Digitalizing material - Appendix.



index.png

Resume

Essays

Resources



nav.png

Home
Up

Appendix.
Building a prototype.


This appendix is a preliminary design for building a IMAT prototype using an IMAT.

Assume we have 5 materials in "cubes" of 50 µm side (Figure A-1) .




Figure A-1


The injectors can have the configuration of Figure A-2, in this configuration the ratio is 14:1 and isolator-1 cannot attach to the other materials. That is the piezoelectric film is free to move. The "x" show the empty conduit for moving the material.


../esp/IMAT102.png
Figure A-2

Other configuration more easily built is in Figure A-3. In this configuration the ratio is 32:1, the only assumption is that isolator-2 cannot bind to the conductor material.

../esp/IMAT103.png
Figure A-3

This figures show in profile the bimorph piezoelectric film, to calculate the approximate length the equation (1) [Pennwalt 87] can be used, where V is the voltage, d 31 is the piezoelectric constant, t is the thickness of the piezoelectric material and L is the length needed.


../esp/IMAT104.png

Now the width of the conduit is selected. If the width is 50 µm, the same as the "cubes",then the "cubes" can have a difficult time moving, because of this, a postprocess which can expand the width in 10-20% is needed, perhaps using an etching procedure or build a machine to grind the conduits. That is feasible because the conduits are accessible for one side of the machine. Additionally this machine is probably needed for periodic cleaning maintenance of the conduits.

Taking the final width of the conduit as 60 µm, then Æx = 60 µm, suppose V = 100 volts and d 31 = 50 x 10 -12 C/N this value is in between PVDF (plastic, Kynar) and BaTiO 3 (ceramic) . With these values L = 6.32 mm. This value assumes that the film closes all of the conduit. That is not necessary if the tip of the film touches the side of the last cube and presses into the wall then the column of cubes stop moving. For this case a tip displacement of only 10 µm is required and L = 2.58 mm; using L = 5 mm is a conservative approximation.


Assuming a work area of 10 cm per side, a ratio of 100:1 (symmetric in X and Y) and using 10 materials at the time; with these values 20 sections of 5 mm each are required, each section with 10 transversal injectors (one for each of 10 materials) and 200 longitudinal injectors for a total of 2000 injectors per section.

For controlling 2000 injectors, a matrix of 50 by 40 electrodes is proposed, that is 90 conductors per section. Because the piezoelectric film is a capacitor, cycles of charge/discharge of the capacitors are needed, using 100 Hz as a work frequency and 10 times more, that is 1 kHz as a multiplexing frequency for 40 lines, then a 40 kHz frequency using 50 control lines can control the 2000 injectors.


Figure A-4 shows how the top of the piezoelectric film can join a reservoir of material (here 2x3) in top of this is a feeder channel (here 4x3) , the film extend two electric connections to the top of the assembly. Two transversal cuts are show also.



Figure A-4


To complete the electrical connection 4 injectors are grouped at the time, Figure A-5 shows this grouping for one material and Figure A-6 for 10 materials.

../esp/IMAT112.png
Figure A-5




Figure A-6

This last figure shows also the 40 multiplexed lines. This can be above and connected to the M lines. The Common line can be connected with one of the 50 control lines, see Figure A-7. Take notice that in figure A-5 the darker parts correspond to the top of figure A-4, the extra space is needed for maintaining the 100:1 ratio (and simplifying the electrical connections) . The Figure A-6 is drawn with a smaller ratio (27:1) to simplify the drawing.



Figure A-7

In figure A-7 the lines are apart 2 mm, the control lines can go down in straight lines, while the multiplexed lines can arrange in a horizontal "bus" of 4 mm of width.

The feeder channels can be connected with a larger duct, which is also connected to external tubing to maintain a flow of air and "cubes" as show in Figure A-8. Problems which can arise are erosion in the walls and fragmentation of the "cubes". Also this method does not maintain orientation.



Figure A-8


Finally it is possible that the tip of the film cannot touch the side of the last cube because the cubes are not uniform. A 5-10% variance in the length of the cubes can probably be expected. This problem can be resolved if there exists a flexible interface between the piezoelectric film and the cubes. For example, the conductor facing the cubes can be an elastomer or an extra layer of flexible material can be added. Then the film can touch the last 10-20 "cubes" of the conduit when it is activated (that is 10-20% of the full length of 100 cubes or 5 mm) . This method can also reduce erosion problems at the tip of the piezoelectric film.

Because this is a very preliminary design, it is not possible to define materials, dimensions, preprocess and post processes, thermal and electrical factors and other details. However of the analyzed data, building a machine with the features described looks viable.


© 2000-2003 Javier Diaz R. Up