June 2004: Part 8
Previous  Top  Next

27/6/2004
Proposal for first coating products
We are now in the position to develop some first coating prototypes. With our oxides and silver it should be possible to produce low-e and solar control coatings. We will use GenetiCode to find the coating designs, and then try to realize coating prototypes with DD1. In the case of success, we will try to transfer the coating production to a large area coating equipment to manufacture our first real products.

A new GenetiCode configuration based on the CODE configuration window_design.wcd is prepared. A double pane window using 4 mm Float A glass and a 16 mm spacing with Ar gas (90% Ar, 10% air) is used, and the best design for a low-e coating with 3 layers on the bottom pane is searched for. The genetic algorithm tries to reach the target value1 1.0 for the light transmittance (D65 light source) and 1.1 for the U-value (EN 673). The following best design (with a fitness of 21.8) is found:

clip0057

We re-fine the design a little bit by setting a target value of 0.0 for both a* and b* of the reflectance spectrum with weight factor 0.001, in addition to the previous targets (1.0 for light transmittance, 1.1 for U, both with weight factor 1.0). This way we try to obtain a coating which leads to a neutral appearance when viewed from the outside. The result of the optimization is very similar to the one shown above, basically because the reflectance color of the first design was already quite neutral:

clip0058

The virtual office tower equipped with our first window product really looks neutral:

clip0059

This design is saved as our first coating in the database (subfolder coatings, filename: Low-e_neutral_prototype_1.ls). The GenetiCode configuration is save in the subfolder software configurations/GenetiCode using the filename double_pane_low_e_3layers_neutral.gcd. Soon we will try to produce a coating like this and we hope to verify the design experimentally.

Now we try to obtain a more colorful appearance of the building. We change the target value for b* (reflectance) from 0.0 to 30.0 while all other target values remain unchanged. Formally the best result (fitness 8.3) is this:

clip0060

However, the unprotected silver layer on top of the coating will not be stable and this design is rejected. We should not include silver as a possible material for the top layer in the next GenetiCode runs in order to avoid solutions like this.
The second best design (fitness 8.1) looks more reasonable:

clip0061

The goal to achieve a more colorful building is clearly reached:

clip0062

This coating design is also saved to the database, using the filename Low-e_blue_prototype_1.ls.