2. CAE project lifecycle

An automotive CAE project is not a single simulation task; it is a sequence of engineering decisions that must be synchronized with the vehicle development program. The CAE lifecycle mirrors the product development lifecycle but operates on a different currency: instead of physical parts, it produces predictive evidence that a design will meet its performance targets. Understanding where each CAE activity sits, what it must deliver, and what gate it must pass is the foundation of CAE project management.

The lifecycle is typically divided into five phases: Concept, Architecture, Detailed Design, Validation, and Launch. Each phase has a distinct CAE objective, a characteristic model fidelity, and a set of deliverables that feed the program's stage gates. The critical management principle is that CAE maturity must lead design maturity — analysis must be ahead of the release of the geometry it evaluates, otherwise the results arrive too late to influence the design.

Phase 1: Concept — Establishing Feasibility

In the concept phase, the vehicle's overall targets are set: mass, stiffness, crash performance, NVH, thermal, and durability. CAE's role here is target cascading and feasibility assessment. The models are coarse — simplified beam-and-shell representations, lumped-mass models, or 1D system models. The goal is not precise prediction but rapid exploration of the design space.

Typical CAE deliverables in this phase include:

• Global body stiffness targets — e.g., a BIW torsional stiffness target of 25–35 kN·m/deg for a C-segment vehicle, derived from benchmarking and vehicle class requirements.
• Crash energy management concept — a 1D or simplified lumped-mass model showing how crush energy is distributed across the front structure, using assumed force-displacement curves for the crash rails.
• Modal target allocation — first bending and torsion modes of the body-in-white, typically 45–60 Hz for bending and 50–70 Hz for torsion, allocated to avoid resonance with powertrain and road inputs.
• Thermal and aerodynamic feasibility — 1D cooling system models and CFD on a simplified exterior surface to confirm drag coefficient targets.

The gate at the end of concept is the Concept Gate (often called G8 or equivalent in OEM stage-gate systems). CAE must demonstrate that the chosen architecture can plausibly meet the performance targets. If the concept fails a target, the architecture is revised before significant tooling investment.

Phase 2: Architecture — Freezing the Load Paths

Once the concept is approved, the architecture phase freezes the major load paths, joint locations, and material strategy. CAE models become more detailed: full BIW shell models with representative thicknesses, suspension pickup points, and powertrain mounting locations. This is where stiffness, modal, and crashworthiness predictions become quantitative.

A key CAE deliverable here is the BIW stiffness and modal report. For example, a full-vehicle model with 1.2 million elements may be solved for static torsion and bending, producing a torsional stiffness value that is compared against the target. If the predicted value is 22 kN·m/deg against a 28 kN·m/deg target, the CAE engineer identifies the weak joints — often the rear shock tower or the front strut tower — and proposes local reinforcements. The design team then updates the CAD, and the CAE model is re-meshed and re-solved.

Crash CAE in this phase uses full frontal, offset deformable barrier (ODB), and side impact models. The deliverable is a crash performance assessment showing intrusion, deceleration pulses, and occupant load cases. The gate — Architecture Gate — requires that the architecture meets stiffness, modal, and crash targets within an agreed tolerance, typically ±10% for stiffness and ±15% for crash intrusion.

Phase 3: Detailed Design — Optimization and Verification

The detailed design phase is the longest and most CAE-intensive. Geometry is now nearly production-intent, and CAE shifts from feasibility to optimization and verification. Models include all major subsystems: body, closures, chassis, powertrain, and interior. Mesh sizes are refined, and nonlinear material models are used for crash and durability.

CAE deliverables in this phase are numerous and must be tightly scheduled:

• Crashworthiness verification — full frontal, ODB, side pole, rear impact, and pedestrian protection, each with a formal report and pass/fail status against regulatory and consumer-test targets (e.g., Euro NCAP, C-NCAP).
• NVH analysis — body modal, acoustic cavity modal, and powertrain mount optimization. A typical deliverable is a mode map showing body modes, suspension modes, and powertrain modes with adequate separation.
• Durability and fatigue — hot spot stress or strain-life analysis on the body and chassis, using load spectra from proving ground or customer-correlated road load data.
• CFD and thermal — underhood thermal management, cabin comfort, and brake cooling, with results compared against targets.
• Optimization studies — topology, topography, and gauge optimization to reduce mass while maintaining performance. A typical result is a 5–8% mass reduction on a body panel with no stiffness loss.

The Detailed Design Gate (often the final data freeze gate) requires that all CAE reports show compliance with targets, or that any non-compliance is documented with a deviation request and a mitigation plan. This gate is the last point at which major design changes are economically feasible.

Phase 4: Validation — Correlating with Physical Tests

Validation is where CAE meets reality. Physical prototypes are built and tested, and CAE models are correlated against measured data. The objective is not to re-run every simulation but to confirm that the CAE predictions are trustworthy and to update models where correlation is poor.

Key activities include:

• Modal correlation — comparing measured and predicted mode shapes and frequencies using MAC (Modal Assurance Criterion). A MAC value above 0.8 for the first five modes is a common acceptance criterion.
• Crash correlation — comparing deceleration pulses, intrusion, and deformation patterns from sled and full-vehicle crash tests. A typical correlation target is a peak deceleration difference within 10% and intrusion difference within 15 mm.
• Stiffness correlation — comparing measured static torsion and bending with CAE predictions. A difference within 5% is considered good correlation.

The Validation Gate requires that CAE models are correlated and that any remaining discrepancies are understood and documented. This gate also confirms that the vehicle meets regulatory and consumer-test requirements based on both CAE and physical test evidence.

Phase 5: Launch — Supporting Production and Field Issues

In the launch phase, CAE support shifts to production variation, field issue resolution, and continuous improvement. The models are mature, and CAE is used to investigate manufacturing deviations, supplier changes, and field failures. Deliverables include:

• Production variation studies — e.g., the effect of a 0.2 mm thickness variation in a crash-critical rail on intrusion.
• Field failure root-cause analysis — using CAE to reproduce a durability or crash issue reported from the field.
• Change impact assessments — evaluating the performance impact of a late design change, such as a material substitution or a supplier-driven geometry modification.

The Launch Gate (or SOP gate) confirms that all CAE deliverables are closed, that no open CAE issues threaten production, and that the CAE evidence supports the final vehicle release.

Visualizing the Lifecycle and Gate Flow

The following diagram shows the five phases, their primary CAE deliverables, and the gates that control the transition between them. The arrows indicate the direction of information flow: CAE evidence feeds the gate decision, and the gate decision authorizes the next phase.

Managing the Lifecycle: Practical Considerations

Two management challenges dominate the CAE lifecycle. First, schedule alignment: CAE must deliver results before the design gate, not after. A common practice is to define a CAE "delivery date" that is two to three weeks ahead of the design gate, allowing time for review and design iteration. Second, model maturity tracking: each CAE model must be tagged with its maturity level — concept, architecture, detailed, or validated — so that decision-makers know how much confidence to place in the results.

A practical tool for this is a CAE deliverable matrix that maps each phase to its required analyses, the responsible engineer, the due date, and the gate it supports. For example:

In summary, the CAE lifecycle is a structured progression from coarse feasibility to validated prediction. Each phase has a clear objective, a defined set of deliverables, and a gate that controls the transition. Managing this lifecycle effectively requires early CAE involvement, disciplined model maturity tracking, and a clear understanding that CAE evidence is the currency that buys design decisions.