Avionics Unit Vibration Analysis

Overview

This case study presents a random vibration analysis of a fictional avionics line replaceable unit (LRU) mounted inside a tactical aircraft equipment bay. The goal of the analysis is to demonstrate that the unit and its internal components will survive the random vibration environment defined in MIL-STD-810H, Method 514.8, without fatigue failure or functional degradation.

The analysis follows a combined analytical and finite element approach: a finite element modal analysis is used to extract natural frequencies and mode shapes, and the resulting modal data feeds into a frequency-domain response analysis to predict structural response under the specified random vibration input. Component-level fatigue life is then assessed using the Steinberg three-band technique.

Input Environment

The input power spectral density (PSD) profile is defined per MIL-STD-810H Category 7 (jet aircraft, fixed wing, under-wing store). The profile is applied as a base excitation at the unit mounting interface in each of three orthogonal axes. The overall GRMS level is computed by integrating the area under the PSD curve across the frequency bandwidth.

Finite Element Model & Modal Analysis

A finite element model of the avionics LRU chassis is constructed in ANSYS Mechanical using shell and solid elements. The chassis is modeled as Al 6061-T6 with appropriate elastic modulus, Poisson's ratio, and density. PCB assemblies are represented as smeared orthotropic plates to capture their mass and stiffness contribution without detailed component-level modeling.

A free-free modal analysis is first performed to verify model integrity, followed by a fixed-base modal analysis with all mounting bolt locations constrained. The first ten natural frequencies and corresponding mode shapes are extracted. The first three modes of interest are summarized below.

Response Analysis & GRMS Check

A random vibration response analysis is performed using the modal superposition method with 2% critical damping applied uniformly across all modes. Response PSD curves are generated at critical locations including chassis mounting feet, PCB card guides, and connector attachment points.

Miles' equation is applied as a quick-check tool to estimate the peak 3-sigma GRMS response at each resonance, assuming a single-degree-of-freedom (SDOF) approximation. The transmissibility (Q factor) at each mode is used to amplify the input PSD at the natural frequency and compute the response GRMS.

Steinberg Fatigue Assessment

Component-level fatigue life on the PCB assemblies is assessed using the Steinberg three-band technique. This method distributes the response GRMS into three stress bands — 1σ (68.3% of time), 2σ (27.1% of time), and 3σ (4.33% of time) — and applies Miner's cumulative damage rule to estimate fatigue life relative to the required test duration.

The allowable PCB deflection at the component center is computed per the Steinberg criteria as a function of board length, component length, and lead wire configuration. Components with deflection ratios exceeding 1.0 are flagged for design review or conformal coating addition.

Conclusions & Recommendations

The avionics LRU chassis structure demonstrates positive margins under the MIL-STD-810H random vibration environment across all three input axes. First natural frequency of 187 Hz falls within the flat portion of the input PSD, resulting in moderate amplification levels at primary structural modes.

The PCB assembly (CCA-1) shows a response GRMS of 11.4 G at its first bending mode, which is within acceptable structural limits for the chassis but warrants component-level review. The transformer (T1) at board center exceeds the Steinberg allowable deflection criterion with a ratio of 1.12 and is recommended for conformal coating to improve fatigue life. All other components pass the Steinberg criteria with margin.

No structural redesign is required. The primary action item is addition of conformal coating to the PCB assembly in the vicinity of T1 prior to qualification testing.