Liquid Cooling Validation

Cold Plate Leak Testing Methods and Specifications

How to select and document leak, proof-pressure, flow, and related functional tests for cold plates and liquid cooling manifolds.

“Leak-Free” Is Not a Test Specification

A useful requirement defines test medium, pressure or vacuum, stabilization time, test time, temperature, acceptable leak rate or pressure change, port configuration, fixture, calibration, and product state. Without those inputs, different methods can produce results that are not comparable.

Common Leak Test Methods

MethodStrengthKey limitation
Bubble or immersionSimple visual location of larger gas leaksSensitivity, wetting, observation, and drying must be controlled
Pressure decayClean, automatable pass/fail testingTemperature, volume, stabilization, fixture leakage, and part elasticity affect results
Differential pressureHigher resolution by comparison with a reference volumeRequires stable equipment, validated reference, and suitable cycle design
Tracer gas or heliumHigh sensitivity and possible leak localizationCost, background gas, handling, calibration, and correlation to service conditions
Vacuum testingUseful for defined sealed-volume applicationsMust reflect pressure direction and realistic leak-path behavior

Choose a Method From Product Risk

The best method depends on allowable leakage, internal volume, test pressure, material, joining process, seal type, production rate, cleanliness, and whether leak location is required. Gas testing can reveal smaller paths than liquid testing, but correlation to the service fluid and pressure must be understood.

Agree whether the test applies to the machined body, joined cold plate, installed plugs, coating, seals, or final assembly. Testing too early can miss later defects; testing only at the end can make root-cause analysis harder.

Leak Test, Proof Test, and Burst Test Are Different

  • Leak test: determines whether fluid or gas escapes above an allowable rate.
  • Proof-pressure test: verifies that a part withstands a specified pressure without permanent damage or unacceptable leakage.
  • Burst test: characterizes ultimate failure pressure and mode; it is destructive and requires controlled safety measures.

Never infer proof or burst pressure from operating pressure without engineering approval and an applicable standard or customer requirement.

Control Temperature, Volume, and Fixtures

Gas temperature changes can mimic pressure loss or gain. The part, fixture, and test medium need adequate stabilization. Flexible walls, O-rings, trapped volume, valves, hoses, and connector seals can also influence measured decay.

Fixture leakage should be quantified and separated from product leakage. Use calibrated references, defined maintenance, and measurement-system studies appropriate to the acceptance threshold.

Add Flow and Pressure-Drop Testing When Needed

A component can pass a leak test while still having a blocked channel, debris, wrong plug, or excessive restriction. Flow or pressure-drop testing can verify passage connectivity and hydraulic behavior under specified fluid, temperature, flow, and port conditions.

Thermal testing is a separate system-level activity. It requires controlled heat input, contact assembly, coolant conditions, sensors, and acceptance criteria.

Protect Cleanliness After Testing

Wet testing introduces fluid that may need draining, flushing, and drying. Define compatible media, water quality, corrosion prevention, residual-fluid limit, drying method, port caps, and time to packaging. For clean gas testing, prevent test fixtures from adding particles or oil.

What a Test Record Should Include

  • Part number, revision, serial or batch identification, and test date
  • Equipment and calibration status
  • Method, medium, pressure or vacuum, temperature, stabilization and test time
  • Fixture and port configuration
  • Acceptance criterion, actual result, disposition, and operator
  • Any retest, repair, deviation, or nonconformance reference

Connect test records to the dimensional control plan described in our liquid cooling quality guide.

Test Specification Checklist

  • Product state and features included in the test
  • Service fluid, operating pressure, temperature, and pressure direction
  • Test method, medium, setpoint, stabilization, hold time, and leak limit
  • Proof or burst criteria if separately required
  • Flow, pressure drop, thermal, cleanliness, drying, and packaging requirements
  • Sampling or 100% frequency, report format, traceability, and retention

For manufacturing support, see AI server liquid cooling component machining and the cold plate design guide.

Need machining and leak testing reviewed together?

Send the cold plate or manifold drawing, operating conditions, quantity, test method, and acceptance limits.