it is not a simple task but I need to fabricate the following structure:
Channel:
Enclosed view of the 50x50um part:
It is a microfluidic channel (master, it will be molded later), with various width, average thickness is 0.4mm. There is a small part of 0.05x0.05mm cross section, and it is the tricky point. All the others parts of the channel are easily achieved at the correct value.
Normally I will have to print it on a substrate (either glass or silicon) but I also tried to make it directly on the build plate.
I used the solus proto with 42xy. I tried exposure values from 2800 to 5000 but I was not able to get the small section of the channel on a substrate or the build plate. I tried to 10 to 25um for z thickness
I print the structure flat without supports (can't be changed)
I was quite easily able to get it printed with a layer below of 1mm, but normally I can't accepted a layer below and I need at leat 1cm flat area around the part.
Some comments:
- I see sometimes an error on the total thickness (not only this process) of about 0.1mm, I don't know if it the 0 of the z. I tried to calibrate many times the build plates, to shift the part in z by 0.1mm (with that I was able to get a part of the channel but in general I had poor adhesion on the build plate) or just to increase the pieces in z by 0.1mm (most efficient to get high accuracy parts so far for me). So far, none of these solutions had worked.
- As in semiconductor photolithography, I know there can be some change with polymerization due to change of reflectivity, so I tried to add as the 2nd design a thin layer 0.05 to 0.5mm (tried different value) but with the needed surface (see below). I was not able to get the channel, still with high subtrate value. I saw with this design that I had never get the right total thickness value for the channel (the thick part at 0.4mm) it is in general thinner. I then increase the exposure value to 10000ms and then I get the small part of the channel appearing but still the thickness is not right.
Maybe I did something wrong with the setup of the solus, maybe it is not feasible... I also feel that the part with the large surface is sticking to the peeling film, explainning eventually the problem of thickness error.
Any others ideas to test or comments?
Sebastien