On Fri, Nov 4, 2011 at 11:26 AM, Jones Beene <[email protected]> wrote: > It seems you are conflating two processes when only one will suffice. And one > of them is absurd from the start. > > Why pump the liquid at all? Why use a magnetic field with pumping, when a > simpler route exists? Calutrons were a gigantic waste of money in the > Manhattan project and were only used for a few years as an expedient.
Because you need to have v > d * q * B/(m1 - m2) where v is the speed, d the desired separation distance, q the charge, B the magnetic field and m1 and m2 the respective masses, and if you don't pump, you'll have to rely on the acceleration provided by the electrical field, which may be too low. > > Centrifugal acceleration (even the common lab centrifuge) should give similar > or better results, if what you want is enrichment by density gradient in a > NiCl solution. > > In fact the chloride is ready-made for this since by varying the water > content and temperature (solubility) - the heavier fraction can be solidified > by chilling - while the light fraction remains a liquid and is more easily > removed at the early stages (to automate the process). > > If you are going for enriching an isotope that is 10% denser, it will take at > least seven stages for every doubling (not counting losses). This is the > "rule of seventy" (similar to formula used in compound interest). Therefore, > to increase a 1% isotope to 16% might require a minimum of 28 stages of > progressive enrichment, but when losses are included, it is probably closer > to 50 stages. Automation makes a big difference with this many stages. > > For the NiCl solution (hexa-hydrate) the solubility is 254 g/100 mL at 20 °C > - and 600 g/100 mL at 100 °C. That difference could help a lot in automating > the processing, so that even 50 stages in a continuous centrifuging would not > be a insurmountable problem to get 64Ni enriched to a level in the mid-teens > at an affordable cost. > > At least this is doable, but - as for final cost - that is another question > based on many issues. But if the enrichment percentage can be kept to a low > level, it need not be too expensive for the numbers Rossi is throwing around. > > This is because with NiCl - the rejected isotopes are of the same value as > the feedstock, and this makes the processing simply a matter of overhead, > efficiency and labor. The bulk nickel is no less useful in industry - with > the 64Ni removed as with it there. In effect, you only "rent" the feedstock, > removing very little. > > That is a huge difference compared to what we look to as the "model" for > isotope enrichment. With uranium enrichment - in contrast, the feedstock cost > must be 100% absorbed in the cost of the enrichment (since the depleted U has > almost no value) so that factor alone grossly inflates the net cost by > several orders of magnitude (compared to nickel). > > Enrichment cost alone, for even the heavy metals - is not outrageous so long > as there is a large market for the depleted feedstock. That is key. > > There seldom is a market, but since nickel has that as its major feature, > then an enriched isotope on a mass production scale, for a NiH energy system, > is not out of the question. > > > -----Original Message----- > From: Berke Durak > > On Thu, Nov 3, 2011 at 2:52 PM, Peter Heckert <[email protected]> wrote: >> The ion diffusion speed in an electrolyte is only some centimeters >> per minute at best, while the speed in a Calutron is probably some >> 100 to some 1000 kilometres per second. >> >> Therefore the mass inertia of the nucleus at this low speed has no >> effect. The electrolyte vessel must be some 1000 km long for this >> to work. > > Yes, but can't the liquid be accelerated to a sufficient velocity > using pumps? > > A quick search reveals that the radius of the circular path described > by a charged particle subject to a transverse magnetic field is R = > mv/qB where m is the mass, v is the velocity, q is the charge and B is > the field in tesla. > > Assume we want to separate two isotopes of masses m1 and m2, we'll > want R1 - R2 > d for some sufficiently large d. Take d = 1cm, m1 = 58 > amu and m2 = 64 amu, and q = 2 x 1.6e-19 C (for Ni 2+), then we need v >>= qB/(m1 - m2) = 32e6 m/s/T. For a 100 nano tesla field, this gives > 3.2 m/s and R1 = 9.6 m and R2 = 10.6 m. I suppose 3.2 m/s is a > reasonable velocity. > > If we pump the solution so that the Ni2+ ions reach a velocity of 3.2 > m/s while keeping the magnetic field around 100 nanotesla, we might be > able to separate them. > > By properly orienting the setup with respect to the Earth's magnetic > field, some mu-metal shielding or using some active cancellation > technique, it might be possible to obtain a 100 nT field. > > The problem might be that you will also have whatever cations are > present swirling in the opposite direction. I don't know how that > would affect the Ni2+ ions. > > Any physicists / electrochemists in the room? > -- > Berke Durak > > > >
