For my solar tracking sensor I need transparent domes with good optical performance.
The classic thermoformin method is vacuum forming, where once the plastic sheet is softened wit heat, it is pushed against a dye (that wooden semi sphere you see on the table, near the tape) and vacum between the table and the frame holding the plastic sheet will make the plastic adhere to the dye. This technique results in good geometry, but leaves some marks on the dome.
Therfore I went the other way around and used compressed air to blow the dome from below. You see the heater “blowing table” with a hole in the middle and four seals around it. The frame holding the plastic is on the heater. Once the plastic sags, it is ready to be pushed on the blowing table. It takes some trial&errors to figure out how much time to stay on the heater, and how much to push on the blow table. In fact, the dome form quickly but some pressure is needed to hold it in shape, or it will “deflate” while is still warm. On the other hand, pushing too much on the rubber seals, the dome will come out too tall, or explode.
A few tests with 2mm plexiglass before getting decent results.
Note how the top of the dome is flat, as the air jet from the hole of the blowing table cools the plastic in the middle faster than other areas, resulting in less deformation where colder.
It is possible to make rectangular domes (picture below, left), by using a frame with a rectangular hole and rearranging the seals position.