4. Simulation model setup for airbag deployment

An airbag deployment model is only as credible as the gas representation behind it. The same folded bag, inflator curve, and venting geometry can produce wildly different pressure histories depending on whether the gas is treated as a uniform pressure volume, a cloud of discrete particles, or a coupled fluid mesh. This chapter explains how to choose among the five mainstream approaches — Uniform Pressure (UP), Corpuscular Particle Method (CPM), Arbitrary Lagrangian-Eulerian (ALE), Finite Pointset Method (FPM), and Corpuscular Particle Method with fabric leakage (often labeled CPM-leak or simply the particle-based family) — based on bag geometry, venting behavior, and fold complexity.

What Each Method Actually Solves

The core question is: how does the gas inside the bag exchange momentum and energy with the fabric? Each method answers this differently.

• Uniform Pressure (UP): The bag interior is a single control volume. Pressure is computed from the inflator mass flow rate, temperature, and volume via the ideal gas law. The fabric receives a normal traction proportional to the instantaneous pressure. No gas inertia, no jetting, no local pressure gradients.
• Corpuscular Particle Method (CPM): Gas is discretized into a large number of particles (typically 10,000–100,000) that carry mass, momentum, and energy. Particles collide with fabric elements and with each other statistically. Pressure emerges from particle-wall momentum transfer. This captures gas inertia, jetting, and non-uniform loading.
• ALE (Arbitrary Lagrangian-Eulerian): The gas domain is a structured or unstructured Eulerian mesh that can move with the structure. Conservation equations for mass, momentum, and energy are solved on this mesh. The fabric is typically Lagrangian, coupled via penalty or tied contact. ALE captures shock waves, supersonic jets, and complex flow fields but requires careful mesh design and is computationally expensive.
• FPM (Finite Pointset Method): A meshless Lagrangian particle method for the gas. Points move with the flow, and derivatives are approximated via weighted least squares over local neighborhoods. FPM handles large deformation and fragmentation of the gas domain naturally, making it attractive for bags that unfold into complex shapes.
• CPM with leakage (CPM-leak): An extension of CPM where a fraction of particles escape through fabric porosity and seam gaps based on a permeability model. This is essential for realistic venting and pressure decay.

Selection Criteria: Geometry, Venting, and Fold Complexity

The decision is not about which method is "best" in general. It is about which method captures the dominant physics for a given bag.

For a driver-side bag with a deep, complex fold and a central inflator, CPM or CPM-leak is usually the right starting point. The gas jet from the inflator impinges on the folded fabric, creating local pressure peaks that UP cannot represent. For a passenger-side bag with a large volume and relatively simple fold, UP may be adequate for early design, but CPM becomes necessary when OOP (out-of-position) occupant interaction is studied. For a side curtain airbag with a long, narrow gas channel and multiple vents, ALE or FPM may be required to capture the pressure wave propagation along the channel.

Venting: The Deciding Factor

Venting is where UP most often fails. In UP, venting is modeled as an orifice flow using a discharge coefficient. This works when the bag pressure is nearly uniform. But in a real bag, the pressure near the vent is lower than the pressure near the inflator, especially during the first 10–20 ms. CPM and FPM naturally capture this gradient because particles must physically travel to the vent. ALE resolves the vent as a boundary condition with local flow. The practical rule: if the vent is large and the bag is small, UP may overpredict pressure; if the vent is small and the bag is large, UP may underpredict the initial pressure peak.

Fold Complexity and Mesh/Model Fidelity

Fold complexity affects the choice in two ways. First, the gas domain shape changes dramatically during deployment. UP ignores this entirely. CPM and FPM handle it naturally because particles and points move with the fabric. ALE requires remeshing or a very fine mesh to avoid negative volumes. Second, the fold creates local gas trapping. In a tightly folded bag, gas may be trapped in pockets, leading to delayed pressure equalization. CPM captures this because particles collide with the fabric and each other; UP cannot.

For a complex fold (e.g., a driver bag with a star pattern), CPM with 50,000–100,000 particles is a reasonable starting point. For a simple fold (e.g., a passenger bag with a single roll), UP may be sufficient for concept studies, but CPM is recommended for final validation.

Model Setup Workflow

The following workflow applies to CPM and CPM-leak, which are the most common choices for production airbag correlation.

1. Define the gas domain: Start from the folded bag geometry. Create a closed volume that represents the initial gas space. For CPM, this is the particle initialization region.
2. Set the inflator: Provide a mass flow rate vs. time curve (kg/s) and a gas temperature vs. time curve (K). These come from tank tests or inflator characterization.
3. Initialize particles: Distribute particles uniformly within the initial gas volume. Use a particle count that balances resolution and cost. A common starting point is 20,000 particles for a 60 L bag.
4. Define venting: For CPM-leak, specify the fabric permeability (e.g., in L/dm²/min at 500 Pa) and the vent hole area. For discrete vents, define the orifice area and discharge coefficient.
5. Couple to fabric: Use a penalty contact between particles and fabric elements. The penalty stiffness should be high enough to prevent particle penetration but not so high that it causes numerical instability.
6. Run and monitor: Track bag pressure at a reference point (e.g., center of the bag), particle count inside the bag, and fabric stress. Compare with test data if available.

Example: Driver Bag with CPM

Consider a 60 L driver bag with a single central inflator and two 25 mm vents. The inflator provides 120 g of gas in 30 ms. The bag is folded in a star pattern. A CPM model is set up with 50,000 particles. The initial particle temperature is 800 K. The vent discharge coefficient is 0.8. After 10 ms, the bag pressure at the center reaches 250 kPa. The particle count inside the bag drops from 50,000 to 42,000 due to venting. The fabric stress at the fold peaks at 120 MPa. These values are consistent with typical driver bag behavior and can be used as a sanity check before correlation with test data.

Method Selection Decision Flow

The following diagram summarizes the selection logic based on the three primary factors: bag geometry, venting, and fold complexity.

Summary

Selecting a gas model is a trade-off between fidelity and cost. UP is fast but blind to gas inertia and local pressure gradients. CPM and CPM-leak capture particle-level physics and are the workhorses for driver and side bags. ALE is the reference for shock-dominated flows but demands significant meshing effort. FPM is a promising meshless alternative for highly deformable gas domains. The decision should be driven by the bag's geometry, venting mechanism, and fold complexity — not by a single "best" method. Start with the simplest model that captures the dominant physics, then increase fidelity only when correlation with test data demands it.