1 Introduction
5
Fig. 1.2 Passive droplet microfluidic network proposed in [13]
to manipulate droplets, while passive modules rely on the variation of applied
pressures, geometries, and fluid properties to manipulate droplets. For example,
the microfluidic network shown in Fig. 1.2 implements a passive trapping well,
which allows to trap, merge, and mix two different kinds of droplets (i.e., white and
gray droplets). This microfluidic network can be applied to screen drug compounds
that inhibit the tau-peptide aggregation [13]. Overall, droplet microfluidic networks
support a diverse range of experiments including the synthesis of biomolecules, drug
delivery, and diagnostic testing [117].
However, designing microfluidic networks is a complicated task. In fact, in the
design process of a microfluidic network, the designer has to conduct several tasks
for determining, e.g., the applied modules and their connectivity, the dimensions of
channels, the applied pressures, the used phases, etc. This requires the consideration
of a large number of design parameters, which all affect the functionality of
the resulting device. Thus far, designers often rely on their expert knowledge
and derive the design based on manual calculations, simplifications, as well as
assumptions. For example, designers frequently simplify or ignore time-dependent
effects on the hydrodynamic resistance of channels caused by droplets because it is
simply infeasible to manually consider all droplet states and positions. Due to the
complexity of these tasks, designers often end up with situations where they cannot
predict the consequences of their design decisions.
Therefore, in order to test and validate the respective designs, usually prototypes
are fabricated, which are used to check whether they correctly realize the desired
functionality, i.e. correctly implement the experiments. In case that the functionality
is not implemented as desired, the designer has to go back, revise the design, and
repeat the prototyping before he/she can conclude whether the revisions eventually
lead to the desired result. This often results in multiple iterations for prototyping
(even dozens of iterations can be necessary depending on the complexity of the
microfluidic network), where each iteration requires the generation or modifications
of the design, the fabrication of the prototype, and the execution of the physical
5
Fig. 1.2 Passive droplet microfluidic network proposed in [13]
to manipulate droplets, while passive modules rely on the variation of applied
pressures, geometries, and fluid properties to manipulate droplets. For example,
the microfluidic network shown in Fig. 1.2 implements a passive trapping well,
which allows to trap, merge, and mix two different kinds of droplets (i.e., white and
gray droplets). This microfluidic network can be applied to screen drug compounds
that inhibit the tau-peptide aggregation [13]. Overall, droplet microfluidic networks
support a diverse range of experiments including the synthesis of biomolecules, drug
delivery, and diagnostic testing [117].
However, designing microfluidic networks is a complicated task. In fact, in the
design process of a microfluidic network, the designer has to conduct several tasks
for determining, e.g., the applied modules and their connectivity, the dimensions of
channels, the applied pressures, the used phases, etc. This requires the consideration
of a large number of design parameters, which all affect the functionality of
the resulting device. Thus far, designers often rely on their expert knowledge
and derive the design based on manual calculations, simplifications, as well as
assumptions. For example, designers frequently simplify or ignore time-dependent
effects on the hydrodynamic resistance of channels caused by droplets because it is
simply infeasible to manually consider all droplet states and positions. Due to the
complexity of these tasks, designers often end up with situations where they cannot
predict the consequences of their design decisions.
Therefore, in order to test and validate the respective designs, usually prototypes
are fabricated, which are used to check whether they correctly realize the desired
functionality, i.e. correctly implement the experiments. In case that the functionality
is not implemented as desired, the designer has to go back, revise the design, and
repeat the prototyping before he/she can conclude whether the revisions eventually
lead to the desired result. This often results in multiple iterations for prototyping
(even dozens of iterations can be necessary depending on the complexity of the
microfluidic network), where each iteration requires the generation or modifications
of the design, the fabrication of the prototype, and the execution of the physical