Fillet For Silicon Micro cantilever beam

[san07gita]

Is there any desing guideline for fillet radius for silicon micro beam used for MEMS application. Any idea?
Regards,
San07gita

 

[jedward]

Why do you need the radius? The usual lithographic processing steps would make it difficult to generate such a fillet. Also, consider that that silicon is very near to the perfect material, on the atomic level, and is very strong. It does not seem appropriate to apply some macroscopic concepts of (presumably) fatigue to objects that are using MEMs fabrication concepts.

 

[RPstress]

Single crystal pure Si is very strong and fatigue insensitive. (For essentially flaw free material, static strength in the (110) direction can be around 5000 - 7000 MPa, 700 - 1000 ksi.)

However, it might still pay to stress the beam properly. Even at MEMS scales classical stress concentrations occur.

If so, then you will need an allowable stress for whatever operations the beam is performing in whatever crystal direction is along the beam, and then assess what permissible stress concentration (times the basic beam stress) can occur.

As jedward points out, manufacturing such a feature accurately is another issue.

 

[IRstuff]

Generally, there wouldn't be a fillet even created in the process. A typical beam is a deposited or doped material that is then undercut from the underlying material. There wouldn't even be the possibility of creating a true fillet vertically.

Horizontally, you're limited by the as-etched radius, which is usually a sizable fraction of the smallest linewidth already. Most beams are designed to accommodate the stresses by reducing the flex through a longer arm.

TTFN

Eng-Tips Policies FAQ731-376