Air Starter Motor Test Procedure: Performance Testing and Curve Generation

The SAE J2437 Recommended Practice provides standardized procedures for testing the output performance of air starting motors and plotting their performance curves. This guide explains the key elements of the standard, including the two test methods, instrumentation accuracy requirements, and practical insights for applying the results to engine starting systems. Although the standard has been stabilized due to mature technology, its procedures remain essential for validating starter motor performance.

Understanding the Standard and Its Test Procedures

SAE J2437 establishes a uniform method for testing air starting motors used in various vehicle and industrial applications. It describes how to mount the starter in a test stand, maintain constant air pressure at the inlet, and record torque, speed, air flow, and pressure under controlled conditions. The standard does not specify pass/fail criteria but provides a consistent basis for generating performance curves that can be used for design, comparison, and quality assurance.

Key Requirements:

• Mount the starter per the schematic in Figure 1 of the standard.
• Measure air pressure at the starter inlet port and maintain it constant throughout the test.
• Do not exceed the manufacturer’s recommended inlet pressure.
• Record ambient temperature and note supply air moisture and lubricant content.
• Reflect all torque measurements to the starter output shaft.
• Reduce all data to standard conditions of 1 bar (14.7 psig) and 21 °C (70 °F).

The standard defines two methods for generating performance curves. Both produce identical results when performed correctly but differ in their approach to data collection.

Feature	Method A (Discrete Points)	Method B (Continuous Operation)
Procedure	Run the starter at a series of discrete torque loads. Record air pressure, temperature, torque, air flow, and speed at each point.	Operate the starter continuously while decreasing torque load from the stall point. Data are recorded automatically as speed increases.
Data Processing	Use linear regression on the discrete data points to generate smooth curves for torque, power, and flow versus speed.	The continuous recording directly produces the curves; no additional curve fitting is required.
Typical Use	Suitable when manual data collection is preferred or when automated recording is not available.	Ideal for production testing or when fully automated data acquisition systems are in place.
Note	Sufficient data points must be collected to ensure a good regression fit.	Deviation from Method A must be noted on the performance curve.

Both methods require that air pressure be held constant and that measurements be taken across the full performance range from stall to no-load speed. The resulting performance curves typically include torque, power, and air flow plotted against speed, as shown in Figure 2 of the standard.

Instrumentation, Test Conditions, and Data Reduction

To ensure reliable and repeatable results, the standard specifies minimum accuracy levels for test instruments:

• Air pressure sensor and air flow meter: Accuracy of ±1% of the full scale reading. The full scale value must not exceed the maximum performance reading by more than 50% (i.e., full scale ≤ 1.5 × maximum reading).
• Speed sensing and torque measuring devices: Accuracy of ±1% of the actual reading.

Performance curves can be generated at various ambient temperatures and inlet pressures. The standard recommends a set of standard conditions for reference:

• Standard temperature: 21 °C (70 °F)
• Standard pressure: 1 bar (14.7 psig)
• Other optional temperatures: –40, –29, –7, 21, and 82 °C
• Optional inlet pressures: 2, 4, 6, 8, 10, 12, 15, 20, and 30 bar

Whenever tests are conducted at non-standard conditions, data must be corrected to the standard conditions for comparability. This is typically done using gas laws and empirical factors.

Applying the Results: Cranking Speed Matching and Frequently Asked Questions

One critical application of the performance curves is selecting the appropriate starter pinion and ring gear ratio to achieve the required engine cranking speed. The standard provides guidance: The starter speed at maximum power shall be equal to or greater than the required reflected engine cranking start speed. Ambient conditions at the application should, of course, be considered.

Engineering Design Insight: Use the performance curves to determine the speed at which the starter produces maximum power. Then, using the candidate gear ratio, calculate the reflected engine speed at that point. Adjust the ratio if necessary so that the reflected speed meets or exceeds the engine’s minimum cranking speed requirement, typically specified by the engine manufacturer. Ambient temperature and pressure should be factored into the matching process as they directly affect starter output.

Frequently Asked Questions

1. What is the difference between Method A and Method B?
Method A involves running the starter at discrete torque loads and using linear regression to plot smooth curves. Method B operates the starter continuously from stall while automatically recording data. Both methods produce identical performance curves; the choice depends on available test equipment and preferences.

2. What instrumentation accuracy does SAE J2437 require?
Pressure sensors and flow meters must have an accuracy of ±1% of full scale, with the full scale value not exceeding 1.5 times the maximum reading. Speed and torque devices must be accurate to ±1% of the actual reading.

3. Why must test data be reduced to standard conditions?
Reducing data to 1 bar and 21 °C provides a common baseline for comparing performance across different tests and ambient conditions. This correction accounts for variations in air density and ensures that curves are representative of the starter’s inherent capability.

4. How do I select the correct pinion and ring gear ratio?
Refer to the starter’s performance curve and identify the speed at which maximum power occurs. Then, using the gear ratio, compute the reflected engine cranking speed. Ensure this reflected speed is equal to or greater than the engine’s minimum cranking speed. Also consider the ambient temperature and pressure conditions expected in the application, as they affect starter output.