Supporting Structure and Bonding Methods of Strain Gauges

Wednesday, November 22nd, 2017 - Force, Piezoelectric, Resistive Transducers

Supporting Structure and Bonding Methods of Strain Gauges

There are two main structures for holding piezoresistive sensing elements: membrane type (Figure 1.a, typically in pressure and flow sensors) and cantilever beam type (Figure 1.b, typically in acceleration sensors).

Supporting Structure and Bonding Methods of Strain Gauges

Figure 1. Cross sections of (a) membrane-type and (b) cantilever beam-type of piezoresistive sensors

A membrane-type sensor consists of a thin silicon membrane or plate with a load (e.g., pressure) differential across the plate. The resulting deformation causes strain along the edges of the plate. Usually the plate, supported by a thicker silicon rim, is fabricated by etching away the bulk silicon in a defined region until the required thickness is reached. Both monocrystalline and polycrystalline silicon can be used for membranes. With careful design, regions along the edge of the plate can be “doped” to create resistors, which will subsequently exhibit resistance change in proportion to the applied strain. Most sensors with square membranes have four or more piezoresistors positioned at the four edges of the membrane and as close to the edge center as possible (see Figure 2) where the stresses are maximal. From this central point, stress drops more rapidly toward the center of the membrane than toward the corners. Thus perpendicularly placed (relative to the edge) piezoresistors are less sensitive than the parallelly placed ones.

In beam-type sensors, the stress caused by deflection of the inertial mass under a measurand (e.g., acceleration) is concentrated on the surface of the beam, where piezoresistors are placed.

Various layouts of piezoresistors on a square membrane

Figure 2. Various layouts of piezoresistors on a square membrane

Figure 3 shows four bonding methods for strain gauge:

  1. adhesive bonding with backing,
  2. adhesive bonding without backing,
  3. deposited molecular bonding, and
  4. diffused molecular bonding.
Four strain gauge-bonding methods

Figure 3. Four strain gauge-bonding methods: (a) adhesive bonding; (b) diffused bonding; (c) molecular bonding; (d) embedded bonding

In Figure 3.a, a metallic foil strain gauge is directly bonded to a strained surface through a thin layer of epoxy resin. The backing and the adhesive agent work together to transmit strain. The adhesive also serves as an electrical insulator between the foil grid and the strained surface. A semiconductor gauge usually has no backing; therefore, it is diffused to the substrate and is then bonded to the strained surface with a thin layer of epoxy (Figure 3.b). It is smaller and lower in cost than a metallic foil sensor. The epoxy used to attach foil gauges can also be used to bond semiconductor gauges. A thin-film metallic strain gauge is molecularly bonded to the specimen (Figure 3.c) by first depositing an electrical insulation (typically a ceramic) onto the stressed metal surface, and then depositing the strain gauge onto this insulation layer. This results in a much more stable installation with less resistance drift. The errors due to creep and hysteresis therefore are also eliminated. The diffused (or embedded) semiconductor strain gauge also eliminates the need for an adhesive bonding agent (Figure 3.d). It uses photolithography masking techniques and solid-state diffusion of boron to molecularly bond the resistance elements.

The resistance change of piezoresistive gauges is typically measured using a balanced Wheatstone bridge. This allows a small change in resistance to be measured relative to an initial zero resistance value in a balanced bridge, rather than to a large initial resistance value, which greatly improves sensitivity, accuracy, and resolution. Chapter 6 gives the details of a piezoresistive sensor measurement using bridge circuits.

I hope this information about “Supporting Structure and Bonding Methods of Strain Gauges” is useful.