Design engineers for today's cutting-edge technology companies face a complex set of challenges in the marketplace. A common theme is the requirement to pack greater levels of performance into smaller and lighter packages, all while maintaining or even lowering costs. Faced with such challenges, designers and engineers are increasingly looking to advanced materials and micromachining processes to achieve product performance and cost breakthroughs. That push for miniaturization has translated into an increased demand for the production of high-quality micromachined features and parts in advanced materials.

Advanced materials, including engineered ceramics, glass, sapphire, etc., offer many attractive properties for a range of applications, including structural, semiconductor, microelectromechanical systems (MEMS), medical, defense, aerospace and electronics. However, machining these materials remains technically challenging.

Driven by MEMS and the capability to fabricate silicon-based sensors and systems at the wafer level, previously cost-prohibitive or technologically unattainable devices can now be cost-effectively fabricated with precision. New applications are being introduced daily, resulting in explosive growth in demand for the products that employ them. Similarly, the ability to fabricate precision glass, ceramics and advanced materials with a scalable machining process provides designers and engineers with the cost-effective solutions they need to develop high-performance, miniaturized devices. Ultrasonic machining (UM) is one such scalable machining process.

How Does UM Work?

As shown in Figure 1, during ultrasonic machining, a low-frequency electrical signal is applied to a transducer that converts the electrical energy into high-frequency (~ 20 KHz) mechanical vibration. The mechanical energy is then transmitted to a horn and tool assembly, resulting in a unidirectional vibration of the tool at the ultrasonic frequency with known amplitude.

As a stream of abrasive slurry passes between the tool and the workpiece, the vibrating tool, combined with the abrasive slurry, uniformly abrades the material to create a precise reverse image of the tool shape. Since only a very low force needs to be applied to the abrasive grain, material requirements are reduced and minimal to no damage to the surface occurs. Material removal during the UM process can be classified into three mechanisms:

• mechanical abrasion by the direct hammering of the abrasive particles into the workpiece (major)
• micro-chipping through the impact of the free-moving abrasives (minor)
• cavitation-induced erosion and chemical effect (minor)(1)

Since it is non-thermal, non-chemical and non-electrical, the UM process does not change the workpiece's chemical composition, material microstructure or physical properties. The process can be used for machining conductive and non-metallic materials with hardnesses of greater than 40 HRC (Rockwell Hardness measured in the C scale). Material properties and process parameters, such as the type and size of abrasive grain employed and the amplitude of vibration, as well as material porosity, hardness, and toughness, determine material removal rates and the machined piece's surface roughness.

What Can UM Achieve?

The UM process can be used to generate intricate features in many types of hard and brittle materials, including engineered ceramics and piezoceramics, quartz, single crystal materials, composites, ferrites, and glass. Possible feature shapes include (but are not limited to) round, square and odd-shaped thru-holes and cavities of varying depths, as well as OD/ID features. UM-machined features exhibit hole/cavity walls without taper (draft angle < 0.5°) and rectangular features with square corners approaching 90°.

For small workpieces, wafers, larger substrates and material blanks, features ranging in size from 0.008 in. (200 μm) up to several inches are possible. Depending on the material type and feature size, the process is capable of producing aspect ratios as high as 25:1. While tighter tolerances are possible depending on application requirements and process parameters, variations of feature size, shape and cavity depth are typically held to within a tolerance of ± 0.002 in. (50 μm). Additional benefits include:

• locational tolerance of features relative to fiducials or preexisting features typically held to within ± 0.002 in. (50 μm); tighter tolerances are possible, depending on the application
• little or no sub-surface damage and no heat-affected zone
• reduced stress and a lower likelihood of fractures
• improved performance in downstream machining processes (depending on the application)

Scalability and Parallel Processing

UM has become the machining method of choice for glass and Si constraint wafers used in the fabrication of MEMS pressure sensors for medical, industrial, and automotive applications. Glass/Si constraint wafers are machined with thru hole arrays at the wafer level and supplied to MEMS sensor die manufacturers. The process produces precise features with excellent location and dimensional accuracy and, because multiple features can be machined simultaneously, offers significant cost advantages over competing step-and-repeat machining technologies.

Following a bonding and dicing step, each micromachined hole on the glass/Si wafer becomes part of the final pressure sensor die. Due to the parallel processing capabilities of ultrasonic machining, the process for machining a 500-hole array on a 2-in. wafer or a 2000-hole array on a 4-in. wafer are fundamentally the same. The fabrication costs for the glass and Si constraint wafers do not rise in direct proportion to the size of the wafer or the number of features machined.

For MEMS pressure sensor manufacturers targeting greater numbers of sensors on a wafer and scaling to larger wafer sizes to meet demand, the process results in lower per-die input costs. In fact, for every doubling in wafer or substrate size, there is a fourfold increase in the number of die in a given layout, while the cost differential may be a factor of two or less.

By scaling the machining process accordingly, machining efficiency can be increased and cost-per-die levels decreased in order to achieve application-specific requirements. The scalability of the UM process has been critical in allowing manufacturers to ramp up production while controlling costs, assisting many to transition to lower-cost volume production.

These advantages apply to smaller parts that are not machined at the wafer or substrate level as well. Through the use of precision fixturing and tooling, parts can be fixtured in an array and machined in parallel in a multi-up machining process. By eliminating the costs associated with multiple setups and machining parts simultaneously, the total solution cost can be reduced. In one recent example, a fixtured array process was developed that reduces the total solution cost by 33%.

The process works well for the machining of small parts from advanced materials. With a workpiece or substrate at thickness, precision tooling can be employed to core or stamp out precise parts from hard, brittle materials. Again, the benefits or parallel processing are significant when compared to what can be achieved with conventional milling, laser process and other such step-and-repeat machining processes.

For more information regarding glass and ceramic machining, contact BULLEN at 1301 Miller Williams Rd., Eaton, OH 45320; (937) 456-7133; fax (937) 456-2779; or visit www.bullentech.com.

Links

• BULLEN