3.4 Case Study
43
designs are explored. However, probably different designs would be even more
suited, e.g. would be more robust, would be smaller, would increase the throughput,
or would have a positive effect on the application.
For the case study, no design has been explored to answer the following
questions:
1. Question 1: What is the minimal bypass channel length so that droplets still get
trapped?
2. Question 2: What is the maximum pressure over five sets of trapping wells so that
no objective is violated?
3. Question 3: How many trapping wells can be cascaded and loaded by droplets
in a given time, i.e. what is the maximal throughput?
Why these questions are important and how to address them will be discussed later
on page 46 (in part “Utilizing the Simulator for Design Exploration”).
Overall, it can clearly be seen that the design process as conducted thus far
is certainly not ideal. Deriving a working specification requires several iterations
of physical design and prototyping, which becomes a time-demanding and costly
process. Moreover, as a consequence, often the first working design is eventually
used even if better and more advanced solutions would, in principle, be possible.
3.4.3 Improved Design Process Using Simulation
In this case study, the design process is re-visited but now employing the simulation
framework proposed in Sect. 3.3. Simulations allow to predict the behavior of
microfluidic specifications before physical experiments are conducted. One simulation result for the device of [13] is shown in the video available at http://
iic.jku.at/eda/research/microfluidics_simulation/. Additionally, Fig. 3.11 provides a
comparison of the simulator and the physical experiment (images of the physical
experiment were captured with a frequency of 50 fps). On the left-hand side of this
video and figure, the graphical output of the simulation including the current droplet
positions is shown. On the right-hand side of this video and figure, the simulation
is compared to a physical experiment. When comparing the simulation with the
experiment, equal droplet paths as well as similar droplet speeds and residence
times can be observed. Furthermore, the simulator constantly checks the objectives
introduced in Sect. 3.4.1.
The simulator and its features can now be utilized in order to address major
shortcomings of the current design process for microfluidic networks as reviewed
in Sect. 3.4.2. Recall that the designer derives the specification based on manual
calculations, simplifications, as well as assumptions and can, thus far, only validate
the specification by fabricating and testing the resulted prototype. Utilizing the
simulator, many of these tests can now be conducted much earlier in the design
process and without the need for either a physical design or a fabricated device. This
additionally allows for a much more elaborated design exploration as variances of
43
designs are explored. However, probably different designs would be even more
suited, e.g. would be more robust, would be smaller, would increase the throughput,
or would have a positive effect on the application.
For the case study, no design has been explored to answer the following
questions:
1. Question 1: What is the minimal bypass channel length so that droplets still get
trapped?
2. Question 2: What is the maximum pressure over five sets of trapping wells so that
no objective is violated?
3. Question 3: How many trapping wells can be cascaded and loaded by droplets
in a given time, i.e. what is the maximal throughput?
Why these questions are important and how to address them will be discussed later
on page 46 (in part “Utilizing the Simulator for Design Exploration”).
Overall, it can clearly be seen that the design process as conducted thus far
is certainly not ideal. Deriving a working specification requires several iterations
of physical design and prototyping, which becomes a time-demanding and costly
process. Moreover, as a consequence, often the first working design is eventually
used even if better and more advanced solutions would, in principle, be possible.
3.4.3 Improved Design Process Using Simulation
In this case study, the design process is re-visited but now employing the simulation
framework proposed in Sect. 3.3. Simulations allow to predict the behavior of
microfluidic specifications before physical experiments are conducted. One simulation result for the device of [13] is shown in the video available at http://
iic.jku.at/eda/research/microfluidics_simulation/. Additionally, Fig. 3.11 provides a
comparison of the simulator and the physical experiment (images of the physical
experiment were captured with a frequency of 50 fps). On the left-hand side of this
video and figure, the graphical output of the simulation including the current droplet
positions is shown. On the right-hand side of this video and figure, the simulation
is compared to a physical experiment. When comparing the simulation with the
experiment, equal droplet paths as well as similar droplet speeds and residence
times can be observed. Furthermore, the simulator constantly checks the objectives
introduced in Sect. 3.4.1.
The simulator and its features can now be utilized in order to address major
shortcomings of the current design process for microfluidic networks as reviewed
in Sect. 3.4.2. Recall that the designer derives the specification based on manual
calculations, simplifications, as well as assumptions and can, thus far, only validate
the specification by fabricating and testing the resulted prototype. Utilizing the
simulator, many of these tests can now be conducted much earlier in the design
process and without the need for either a physical design or a fabricated device. This
additionally allows for a much more elaborated design exploration as variances of