Although the internal optimizer treats only the inequality and equality constraints of types shown in Equation 36–2 and Equation 36–3, respectively, it is possible to take into account constraints of the following type:
 | (36–45) |
This is because Ansys Polyflow can transform them to be:
 | (36–46) |
The parameters for a constraint include the following:
name
Short names are recommended for constraints, for the sake of readability in the output files.
goal of the constraint
The goal of the constraint can be:
an equality of the form
, where 
represents 
in Equation 36–3
an inequality of the form
or 
, where 
represents 
in Equation 36–2
form of the function for the constraint
The predefined functions available for the objective functions (Equation 36–38 – Equation 36–44) are also valid for the constraints.
If you want to define a constraint of the type
(36–47)
where the value of limit
is not zero, you have to select a predefined function with
a constant term like parameter 
in Equation 36–39 or parameter 
in Equation 36–44, and set this constant term to be equal to
. extracted value
extracted value
, if required parameters
, 
, 
, 
, and 
, if required
The equality constraints can be difficult to satisfy when the initial value of
such a constraint (
) is too small (for example, 
). In such circumstances, the number of iterations needed to bring
the ALM method to convergence is greatly increased. It is therefore recommended that
the values of parameters 
, 
, 
, 
, and 
in Equation 36–38 – Equation 36–44 be selected so as to yield a reasonable value
for 
. This discussion also impacts optimization problems that minimize
multiple objective functions using the multi-priority scheme: as described earlier,
when the objective function of a higher priority (for example, 
) is optimized to yield a minimum value (for example,
), it is then converted into an equality constraint (for example,
) for subsequent optimization solutions. Since it would be too
difficult to try and predict what the initial values of such converted constraints
would be, the best practice you can follow is to define the multiple objective
functions
such that they are all of the same order of magnitude numerically. For
example, it is not advisable to have one objective function that has a value of
(which is typical of a pressure drop) while another objective
function has a value of 
(which may be typical of the flow balance parameter).