Page 342 - High Temperature Solid Oxide Fuel Cells Fundamentals, Design and Applications
P. 342

3 18  High Temperature Solid Oxide Fuel Cells: Fundamentals, Design and Applications

        where f3 is the CTE, E is the Young’s modulus, v is the Poisson’s ratio, and d  is
        the half-thickness  of  the plate.  Assuming that the lattice expansion-induced
        strain can be treated the same way as the thermal strain, PT, Eq.  (32) can be
        recast into a stress-isothermal  strain relation.  The experimental data for the
        isothermal  expansion  characteristics  can  then  be  used  to  obtain  the
        nonstoichiometry-induced  strain.  Consequently, the isothermal  stress  due to
        nonstoichiometry  can  be determined.  Delamination  will  occur  if  the  elastic
        energy  release  rate  by  crack  formation  exceeds  a  critical  value  where  the
        critical value is an interface property and must be determined experimentally.
        Alternatively,  mechanical  failure  occurs  if  the  thermal  stress  exceeds  a
        certain value.



         11.8 Electrochemical Models at the Electrode level
        Electrode-level models describe the performance  of  SOFC electrodes in detail.
         They  take  into  account  the  distribution  of  species  concentrations,  electric
         potential, current, and even temperature in the electrode. Their purpose is to (i)
         interpret  the performance  (polarisation curve) of  electrodes in terms of  rate-
         limiting  resistances  such  as  kinetic  (activation),  mass  transfer,  and  ohmic
         resistance:  and  (ii) predict  the local  polarisation  in full-scale  cell  and  stack
         models.
           To predict the local polarisation  in a full-scale cell or stack at any point, its
         dependence on composition, pressure, and temperature of the gas flowing in the
         gas channel contacting the electrode must be known. In a large cell, these bulk
         gas  properties  vary  from  one  point  to  the  next.  Electrode  polarisation  or
         overpotential - the difference between the local potential of the electrode under
         load and the potential at open circuit (equilibrium potential) - is also a local
         quantity because it depends not only on the bulk gas composition but also on the
         current density. In a large cell the current is usually distributed nonuniformly,
         as discussed in Sections 11.2-11.5.  Similar to Eq.  7, one can express the local
         cell voltage under load, i.e., when current is passed, as the thermodynamic cell
         potential minus three loss terms: the ohmic loss, the cathode polarisation, and
         the anode polarisation:

             V(i)  = E,,  - iRi - qc - g~                                 (33)

           As discussed in Section 11.2, the total polarisation of each electrode consists of
         two contributions, activation polarisation  (due to electrode kinetic resistance)
         and concentration polarisation (due to mass transfer resistance), so

             VC  = VCa + VCc  and  VA   VAa + VAc                         (34)

         For cell- and stack-level modelling it is necessary to have reliable values of the
         total polarisation of cathode, qc, and anode, qA, as a function of  local bulk gas
         composition, pressure, and temperature, as well as the local current density.
   337   338   339   340   341   342   343   344   345   346   347