kerleytechnical.com

Several years ago I began writing a tutorial on applications of quantum mechanics to equation-of-state theory. I had expected that a discussion of the Born-Oppenheimer approximation (BOA) would be relatively straightforward, showing it to be the leading term in a systematic perturbation expansion.

However, I soon learned that the theoretical framework I planned to discuss actually didn’t exist. The problem of deriving the BOA and calculating corrections to it was still basically unsolved. I also discovered an enormous body of literature on the subject. As I read and studied these papers, I began to develop some new ideas of my own.

During the last six months I have taken the time to turn these ideas into a methodology that I believe to be substantially new and suitable for application to real problems in the chemistry and physics of molecules and condensed matter.

The method includes the standard BOA as a special case but treats two kinds of corrections using perturbation theory.
—The adiabatic corrections include all terms that do not explicitly involve the nuclear wavefunctions, so that the nuclei move on a single electronic potential surface.
—The non-adiabatic corrections, which allow the nuclei to move on more than one potential surface, include coupling between the electronic and nuclear wavefunctions.

This approach is discussed in a new report, KTS11-1, “On Corrections to the Born-Oppenheimer Approximation,” available on this website. It gives a simple example of the technique, as well as a review of previous work.

The report is currently marked “draft,” in case I choose to modify it after getting comments from reviewers.

- Jerry Kerley

P.S. The quantum mechanics tutorial still hasn’t been finished, although some of the ideas are discussed in my website presentations.

It has recently become fashionable to employ the term “high-fidelity” when referring to one’s scientific work. We now hear about high-fidelity calculations, simulations, even experiments, in shock wave physics and material modeling.

Here’s my opinion: “High-fidelity” is not a meaningful scientific term. It is a marketing term. It is useful for promoting your methods—while planting the idea that other methods are obsolete.

This is the kind of catchy expression that advertising wordsmiths use to sell everything from beer, to pharmaceuticals, to politicians, etc. And it works. People frequently buy things they don’t really need or want, things that don’t even work the way they are supposed to. (Just look at our elected officials.)

Maybe this is smart business, but it is not good science.

Let’s take a closer look. “Fidelity” means “faithfulness” or “faith.” It is considered a virtue. “Semper Fidelis” is the motto of the USMC (and a march by J. P. Sousa). It is a good thing. If you can convince others that your work has something to do with fidelity, you have already scored points in the marketplace.

But the term also has connections with technology. Hi-fi sound equipment is supposed to be faithful to an original performance, i.e., to record and replay it accurately. Used in this context, the term has acquired secondary meanings over the years, e.g., “high-quality” or “high-tech.” Salesmen of science can exploit these nuances when they apply the term to their own work.

The term “high-fidelity simulations” is also used in various training programs, e.g., the use of realistic mannequins and other props in medicine, emergency response, etc. Those kinds of simulations have nothing to do with the numerical simulations used in material modeling, but they carry positive associations that can be exploited.

The Oxford dictionary gives the following (secondary) definition of fidelity: “the degree of exactness with which something is copied or reproduced.” Just what are these so-called hi-fi codes and theories supposed to copy or reproduce?

Surely it is not enough to reproduce some set of experimental data, which would only qualify as an “accurate fitting method.” Fitting a Hugoniot curve isn’t hard to do, and it doesn’t ensure that a material model gives a good description of the EOS surface.

Okay, maybe a hi-fi code or model is supposedly “faithful” in the sense that it simulates the real properties of the material (with or without experimental data). Here’s the problem with that meaning: Claiming that your theory is accurate doesn’t make it so. It is not intellectually honest to make this claim unless you have the evidence to back it up. And picking the evidence that makes your theory look good doesn’t count. [1]

I am not at all opposed to using theoretical methods different from my own. But they should not be touted as “high-fidelity science.” Such claims are false, unproven, empty.

This term is just starting to get widespread use, but I’m sure that it will become more trendy and popular—until it becomes commonplace and loses its power to impress others. Then we will see another “paradigm shift,” and people will look elsewhere for sources of “synergy.”

For the record, I think the term “blue-chip science” is more appropriate for my own work.

- Jerry Kerley

[1] Proving that a theoretical model is accurate, sometimes called “validating” it, is very difficult, to say the least. See Sec. 2.4 of my report KTS09-1, available from my website.

In my 9-11-10 blog article, I said that a 1995 paper [Astrophys. J., Suppl. Ser. 99, 713 (1995)], had made an erroneous claim—that my D2 EOS had a plasma phase transition (PPT), that this error had been picked up and repeated in the literature by others.

One of the authors has pointed out that I, in fact, have misrepresented them. This posting is both a retraction and apology.

The authors correctly observed that my EOS and theirs both showed first-order phase transitions from the molecular to the metallic phases, at similar pressures. However, they also observed that the transitions arose from completely different mechanisms—a PPT to the liquid in their case, an ordinary solid-solid transition in my case. The relevant text, given in their paper, follows:

“This comparison shows that the LANL EOS agrees quite well with the
SC EOS, the PPT being revealed by the density discontinuity on each
of the lowest two isotherms.  Breaks in the lowest two LANL curves are
similarly due to a transition to a metallic solid hydrogen phase, although
the transition itself is not explicitly plotted here, unlike the SC(P) case.
This is qualitatively different from the PPT of SC, since the SC EOS
predicts that a metallic hydrogen is in a fluid state above the PPT.”
(See page 735 of the original article, just below Fig. 21b.)

My mistake is regrettable, but it doesn’t change the main point of my article—that people often make incorrect statements about my work because they quote others without reading my papers and reports for themselves. In particular, certain papers have referred to a PPT in my D2 EOS. But this 1995 paper cannot be blamed for that.

- Jerry Kerley

After completing my thesis and getting my Ph.D., I began looking for new research topics to pursue. My search ended when I went to Los Alamos to work on EOS modeling. EOS theory appealed to me for two reasons:
1. It was very challenging, with plenty of opportunities to do good science.
2. It was useful for practical problems. People might actually care about the work.

That is why I have spent 40 years of my professional life developing better EOS theories and models, applying my methods to practical problems, and writing software to make EOS models easy and efficient to use in hydrocodes.

It is hard to get supervisors and project managers to invest in the future, to support research that won’t have an immediate payoff but will be useful a few years down the road. So I’ve had to do it without their permission, bootlegging time from programmatic projects. That is why I can construct much better EOS now than I could years ago.

But an even greater frustration is that some people often refuse to use a good EOS even after it finally does become available. There is a long list of excuses I’ve heard for not using my EOS in the last 40 years. I thought it would be interesting to compile them and put them all in one place. Here’s my list.

Legacy issues
—We never used fancy EOS in the past; the old ones were good enough, and we’re not going to change now.
—Our codes and models are calibrated to the old EOS. If we switched to new ones, we would have to recalibrate everything.
—We really didn’t want to see this problem solved. Our group needs continued funding to study this problem in perpetuity.

Arguments from authority/majority
—We don’t need a better EOS because I say so, and I am your supervisor.
—We don’t need a better EOS because Dr. X says so, and he is my supervisor.
—We don’t need a better EOS because Dr. Y says so, and he is a world-class expert.
—We don’t need a better EOS because Dr. Z says so, and he is funding the project.
—We took a poll, and nobody thinks we need a better EOS except you.

Other things are more important
—Good EOS aren’t needed because the effects of material strength and fracture dominate everything anyway.
—Good EOS aren’t relevant because everything is out of equilibrium and your models don’t apply anyway.
—We are using coarse zoning, so EOS differences get smeared out anyway.
—We don’t have time to develop or use a good EOS because we have to run hundreds of hydrocode calculations in support of this important project.

I question your approach
—We only use EOS that were developed here, by our own home-grown theoreticians.
—Your models are too complicated and theoretical. I only trust EOS based on models that I can understand.
—Your models are too simplistic. Dr. W’s models must be better because they have more equations and more input parameters.
—Your models aren’t rigorous enough. Numerical calculations, like DFT/MD, are the only way to go.
—Your EOS doesn’t give a linear US-UP curve, so it can’t be right.
—I plotted your EOS, and it has unphysical oscillations, so it isn’t any good.
—I plotted the derivatives of your EOS, and they aren’t nice and smooth. So your EOS isn’t any good.

Miscellaneous
—I won’t use your EOS unless you first prove that it would make a difference. (And maybe not even then.)
—I tried out your EOS on another problem and it didn’t make any difference. That shows that EOS differences aren’t important.
—We think that the EOS are okay as long as the code runs the problem to completion and doesn’t put out (too many) error messages.
—Using your EOS makes the code spit out error messages, so it can’t be any good.
—Your EOS doesn’t give the same Hugoniot as the old one, so it must be wrong.
—Your EOS gives the same Hugoniot as the old one, so it can’t be all that different.

Let me know if I’ve omitted your favorite reason, and I will add it to this list.

- Jerry Kerley

P.S. I forgot one of the most important ones: We can’t afford to use tabular EOS because they take too much computing time. (See my 8-19 blog.)

Ignorance is an ugly thing, a sad thing. I suspect it is responsible for much of the misery that human beings inflict on one another.

It is not always fair to condemn someone for being ignorant. People can’t choose the conditions of their birth, what they are taught as children, etc. I have been blessed with many opportunities not available to others and want to be thankful, not proud.

However, I have little respect for those who are ignorant by choice. It seems that some people are not just ignorant; they are proud of it and determined to stay that way.

This attitude is often found among people in authority, e.g., managers and other testosterone addicts. These individuals would rather stay ignorant than admit they don’t know it all. They would rather stay ignorant than learn from people who they regard as inferior to them.

A colleague once told me that he disliked a certain person so much that he wouldn’t use his ideas no matter how good they were. This attitude has no place in science.

Let me put it this way: If you reject a person’s ideas because you don’t like him/her, you are a jerk. I don’t care how many publications you have, how many awards and honors you have received; you are still a jerk.

That’s all I have to say on this matter for now. I rather enjoyed getting this off my chest.

- Jerry Kerley

My last three blogs have discussed misconceptions and misinformation about my work on H2 and D2. This fourth and final article discusses my most recent work.

Most Recent Work (2003-04)

I have already noted that my 1972 EOS agreed quite well with experimental data obtained 25 years later. However, the new shock-wave and static measurements showed that the model could be significantly improved. I completely revised the model in 2003, retaining the basic approach but incorporating many improvements to the liquid model, the compressibility of the molecular solid at high pressures, and the treatment of dissociation and ionization. I also generated a table for H2 as well as D2. This work was documented in my report SAND2003-3613. I also used the new H2 EOS to calculate the structures of the planets Jupiter and Saturn, as described in my report KTS04-1.

This work is far too complicated to be summarized here. However, I will call attention to the following points:
—The anharmonicity and density corrections to the rotational and vibrational degrees of freedom (see my 9-14 blog) were shown to be very important in matching the Hugoniot in the dissociation region.
—The new mixture model, which assumed equal pressures and temperatures, gave better results than linear mixing (equal densities and temperatures).
—The zero-Kelvin curve for the molecular phase was softened to obtain better agreement with static compression data. The new EOS predicts a much higher molecular-metallic transition pressure, giving a much smaller radius for the metallic core in the Jovian planets.
—Corrections for nuclear spin statistics (ortho- and para-H2) had a significant effect on calculations of the planetary adiabats. Including these terms gives higher temperatures, an important issue for understanding the miscibility of H2 and He in the planetary interiors.
—Agreement with the Jupiter and Saturn moments required the assumption of a dispersed heavy element core, which deviates from “conventional wisdom” but is consistent with planetary accretion models.

As far as I know, my recent work has not (yet) generated any new misinformation in the literature. In fact, it seems to be under an information blackout, as previously happened with my 1972 model. (See my 9-14 blog.) Astrophysical researchers seem especially inclined to ignore my recent work, even though they continue to cite my older model. I suspect this strategy results from unwillingness to face the fact that my new EOS requires significant changes to older models of the Jovian planets—a smaller metallic core, a dispersed heavy element core, higher temperatures on the adiabat, etc.

I also want to comment on the current tendency to dismiss “chemical models” as naïve and simplistic, at least in comparison with “physical models.” The “chemical picture,” which is the basis of my EOS models, treats matter as a mixture of individual chemical species—molecules, atoms, ions, and free electrons. The “physical picture,” which is the basis of ab initio numerical methods, assumes only the existence of electrons and nuclei, relying on (presumably) exact calculations to define the material properties.

In recent years, some ab initio calculations have predicted that H2 molecules “lose their identity” at high pressures, suggesting that the chemical picture is unrealistic. I believe that these calculations and conclusions are erroneous.
—The most common numerical method, density-functional theory (DFT), underestimates the ionization energy (band-gap, if you prefer) by a factor of 2. As a result, DFT calculations predict that pressure ionization occurs at much too low a density. Molecules dissociate more easily when the atoms are pressure ionized, so the DFT calculations underestimate their stability.
—DFT calculations, which treat only the electronic structure, are normally paired up with classical molecular dynamics calculations of the nuclear motion. This method cannot be expected to give satisfactory results for the molecular vibrations, which are quantized. (Correction terms have been derived but have never been included in the calculations.)
—I refer you to Sec. 2 of my 2003 SAND report for a more detailed discussion of these issues.

It does not follow, from the above arguments, that all chemical models are good models. On the contrary, a good chemical model requires careful attention to the details. But it seems to me that a model that gives reasonable predictions of experiments 25 years in advance should be worth a second look. And an updated version of that model even more so.

The Livermore laser group recently presented data [Phys. Rev. B 79, 014112 (2009)] showing the D2 Hugoniot above 110 GPa to be softer than predicted by my model and by the ab initio calculations—though not as soft as claimed in their original work. The Sandia and Russian techniques have not yet reached pressures in this regime. But a subsequent Sandia paper [Phys. Rev. Lett. 103, 225501 (2009)] claimed that this discrepancy disappears when a better EOS for the quartz anvil is used to analyze the data.

To be continued

I say again, as I have said many times, that no EOS model—mine included—is ever the final answer; refinements to the theory are always possible and desirable. Sec. 11 of my SAND report discusses a number of issues that need further investigation.

- Jerry Kerley

My last two blogs have discussed misconceptions and misinformation about my early work (1969-72) on H2 and D2. This article begins discussion of later work.

Intermediate Period (1980-83)

I did not wish to spend my entire career working on H2 and D2, so I soon moved on to other problems—including the Pu EOS, the Sesame library, the Panda code, improvements to my liquid model, and the theory of electron correlation.

I revisited the H2-D2 problem in 1980, to examine the effects of changes in my liquid model and to compare with some new experimental data that had appeared in the literature after 1972. These studies were concerned entirely with the molecular fluid. Agreement with the new data was good, and the model changes did not have a major effect on the calculations for H2 and D2. There was no compelling reason to redo the model for dissociation and the metallic fluid, especially since there were still no data to test the model in that regime. Therefore, I did not put out a new EOS table at that time.

This work was discussed in a joint paper with a Los Alamos colleague [J. Chem. Phys. 73, 5264 (1980)], and later in a review article [Molecular-Based Study of Fluids, (ACS, Wash., D.C., 1983), pp 107-138]. The review article, which also discussed calculations for the rare gases, methane, and iron, did give a brief summary of the 1972 D2 EOS work.

Shortly after publication of the review article, someone began using it as the primary reference to “the Sesame EOS,” ignoring the existence of the earlier reports and papers. And those who copy references out of other papers, without reading them, soon began following suit.

One consequence of this incorrect citation, other than shifting the date of origin by a decade, was that certain people started claiming that my 1972 EOS didn’t include the effects of dissociation. Of course, they would have known better if they had actually read the review article, since it explained what had gone into the model. But, as I have already pointed out, most people simply copied the reference into their own papers without reading it.

Fortunately, a couple of Los Alamos scientists recognized that these people had made a mistake, as they pointed out in their own publications. I thank them for that—both for searching out the truth and making an effort to set the record straight. I hope the above discussion straightens the matter out for anybody who still hasn’t gotten the word.

High-Pressure Experiments (1997-2003)

My EOS model predicts the onset of dissociation at ~20 GPa on the D2 Hugoniot. No experimental data to test the EOS in this regime appeared until 1997.

In 1997-98, a group at Livermore used the Nova laser to generate shocks in D2 in the pressure range 25-400 GPa, where dissociation was expected to occur. Those experiments gave shock compressions as much as 45% higher than predicted by my 1972 EOS. Later experiments using the Nike laser at the Naval Research Laboratory gave similar results.

Having virtually ignored my 1972 EOS for 25 years, Livermore now began featuring it in every experimental paper, emphasizing the discrepancy between my prediction and their data. (The EOS that had been generated at Livermore also failed to agree with their data, but they only discussed that fact in their internal laboratory documents, not in their more high-profile published papers.)

The groups at LLNL, NRL, and elsewhere, concluded that the discrepancy was due to inadequacies in my treatment of dissociation. Many theoretical papers also appeared to offer explanations for the discrepancy and to present models that were in better agreement with the laser data.

It is now clear that the problem was with the experimental data, not with my theory. Subsequent Hugoniot measurements, using magnetically-driven flyers generated by the Z-machine at Sandia (2001-03) showed the laser data to be in error and agreed rather well with my 1972 model—better, in fact, than models that had been developed much later. Russian shock data, obtained using a different approach, were also consistent with the Sandia data and my calculations. Ab initio numerical calculations also agree with the Sandia and Russian data.

It is clear that the Nova/Nike measurements are not consistent with an equilibrium response of the material; they either contain systematic errors or non-equilibrium effects that have not yet been identified.

You can refer to my report on H2 and D2, SAND2003-3613, for more information on this matter. I will discuss that and other recent work in my next blog article.

- Jerry Kerley

My last blog article began a discussion of misconceptions and misinformation surrounding my work on the EOS of H2 and D2. I will continue that discussion here, with additional comments on The Early Work (1969-72).

The Sesame EOS?

My 1972 D2 EOS table eventually came to be known as “the Sesame EOS,” even though it was created a few years before the Sesame library came into being. This might not seem to be an egregious offense, but it is misleading, laying the foundation for more serious misconceptions. In particular:
—Lumping my D2 EOS in with other Sesame tables obscures the fact that it was derived from a much more sophisticated model than the others, which were based on the simplistic three-term, average atom model. (See my 8-26 blog article, for example.)
—In fact, a later paper did refer to my EOS as “Thomas-Fermi-Dirac based,” a completely inaccurate description.
—Calling it the Sesame EOS obscures the fact that a Sesame database allows more than one EOS table for any material. This fact is particularly relevant now, because I have generated completely updated EOS, for both H2 and D2, also available in the Sesame format. I have also created another Sesame database that is completely separate from the one at Los Alamos.

The information blackout

It is well-known that the peer review system can be misused to block publication of a colleague’s work in journals, making it harder to correct misconceptions and misinformation. If that colleague stubbornly insists on writing reports—as I have done—you can simply ignore them in your own publications. That creates the impression that the work either doesn’t exist or isn’t good enough to warrant a mention. (Of course, you can still criticize the work in private.)

In 1977, a report appeared claiming to survey existing work on molecular and metallic hydrogen [1]. Consisting of 135 pages, with 128 references, it was commendably complete, except for making no mention whatsoever of my work. These authors were not ignorant of my work, since they did refer to the work of two of my Los Alamos colleagues. Besides, I had personally discussed my work with the first author (who worked at Livermore).

I am sad to say that my Livermore “friends” continued to avoid mention of my work in this way for many years, until they obtained shock-wave data that they thought proved my EOS to be erroneous. I’ll talk about that later.

Linear-mixing model

As noted in my last blog article, my 1972 EOS used the theory of chemical equilibrium to model the molecular-metallic transition in the fluid phase. This theory involves two steps: set up a model for a mixture of molecules and atoms, then minimize the free energy to compute the concentration as a function of density and temperature.

To simplify the calculations, I treated the two mixture components as having the same density and temperature. This model was later “rediscovered”—by one of the authors on the 1977 survey mentioned above—and came to be known as “the linear mixing model.” The fact that I had already come up with this model, years earlier, was not acknowledged. You can get away with this kind of thing when you put a blackout on somebody else’s work.

BTW, my updated EOS uses a better mixture model, which makes linear mixing obsolete.

Rotational & vibrational degrees of freedom

I will conclude today’s post by discussing one more issue, which seems to have been ignored. (Not surprising, since so few people have actually read my reports.) My 1972 EOS used a rather sophisticated model for the contributions from internal rotation and vibration of the molecules. My most recent EOS uses an updated version of that model.

Virtually everyone else in the field treats the internal degrees of freedom using the rigid rotator/harmonic oscillator (RRHO) approximation. A serious problem with RRHO is that it does not account for anharmonicity, which limits the amount of energy that can be put into these degrees of freedom. Anharmonicity is also affected by perturbations from neighboring molecules, resulting in destabilization of the vibrational modes at high densities.

My 1972 model included the effects of anharmonicity, centrifugal distortion, vibration-rotation coupling, and destabilization of vibrational motion. These effects turned out to be very important in reproducing the Hugoniot data for D2 at high pressures, during the onset of dissociation, as I will mention later.

To be continued

This article completes my discussion of the early work. I will discuss subsequent work in the next articles.

- Jerry Kerley

[1] I’ll let you track this report down. It can be downloaded from the internet.

My work on hydrogen and deuterium has suffered more misconceptions and misinformation than any of my other projects. I will try to straighten things out in my next four blog articles. The discussion is divided into four parts: Early Work, Intermediate Period, High-Pressure Experiments, and Recent Work.

Early Work (1969-72)

Los Alamos hired me in 1969, specifically to develop a new EOS for D2, which was badly needed by the laboratory at the time. Anyone who has given any real thought to this problem knows that D2 is a surprisingly complicated material, even though it has the lowest atomic number, Z=1. It has both molecular and metallic forms; it exhibits a solid-solid phase transition, melting, dissociation, and ionization. And there were very few experimental data to help me at the time [1].

I spent three years completing this project, much of that time spent learning new things and developing theories and models that I later extended and applied to other materials. The work was fully documented in two Los Alamos reports—LA-4760 (1971) & LA-4776 (1972)—and a short paper, Phys. Earth Planet. Int. 6, 78 (1972). My model was very sophisticated, especially considering the state of EOS theory at that time. It was also very successful, essentially resolving the problems that had motivated Los Alamos to hire me.

Enter the predators

Unfortunately, success often exposes one to predators, cowards who tear down the work of others in order to gain an advantage for themselves. For years afterwards, I would occasionally receive visits from senior members of the lab, asking questions about my D2 model—what I had and hadn’t included, what approach I had used, etc. I suspected, even then, that others had been whispering in their ears, moving behind the scenes, looking for vulnerable spots, planting seeds of doubt about my work.

My detractors became more visible in later years. With one exception, they always attacked behind my back, so that I could not defend myself. But I learned about some of their activities from third parties.

An empirical EOS?

My 1972 EOS model has often been described using inaccurate terms. One of the most fatuous comments to appear in print was to call it “an empirical EOS.” Now, I am not against using empirical methods, if that is what it takes to solve the problem. In fact, I had used experimental data to test and refine the model—the very same thing everybody else was doing. Everybody else.

One reason that the term “empirical” is so inappropriate in this case is that so few data were available to use in any “fitting” procedure [1]. It is far more accurate to say that I had developed a theoretical model and used experimental data to determine some of the parameters: 80-90% theory, 10-20% experiment. Intellectual honesty requires that the model be considered a priori in those regions where no data were available at the time. This fact is important, in view of how well the EOS agreed with experimental data obtained much later.

Plasma phase transition?

Another misconception has to do with my treatment of the transition between the molecular and metallic states. In the solid region, the molecular and metallic phases are presumably immiscible, and the transition is first-order. In the liquid region, the molecular and metallic forms are expected to be at least partly miscible, so the transition is smeared out over a range of pressures and temperatures. My 1972 EOS treated these two regions using a phase transition model and a chemical equilibrium model, respectively. (The methods were similar to those now used in Panda and EOSPro. Also see my 8/26 blog article.)

However, it is possible that the transition could generate a region of thermodynamic instability in the liquid, leading to phase separation. This behavior, which is similar to vapor-liquid phase separation, is sometimes called a plasma phase transition (PPT). I did look for such a transition in my work but did not find one—as reported in my 1972 paper. (My most recent EOS does not predict a PPT either.)

Nevertheless, a 1983 paper [Astron. Astrophys. 120, 227 (1983)] asserted that my EOS might have a PPT, and a 1995 paper [Astrophys. J., Suppl. Ser. 99, 713 (1995)] apparently concluded that it did. (Both sets of authors had observed a PPT in their own work.) [Note added, 11-13-10: My comment about the 1995 paper is not correct. Please refer to my 11-13 blog article.] I don’t remember being consulted by the authors about this issue, which I could easily have straightened out. Unfortunately, this error has since been repeated in the literature—the consequence of quoting somebody else’s comments without reading the original papers for oneself.

To be continued

I have much more to say on the subject of H2 and D2 but will postpone further discussion to later posts.

- Jerry Kerley

[1] At the time of this work, static compression data were only available to 2 GPa in the solid and 0.24 GPa in the liquid. There were only five Hugoniot data points, the highest pressure being 13 GPa, and six reflected shock points, the highest pressure being 34 GPa.

If you’ve been reading my recent blogs, you may have noticed that I get frustrated when people talk about my work without bothering to read my papers and reports.

Perhaps I should try harder not to exceed peoples’ attention spans. I heard somewhere that an e-mail is too long if one has to use the scroll bar to read it all. I think I’m in trouble. My blog posts are probably too long too.

Perhaps I should look into using Hip-Hop as a communication tool. A typical offering might look something like this:
Dude I wanta tellya ‘bout my code E-O-S-Pro.
Listen up ya gonna find out stuff ya needta know.
What do you think? Am I onto something here? Should I put it out as an MP3 file or create a video for YouTube?

Perhaps I ought to check out other options as well—Facebook, MySpace, Twitter?

Stay tuned.

- Jerry Kerley