Professional Documents
Culture Documents
Abstract
A quick and low-cost process chain for bottom grooved micromixer is presented. The process chain consists of
abrasive water jet and pure water jet cutting of two copper electrodes, die-sinking EDM with said electrodes to
produce the steel tool insert and casting of PDMS onto the tool insert to produce the micromixer. The obtained
characteristic dimensions on electrodes, tool inserts and PDMS products were measured and the mapping of
dimensions from one process to the next one was examined. The results show that the precision obtained mainly
depends on the precision of the first process, namely WJ and AWJ machining. The obtained micromixer channel
width was 764 µm and average groove width was 413 µm with relatively high deviations of dimensions in a range
of ± 50 µm.
Keywords: AWJ, EDM, casting PDMS, process chain, micromixer, lab on a chip.
2.3. Casting of PDMS Fig. 6. Measured widths of the features on the tool insert
machined by EDM
A special holder was manufactured and the tool
insert was mounted on it in order to form a pool where 4. Process chain
the PDMS material was casted. Due to the mechanical
properties and low price, QSil216 (ACC Silicones LTD, WJ machining can produce smaller kerf width, but
UK) was used. In order to eliminate micro bubbles that the variation of dimensions is greater that in the case
are present in the liquid PDMS due to mixing and of AWJ machining. Since the width of mixing features
casting, the tool insert, holder and liquid PDMS were should be in the range of 150 µm [6], it is feasible to
exposed to under pressure of 70 mbar for half an hour. use WJ technology to manufacture electrode B for
Total curing time at 25 °C was 20 hours. machining of mixing features and AWJ technology to
manufacture electrode A for machining main channel.
3. Analysis of process performances To further improve mixing capabilities of micromixer,
the mixing features were machined as close as
As expected, kerf width is much smaller in the possible. Therefore, the final process chain is as
case of WJ machining, but standard deviations of follows: AWJ machining of electrode A, WJ machining
dimensions are smaller in the case of AWJ machining of electrode B, EDM machining of main channel by
(Fig. 5). One can notice, that WJ and AWJ machining electrode A and EDM machining of mixing features on
are repeatable since electrode A and B have mean main channel by electrode B. The final product is
value of dimensions within the standard deviations produced by casting of PDMS.
(represented by error bars) of all measurements. Thus, At each process step, the dimensions were
the measurement results of both electrodes can be evaluated and the results are gathered in Fig. 7. Due
merged together and evaluated together. to the taper and rough cut surface on the electrode, a
relatively high wear occurs on the electrode during
EDM process.
1 El. before EDM
0.8 El. after EDM
Tool insert
Width, mm
0.6 Micromixer
0.4
0.2
0 Main Mixing
Fig. 5. Kerf width measured on the bottom of the electrodes
channel features
machined by WJ and AWJ technology. Electrode A is used
to machine main channel channel whereas electrode B is Fig. 7. Characterisation of dimensions obtained at each
used to machine mixing features in the channel. process included in the process chain
From the electrodes the shapes are transferred to The width of main channel increases for 80 µm.
the tool insert by die-sinking EDM process. The Although the taper is greater on electrode B, which is
difference in dimensions are only due to the side gap machined by WJ, the increase of the mixing features
between the electrode and the workpiece, which is width due to EDM is insignificant. The reason is the
defined by EDM machining parameters. Discharges EDM machining time required to produce mixing
with greater energy cause greater gap. Since the features, which is only a half of the machining time
machining parameters with lower energy were used required to machine main channel (Table 1).
when machining the mixing features on the bottom of Comparing the width on the electrode and on the tool
the main channel, these features are wider than the insert, the difference is greater in the case of the main
main channel. Again, the deviation of width is smaller channel manufacturing. Rough and fine machining
when applying AWJ machining (Fig. 6). was applied without orbital motion of the electrode
when machining main channel, whereas only fine rough and fine EDM machining was applied. The
machining was applied when machining mixing required machining time to produce the tool insert is
features. Additionally, fine machining parameters for significantly shorter compared to the machining time
mixing features machining determined lower the other micro technologies require, e.g. laser
discharge energy then fine machining parameters for ablation or micro milling. To further improve process
main channel machining, and hence smaller gap in the performances, especially the obtained precision of the
former case. The difference in mixing features machined features, the machining with orbital motion
dimensions is 18 µm whereas in the case of the main of the tool electrode should be used.
channel, the difference is 190 µm. The shapes and Casting can be used for a serial production, but it
dimension are transferred to the casted micromixer is not really a technology for mass production. The
and width reduction of 25 µm is observed on the main results show that the process parameters were not
channel and mixing features due to the shrinkage of completely under control and hence the dimensions of
material during curing. five micromixers vary significantly. Further research
Finally, five micromixers were casted and their will be focused on technologies for mass replication
dimensions measured. The mean values of main such as micro injection moulding and hot embossing.
channels are varying significantly (764 µm ± 42 µm).
The same is valid for the widths of mixing features (413 Acknowledgements
µm ± 45 µm). For all five micromixers, the same tool
insert was used, thus the process chain has no This work was supported by the Slovenian
influence on the obtained result, but only the process Research Agency (ARRS) (Grant No. P2-0248 (B))
of casting. Before casting, the resin was prepared from and EU EC H2020 funded Era Chair of Micro Process
two components and the percentage of hardener Engineering and Technology – COMPETE (Grant No.
influence on the shrinkage of the resin during curing. 811040).
The variation in percentage of hardener might be the
source of width variation. References