ENGINEERING: MEMS
lateral dimensions of tens to hundreds of micrometers, significant forces might be needed to pull apart two surfaces in contact and initiate motion. These and other factors make the
design and modelling of MEMS a unique engineering discipline. At small scales, adds Bernt Nilsson, senior VP of marketing at Comsol, resonators, sensors, actuators, piezoelectric and microfluidic system designs must consider the effects of physical phenomena such as electromagnetic- structural, thermal-structural and fluid- structure interactions. He says that, even though the same laws of physics apply both in the macro and micro world, certain effects predominant in the micro scale and adds the example of film damping, which often appears in microsystems. Here, a narrow gas film surrounds vibrating structures; this is an important factor in accelerometers and resonators. And while film damping can be unwanted, it can also be used for adjusting the transient operation of a component. In general, if gravity plays a far smaller
role in the micro domain, virtually all other physical, electromechanical and thermal effects take on major significance, and you can’t ignore any of them if a model is to come close to mimicking reality. In other words, you must consider MEMS design as a multiphysics problem (see page 26).
The multidomain challenge It’s true that multiphysics software lies at the heart of MEMS design; you need it to find out how to construct a given design, what the optimum physical dimensions are and how it performs. And while multiphysics accounts for the various underlying physics that are so critical for MEMS operations, comprehensive MEMS design software takes that all-inclusive concept a big step further because it must also be multi-disciplinary. That’s because MEMS devices are fabricated at the silicon level, so engineers must also closely consider process, fabrication and packaging aspects, as well. Almost all MEMS devices are tightly integrated with electronics, either on a common silicon substrate or in the same package, yet MEMS design has traditionally been separated from IC design and verification. According to one analyst report, it takes four years of development and $US 45m to bring a MEMS product to market – a
www.scientific-computing.com
In this triple-piston micro steam engine, water inside three compression cylinders is heated by electric current and vaporises, pushing the piston out; capillary forces then retract the piston once current is removed. (Image courtesy Sandia National Laboratories)
situation that is untenable in the fast-paced, cost-sensitive consumer market. A critical key in the ‘democratisation’ of MEMS design is to create an integrated design flow for MEMS devices and the circuits they interact with using a structured design approach that avoids manual handoffs. Because MEMS and the accompanying IC must be designed and optimised together, MEMS tools must be compatible with widely used IC design and verification tools.
‘Even though the same laws of physics apply both in the macro and micro world, certain effects predominant in the micro scale’
Thus, MEMS software has developed dramatically. ‘When I joined Analog Devices 16 years ago,’ comments Michael Judy, ‘we had hand-cobbled Matlab code, and some vendors had modified code for capacitance extrapolation.’ Since then, software suppliers have come up with packages to handle these tasks, and Judy points out Coventor and Ansys as being two big tools used for coupled domains, while LabView is used extensively in the evaluation phase of the design for testing and benchmarking.
Mathcad from PTC is also used for the design of 1st-order models, which are then merged into 3D tools. Finally, when it comes time to bring these devices into silicon with support circuitry, popular choices are Spice (plus its own Adice) plus chip-design tools from Cadence and Synopsis. Judy adds that a big challenge in MEMS is
that they don’t fit into the normal primitives of capacitors or resistors; MEMS and supporting circuitry have traditionally been worked on in parallel and then manually stitched together, but these areas are slowly moving together into a single design flow. Getting everything to work together where the circuits are exposed to stresses and other factors have contributed to MEMS design cycles being longer than traditional system design cycles. ‘We believe that the co-design of MEMS
and electronics together will be the key determinant of success for MEMS companies and the industry,’ says Dan Hamon, VP of business development at SoftMEMS. ‘We see some MEMS start-ups failing because of their lack of circuit experience, not because of their MEMS expertise. This is because MEMS-based systems are getting more complex with higher levels of integration and with more value-add beyond the MEMS device. In addition, packaging technologies have the potential to improve product design but add additional complexity.’
SCIENTIFIC COMPUTING WORLD OCTOBER/NOVEMBER 2010 45
➤
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56