4. Inflator modeling and gas dynamics

The inflator is the energy source of the entire restraint system. In an airbag deployment simulation, the inflator model determines how much gas enters the cushion, how fast it enters, and at what temperature. Get the inflator wrong, and no amount of careful fabric modeling will produce a credible result. This section covers the three pillars of inflator modeling — mass flow rate curves, tank test characterization, and temperature effects — and then shows how they couple to the airbag chamber model in a CAE solver.

Mass Flow Rate Curves: The Primary Input

An inflator is characterized primarily by its mass flow rate curve, usually expressed as ṁ(t) in kg/s or g/ms, together with the total gas mass and the gas temperature. For a pyrotechnic inflator, the curve rises sharply within 2–5 ms, peaks, and then decays as the propellant is consumed. For a hybrid inflator, the curve has two contributions: the rapid combustion of the pyrotechnic initiator and the subsequent release of stored compressed gas, which produces a longer, flatter tail.

In most solvers (LS-DYNA, PAM-CRASH, Radioss), the inflator is defined through a *AIRBAG_... keyword family. A typical pyrotechnic definition uses a mass flow rate versus time table:

*DEFINE_CURVE
         1         0       1.0       1.0       0.0       0.0
     0.000     0.000
     0.001     0.850
     0.003     1.420
     0.006     1.180
     0.010     0.620
     0.020     0.180
     0.040     0.020
     0.060     0.000

Here the abscissa is time in seconds and the ordinate is mass flow rate in kg/s. The integral of this curve gives the total gas mass — a useful sanity check. For the table above, numerical integration yields approximately 0.0125 kg, which should match the inflator datasheet. If it does not, either the curve is wrong or the units are inconsistent.

Two additional parameters matter. The gas temperature (often 800–1200 K for pyrotechnic inflators) controls the energy content and therefore the cushion pressure. The gas constant or the specific heat ratio γ determines how the gas expands. For a typical combustion gas mixture, γ ≈ 1.25–1.35. Using air properties (γ = 1.4) is a common but avoidable error.

Tank Tests: How the Curve Is Actually Measured

The mass flow rate curve is not a theoretical quantity — it is measured. The standard method is the tank test (also called a pressure vessel test). The inflator is fired into a rigid, known-volume tank, and the internal pressure is recorded versus time. From the pressure history and the tank volume, the mass flow rate is back-calculated using the ideal gas law and an energy balance.

A typical tank test setup uses a 60 L or 100 L steel vessel. The procedure is:

1. Evacuate the tank to a known initial pressure (often near vacuum or a controlled back-pressure).
2. Fire the inflator inside the sealed tank.
3. Record pressure with a high-frequency transducer (≥ 50 kHz sampling).
4. Apply the reduction equation to obtain ṁ(t).

The reduction is not simply P·V/(R·T). Because the gas entering the tank is hot and the tank walls absorb heat, a proper energy balance is required. A simplified form often used in practice is:

ṁ(t) = (V / (R·T_gas)) · dP/dt + (P / (R·T_gas)) · (dV/dt) + heat_loss_term

In a rigid tank, dV/dt = 0, so the first term dominates. The heat loss term is estimated from a calibration firing with a known gas mass. This is why tank tests are always paired with a calibration shot — without it, the derived mass flow rate can be off by 10–20%.

Temperature Effects

Inflator performance is strongly temperature-dependent. Automotive specifications require operation from −35 °C to +85 °C. At low temperature, the propellant burns slower, the peak mass flow rate drops, and the total gas mass may be slightly reduced. At high temperature, the opposite occurs — the burn rate increases, and the peak pressure in the cushion can rise significantly.

In CAE, temperature effects are handled in one of three ways:

• Temperature-scaled curves: Define separate mass flow rate tables for hot, nominal, and cold conditions, and interpolate or select based on the specified ambient temperature.
• Burn rate modifier: Apply a multiplicative factor to the mass flow rate curve as a function of temperature, derived from tank tests at multiple temperatures.
• Full gas dynamics model: Model the propellant combustion with an Arrhenius-type burn rate, which naturally captures temperature dependence. This is more accurate but requires propellant-specific constants that are rarely available.

For most production CAE work, the temperature-scaled curve approach is standard. A typical dataset might show a peak mass flow rate of 1.42 kg/s at 23 °C, 1.55 kg/s at 85 °C, and 1.18 kg/s at −35 °C. These are example values for illustration; actual values come from the inflator supplier's tank test matrix.

Coupling to the Airbag Chamber Model

The inflator model does not act in isolation. In a control-volume (CV) airbag model, the solver solves a set of coupled ordinary differential equations for each chamber: mass, energy, and volume. The inflator provides the mass and energy source terms. The chamber provides the back-pressure that opposes the flow.

The coupling is governed by the orifice equation. If the inflator is modeled as a nozzle discharging into the chamber, the mass flow rate is:

ṁ = C_d · A · P_inflator / sqrt(R · T) · Φ(P_chamber / P_inflator)

where C_d is the discharge coefficient, A is the effective orifice area, and Φ is the pressure ratio function (choked flow when P_chamber/P_inflator < 0.528 for γ = 1.4). In many solvers, the user supplies ṁ(t) directly and the solver applies a correction based on the instantaneous chamber pressure. This is the pressure-coupled inflator model.

The practical consequence: if you use a tank-test curve without pressure coupling, you will over-predict the early cushion pressure, because the tank test curve assumes the inflator is discharging into a near-vacuum. In a real airbag, the chamber pressure rises within 1–2 ms and chokes the flow. The difference in peak cushion pressure can be 15–30%.

Practical Workflow and Verification

A defensible inflator model follows a clear sequence. First, obtain the tank test data from the supplier — pressure history, tank volume, initial conditions, and calibration shot. Second, reduce the data to a mass flow rate curve, verifying that the integral matches the reported total gas mass. Third, define the curve in the solver with the correct gas properties. Fourth, run a single-chamber validation case with a rigid, non-deploying cushion and compare the predicted chamber pressure against the tank test pressure. If the two do not match within a few percent, the problem is in the inflator model, not the fabric.

Common failure symptoms and their causes:

The inflator model is the boundary condition that drives the entire deployment. Treat the tank test as ground truth, respect the pressure coupling, and account for temperature. With those three elements in place, the chamber model has a physically meaningful input, and the subsequent discussion of fabric behavior, venting, and folding rests on a solid foundation.