Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Low-Temperature Cranking Load Requirements of an Engine – Air Starter Method (SAE J2438)
When temperatures drop, engine cranking loads increase significantly—making reliable starts a serious engineering challenge. The SAE J2438 standard (stabilized as of 2023) provides a structured, reproducible method for measuring these loads using an air starter. This article breaks down the essential steps, common pitfalls, and design insights engineers need to know when applying J2438 for low‑temperature air starting systems.
🔍 Standardized Test Preparation and Execution
The J2438 procedure focuses on obtaining torque‑versus‑speed data under controlled cold conditions. Proper engine and motor preparation is critical to ensure repeatable results.
Engine Preparation
- Equip the engine with all production accessories that add parasitic loads (power steering pump, automatic transmission, etc.).
- For new engines, run in for the equivalent of 2400 km or 18 hours to stabilize friction.
- Winterize the cooling system with antifreeze suitable for the test temperature.
- Use engine oil that represents the high‑limit viscosity within the SAE grade (per SAE J300). Obtain a single batch for the entire test program.
- Perform the cranking test with no fuel in the engine—either drain the carburetor or cut off the fuel system—to avoid fuel dilution of the oil, which can lower viscosity and skew results.
Cold Soak and Motor Calibration
- Install a thermocouple in the oil sump to monitor soak temperature.
- After warming the engine to circulate oil, soak the assembly at the target test temperature for 16 to 24 hours.
- The cranking motor itself must be a calibrated unit. A new motor should be run in until its performance stabilizes, then calibrated to determine the relationship between air pressure, speed, and torque.
- Re‑calibrate the motor after testing to confirm that no performance shift occurred.
Running the Tests
- Vary the air inlet pressure to cover a speed range appropriate for the engine type (see Table 1).
- Keep the air supply temperature within ±5 °C of the calibration temperature; this avoids variations in energy supplied to the motor.
- Crank for about 5 seconds per test point, using readings taken between 0 and 5 seconds.
- Allow at least 30 minutes of additional soak before the next test.
⚠️ Critical – Fuel Contamination
Fuel dilution dramatically reduces oil viscosity. Always run cranking tests with the fuel system shut off or the carburetor dry. Even a small amount of fuel can invalidate the torque curve.
Fuel dilution dramatically reduces oil viscosity. Always run cranking tests with the fuel system shut off or the carburetor dry. Even a small amount of fuel can invalidate the torque curve.
Table 1 – Typical Cranking Speed Ranges by Engine Type
| Engine Type | Target Cranking Speed Range (rpm) |
|---|---|
| Gasoline engines | 30 – 120 |
| Direct‑injection diesel engines | 50 – 150 |
| Small indirect‑injection diesel engines | 120 – 220 |
Minimum unaided cranking speed is generally 100–120 rpm; engine manufacturers typically recommend starting aids below 4 °C (40 °F).
🛠️ Engineering Design Insights from J2438
The torque measured by the cranking motor is not the true engine torque because gear inefficiencies are not accounted for. However, it represents the load as seen by the starter motor, making it the correct value for system sizing and component selection.
A properly sized air cranking system should provide at least 5 seconds of cranking above the manufacturer’s minimum unaided cranking speed at the design temperature. This duration is based on the usable capacity of a typical 60‑gallon air tank on a heavy‑duty vehicle. Engine manufacturers often recommend starting aids below 4 °C, but end users expect unaided starts down to –7 °C (20 °F).
💡 Design Tip – Oil Viscosity Impact
Using oil with the high‑limit viscosity for its SAE grade ensures that the measured cranking load represents worst‑case conditions. If the oil is too thin at the test temperature, the cranking torque will be underestimated, risking an undersized air system.
Using oil with the high‑limit viscosity for its SAE grade ensures that the measured cranking load represents worst‑case conditions. If the oil is too thin at the test temperature, the cranking torque will be underestimated, risking an undersized air system.
❓ Frequently Asked Questions
How is engine torque calculated from the air starter test data?
Engine torque is derived by first reading the cranking motor’s torque from its calibration curve at the observed air pressure and speed. This value is then multiplied by the ring‑gear‑to‑pinion gear ratio. Remember that the result is the torque seen by the motor, not the absolute engine torque, but it is correct for system design.
Why must a new engine be run in before testing?
New engines have higher and unstable friction loads. Running the engine for 2400 km (or 18 hours) stabilizes the friction characteristics, ensuring that the measured torque is representative of a broken‑in engine and that results are repeatable over time.
How long must the engine soak at the test temperature?
A minimum of 16 to 24 hours of cold soak is required before the first cranking test. This ensures all engine components reach thermal equilibrium. Between successive tests, allow at least 30 minutes of re‑soak to maintain consistent conditions.
For further details, refer to the full SAE J2438 document (stabilized August 2023).
