Hacker News Reader

Comment by entrep | Hacker News Reader

Mirror deployment was done before insertion, moving the 18 mirrors up from their locked position by 12.5 mm. Now, 3 months of mirror alignment will follow. They will align with a reference star having to first move one mirror a little bit to identify how the mirror affects the alignment, then calculate the alignment and move the mirror. Each mirror's position can be controlled at 10 nm resolution.

View on HN

It's not only position, also deformation by pulling or pushing on the center of the mirror changing it's curvature.

I can't fathom how they have a servo, a linear actuator, or anything else, that can apply pressure to the mirror and keep it there, for years, in the most extreme environments.

Every component of this thing is mind bending.

Here's a good paper on the mirror actuators [1]. Skimming, they look like hexapod rotary stepper architecture that would not need power to hold a position, only to change position.

[1] https://www.esmats.eu/amspapers/pastpapers/pdfs/2006/warden....

Steppers make good sense for that application, what a beautiful design. 7 nm step size with a repeat accuracy of 1 nm is insane precision.

It is a beautiful design, isn't it?

I'll speculate that the engineers who designed this and other ground-breaking aspects of JWST feel the same euphoria that Beethoven might have felt on completing his 9th symphony.

I think beauty wears many masks, but she's in every human effort to bring something new into the world.

"Cryogenic Nano-Actuator for JWST" https://www.esmats.eu/amspapers/pastpapers/pdfs/2006/warden....

Electromagnet with a constant current running through it would do the job nicely I would think? Keep in mind that with the sun always shining where Webb is that power is not really their first constraint.

Correct. And since the beryllium mirrors are 8 times harder than steel, it's probably a tough job.

Not really. They are quite thin and the deformation is minute, on the order of less than a micron at the largest point.

Edit: that should have been 10 micron.

Reply on news.ycombinator.com