Posts by jeff

Clearly alumina does not work as a containment material when molten Li is present. See the attached photo. Based on a URL that I posted earlier today, the results were not unexpected. I'm just amazed by how reactive molten Li really is. I'll need to line the cell with a metal that is compatible. The previously mentioned article suggests pure Fe, ferritic stainless or Mo, Ta, W. Ferritic stainless is probably the easiest to get.

Today I loaded Ni and alumina powder and a 50 mg slug of Li into an alumina tube and under vacuum applied power sufficient to reach ~500C. Afterwards I noticed a black stain on the exterior of the alumina tube corresponding to where the metallic Li was loaded. Clearly there is a reaction occurring between the Li and what I had assumed was nonreactive alumina. A bit of research yielded the following reference from the IAEA:

http://www.iaea.org/inis/colle…Public/09/410/9410560.pdf

Based on this report, it appears there are very few materials that are compatible with liquid metallic Li. This raises another question: does LiAlH decompose into metallic Li to the extent that it reacts with the vessel wall?

Jeff

Any news on when this report will be (or has been) published? I'm considering running DFT simulations on Ni/Li/H systems, but there are a ton of user specified conditions that must be specified, and I do not want to re-invent the wheel. I would also be interested in what the author(s) means when he states the material "satisfies all the conditions". In other words, what criteria are being used to define an NAE environment?

Louis DeCharios news could be most interesting for you! You should read it."We have run materials simulations (also known as Density Functional Theory simulations) on our best guess of Rossi’s alloy material. It satisfies all the conditions given above, while pure Nickel does not."

Jeff

Dave,

Most of the questions or suggestions you posed are addressed in some of my previous posts under "Replication Attempts". There I posted some .pdf files detailing the apparatus as well as calibration plots for the calorimeter. BTW, you are preaching to the choir when it comes to accurate calorimetry. The calorimeter I'm using operates in a closed loop mode where airflow is controlled to maintain a constant thermal mass/time rate. It yields nearly a perfect linear fit with a small nonzero Y-intercept that is due to an offset between the two thermal sensors.

Regarding Al2O3, I plan to mix it 50/50 by weight into the Ni to prevent sintering, as you recommended.

Jeff

The apparatus I built is designed such that H2 is furnished eternally from a hydrogen generator (much safer than bottled H2). The metallic species are placed in an alumina tube where both ends are stuffed with quartz wool to keep things in place, while permitting hydrogen to permeate the metals. The outer 1/3 of each end of the tube is filled with Al2O3 powder that acts as a filler and will absorb any molten metal that might tend to leak out the ends. External heating is provided by Kanthal wire wrapped around the exterior of the alumina tube.

Before proceeding, I have a few questions that hopefully someone out there can answer.

1. I'm aware that metallic Al will evolve as a consequence of the thermal decomposition of LiAlH. At elevated temperatures Al in a liquid state will wet Li2O and reduce it to Li (gas or liquid) + Al2O3. So Al will act as an oxygen scavenger, but does it play any other role in LENR reactions?

2. Is there a known protocol for preconditioning the Ni + Al + Li before attempting to initiate LENR by adding hydrogen? I have read accounts that the mixture should be heated to ~800C but am not sure for how long.

3. At what temperature have LENR effects been observed? I'm not looking for high COPs for the first experiment, just reliable and repeatable indications of LENR activity.

Jeff

I took chance and purchased a used H2 generator off Ebay and got lucky. Other than needing a few filters replaced, the only other problem was a defective float switch which I temporarily bypassed. The unit generates H2 at pressures up to 100 PSI. To make sure the gas was indeed H2, I lit it. Now to hook the unit up to the gas/vacuum manifold.

Jeff

What you state in your last sentence pretty much describes what I intend to do using MM software. Most MM packages support DFT which solves exactly the electrons' correlation and exchange components of the potential. It should not be that difficult to construct a 1-D model of alternating metal and hydrogen cores and utilize a realistic potential, either an equation-based potential like L-J or Morse potentials or one empirically measured for each species of atom. Most MM packages also comprehend phonon interactions, so DFT + the implicit B-O approximation should be sufficient to model discrete breathers.

Jeff

I have found several sources for ASTM II DI water. It costs about as much per gallon as really bad wine. Since I'm used to drinking good Burgundy, the DI water is a real bargain. BTW, normal store bought distilled water has too high an ion content due to the metals used in the distillation process. A friend of mine who works in semiconductor fabrication cautioned me not to let the DI water come into contact with any metals.

I have been trying to get someone here in the Peoples Republic of Kalifornia to sell me a small bottle of H2. Apparently the powers that be consider bottled H2 in the same hazard class as Polonium. The welding shop will only sell it if I have a properly outfitted truck. They won't sell it if I plan to carry it in a car. Never mind that it's a total of only 25 cu ft. So I broke down and bought a laboratory hydrogen generator. Now, with my luck California will refuse to sell me de-ionized water.

LENR science is difficult enough without politicians and regulatory agencies making it ever more so.

Jeff

I have a graduate background in physics, although I have worked in EE, rather than physics, for many years.

Much of the difficult work has already been done in mathematically describing discrete breathers and in general, making molecular modeling computationally tractable. Over the last 20 years the availability of ever more inexpensive compute power has made it possible to run MM programs on a PC. This should also be the case for modeling DBs.

Two key conditions must be met before DBs can be produced: these include a DB oscillation frequency that lies outside the metal matrix phonon spectrum, and a de-localized potential that takes into consideration the neighboring atoms. I believe the best (and perhaps the only) way to proceed is to utilize molecular modeling techniques to ascertain if a Ni/H + other metals system can generate the appropriate NAE environment. By NAE, I mean that P-P spacing can transiently approach 1E-12 meters.

Molecular modeling is a huge field, something that is becoming increasingly obvious as I attempt to find a modeling package that is suitable for simulating H2/metal systems. So called ab-initio modeling is not easy since all modeling tools must make approximations to be mathematically solvable and computationally tractable. Even constructing and simulating an isolated molecule and getting proper bond lengths is tricky. There are many parameters that need to be specified. Modeling must first define a basis set, which is, by necessity, incomplete and will inevitably produce some error. A complete basis set has an infinite number of vectors. For transition metals both the s and p electrons must be considered which increases the number of orbitals that must be evaluated. A quick literature search reveals literally dozens of basis sets, and there does not seem to be an easy way of selecting which set will give the best results.

The next step is to select a so called pseudo-potential. Again, there are a huge number of these to choose from. DBs require a potential model that properly models the higher order (cubic and quartic) non-linearities and properly describes large amplitude motion of a pair of hydrogen nuclei. The so called Morse potential or a variant thereof is a good place to start. I could go on, but it should be clear that molecular modeling, while I think is the only method of obtaining meaningful predictive results, entails a good bit of black magic. Nevertheless, if time permits, I plan to get hold of an open source software package like CP2K and see if I can get any meaningful results.

Jeff

Over the last week or so I built a 4'x6' table to accommodate all the components for the setup. Everything but the hydrogen source is now ready to go. Lots of little fabrication jobs such as making the holder for the Lumasense sensor body. Vacuum pump achieves <1 milli Torr pressure, which is sufficient to remove air before introducing hydrogen. It also is a good check for system integrity for positive pressurization. I checked the calorimeter temp rise with the cell evacuated and at atmospheric pressure, and the results were the same. (They should be.) Also noticed that there was very little difference in the heating element temperature for evacuated vs. non-evacuated cell. A bit surprising, but probably indicates that most of the heat transfer is radiative rather than conductive.

Jeff

LAVs exhibit two unusual behaviors compared to normal atoms in a periodic lattice. The first is the ability to accumulate very large energies (hundreds of eV); the second is the ability to retain that energy without dissipating it to neighboring atoms, and these two phenomena are related. In some ways there equivalent to the Q of a tuned circuit. A small excitation voltage can produce very high voltages in the tuned circuit.

Here is where a bit of quantum mechanics comes into play. Phonons, although not real particles, may be treated as such in terms of QED theory. Just as a simple system like the hydrogen atom can only absorb or emit energy at specific wavelengths, the same holds true for phonons. The anharmonic property of LAVs means that they exhibit different oscillatory frequencies than the host lattice and cannot transfer energy to the host lattice at its phonon wavelengths. As a result, an ensemble of hydrogen atoms embedded in a metal matrix will possess a characteristic oscillation frequency substantially different from the metal lattice. This is what is meant by the phonon wavelength gap. Such a system of hydrogen atoms can accumulate energy from thermal excitation but cannot readily dissipate it to the metal lattice, so the amplitude of oscillation can become orders of magnitude greater than the ambient thermal energy.

After doing a bit of a web search there are indeed public domain molecular dynamics tools. One that looks particularly interesting is called CP2K. They have a website that lists some of the applications to which the tool was applied. Some of these include metal-hydrogen systems. It supports many of the algorithms and models commonly used in molecular dynamics, perhaps most importantly phonons and the B-O approximation. I may download the tool and see if I can get it to work on simple problems such as a single water molecule or H2 dissociation on a metal surface.

Localized anharmonic vibrations are an interesting and phenomena and offer something that is rare among LENR theories: namely the ability to simulate multi-atom configurations and generate quantitative results based on well understood quantum mechanical principles. Ab initio DFT calculations have been confirmed against measurements for such key characteristics as inter-atomic distances, electron density, and molecular dissociation energy. It would be interesting indeed to simulate Ni/H or Ni/H/Li systems and see if they yield any additional insights regarding inter-nuclear distances and tunneling enhancements. I also wonder if there are any public domain simulation tools available. I have heard that VASP is sort of public domain but do not know much about its capabilities. In any case, I suspect such tools require compute capabilities well beyond that of a PC.

There are other tantalizing aspects to LAVs. They have a long lifetime and may undergo thousands of oscillations during that time. In his recent book Edmund Storms postulates that the lack of high energy products during LENR is due to a gradual release of energy over many cycles. This would accord with the periodic and long lived nature of LAVs.

It is not clear to me whether there exists a physical theory for the gradual release of nuclear scale energy over a long time frame and at distances that are ~100x greater than the nuclear radius. Typical high energy explanations may not be of much use because they assume near instantaneous reaction times occurring over distances that are on the order of the dimensions of a nucleus. Does anyone out there have knowledge of how to extend Fermi’s Golden Rule Equation to the time and distance scales found in LAVs? Are there other methods of analyzing nuclear interactions?

Good luck with this new approach. I spent approx two years on and off attempting to get positive COP from a Piantelli-type experiment. During that time I learned a lot, especially about the need for accurate calorimetry. Based on the 1999 S.Romanowski paper in "Langmuir" dealing with H2 dissociation, I elected to use Ni/Cu thermocouple wire, but pure Ni should work also. The H2 dissociation energy is 2.444 eV, and that of the Ni surface is >3 eV. A 50/50 Ni/Cu alloy has an even higher dissociation potential, which is why Piantelli chose to use it.

I cannot confirm that I was ever able to conclusively obtain a COP >1, although there were a few runs where ~5W excess power was indicated.

Jeff

By applying a series of known voltages and currents across the heating element of a cell placed in the airflow calorimeter it is possible to get a graph of applied power vs. temperature rise. The result is a nearly linear plot running through the origin. A max voltage of 28V was applied that, from a previous calibration run, gave a temperature of ~1100C. That temperature is as high as I believe is necessary to initiate some LENR activity. These results demonstrate that feedback based on thermal mass flow does yield a linear power vs. temperature behavior.

The next step is to tear down the setup and connect the gas/vacuum manifold and IR thermometer. I'll make one final in situ calibration, where internal cell temperature (measured by a TC), is plotted against the temperature reported by the IR thermometer.

Then it will be time to load the cell with active ingredients.

Jeff

Now that the control and signal conditioning electronics are working I have put an empty cell into the calorimeter and started taking temperature readings for different power inputs. Power is furnished by a DC supply wired with a 2.5 mOhm shunt and a 4-wire Kelvin voltage measuring configuration at the thermal boundary. Each measurement requires approx 2 hours because of the long thermal time constant. The end result will be a power vs. temperature rise plot that should be linear, passing through the origin. Then it will be time to test with an active cell.

Jeff

The circuit boards arrived yesterday, and I assembled one last night. After making one minor circuit change and chasing down a cold solder joint (Do not attempt to assemble surface mount components with a soldering iron, no matter how fine a tip you may have.) the circuits work as designed. The thermal mass controller measures airflow by means of a pair of diodes, one with a resistive heater and one without. If you touch the cold diode it warms up, and the fan speed decreases. Similarly, if the heated diode is removed from the airstream the fan speed increases.

The next step is to connect all the sensors and the DAQ module and start performing calibration on an empty cell. This will give a baseline for power applied to the heater vs. air temperature rise through the calorimeter chamber.

Jeff

Quote from PetaLue

Your circuits are just looking fine. Can you please tell me the function of the IC IN11AP? Also tell me how you calculated the value of all resistors,i mean what parameters you consider here? What is the threshold level of all of your sensors? Are these circuits working efficiently?

The INA114 is a so called instrumentation amplifier. Internally it consists of three opamps with precision resistors in the internal feedback paths. This type of amplifier exhibits excellent common mode rejection, very high input impedance, and has the ability to set gain with a single external resistor.

Calculating the values of all the resistors would entail quite a bit of work on my side, since I did not commit all my design calculations to writing. Most of what you want can be obtained by examining the data sheets for the various integrated circuits. The same holds true for the sensor thresholds. Take a look at the LM35 data sheet. Regarding the temperature sense diodes, any book on solid state theory will give the diode junction vs. temperature relation. Gain for the two INA114p devices connected to the LM35 sensors was set to yield a ~5V dynamic range to the DAQ module. Gain for the other INA114s was set extremely high since they operate in a servo mode, where feedback is provided by an external input. The accuracy of the closed loop feedback system, in these cases, is a function of the difference between open and closed loop gains. Again, consult any good textbook on feedback theory for a more complete explanation.

I'm not sure what you mean by efficiently, but I have working examples of similar circuits used in a previous experiment. The circuits I posted have been simulated using HSPICE and appear to behave correctly. Some minor adjustments to resistor values may be required.

Jeff

This evening I ran a series of voltages through a Kanthal wound alumina tube in free air, yielding temperatures from 350 to 1100 C. The tube was instrumented with a Lumasense IR thermometer focused on the heater windings and a type K TC inserted into the alumina tube. The Kanthal wire was not covered with ceramic cement and presented a dark gray appearance. The Lumasense thermometer was operated in the 4-20 ma mode into a 250 ohm resistor, yielding a 1V (350 C) to 4V (1800 C) span. Emissivity was set at 100%, for now. Results showed a relatively poor match between the the TC temperature and the output voltage of the IR thermometer based on the span temperature limits. It was, however, possible to obtain an excellent match if the IR thermometer voltage readings were fit with a 1st order polynomial. See the attached figure for details. My recommendation for anyone using an IR thermometer is to calibrate it against a reliable standard such as a TC.

Jeff