is unlikely that any of them would have imagined the path to lead through the experiments of Ampere, Biot, Oersted and Faraday, four equations of vector calculus, and the judgement to preserve the displacement current in a vacuum. They would, I think, have gotten nowhere. Meanwhile, on his own, driven only by curiosity, costing the government almost nothing, himself unaware that he was laying the ground for the Westminster Project, 'Dafty' was scribbling away. It's doubtful whether the self-effacing, unsociable Mr Maxwell would even have been thought of to perform such a study. If he had, probably the government would have been telling him what to think about and what not, impeding rather than inducing his great discovery.
Late in life, Maxwell did have one interview with Queen Victoria. He worried about it beforehand - essentially about his ability to communicate science to a non-expert - but the Queen was distracted and the interview was short. Like the four other greatest British scientists of recent history, Michael Faraday, Charles Darwin, P.A.M. Dirac and Francis Crick, Maxwell was never knighted (although Lyell, Kelvin, J.J. Thomson, Rutherford, Eddington and Hoyle in the next tier were). In Maxwell's case, there was not even the excuse that he might hold opinions at variance with the Church of England: he was an absolutely conventional Christian for his time, more devout than most. Maybe it was his nerdishness.
The communications media - the instruments of education and entertainment that James Clerk Maxwell made possible - have never, so far as I know, offered even a mini-series on the life and thought of their benefactor and founder. By contrast, think of how difficult it is to grow up in America without television teaching you about, say, the life and times of Davy Crockett or Billy the Kid or Al Capone.
Maxwell married young, but the bond seems to have been passionless as well as childless. His excitement was reserved for science. This founder of the modern age died in 1879 at the age of 47. While he is almost forgotten in popular culture, radar astronomers who map other worlds have remembered: the greatest mountain range on Venus, discovered by sending radio waves from Earth, bouncing them off Venus, and detecting the faint echoes, is named after him.
Less than a century after Maxwell's prediction of radio waves, the first quest was initiated for signals from possible civilizations on planets of other stars. Since then there have been a number of searches, some of which I referred to earlier, for the time-varying electric and magnetic fields crossing the vast interstellar distances from possible other intelligences - biologically very different from us - who had also benefited sometime in their histories from the insights of local counterparts of James Clerk Maxwell.
In October 1992, in the Mojave Desert, and in a Puerto Rican karst valley, we initiated by far the most promising, powerful and comprehensive search for extraterrestrial intelligence (SETI). For the first time NASA would organize and operate the programme. The entire sky would be examined over a ten-year period with unprecedented sensitivity and frequency range. If, on a planet of any of the 400 billion other stars that make up the Milky Way galaxy, anyone had been sending us a radio message, we might have had a pretty fair chance of hearing them.
Just one year later, Congress pulled the plug. SETI was not of pressing importance; its interest was limited; it was too expensive. But every civilization in human history has devoted some of its resources to investigating deep questions about the Universe, and it's hard to think of a deeper one than whether we are alone. Even if we never decrypted the message contents, the receipt of such a signal would transform our view of the Universe and ourselves. And if we could understand the message from an advanced technical civilization, the practical benefits might be unprecedented. Far from being narrowly based, the SETI programme, strongly supported by the scientific community, is also embedded in popular culture. The fascination with this enterprise is broad and enduring, and for very good reason. And far from being too expensive, the programme would have cost about one attack helicopter per year.
I wonder why those members of Congress concerned about price tags don't devote greater attention to the Department of Defense, which, with the Soviet Union gone and the Cold War over, still spends, when all costs are tallied, well over $300 billion a year. (And elsewhere in government there are many programmes that amount to welfare for the well-to-do.) Perhaps our descendants will look back on our time and marvel at us, possessed of the technology to detect other beings, but closing our ears because we insisted on spending the national wealth to protect us from an enemy that no longer exists.*
[* The SETI programme was briefly resurrected, using $7 million in private contributions, in 1995 under the appropriate name Project Phoenix.] David Goodstein, a physicist at Cal Tech, notes that science has been growing nearly exponentially for centuries and that it cannot continue such growth, because then everybody on the planet, would have to be a scientist, and then the growth would have to stop. He speculates that for this reason, and not because of any fundamental disaffection from science, the growth in funding of science has slowed measurably in the last few decades.
Nevertheless, I'm worried about how research funds are distributed. I'm worried that cancelling government funds for SETI is part of a trend. The government has been pressuring the National Science Foundation to move away from basic scientific research and to support technology, engineering, applications. Congress is suggesting doing away with the US Geological Survey, and slashing support for study of the Earth's fragile environment. NASA support for research and analysis of data already obtained is increasingly constrained. Many young scientists are not only unable to find grants to support their research; they are unable to find jobs.
Industrial research and development funded by American companies has slowed across the board in recent years. Government funding for research and development has declined in the same period. (Only military research and development increased in the decade of the 1980s.) In annual expenditures, Japan is now the world's leading investor in civilian research and development. In such fields as computers, telecommunications equipment, aerospace, machine tools, robotics, and scientific precision equipment, the US share of global exports has been declining, while the Japanese share has been increasing. In that same period the United States lost its lead to Japan in most semiconductor technologies. It experiences severe declines in market share in colour TVs, VCRs, phonographs, telephone sets and machine tools.
Basic research is where scientists are free to pursue their curiosity and interrogate Nature, not with any short-term practical end in view, but to seek knowledge for its own sake. Scientists of course have a vested interest in basic research. It's what they like to do, in many cases why they became scientists in the first place. But it is in society's interest to support such research. This is how the major discoveries that benefit humanity are largely made. Whether a few grand and ambitious scientific projects are a better investment than a larger number of small programmes is a worthwhile question.
We are rarely smart enough to set about on purpose making the discoveries that will drive our economy and safeguard our lives. Often, we lack the fundamental research. Instead, we pursue a broad range of investigations of Nature, and applications we never dreamed of emerge. Not always, of course. But often enough.
Giving money to someone like Maxwell might have seemed the most absurd encouragement of mere 'curiosity-driven' science, and an imprudent judgement for practical legislators. Why grant money now, so nerdish scientists talking incomprehensible gibberish can indulge their hobbies, when there are urgent unmet national needs? From this point of view it's easy to understand the contention that science is just another lobby, another pressure group anxious to keep the grant money rolling in so the scientists don't ever have to do a hard day's work or meet a payroll.
Maxwell wasn't thinking of radio, radar and television when he first scratched out the fundamental equations of electromagnet-ism; Newton wasn't dreaming of space flight or communications satellites when he first understood the motion of the Moon; Roentgen wasn't contemplating medical diagnosis when he investigated a penetrating radiation so mysterious he called it 'X-rays'; Curie wasn't thinking of cancer therapy when she painstakingly extracted minute amounts of radium from tons of pitchblende; Fleming wasn't planning on saving the lives of millions with antibiotics when he noticed a circle free of bacteria around a growth of mould; Watson and Crick weren't imagining the cure of genetic diseases when they puzzled over the X-ray diffractometry of DNA; Rowland and Molina weren't planning to implicate CFCs in ozone depletion when they began studying the role of halogens in stratospheric photochemistry.
Members of Congress and other political leaders have from time to time found it irresistible to poke fun at seemingly obscure scientific research proposals that the government is asked to fund. Even as bright a senator as William Proxmire, a Harvard graduate, was given to making episodic 'Golden Fleece' awards, many commemorating ostensibly useless scientific projects including SETI. I imagine the same spirit in previous governments - a Mr Fleming wishes to study bugs in smelly cheese; a Polish woman wishes to sift through tons of Central African ore to find minute quantities of a substance she says will glow in the dark; a Mr Kepler wants to hear the songs the planets sing.
These discoveries and a multitude of others that grace and characterize our time, to some of which our very lives are beholden, were made ultimately by scientists given the opportunity to explore what in their opinion, under the scrutiny of their peers, were basic questions in Nature. Industrial applications, in which Japan in the last two decades has done so well, are excellent. But applications of what? Fundamental research, research into the heart of Nature, is the means by which we acquire the new knowledge that gets applied.
Scientists have an obligation, especially when asking for big money, to explain with great clarity and honesty what they're after. The Superconducting Supercollider (SSC) would have been the preeminent instrument on the planet for probing the fine structure of matter and the nature of the early Universe. Its price tag was $10 to $15 billion. It was cancelled by Congress in 1993 after about $2 billion had been spent - a worst of both worlds outcome. But this debate was not, I think, mainly about declining interest in the support of science. Few in Congress understood what modem high energy accelerators are for. They are not for weapons. They have no practical applications. They are for something that is, worrisomely from the point of view of many, called 'the theory of everything'. Explanations that involve entities called quarks, charm, flavour, colour, etc. sound as if physicists are being cute. The whole thing has an aura, in the view of at least some Congresspeople I've talked to, of 'nerds gone wild' - which I suppose is an uncharitable way of describing curiosity-based science. No one asked to pay for this had the foggiest idea of what a Higgs boson is. I've read some of the material intended to justify the SSC. At the very end, some of it wasn't too bad, but there was nothing that really addressed what the project was about on a level accessible to bright but sceptical non-physicists. If physicists are asking for $10 or $15 billion to build a machine that has no practical value, at the very least they should make an extremely serious effort, with dazzling graphics, metaphors and capable use of the English language, to justify their proposal. More than financial mismanagement, budgetary constraints and political incompetence, I think this is the key to the failure of the SSC.
There is a growing free-market view of human knowledge, according to which basic research should compete without government support with all the other institutions and claimants in society. If they couldn't have relied on government support, and had to compete in the free-market economy of their day, it's unlikely that any of the scientists on my list would have been able to do their groundbreaking research. And the cost of basic research is substantially greater than it was in Maxwell's day -both theoretical and, especially, experimental.
But that aside, would free-market forces be adequate to support basic research? Only about ten per cent of meritorious research proposals in medicine are funded today. More money is spent on quack medicine than on all of medical research. What would it be like if government opted out of medical research?
A necessary aspect of basic research is that its applications lie in the future, sometimes decades or even centuries ahead. What's more, no one knows which aspects of basic research will have practical value and which will not. If scientists cannot make such predictions, is it likely that politicians or industrialists can? If free-market forces are focused only towards short-term profit - as they certainly mainly are in an America with steep declines in corporate research - is not this solution tantamount to abandoning basic research?
Cutting off fundamental, curiosity-driven science is like eating the seed corn. We may have a little more to eat next winter, but what will we plant so we and our children will have enough to get through the winters to come?
Of course there are many pressing problems facing our nation and our species. But reducing basic scientific research is not the way to solve them. Scientists do not constitute a voting bloc. They have no effective lobby. However, much of their work is in everybody's interest. Backing off from fundamental research constitutes a failure of nerve, of imagination and of that vision thing that we still don't seem to have a handle on. It might strike one of those hypothetical extraterrestrials that we were planning not to have a future.
Of course we need literacy, education, jobs, adequate medical care and defence, protection of the environment, security in our old age, a balanced budget, and a host of other matters. But we are a rich society. Can't we also nurture the Maxwells of our time? To take one symbolic example, is it really true that we can't afford one attack helicopter's worth of seed corn to listen to the stars?

martes, 20 de agosto de 2019

Comparison of results of the Ewald Particle Grid method with the standard Ewald method.


Comparison of results of the Ewald Particle Grid method with the standard Ewald method.

Project proposal PAPIIT, DGAPA, or CONACYT 2003.



MS. Roberto Pérez


INTRODUCTION

In a typical MD simulation, about 95% of the CPU time, it is spent examining the complete set of N (N-1) / 2 pairs, identifying these pairs separated by at least the cutting distance rc, and computing the internal forces for this subset. The interactions of remaining pairs do not influence the dynamics of the system. For each molecule the set of neighbors within the cutting distance changes over time; The problem of identifying and rejecting the others is CPU time consumed. Any other increase in speed can reduce the time spent currently evaluating peer forces within the cutting radius.

Multi-step temporary method
(MULTIPLE TIME STEP METHOD)


The multi-step method (multi-step method), was designed for this, the neighborhood of an atom i is split into 2 groups, which are the primary neighbors, which rest at a distance rp from the atom i, and the secondary neighbors , which are at a distance between rp and rc of i. In this way, the total fixed force at i is separated into a primary fip component, due to the closest neighbors and a second fis component due to the furthest neighbors. Typically for a Lennard Jones fluid, rp will be chosen to be between  and 1.5.

Where is the length parameter in a potential pair. 

Comparación de resultados de el método de Rejilla de Partículas de Ewald con el método de Ewald estándar.

Propuesta de proyecto PAPIIT, DGAPA, ó CONACYT 2003.



MS. Roberto Pérez












INTRODUCCIÓN

En una simulación típica de MD, cerca del 95% del tiempo de CPU, se gasta en examinar el conjunto completo de N(N-1)/2 pares, identificando estos pares separados por lo menos la distancia de corte rc, y computando las fuerzas internas para este subconjunto. Las interacciones de pares remanentes no influyen en la dinámica del sistema. Para cada molécula el conjunto de vecinos dentro de la distancia de corte cambia con el tiempo; el problema de identificar y rechazar las otras es tiempo de CPU consumido. Cualquier otro incremento en la velocidad puede alcanzarse reduciendo el tiempo gastado actualmente evaluando fuerzas entre pares dentro del radio de corte.

Método de múltiples pasos temporales
(MULTIPLE TIME STEP METHOD)


El método de múltiples pasos  (multiple time step method), fue diseñado para esto,  la vecindad de un átomo i es partida en 2 grupos, los cuales son los vecinos primarios, que descansan a una distancia rp del átomo i, y los vecinos secundarios, los cuales se encuentran a una distancia entre rp y rc de i. De esta manera, la fuerza total fi en i es separada en una componente primaria fip, debida a los vecinos más cercanos  y una segunda componente fis debido a los vecinos más lejanos. Típicamente para un fluido de Lennard Jones, rp se escogerá para encontrarse entre s  y  1.5s .

Donde s  es el parámetro de longitud en un potencial par.

El movimiento del átomo i es dominado por un cambio rápido de la fuerza primaria resultado de las colisiones con el vecino más cercano. La fuerza secundaria es más pequeña y cambia lentamente con el tiempo. El método de múltiples pasos obtiene ventajas de esta separación, y tiene el fin de calcular las interacciones pares secundarias de manera menos frecuente que las primarias que son más importantes.

El método es mejor si se usa conjuntamente con el predictor de Grear el cual es un algoritmo corrector de distancias. Al tiempo t las fuerzas primarias y secundarias en cada átomo i son calculadas de manera usual. Al mismo tiempo se calculan las derivadas temporales de las  fuerza secundarias, ,
hasta el orden m, son calculadas, y una lista de vecinos primarios es compilada.  A cada uno del siguiente tmts-1 pasos, la fuerza primaria es calculada explícitamente. La fuerza secundaria es estimada usando la serie de Taylor de orden m.

Siguiendo estos tmts pasos (uno que envuelve el cálculo completo de fuerzas y el segundo las derivadas temporales, y tmts-1 usando la ecuación anterior para las fuerzas secundarias), el proceso es repetido íntegramente, empezando con un nuevo cálculo de la lista primaria. Entonces el método usa  dos pasos temporales, dt para las interacciones primarias y  tmts dt para las secundarias.

Expresiones convenientes para las derivadas de fuerza pueden obtenerse en términos de las derivadas del tiempo de las posiciones de las partículas y las derivadas especiales del potencial. Por ejemplo,


Y las sumas son sobre todos los vecinos secundarios j. Estos cálculos toman la forma en el loop de fuerza; derivadas temporales como

Son calculadas de las derivadas temporales de posiciones atómicas disponibles para el predictor de escenario de algoritmos de Gear.

En el estudio de Streett et al, una serie de tercer orden de Taylor, con tmts =10, rc =  2.5 y rp = 1.1 fueron usados.  En la simulación de un fluido de Lennard Jones a una densidad reducida  r = 0.8, los números promedio de vecinos primarios y secundarios son del orden de 1.4 y 24, respectivamente.  El método de múltiples tiempos en este caso resultó en una mejora de velocidad de un factor entre 7 y 10 sobre un programa convencional de MD  que no tenían listas de vecinos y un factor de 3 a 5 sobre un programa usando la lista de vecinos de Verlet. Estos tiempos relativos dependerán de detalles precisos de operación llevados a cabo en el loop de fuerza, y será dependiente del tipo de máquina.

El método ha sido felizmente aplicado a fluidos moleculares [Swindoll y Haile 1984]. Se ha extendido al caso cuando hay 3 regiones dentro de la esfera de radio rc [Nicolas et al].  Una aplicación interesante de este método es en la simulación de fluidos con fuerzas de tres cuerpos [Haile 1978]. Para el potencial de Axilrod-Teller la contribución de tres cuerpos a la fuerza varía lentamente con el tiempo, y es posible tratar la componente entera de tres cuerpos como una interacción secundaria. De esta manera la sumatoria prohibitiva y cara sobre el triplete puede ser evitada, para la mayoría de los pasos en la simulación. Todavía existen dificultades para este método.


Primero.-  Hay un esfuerzo de programación considerable requerido para incorporar el código relevante en el programa convencional.

Segundo.- El método ha sido usado solamente con un tiempo dt más pequeño que el usado en simulaciones convencionales   (5 X 10 -15 s) en oposición al más usual  10 -14. Se encuentra que usar un tiempo más grande requiere un incremento en rp y un decremento en el intervalo tmts, entonces se compensaran algunas de las ventajas ganadas usando el método de múltiples pasos. La eficiencia de este método es claramente sensitiva a la elección de rp, sin embargo, la mejor manera de probar este método es fijar el tiempo deseado, decidir sobre un nivel de energía satisfactorio, y ajustar rp y el intervalo tmts, para maximizar la eficiencia de la simulación sujetas a estas variables. Para fluidos moleculares, donde un tiempo menor se requiere, estas dificultades se hacen menos evidentes.

COMO LIDIAR CON FUERZAS DE LARGO ALCANCE

En las secciones previas, se ha discutido el núcleo del programa cuando las fuerzas involucradas son de corto alcance. En esta sección cambiamos nuestra atención para lidiar con fuerzas de largo alcance. Una fuerza de largo alcance es frecuentemente definida como una fuerza en la cual la interacción espacial decae no más rápido que r-d donde d es la dimensionalidad del sistema. En  esta categoría se encuentran las interacciones carga-carga  entre los iones (Vzz(r)~ r -1) y la interacción dipolo-dipolo, la cual es básicamente una interacción molecular (Vmm(r)~ r -3). Estas fuerzas son un serio problema para las simulaciones computacionales, debido a que su rango es mayor que la mitad de la longitud de la caja para las simulaciones típicas de 500 moléculas. La solución por medio de fuerza bruta para este problema podría incrementar el tamaño de la caja central L en cientos de nanómetros, lo cual generaría que el apantallamiento de los vecinos disminuya el rango efectivo de los potenciales.

Incluso con computadoras modernas esta solución es impractica, debido a que el tiempo requerido para ejecutar dicha simulación es aproximadamente proporcional a N2, i.e. L6.

¿Cómo solucionar este tipo de problemas?

El problema es particularmente agudo para Vzz(r).De esta manera el corte esférico del potencial puede desecharse. La esfera resultante alrededor de un ión dado puede desecharse. La esfera resultante alrededor de un ión dado puede cargarse, debido a que el número de aniones y cationes no necesitan balancearse a cualquier instante.  La tendencia de los iones a migrar hacia atrás y adelante sobre la superficie esférica creará efectos artificiales en  r = rc. Esto puede contabilizarse distribuyendo una nueva carga sobre la superficie de la esfera, igual en magnitud y opuesta en signo a la carga neta de la esfera, y esto se efectúa para garantizar la electroneutralidad local. Esto es como cambiar el potencial, Adams [1983b] mostró que los resultados de esta aproximación son sistemas dependientes del tamaño, pero para sistemas de 512 iones los resultados están en buen acuerdo con los resultados obtenidos de las sumas de Ewald.

En contraste, el método básico de mínima imagen correspondiente a un corte de potencial en la superficie del cubo rodeado del ión en cuestión. Este cubo será eléctricamente neutro. De cualquier manera, la desventaja es que iones cargados similares tenderán a ocupar posiciones en las esquinas opuestas del cubo: la estructura de imágenes periódicas será impuesta directamente en lo que debería de ser un líquido isotrópico, y esto tiene como resultado una distorsión en la estructura del líquido. Este efecto podría reducirse en las cajas no cúbicas. De manera similar, y con efectos menos importantes se obtiene al efectuar el corte esférico o de mínima imagen a un sistema polar (Vmm(r)).

En esta trabajo nos concentraremos en 2 métodos que pueden usarse para resolver el problema de las fuerzas de largo alcance. El método de las sumas de Ewald, incluye la interacción de un ión o molécula con todas sus imágenes periódicas. Este método tenderá a enfatizar la naturaleza periódica del fluido modelo.








MÉTODO DE PARTICLE MESH EWALD(PPME) vs. Método Estándar de Ewald[8]

Este método tenderá a enfatizar la naturaleza  continua de un fluido polar y requiere una estimación a priori de la permitividad relativa. Ambos métodos usan las ideas bien conocida de la teoría electrostática. En particular, una distribución de carga con una cavidad esférica polariza el medio que lo rodea. Esta polarización que depende de la permitividad relativa del medio, tiene un efecto en la distribución de carga en la cavidad.

En la siguiente parte del desarrollo de este proyecto se explica como lidiar con las fuerzas electrostáticas de largo alcance y cómo combinar el método de SPME con el multinivel de integración de las ecuaciones de movimiento a fin de obtener algoritmos de simulación eficientes.

La evaluación de interacciones de largo alcance entre partículas es extremadamente  cara si usamos métodos convencionales (es decir las sumas de Ewald), su costo computacional escala típicamente como N2 (siendo N el número de partículas), rápidamente excede cualquier límite razonable. Como se comentó anteriormente, las interacciones de largo alcance  es uno de los mayores problemas con los que hay que lidiar en las simulaciones atomísticas de Dinámica  Molecular. El problema es particularmente agudo en el caso de biopolímeros donde la presencia de carga neta distribuida hace que el potencial local oscile. La naturaleza condicionalmente convergente de las series de energía electrostática para un sistema periódico  tal como la caja de Dinámica Molecular en condiciones de contorno periódicas (Periodic Boundary Conditions) hace de cualquier método posterior de corte es esencialmente incosteable.

El campo de reacción es en principio un método físico que corrige las contabilidades de los efectos de largo alcance y requiere solamente un limitado esfuerzo computacional. La técnica asume interacciones electrostáticas explícitas con una esfera limitada y rodeada por un medio dieléctrico el cual ejerce sobre la esfera una polarización o un  “campo de reacción”. El medio dieléctrico tiene una constante dieléctrica, que es el mismo que el de la esfera interna. El método se ha probado y ha dado resultados idénticos a los obtenidos con el método exacto de Ewald  en una simulación de Monte Carlo de esferocilindros bipolares donde la constante dieléctrica que introdujeron en el campo de reacción es cambiada periódicamente de acuerdo al valor encontrado en la esfera. El método del campo de reacción de cualquier manera sufre de dos grandes desventajas que limita fuertemente su uso.

Hemos descrito como obtener interacciones de largo alcance, ahora en la separación de la escala de tiempo de los potenciales modelo de sistemas moleculares complejos. Adicionalmente, proveemos un subdivisión del potencial general aplicándolo a sistemas biológicos, al igual que en otros sistemas químicos interesantes incluyendo cristales líquidos. Este tipo de sistema son típicamente caracterizados por una gran flexibilidad y por fuertes interacciones Coulombianas intermoleculares. Esquemáticamente, podemos escribir el potencial V debido a dos contribuciones:

                        V = Vbnd + Vnbn

Aquí, la parte ligada o intramolecular Vbnd es rápida y es responsable de la flexibilidad del sistema. La parte no ligada o intermolecular (intergrupal) Vnbn es dominada por las interacciones coulombianas. El animo de la sección siguiente es describir un protocolo general para las subdivisiones de tales formas de potencial de interacción y para mostrar como obtener razonablemente eficientes y transferibles integradores de múltiples pasos temporales válidos para cualquier sistema molecular complejo.

Por el momento nos vamos a concentrar en el potencial Vnbn , que se encarga de determinar y modelar las interacciones coulombianas, este potencial se determina de manera exacta con las sumas de Ewald.

El método de Ewald de partículas en redes suaves

Antes de discutir  la separación de múltiples pasos temporales  no ligadas, es útil describir con detalle una de las técnicas más avanzadas para tratar las fuerzas de largo alcance. En lugar de esto, este tipo de fuerzas no ligadas son las más engorrosas para lidiar y merecen un escrutinio detallado.

En la literatura reciente, pocas técnicas están disponibles para tratar  este problema de las interacciones de largo alcance  en las simulaciones computacionales de partículas cargadas a diferente nivel de aproximación [1, 2, 3]. El método de Ewald provee el resultado exacto para la energía electrostática de un sistema periódico consistente de cajas infinitas replicadas  neutras de partículas cargadas. El método es la elección natural en simulaciones de Dinámica Molecular de sistemas moleculares complejos.

El potencial de Ewald [2] está dado por :


                (1)


       (2)


con

                           (3)

                             (4)

donde, ri es el vector de posición de la carga atómica qi, rij = ri-rj, rn es el vector de la caja,  

es la función de error complementario, erf(x) = 1erfc(x), V el volumen de la celda unitaria, m el vector de la caja reciproca y  a es el parámetro de convergencia de Ewald. En la parte de la distancia directa a la caja,  la ecuación (1),  se omiten los contactos  intramoleculares. En adición, en la Ec. (2) el término Vintra sustrae, en el espacio directo, la energía intra-molecular entre pares ligados, la cual es incluida automáticamente en la parte derecha  de esta ecuación.  En consecuencia, la sumatoria sobre i y j en la Ec. (4) es sobre todos los contactos intramoleculares excluidos. Debemos remarcar que en el potencial de Ewald [7] hemos asumido las condiciones de frontera también llamadas condiciones de contorno laminar:

1)La esfera de Ewald [7] es inmersa en un medio perfectamente conductor y entonces el termino bipolar en la superficie de la esfera de Ewald es cero. Para sistemas grandes el costo computacional de la sumatoria estándar de Ewald, la cual se incrementa como N2, se vuelve muy larga para aplicaciones practicas. Algoritmos alternativos los cuales se incrementan en una potencia menor que N que el estándar de Ewald se han propuesto en el pasado. Entre los algoritmos más rápidos diseñados para sistemas periódicos es el algoritmo de mallas de partículas de Ewald (PME, Particle Mesh Ewald) [4,5], inspirados por el método de malla de partículas desarrollado por Hockney y Eastwood [6]. Aquí una aproximación multidimensional de  piezas discretas interpoladas son usadas para computar la energía de la caja reciproca, Vqr, de la ecuación (2), donde la parte directa, Vqd, es computada directamente. El bajo costo computacional de el método de PME permite la elección de grandes valores del parámetro de convergencia de Ewald a, comparado con el usado en el método convencional de Ewald.  Correspondientemente, menores distancias de corte en el espacio directo de Ewald son adoptados. Si uj es la coordenada de escala fraccional de la i-ésima partícula, el factor de estructura S(m) en la Ec. (3), puede ser reescrito como :

                   (5)

Donde, N es el número de partículas, K1, K2, K3  y  m1, m2, m3 son enteros. La componente a de la coordenada de escala  fraccional del i-ésimo átomo puede escribirse como:

                 (6)

donde , = 1,2,3 son los vectores base de las cajas recíprocas.
S(m) en la Ec. (5) puede mirarse como una transformada discreta de Fourier (FT) de un conjunto de cargas  puestas irregularmente dentro de la celda unitaria. Se han desarrollado técnicas en el pasado para aproximar S(m) con expresiones que involucren la transformada de Fourier de una malla regular de puntos. Tales aproximaciones de factor de estructura son ventajosas desde el punto de vista computacional debido a que estas pueden evaluarse de las transformadas de Fourier rápidas (FFT). Todas estas FFT, son en cierto sentido aproximaciones para juntar las cargas sobre una malla cercana de puntos para producir una distribución  regular de carga.  El método de PME soluciona este problema por interpolación. Entonces, las exponenciales complejas


Computadas en la posición de la i-ésima carga en la Ec. (5), son rescritas como una suma de coeficientes interpolados multiplicados por sus valores en los puntos cercanos a la reja. En la versión suave de PME (SMPE) [5], la suma es multiplicada por un factor apropiado, llamado:

       (7)

donde n es el orden de la interpolación ,   define los coeficientes de la coordenada de interpolación en la coordenada de escala . En la Ec. (7) la suma sobre k, representan los puntos de la reja, es solo sobre un rango finito de enteros, debido a que las funciones son cero fuera del intervalo 0 £ u £ n. Debe hacerse hincapié en que los coeficientes complejos b(mi) son independientes de las coordenadas de las cargas ui y se necesitan computar sólo al inicio de la simulación. Una derivación detallada de las funciones y de los coeficientes ba son dados en la referencia [33]. Insertando la ecuación (7) en (5), S(m) puede rescribirse como:
            (8)

donde  es la transformada de Fourier discreta (FT) en el punto de la reja del arreglo  con 1£ ki £ Ki, i = 1,2,3. El arreglo de carga enrejada , es definida como:
     (9)

Insertando el factor de estructura de la Ec. (2) y usando el hecho de que , la energía de la caja reciproca SPME puede escribirse como:
 
=                                                                                                         (10)

con:

                   (11)                                                               (12)
                                                                                       (13)

Usando el teorema de convolución para FFT la energía (10) puede rescribirse como:


  (14)


Ahora usamos la identidad  

           para llegar a

          (15)

Primero notamos que  no depende de las posiciones de las cargas y que  es diferenciable para n >2 (la cual es siempre el caso en aplicaciones practicas). Entonces la fuerza sobre cada carga puede obtenerse tomando la derivada de la Ec. (15), llamado

 =
                                                                                                (16)

En la práctica, el cálculo  es llevado a cabo de acuerdo al siguiente esquema:
i)                  A cada paso de la simulación se calcula las coordenadas fraccionales de la  reja  y llena un arreglo con Q de acuerdo a la Ec. (9). En este momento, las derivadas de las funciones Mn son también calculadas y puestas en la memoria.
ii)               El arreglo que contiene a Q es entonces sobrescrito con  , i.e. la transformada tridimensional de Fourier de las Q’s.
iii) Subsecuentemente, la energía electrostática es calculada vía  la   ecuación (10). Al mismo tiempo, el arreglo conteniendo es sobrescrito por el producto de el mismo con el arreglo conteniendo BC (calculado al inicio de la corrida).

iv)            El arreglo resultante es entonces la transformada de Fourier para obtener la convolución .
v) Finalmente, las fuerzas son calculadas vía la ecuación (16) usando las derivadas calculadas con anterioridad de las funciones Mn, para formar .

La memoria requerida para el método de SPME es limitada. Las variables  de doble precisión son necesarios para el arreglo de la reja de carga Q, cuando los cálculos de las funciones y sus derivadas requieren solamente 6 X n X N  números de doble precisión. Los enteros   determinan la finura de la reja a lo  largo del vector a-ésimo de la celda unitaria. La precisión de salida  de la energía y fuerzas dependen de los parámetros del SPME:

-El parámetro de convergencia a, el espaciado de la reja y el orden

n de la interpolación. Para un valor típico a » 0.4A-1 una precisión

relativa  entre 10-4 y 10-5 para la energía electrostática son   

obtenidos cuando  el espaciado de la reja es alrededor de 1A  a lo

largo de cada eje, y el orden n de la interpolación B es 4 o 5.










La potencia del algoritmo SPME, comparado con la implementación estándar del método de Ewald, es muy grande. En la Fig 2. se reporta el tiempo de CPU obtenido en una computadora Pentium 3 para la evaluación de la energía de la caja y las fuerzas obtenidas vía SPME como una función del número de átomos en el sistema. Se usaron rutinas de dominio público de la rutina de 3-D FFT fueron usadas. El algoritmo es prácticamente lineal y para    12 000 partículas el SPME toma solamente 2 segundos del CPU para llevar a cabo los cálculos. Una simulación estándar de Ewald para una caja de 64 X 64 X 64 A3 (i.e. con una espaciado de reja en el espacio k de   » 0.01 A-1) para el mismo ejemplo y al mismo nivel de precisión podría tomar varios minutos.

















REFERENCIAS

[1]  P. Ewald. Ann. Phys., 64:253, 1921.

[2] S.W. deLeeuw, J.W. Perram, and E. R Smith. Proc. R. Soc. London A, 373:27, 1980

[3] J.P. Hansen. Molecular-dynamics simulation of coulomb systems in two and three dimensions. In Molecular Dynamics Simulation of Statistical-Mechanics Systems, Proceedings of the International School of Physics “Enrico Fermi”. North Holland Physics, 1986.

[4] T. Darden, D. York, and L. Pedersen. J. Chem. Phys., 98:10089, 1993.

[5] U. Essmann, L. Perera, M. L. Berkowitz, T. Darden, H. Lee, and L.G. Pedersen. J. Chem. Phys., 101:8577, 1995.

[6] R. W. Hockney. Computer Simulation Using Particles. McGraw-Hill, New York, 1989.

[7] M.P. Allan, D.J. Tildesley,  COMPUTER SIMULATIONS OF liquids, Oxford Science Publications.

[8] Massimo Marchiand Piero Procacci , Orac Manual and Guide, 
     A Molecular Dynamics Program to Simulate Complex Molecular Systems with Realistic Interactions.
















FORTRAN




PRINT A4, PARA CARACTERES 4 ES EL NUMERO DE CARACTERES
PRINT i1  , PARA ENTEROS INDICANDO EL NUMERO DE DIGITOS





















zen consultora

Blogger Widgets

Entrada destacada

Platzy y el payaso Freddy Vega, PLATZI APESTA, PLATZI NO SIRVE, PLATZI ES UNA ESTAFA, Platzy and the Clown Freddy Vega, PLATZI SUCKS, PLATZI IS A SCAM

  Platzy and the clowns Freddy Vega and Cvander – Part 1, PLATZI IS A SCAM Hello friends, this post will keep growing as I continue writing...