4. Sensor placement and mounting effects on crash signal quality

In a restraint control system, the accelerometer is not a passive observer. Where it sits and how it is attached determine what the crash signal actually contains. Two vehicles with identical airbag ECUs and identical calibration can deploy differently if one sensor is mounted on a stiff rocker-section bracket and the other on a flexible floor pan. This chapter examines how structural response, mechanical filtering, and resonance shape the signal that reaches the deployment algorithm, and how to reason about placement and mounting as signal-conditioning decisions rather than mechanical afterthoughts.

From Vehicle Motion to Sensor Output

A frontal barrier impact produces a rapid deceleration of the vehicle body. The occupant compartment does not decelerate uniformly: the front structure crushes and absorbs energy, while the passenger cell follows a delayed, lower-amplitude pulse. The accelerometer measures the local motion of whatever structure it is bolted to, not the idealized rigid-body deceleration. The signal chain is therefore:

• Crash event — barrier or vehicle-to-vehicle impact with a defined closing speed and overlap.
• Structural response — crush, load-path redistribution, and local panel vibration at the mounting point.
• Mounting interface — bracket stiffness, fastener torque, and contact area act as a mechanical filter.
• Sensor element — MEMS proof mass and its internal damping convert local acceleration into an electrical signal.
• ECU signal conditioning — anti-alias filtering, gain, and offset correction before the deployment algorithm.

The key insight is that the first two stages are not noise to be removed; they are the physical information the algorithm must interpret. A mounting point that is too compliant attenuates the early crash pulse and delays deployment. A mounting point that is too stiff and lightly damped transmits high-frequency ringing that can confuse discrimination logic.

Structural Response at the Mounting Point

During a 50 km/h full-frontal rigid-barrier test, a typical body-mounted accelerometer records a pulse that rises to roughly 30–50 g within 10–20 ms, followed by a complex decay with oscillations. The exact shape depends on where the sensor is mounted. Consider three common locations:

These are illustrative characteristics, not fixed values. The point is that placement selects which structural modes dominate the measurement. A sensor on the front structure sees the crash earliest but also sees the most severe local deformation, which may not represent the occupant compartment deceleration that the restraint system must respond to.

Mounting as a Mechanical Filter

The bracket and fastener form a spring-mass-damper system between the vehicle structure and the sensor. Its natural frequency and damping ratio determine how the mounting modifies the structural signal. A simplified model treats the bracket as a cantilever or plate with stiffness k and the sensor mass m, giving:

f_n = (1 / 2π) · sqrt(k / m)

where:
  f_n = mounting natural frequency [Hz]
  k   = bracket stiffness [N/m]
  m   = sensor mass [kg]

For a 30 g sensor with a bracket stiffness of 2×10⁶ N/m, the natural frequency is approximately 1.3 kHz. If the bracket is softened to 2×10⁵ N/m, the frequency drops to about 410 Hz. Since the crash pulse energy is concentrated below roughly 300–500 Hz for full-frontal events, a mounting resonance in that range will amplify or attenuate specific frequency components and can introduce phase distortion.

Damping is equally important. A lightly damped bracket (ζ < 0.05) produces pronounced ringing at its natural frequency after the initial pulse. This ringing appears as oscillatory content in the accelerometer output and can cause the deployment algorithm to see multiple threshold crossings where the true structural event had only one. Increasing interface damping — through elastomeric washers, controlled torque, or a damped bracket design — reduces this effect.

Resonance and Signal Quality

Resonance is not inherently a defect. Every mechanical mounting has a natural frequency. The engineering question is whether that frequency and its damping place the resonance outside the band of interest, or whether the algorithm can tolerate its effects. Three practical considerations apply:

1. Keep the mounting resonance above the crash pulse bandwidth. If the structural pulse content is below 500 Hz, a mounting resonance above 1 kHz is generally safe. This is why stiff brackets and high-torque fasteners are preferred for crash sensors.
2. Characterize the mounting with a tap test or shaker sweep. A simple modal check — instrument the bracket, tap it with a calibrated hammer, and record the response — reveals the natural frequency and damping. This is a standard validation step before vehicle-level crash testing.
3. Verify in the vehicle environment. Bracket resonance measured on a bench may shift when the bracket is assembled into the body structure, because the body adds stiffness and mass. Always confirm on the actual vehicle or a representative body-in-white.

Signal quality also depends on the sensor's own internal characteristics. MEMS accelerometers have an internal proof-mass resonance, typically in the 2–10 kHz range, and internal damping that is designed to suppress it. The ECU's anti-alias filter must attenuate content above the Nyquist frequency to prevent aliasing. If the mounting resonance excites the sensor's internal resonance, or if the anti-alias filter is inadequate, high-frequency energy can fold back into the measurement band and corrupt the crash signal.

Practical Placement and Validation Workflow

The following workflow reflects common practice in restraint system development. It is presented as a structured example, not as a claim about any specific vehicle program.

1. Define the sensing requirement. Determine the crash modes the system must detect (full frontal, offset, angular, rear, side) and the required detection time window for each. This sets the frequency band and amplitude range of interest.
2. Survey candidate locations. Use body-in-white modal analysis or existing vehicle data to identify locations with appropriate structural response. Avoid areas with known local flexibility or severe deformation.
3. Design the mounting interface. Specify bracket material, thickness, and geometry to achieve a natural frequency well above the crash pulse bandwidth. Include damping features if needed.
4. Bench-characterize the mounting. Perform a tap test or shaker sweep on the bracket-sensor assembly. Record the frequency response function and extract natural frequency and damping ratio.
5. Validate in vehicle. Conduct sled or full-vehicle crash tests with the sensor in the candidate location. Compare the recorded signal against reference measurements and against the algorithm's expected input.
6. Iterate if needed. If the signal shows excessive ringing, attenuation, or timing shift, revise the bracket or relocate the sensor and repeat validation.

Key Takeaways

• Sensor placement selects which structural response the algorithm sees; it is a signal-conditioning decision, not just a packaging choice.
• The mounting interface acts as a mechanical filter with its own natural frequency and damping; both must be characterized and controlled.
• Mounting resonance above the crash pulse bandwidth and adequate damping preserve signal fidelity; resonance within the band distorts amplitude and timing.
• Validation requires bench characterization followed by in-vehicle confirmation, because the body structure modifies the mounting behavior.
• Signal quality depends on the combined behavior of structure, mounting, sensor element, and ECU filtering — no single stage can compensate for a poorly designed interface.