component of the standard design solution. We can see that for that 300 kA 14 meters long potshell, the
floor maximum vertical displacement is about 30 mm. Figure 6 presents the vertical displacement
component of the retrofitted potshell clearly showing that there is essentially no vertical displacement
component left in the solution.
"almost empty shell" model
the next logical step is to test the retrofitted potshell design idea with the improved "almost empty shell" model.
Figure 7 presents the results obtained, which are almost identical to the one obtained with the "empty shell" model
confirming the validity of the proposed idea. This also highlight the fact that for a large number of potshell retrofit
ideas not affecting the global potshell structure stiffness, the improved "empty shell" model is the most effective
analysis tool because it is the fastest tool.
graphitized cathode blocks with the improved "almost empty shell" model
blocks is changed like it is the case in this second retrofit design proposal, it is not possible to use the "empty shell"
potshell model to analyze the proposed design change, this is why the "almost empty shell" model type has to be
directly use this time.
impact of retrofit design changes affecting the cathode erosion in order to predict the retrofitted cell life expectancy
(assuming that the first failure mode is the attach of the collector bar by the metal after that all the carbon above it
has been removed by erosion). That model predicted a cell life of 2000 for the standard design based on the usage of
45 cm thick graphitized cathode blocks with 26 cm of carbon above the collector bars (see figure 8 for a full cathode
panel solution of that type of erosion model).
possible to increase the cathode block thickness by 10 cm without reducing the height of the insulation under the
blocks or reducing the height of the cell cavity simply by moving the potshell floor 10 cm down. Moving the
potshell floor down this way is reducing the height of the cradles web under the floor by 10 cm, which is of course
reducing the stiffness of the cradles. In [6], using the cathode panel erosion model, it was demonstrated that
combining the usage of 10 cm ticker cathode blocks and the usage of selective rodding would potentially increase
the cell life up to 3500 days. It was also speculated that reducing the stiffness of the cradles would not have negative
impact on the potshell mechanical behavior because a potshells designed to withstand the 4 to 5% sodium swelling
of amorphous cathode blocks have became over-designed now that graphitized blocks with sodium swelling index
less than 1% are used.
hypothesis, but this is of course no longer the case. So the proposed potshell and lining design proposed in [6] has
been analyzed using the "almost empty shell" potshell model. In that model, one of the key parameter is Eo the
setup to 3%, which is a typical value for 20% semi-graphitic cathode blocks. That parameter must be readjusted in
order to match the behavior of graphitized cathode blocks. For that type of cathode blocks, a value of 1% is more
typical so 1% was used to carry out the present analysis.
graphitized cathode blocks will have a potshell that will deflects less laterally than the standard design using 45 cm
thick 20% semi-graphitic cathode blocks even with the cradles web under the potshell floor having 10 cm less
height.