3.5 Conclusion
49
For an early as possible evaluation and validation of the chosen design parameters, simulations on the 1D analysis model can be applied. Based on this model
an advanced simulation framework was presented, which addresses severe shortcomings of state-of-the-art simulators by being dedicated to droplet microfluidics
(cf. directly works on the specification of the microfluidic network) and by
addressing essential physical phenomena, which are required for practically relevant
applications. Furthermore, the open-source implementation allows for a broad
application of the framework and even further extensions.
Finally, this chapter presented a case study demonstrating how simulation can
help in the design process of droplet microfluidic networks; here by means of the
trapping well proposed in [13]. Therefore, this case study compared the design
process when simulation has been used with the design process when no simulation
has been used. When no simulation has been used, the designer validated the
specification by using physical experiments, which resulted in six prototypes, took
one person month, and produced financial costs of USD 1200. If the designer is not
experienced, the costs could even be much higher. Instead, when a simulation has
been used, the designer was able to validate the specification before any prototype
and even before any physical design was made. These simulations allowed the
prediction of the robustness of the respective specifications and, in fact, showed a
clear preference for the specification which was eventually realized in [13]. Hence,
the simulations allowed for selecting the most robust design without the need to
explicitly fabricate and test them. Furthermore, the simulations even allowed to
explore further designs, which were too costly to consider in the current design
process without simulation.
However, before the simulation framework proposed in this chapter can be
applied, the designer first has to derive a specification of the microfluidic network,
i.e. the microfluidic network needs to be dimensioned. For this dimensioning,
designers manually conduct calculations thus far to eventually obtain a specification
fulfilling the objectives. In order to support the designer in this task, the next chapter
proposes methods to validate a specification as well as to conduct the dimensioning
in an automatic fashion. These methods are exploiting the 1D analysis model
reviewed in this chapter. Hence, this chapter provides the basis for the following
presented methods.
49
For an early as possible evaluation and validation of the chosen design parameters, simulations on the 1D analysis model can be applied. Based on this model
an advanced simulation framework was presented, which addresses severe shortcomings of state-of-the-art simulators by being dedicated to droplet microfluidics
(cf. directly works on the specification of the microfluidic network) and by
addressing essential physical phenomena, which are required for practically relevant
applications. Furthermore, the open-source implementation allows for a broad
application of the framework and even further extensions.
Finally, this chapter presented a case study demonstrating how simulation can
help in the design process of droplet microfluidic networks; here by means of the
trapping well proposed in [13]. Therefore, this case study compared the design
process when simulation has been used with the design process when no simulation
has been used. When no simulation has been used, the designer validated the
specification by using physical experiments, which resulted in six prototypes, took
one person month, and produced financial costs of USD 1200. If the designer is not
experienced, the costs could even be much higher. Instead, when a simulation has
been used, the designer was able to validate the specification before any prototype
and even before any physical design was made. These simulations allowed the
prediction of the robustness of the respective specifications and, in fact, showed a
clear preference for the specification which was eventually realized in [13]. Hence,
the simulations allowed for selecting the most robust design without the need to
explicitly fabricate and test them. Furthermore, the simulations even allowed to
explore further designs, which were too costly to consider in the current design
process without simulation.
However, before the simulation framework proposed in this chapter can be
applied, the designer first has to derive a specification of the microfluidic network,
i.e. the microfluidic network needs to be dimensioned. For this dimensioning,
designers manually conduct calculations thus far to eventually obtain a specification
fulfilling the objectives. In order to support the designer in this task, the next chapter
proposes methods to validate a specification as well as to conduct the dimensioning
in an automatic fashion. These methods are exploiting the 1D analysis model
reviewed in this chapter. Hence, this chapter provides the basis for the following
presented methods.