8.2 Droplet Sequence Generation
115
Table 8.2 (continued)
Experiment
# Headers
# Tested candidates
Valid?
Time [s]
Microfluidic network B4 with 15 modules, 101 channels, 12 bifurcations
Exp. 1
1
1
✓
<1
Exp. 2
3
1
✓
<1
Exp. 3
3
1
✓
<1
Exp. 4
2
1
✓
<1
Exp. 5
3
1
✓
<1
Exp. 6
3
1
✓
<1
Exp. 7
1
1
✓
<1
Exp. 8
2
1
✓
<1
Exp. 9
2
1
✓
<1
Exp. 10
3
1
✓
<1
Exp. 11
3
1
✓
<1
Exp. 12
7
1
✓
<1
Microfluidic network B5 with 17 modules, 118 channels, 15 bifurcations
Exp. 1
4
3
✓
2
Exp. 2
1
1
✓
<1
Exp. 3
2
1
✓
<1
Exp. 4
–
–
–
–
Exp. 5
–
–
–
–
Exp. 6
5
11
✓
4
Exp. 7
10
3
✓
65
Exp. 8
–
–
–
–
Exp. 9
2
1
✓
<1
Exp. 10
8
4
✓
25
Exp. 11
–
–
–
–
Exp. 12
7
3
✓
5
Exp. 13
12
4
✓
775
Exp. 14
–
–
–
–
Experiment: the experiment for which a droplet sequence should be determined
# Headers: the number of required headers in the obtained droplet sequence
# Tested Candidates: the number of tested candidates until a valid droplet sequence is found
Valid?: Is the droplet sequence valid in the 1D analysis model?
Time [s]: the required run-time in CPU-seconds to obtain the droplet sequence
blockings of channels, unintended coalescing, etc. Accordingly, the number of
candidates and, hence, the run-time of the proposed method increases. In the worst
case, this may even lead to scenarios where no droplet sequence realizing the desired
experiment at the given network can be generated at all. This is the case for the
largest network B5.
Thus far, all these considerations had to be made manually—which was infeasible in many cases. Using the proposed method, this task gets automated—
significantly easing the generation of droplet sequences. Moreover, through the