Skip to content

Vibration

Verifies a battery pack or system survives the random and sinusoidal vibration spectrum representative of its in-vehicle service life without losing electrical safety.
Clause (method) 8.2.1
Clause (pass criteria) 5.2.1
Object pack / system
Status vs. 2020 revised (requirements and method)
Observation period 2 h at test environment temperature

Pass criteria

After the vibration test, the battery pack or system shall show no leakage, no housing crack, no fire, and no explosion. Insulation resistance after testing shall not be less than 100 Ω/V (DC), or 500 Ω/V if an AC circuit is present.

Source: GB 38031-2025, clause 5.2.1 (PDF p. 12).

Pre-conditions

  • Sample: A battery pack or system. (8.2.1.1)
  • Pre-treatment: Standard, per clause 7.2.
  • SOC: Highest working SOC per clause 6.1.10.
  • Mounting: Per the test object's installation position and fixation method in the vehicle, following GB/T 2423.43. (8.2.1.2)
  • Insulation baseline: Measure before the test per Appendix B (clause 6.1.5).

Test parameters

The parameter set is vehicle-category-dependent. M1 and N1 category vehicles use Table 3 (lower spectrum, 1 h sine dwell). All other categories use Table 2 (higher spectrum, 2 h sine dwell). Packs mounted on the roof use manufacturer-provided parameters that shall be no less than Table 2 / Figure 2.

Common Value Source
Axes x, y, z (random + sine constant-frequency on each) 8.2.1.2
Axis convention x = vehicle driving direction; y = horizontal perpendicular; z = vertical 8.2.1.2
Loading sequence (recommended) random z → sine z → random y → sine y → random x → sine x (lab may re-order to shorten conversion time) 8.2.1.2
Random vibration framework GB/T 2423.56 (Test Fh, wideband random) 8.2.1.2
In-test monitoring Smallest monitoring unit (voltage, temperature, etc.) 8.2.1.5
Observation after test 2 h at test environment temperature 8.2.1.6

For vehicles other than M1 and N1 (Table 2)

RMS levels: x 0.52 g, y 0.57 g, z 0.73 g. (8.2.1.3)

Random vibration PSD (Table 2):

Frequency (Hz) z-axis PSD (g²/Hz) y-axis PSD (g²/Hz) x-axis PSD (g²/Hz)
5 0.008 0.005 0.002
10 0.042 0.025 0.018
15 0.042 0.025 0.018
40 0.0005 — —
60 — 0.0001 —
100 0.0005 0.0001 —
200 0.00001 0.00001 0.00001

Sine wave constant-frequency vibration (2 hours per direction):

Frequency (Hz) z-axis amplitude (g) y-axis amplitude (g) x-axis amplitude (g)
20 ±1.5 ±1.5 ±2.0

For test objects with multiple installation orientations (x/y/z), test in the orientation with the largest RMS. (8.2.1.3)

For M1 and N1 category vehicles (Table 3)

RMS levels: x 0.50 g, y 0.45 g, z 0.64 g. (8.2.1.4)

Random vibration PSD (Table 3):

Frequency (Hz) z-axis PSD (g²/Hz) y-axis PSD (g²/Hz) x-axis PSD (g²/Hz)
5 0.015 0.002 0.006
10 — 0.005 —
15 0.015 — —
20 — 0.005 —
30 — — 0.006
65 0.001 — —
100 0.001 — —
200 0.0001 0.00015 0.00003

Sine wave constant-frequency vibration (1 hour per direction):

Frequency (Hz) z-axis amplitude (g) y-axis amplitude (g) x-axis amplitude (g)
24 ±1.5 ±1.0 ±1.0

For roof-mounted packs

Test parameters per the manufacturer, but no less than Table 2 and Figure 2 above. (8.2.1.3)

Procedure

  1. Mount the test object on the vibration table per its in-vehicle installation position and fixation method, following GB/T 2423.43. (8.2.1.2)
  2. Apply random and constant-frequency vibration loads along each axis. Recommended sequence: random z → sine z → random y → sine y → random x → sine x. (8.2.1.2)
  3. For non-M1/N1 packs, use Table 2 and Figure 2: RMS 0.52 g (x) / 0.57 g (y) / 0.73 g (z), sine 20 Hz at the amplitudes above, 2 h per direction. (8.2.1.3)
  4. For M1/N1 packs, use Table 3 and Figure 3: RMS 0.50 g (x) / 0.45 g (y) / 0.64 g (z), sine 24 Hz at the amplitudes above, 1 h per direction. (8.2.1.4)
  5. For multiple-orientation installations, test in the orientation with the highest RMS. (8.2.1.3)
  6. During the test, monitor voltage, temperature, and other parameters of the smallest monitoring unit. (8.2.1.5)
  7. After all axes/loads complete, observe the test object for 2 h at test environment temperature. (8.2.1.6)
  8. Re-measure insulation resistance per Appendix B. Inspect for leakage, housing cracks, fire, or explosion.

After-test observation

Observe the test object for 2 hours at the test environment temperature (22 °C ± 5 °C) after the last vibration step completes. (8.2.1.6)

In-test monitoring of the smallest monitoring unit (voltage, temperature) is required throughout. (8.2.1.5)

What changed from GB 38031-2020

The 2025 revision changed both the requirements (5.2.1) and the test method (8.2.1):

  • Spectrum is now per-axis with explicit PSD tables (Table 2 and Table 3) rather than a single shared envelope.
  • M1/N1 category vehicles get a separate, lower spectrum (Table 3, 1 h sine dwell).
  • Non-M1/N1 vehicles use Table 2 with 2 h sine dwell per axis.
  • Sine constant-frequency dwell is explicit at 20 Hz (non-M1/N1) or 24 Hz (M1/N1), with per-axis amplitudes.
  • Roof-mounted packs are called out as a special case, with parameters from the manufacturer that shall be no less than Table 2 / Figure 2.
  • In-test monitoring of the smallest monitoring unit (voltage, temperature) is now explicitly required.

Migration impact: Already-type-approved models must comply from 2027-08-01; new type approvals from 2026-07-01. Re-running vibration with the new per-axis PSD is generally required because pre-2025 single-envelope test data does not map cleanly onto the M1/N1 vs non-M1/N1 split. See Re-certification timeline.

Engineering notes (non-normative)

The notes below are practical interpretation, not part of the standard.

Engineering note (non-normative): The 0.73 g (z) RMS for non-M1/N1 packs is roughly 1.6× the M1/N1 z-axis level. For commercial vehicle programs (N2/N3/M2/M3) that previously borrowed M1/N1 vibration data, expect to repeat the campaign — the higher spectrum is not envelope-compatible.

Engineering note (non-normative): The "smallest monitoring unit" in 8.2.1.5 is typically the cell or parallel-cell group monitored by the BMS. Capture cell voltages and at least the BMS-reported temperatures throughout the run; sudden voltage drops mid-test usually indicate intermittent contact at busbars or weld fatigue, which won't necessarily fail post-test insulation but predicts field returns.

Engineering note (non-normative): Roof-mounted packs (some buses, certain commercial EVs) are exposed to a fundamentally different ride spectrum than underbody packs. The "no less than Table 2 / Figure 2" clause is a floor, not a ceiling — manufacturers should justify their chosen spectrum from in-service measurements, not just default to Table 2.