[Isotope-prod] Optimization question

[Pavel Degtiarenko]

A logical extension (and simplification) of the previous argument 
regarding the circular beam rastering: linear rastering.

We could improve thermal and mechanical conditions for all windows, and 
make beam currents up to 5-10 mA accessible, if we switch from the 
circular shape of the raster and the window to the one-dimensional (say, 
vertical) narrow slit shape. We can make the window's horizontal 
dimension small, I guess it could be down to 3-5 mm, and the vertical 
dimension reasonably large (few cm). The narrow beam (say, 200 micron 
sigma) is rastered vertically with some reasonable frequency (say, tens 
of hertz). The power deposited in such window will be spread vertically, 
and it will have a very short thermal conduction path to the cold mass 
of the window holder, both to the left and to the right. The solution 
improves thermal parameters for the window dramatically, and at the same 
time the mechanical stresses in the window are also expected to be 
decreased, as the window span holding vacuum becomes much smaller than 
in the circular case.

Such scenario may actually end up requiring comparable or even smaller 
volume of the target gallium than the present beam sizes - need some 
simulations for a real comparison.

Best regards,
Pavel

On 06/22/2016 03:48 PM, Pavel Degtiarenko wrote:
>
> I'd like to revisit one optimization question that was standing from 
> the very beginning.
>
> We assume presently that minimizing volume of the isotopically pure 
> target is paramount in this problem and we have to design windows, 
> radiators, and targets that would survive extremely high power 
> densities, to the limits of impossible. But there might be an optimal 
> overall solution, in which the target volume is few times larger than 
> the absolute minimum. The target itself, and the Cu67 separation 
> process would be more expensive (most likely less than linearly). But 
> the cost increase would be compensated by avoiding risky and therefore 
> expensive engineering solutions. So may be the optimum solution would 
> be using larger diameter targets and windows that could be cooled 
> easier. I would suggest to consider for the proposal finding the beam 
> diameter that we know could be run safely, base all estimates on that, 
> and include further optimization as one of the goals of the proposal.
>
> One of the solutions to consider could be a circular beam rastering 
> when beam position at the target goes in circles at a constant radius. 
> The target does not fill the whole cylinder volume, but is placed in a 
> cylindrical layer, say, 5-10 mm thick, with the core of the cell made 
> of the same material as the whole cell. The Ga/Zn target volume then 
> goes linearly (not quadratically) with increasing radius, so that 
> might be a comfortable variant for the optimization.
>
> Best regards,
> Pavel