2. CAE discipline taxonomy

In automotive product development, CAE is not a single monolithic activity. It is a portfolio of disciplines, each with its own governing physics, solver technology, modeling conventions, and acceptance criteria. Managing a CAE portfolio effectively requires a clear taxonomy: you must know which discipline owns which attribute, where the boundaries lie, and where the couplings between disciplines create the most risk. This chapter establishes that taxonomy and shows how the disciplines interact across a typical vehicle program.

Why a Discipline Taxonomy Matters

A taxonomy is not an academic exercise. It drives three practical decisions in portfolio management:

• Resource allocation: Crash and NVH models are computationally expensive and require different solver licenses, hardware profiles, and specialist engineers. Knowing the discipline mix lets you plan capacity.
• Data handoff: A crash model and an NVH model may share the same CAD geometry but use different mesh densities, material cards, and boundary conditions. The taxonomy defines what can be reused and what must be rebuilt.
• Coupling risk: Multi-physics problems — such as thermal expansion affecting durability, or aerodynamic loads driving NVH — sit at the boundaries between disciplines. These boundaries are where integration failures occur.

The Six Core Disciplines

The following table summarizes the primary CAE disciplines in automotive development, their governing physics, typical solver families, and representative load cases.

Crash and Occupant Safety

Crash CAE is dominated by explicit dynamics. The solver advances the solution in small time increments, resolving contact between deformable bodies and rigid barriers. The key modeling decisions are mesh size in the crush zone (typically 5–8 mm for body structures), material cards for high-strength steel and aluminum (including strain-rate dependence), and spot-weld or adhesive failure models.

A typical frontal offset deformable barrier (ODB) model contains 3–5 million elements and runs for 120–150 ms of physical time. The output of interest includes intrusion into the occupant compartment, deceleration pulses at the B-pillar, and airbag deployment timing. These results feed directly into restraint system design and are often validated against physical sled or full-vehicle tests.

NVH

NVH analysis operates in the frequency domain or in the time domain for transient events. The most common workflow is a normal modes analysis followed by frequency response or forced response. Modal alignment — ensuring that body, chassis, and powertrain modes do not coincide — is a fundamental design activity.

For a body-in-white (BIW), the first bending mode typically targets above 45 Hz and the first torsion mode above 50 Hz. These targets are set to avoid excitation from road input and engine idle. NVH models are usually linear, which makes them faster to solve than crash models, but the mesh must be fine enough to capture the modes of interest — typically 8–10 mm in the body panels.

Durability and Fatigue

Durability CAE predicts crack initiation and propagation under cyclic loading. The workflow begins with a linear static or modal superposition analysis to obtain stress histories, followed by a fatigue solver that applies a material S-N curve or strain-life curve. The Palmgren-Miner rule is commonly used to accumulate damage.

For a suspension lower control arm, a typical load case might be a 3G vertical bump combined with braking and cornering loads. The fatigue solver then applies a load spectrum derived from proving ground data. The output is a damage contour, and the acceptance criterion is a damage value below 1.0 with a safety factor of 1.5–2.0 depending on the component class.

CFD and Aerodynamics

CFD solves the Navier-Stokes equations on a discretized fluid domain. In automotive applications, the two dominant use cases are external aerodynamics (drag, lift, aeroacoustics) and internal flow (cooling, HVAC, cabin comfort).

External aero models typically use a half-vehicle symmetry model with a virtual wind tunnel. Mesh sizes range from 50 to 200 million cells for a full vehicle. The key outputs are the drag coefficient (Cd), lift coefficient (Cl), and the pressure distribution on the windshield and side glass. A typical target for a passenger car is Cd below 0.30, with a front lift coefficient near zero for stability.

Thermal

Thermal CAE covers conduction, convection, and radiation. In underhood thermal management, the goal is to predict component temperatures during soak and driving conditions. The model includes the engine block, exhaust system, heat shields, and surrounding air.

A common workflow is a conjugate heat transfer (CHT) simulation, where the fluid and solid domains are solved simultaneously. The boundary conditions include heat rejection from the engine, radiation view factors, and fan curves. The acceptance criterion is that no component exceeds its material limit — for example, a plastic intake manifold must stay below 120 °C.

Multi-Physics Coupling

Multi-physics problems arise when two or more disciplines interact strongly. The most common couplings in automotive CAE are:

• Thermo-mechanical: Thermal expansion generates stress, which affects fatigue life. Exhaust manifolds and brake discs are classic examples.
• Aero-acoustic: Turbulent flow over a side mirror generates pressure fluctuations that radiate noise into the cabin.
• Electro-thermal: Battery pack performance and aging depend on temperature, which depends on electrical load.

These couplings are typically handled by co-simulation, where two solvers exchange field data at the interface at each time step. The main challenge is ensuring stable and accurate data mapping between non-matching meshes.

Practical Implications for Portfolio Management

Understanding the taxonomy allows you to make better decisions about model reuse, solver selection, and schedule sequencing. For example, a crash model can often be simplified to a linear model for NVH analysis, but the mesh and material data must be carefully mapped. Conversely, a thermal model may provide boundary conditions for a durability analysis, but the time scales are different — thermal soak takes minutes, while fatigue cycles take hours.

A practical rule of thumb is to maintain a discipline matrix that maps each vehicle attribute to its owning CAE discipline, the required model fidelity, and the coupling interfaces. This matrix becomes the backbone of the CAE portfolio plan and helps prevent the common failure mode where a multi-physics effect is missed because no single discipline owns it.

The taxonomy presented here is the foundation for the rest of this course. Subsequent chapters will dive deeper into each discipline, covering modeling best practices, solver settings, and validation strategies. The key takeaway is that CAE is a portfolio of coupled disciplines, and managing that portfolio requires a clear understanding of what each discipline does, what it needs, and how it interacts with the others.