YOU ARE HERE: HOMEOPTICast UNIT 13: Optimization Strategies

OPTICast UNIT 13: Optimization Strategies

Back to OPTICast Menu

 

There are a number of things to take into consideration when running optimizations, as opposed to simply running a single simulation. These items are presented in this section, in Question and Answer format.
QUESTION: 
How do I get started?ANSWER: You need an initial casting/process design in SOLIDCast, just as if you were running a typical simulation.


QUESTION: 
Is there anything special about the construction of a model for use with OPTICast?

ANSWER: YES! When looking at the results of simulations, OPTICast evaluates only the Casting Material in a model. Gates and risers should therefore be constructed of Riser Material only, so their material will not be included in the evaluation of results.


QUESTION: 
What if I am importing geometry from a CAD system in the form of STL files?

ANSWER: It is necessary for each riser that will be a Design Variable to be a separate shape. This means that each riser and each shape that will vary must be stored and imported as a SEPARATE STL file into SOLIDCast.


QUESTION:
How can I apply a “factor of safety” to the results obtained from OPTICast?

ANSWER: One method is to construct a lower portion of a riser out of Casting Material. Then, when the results are evaluated (for example, Material Density) a “safety zone” will be created by this lower portion of the riser.


QUESTION:
Is there anything special about how the model should be meshed?

ANSWER: In general, it is best to specify “Node Size” rather than “Number of Nodes” when meshing, so that all simulations will be evaluated using the same node size.


QUESTION:
What are we doing when we create an Optimization Project?

ANSWER: We are doing several things. First of all, we’re defining Design Variables. These are any aspects of the model that are allowed to vary while searching for an optimum. These might be Geometric Design Variables, where a shape varies, or Model Data Design Variables, where some data like initial temperature varies.

Secondly, we’re defining Constraints. These are results that we want to specify as a maximum or minimum acceptable condition. For example, we might set a value of 0.995 as a minimum acceptable value for Material Density.

Third, we’re defining an Objective Function. This is the one result that we are trying to achieve. This might be maximization of yield, or minimization of the FCC Criterion (microporosity).

In addition, there are some Optimization Parameters and Simulation Parameters that we can set.


QUESTION: 
Is there a limit on the number of Design Variables in an Optimization Project?

ANSWER: No, but you should in general try to keep the number of Design Variable as small as possible, to reduce run times. If the riser height can’t vary, don’t make it a Design Variable. More DV’s increase the number of simulations required.


QUESTION:
Can a model shape be included in more than one Design Variable?

ANSWER: No. A Specific shape in a model can be included in ONLY ONE Design Variable.


QUESTION:
Does OPTICast save the results of every simulation that it runs?

ANSWER: No. Only the final model and simulation results are available for loading and plotting in SOLIDCast. The intermediate results can be seen with the charts and spreadsheet reports.


QUESTION: 
Will OPTICast always be able to find an optimum solution?

ANSWER: No. In some cases OPTICast may not be able to find a feasible solution, no matter what size the risers are extended to. For example, if Material Density of 0.998 is set as a Constraint, it is possible that given an initial design, OPTICast may not find ANY design which satisfies this Constraint.


QUESTION:
How can I establish a better initial design?

ANSWER: One good way is to run a simulation of the casting without any gates or risers in SOLIDCast, and locate the last areas to solidify in each part of the casting. Feeding these areas will provide the best chance of a successful design. The Gating and Riser Design Wizards included in SOLIDCast provide the best single source to create a good initial rigging design.

QUESTION: 
What factors make the best Constraints and Objective Functions to select?

ANSWER: We have generally found that for ferrous castings, a good starting point is to select Material Density close to 1.000 as a constraint, and maximizing Yield as an Objective Function.

For non-ferrous castings, specifying the FCC Criterion (microporosity) as a Constraint is also a good idea.


QUESTION:
Are there occasions when maximizing Yield is not necessarily the Objective Function to select?

ANSWER: Yield maximization is generally a goal for medium to high-volume production (given a sound casting as a constraint). For low-volume or prototype production, producing a sound part (maximizing Material Density or minimizing FCC Criterion or Niyama) might be a more reasonable objective.


QUESTION:
Can OPTICast move a riser to change its location on the casting?

ANSWER: No. Not in this release. However, we will be considering a method of incorporating this capability in a future release.


QUESTION:
Can OPTICast be used to minimize porosity in just a “zone” in the casting rather than the whole casting?

ANSWER: Since OPTICast considers result data in Casting Material only, you could create a casting model using Casting Material for the zone of interest, and Riser Material for the rest of the casting. This would distort the “Yield” calculation; however, you might be able to use “Minimize Solidification Time” as a replacement objective for “Yield Maximization”.


QUESTION:
What type of computer hardware is required for OPTICast?

ANSWER: The best and fastest that you can get! We recommend a milti -core machine with 4GB or more of RAM and a 500 GB hard drive or larger. Also, we recommend a DEDICATED computer, not one that will be used concurrently for other tasks.

TOP