I gave some samples of the Mizuno mesh (provided by Jed) to Peter at ICCF22. His nice analysis adds additional detail to my initial assessment from August 2019.
I think the analysis is good to have around, never one knows if one of these data points will become important later. This is unprepared mesh, so it serves as baseline.
MR3 Calibration 4 is now running: https://www.youtube.com/watch?v=ASDuoUqZ6es
During previous calibrations, an unexpected continuous rise in pressure above 200°C was seen. This behavior was still evident in Cal3, despite excellent long-term pressure stability at lower temperatures. This cell design uses an axial thermo-well tube into which the cartridge heater is inserted. For today's test, the vacuum pump will be left running, to ensure that the valves at both ends of the reactor cannot be a source of leakage into the cell.
I now suspect that the 304 stainless steel thermo-well tube leaks or out-gasses at the high temperature it sees from heater power of 100 watts and above. If today's test shows the same pressure behavior as previous tests, the thermo-well will be replaced by a blank end cap, and an external heater installed in its place. I have some of the same sheath heaters used by Mizuno for R19(?), to be coiled around the outside of the reactor tube.
Outgassing
To magicsound
It seems that Mizuno had an article on a method to treat stainless steel with a barrier for hydrogen. Is it possible that he uses this method on his reactor?
@Ahifors Thanks for the links. This paper also helped me understand the problem: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5226402/
My problem probably originates at the hot thermowell, where the SS304 tube is in direct contact with the heater cartridge at 450°C or more. While this is well below the full annealing point, grain boundary stress relief happens at around that temperature, allowing escape of entrapped volatiles. Table 1 in the above reference shows measured outgassing rate of SS304 at 450°C of 5E-3 Pa-cm3/sec per cm2 of surface. That is roughly in line with what I observed in the last two calibration attempts.
Tomorrow I'll wind an external sheath heater on the MR4 cell body. If it goes well, assembly and calibration will follow.
I have heard of people using a propane torch to do that...
After extensive testing, it was found that the source of outgassing at 100 watts and above was the SS304 thermowell with its internal heater cartridge. The MR2 cell was then rebuilt without the thermowell tube.
The external heater coil is identical to the one used by Mizuno. Following bake-out the cell reached 1.2E-5 Torr vacuum and no leakage seen following 12-hour soak. Gas plumbing was moved to the far side of the vacuum manifold. Cal1 is now underway, with streaming at https://www.youtube.com/watch?v=xcKIYzOrPZs
Ponder the problems of this system.
Vacuum in particle accelerators.
The CERN accelerator complex contains a total of 104 km of pipes under vacuum. Of these 104 km, 50 km pipes with a total volume of 15 000 m3 with pressures from 10−6
to 10−9 mbar. The other 54 km correspond to the LHC pipes that need to reach a pressure of the order of 10−10 to 10−11 mbar.
After extensive testing, it was found that the source of outgassing at 100 watts and above was the SS304 thermowell with its internal heater cartridge. The MR2 cell was then rebuilt without the thermowell tube.
The external heater coil is identical to the one used by Mizuno. Following bake-out the cell reached 1.2E-5 Torr vacuum and no leakage seen following 12-hour soak. Gas plumbing was moved to the far side of the vacuum manifold. Cal1 is now underway, with streaming at https://www.youtube.com/watch?v=xcKIYzOrPZs
do you agree with this guess: It wasn't a leak in the thermowell or at the weld of the thermowell but it was the high temperature of the thermowell that outgassed lots of gas. In other words, the thermowell gets much hotter than the vessel itself and therefore the stainless steel material of the thermowell does the outgassing.
I'm using a commercial voltage feedthrough (uses ceramic/metal brazed seal) to get power to my heater.
https://www.lesker.com/feedthr…dthroughs/part/eft0523093
as you know, your external heater will reduce the temperature that the nickel mesh experiences vs an internal heater.
Display MorePonder the problems of this system.
Vacuum in particle accelerators.
The CERN accelerator complex contains a total of 104 km of pipes under vacuum. Of these 104 km, 50 km pipes with a total volume of 15 000 m3 with pressures from 10−6
to 10−9 mbar. The other 54 km correspond to the LHC pipes that need to reach a pressure of the order of 10−10 to 10−11 mbar
I found this for the LHC ... note the low temperatures which means the outgassing is much reduced:
A vacuum thinner than the interstellar void
Ultra-high vacuum is needed for the pipes in which particle beams travel. This includes 48 km of arc sections, kept at 1.9 K, and 6 km of straight sections, kept at room temperature, where beam-control systems and the insertion regions for the experiments are located.
In the arcs, the ultra-high vacuum is maintained by cryogenic pumping of 9000 cubic metres of gas. As the beam pipes are cooled to extremely low temperatures, the gases condense and adhere to the walls of the beam pipe by adsorption. Just under two weeks of pumping are required to bring the pressures down below 1.013 × 10-10 mbar (or 10-13 atmospheres).
and ....
Two important design features maintain the ultra-high vacuum in the room-temperature sections. Firstly, these sections make widespread use of a non-evaporable "getter coating" – developed and industrialized at CERN – that absorbs residual molecules when heated. The coating consists of a thin liner of titanium-zirconium-vanadium alloy deposited inside the beam pipes. It acts as a distributed pumping system, effective for removing all gases except methane and the noble gases. These residual gases are removed by the 780 ion pumps.
i am reading currently a paper on reducing oxides from sintered powders.
The reduction under H2 is manifested by swelling of the packed powder from 500 °.
According to the author this corresponds to water vapor removing.
I have a European friend who said that the water had to be removed, by the way.
This process rises up to 90% of melting temperature and must last 1 hour ...
I guess without rising near the melting temperature, it's still necessary to reach this minimum temperature 500 ° then in this case do last reduction for several hours.
I think it's possible with R20 mesh which must be more stable than amorphous Japanese nickel powders, for example.
SS304 has a very wide crystalline structure, don't forget.
I would have to add that hoping for a vacuum just by playing on depression is bullshit because it will never remove all atoms stuck inside the lattice.
After extensive testing, it was found that the source of outgassing at 100 watts and above was the SS304 thermowell with its internal heater cartridge. The MR2 cell was then rebuilt without the thermowell tube.
The external heater coil is identical to the one used by Mizuno. Following bake-out the cell reached 1.2E-5 Torr vacuum and no leakage seen following 12-hour soak. Gas plumbing was moved to the far side of the vacuum manifold. Cal1 is now underway, with streaming at https://www.youtube.com/watch?v=xcKIYzOrPZs
do you agree with this guess: It wasn't a leak in the thermowell or at the weld of the thermowell but it was the high temperature of the thermowell that outgassed lots of gas.
Yes, that was my conclusion, backed by the link I previously posted: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5226402/
Regarding the temperature of the Ni mesh, that is not known. But because of the high thermal conductivity of Hydrogen, it will be close to the temperature of the steel cell. What will be absent is the possible effect of IR radiated from the hot thermowell, which could have caused the inner mesh surface to run hotter than the shell.
I'm using a commercial voltage feedthrough (uses ceramic/metal brazed seal) to get power to my heater.
Yes, that approach is one I've considered. The related issue of what kind of heater and how to securely mount it still needs resolving. Can you supply some details in that regard?
My current thinking is Kanthal wire coil wound on an alumina support rod, with the far end resting on a 304 alloy support inside the cell. Crimping the Kanthal to the copper feed-through wires is also potential weak point as the temperature rises.
The coating consists of a thin liner of titanium-zirconium-vanadium alloy deposited inside the beam pipes
If the LHC injects a little deuterium and heats it up a bit.. perhaps it could become an LENR reactor
Yes, that approach is one I've considered. The related issue of what kind of heater and how to securely mount it still needs resolving. Can you supply some details in that regard?
My current thinking is Kanthal wire coil wound on an alumina support rod, with the far end resting on a 304 alloy support inside the cell. Crimping the Kanthal to the copper feed-through wires is also potential weak point as the temperature rises.
if you double over and twist the Kanthal before and to the copper crimp, the heat and stress at that connection will be greatly reduced. ....As long as the Kanthal is not too scarred up from the fold and twist process, anyways.
Results of MR4 Cal1 4-5 September 2020. Calibration was done with Deuterium in the cell but no Ni-Pd mesh. Note the pressure increase rate from outgassing. Further bake out at 300°C under vacuum will attempt to reduce this sufficiently for testing with prepared mesh.
The nickel mesh acts as an insulator and more heat will exit both ends of the vessel (as opposed to going through the middle of the vessel having no mesh inside). I haven't yet figured out how much the mesh changed my calibration constant for the vessel (I have 8 thermocouples clamped to the outside of the vessel) but will do that this week. But I'm guessing it changed the calibration constant >15%.
Thorough bake out of the MR4 cell for 8 hours at up to 300 watts/350°C substantially reduced the out gassing from the stainless steel. The pressure increase measured after cooling was reduced from 60 Pa (Cal2) to just 8 Pa (Cal3), and appears to be close to standard gas law behavior. See the attached graph for details.
The protocol originally published by Mizuno and Rothwell specifies evacuating the cell for 1-2 hours at 200°C to remove water vapor. Based on the behavior seen in MR4, that would not be enough to reduce out gassing from the metal cell wall. Out gassing at 350°C is well below the annealing point of the Ni mesh (700°C) or decomposition of any CaCO3 present, but there could be changes in the Pd-Ni interface. It also seems possible that addition of such gas of unknown and uncontrolled composition and density might make the experiment hard to reproduce, as has apparently been the case. Further thought and discussion is certainly needed.
The protocol originally published by Mizuno and Rothwell specifies evacuating the cell for 1-2 hours at 200°C to remove water vapor. Based on the behavior seen in MR4, that would not be enough to reduce out gassing from the metal cell wall.
I am sure we also said you need to monitor the gas with a mass spectrometer to be sure the water is gone. As you say, it might take longer than Mizuno estimated.
I am sure we also said you need to monitor the gas with a mass spectrometer to be sure the water is gone.
Yes, you did, and I did as you recommended. Here's an analysis from an earlier bake out. The primary constituent coming out of the steel seems to be Nitrogen, followed by water vapor as expected. My test documented above shows that out gassing continues and accelerates well above 200°C, when water vapor from surface adhesion should be long gone. I suggest that substantial N is trapped in the steel grain boundaries, and is only released as the boundary stresses are relaxed at 300°C and above. Nitrogen is commonly added to austenitic stainless steels like 304 alloy, up to 0.1%. For details, see
https://www.totalmateria.com/p…cle&LN=EN&site=kts&NM=202
In this earlier test there was also some atmospheric leakage that might account for some of the gas detected. For that reason, the analysis should not be considered conclusive. The amount and temperature-dependence of out gassing measured in the latest test is clear, and its reduction after extended bake out confirms that possible atmospheric leakage has been eliminated.
