Single-Phase Direct-to-Chip Cold Plate Test Specification, Version 1.0.0 (OCP Cooling Environments: Cold Plate Sub-Project, effective September 3, 2026)

Overview

Written by CoolIT Systems with PFX Group, Fabric8Labs and INFICON, the specification is described as the Test layer within the OCP modular framework. It states how a characteristic is measured, not how it is achieved, and does not set dimensions, materials, connector types or performance targets unless stated. Its purpose is to replace supplier-specific qualification practice with one shared reference, so suppliers validate once and operators can qualify second sources against one criterion. Each test uses the same structure: description, setup, test fluid, temperatures, timing and acceptance criteria. Methods rely on ASTM, ISO and IEC standards so any laboratory can run them, and a subset may be used early in design while the full set supports qualification.

The scope is design and prototype qualification; production and factory acceptance testing, including sampling plans and in-line screening, are outside it. Leakage and pressure tests start with tracer gas leak detection using hydrogen as 5% H2 in 95% N2 or helium, with an example criterion below 0.6 SCCM for water-tight, then gas pressure decay with an example limit of under 0.5% loss over a 5-30 minute dwell. Hydrostatic testing holds maximum operating pressure for 5 minutes and then three times that pressure for 2 minutes, checked under UV light with dimensions measured before and after. Burst testing proceeds to rupture or three times minimum proof pressure, at 1-5 bar/min per ISO 18869, with an example threshold of more than 20 bar.

Thermal and flow tests mount the plate on a thermal test vehicle or processor with the recommended thermal interface material, at fluid temperature typically 45 °C +/- 2 °C, and compute thermal resistance as R = (Ts - Tl)/P across several flow rates. Pressure drop is measured at no fewer than six flow rates from 25% to 150% of rated flow, relative to a tare loop, and reported as a curve fit with a minimum Cv. Flow rate testing goes up to velocities below 1.5 m/s to avoid erosion. Mechanical tests cover retention force (typically 30 N +/- 10% for 5 seconds), flatness per ISO 12781-2 and stiffness. Many acceptance values defer to the product specification.

Reliability tests include operational temperature cycling between 10 and 65 °C with 72-hour dwells, shipping cycling between -40 and +70 °C with a helium leak rate of at most 1x10^-5 mbar-L/s, shock and vibration on a representative board, salt spray per ASTM B117 (5% NaCl, 35 °C, 8 hours), humidity at 40 °C with 93% and 10% RH for 16 hours each, and fluid compatibility using ASTM D8040 glassware corrosion, D471 for elastomers, D543 for plastics and D7933 for brazed joints, followed by fluid analysis. Cleanliness requires flushing in an ultrasonic bath with particulates below 50 micrometres. Because limits are mostly examples or set by the customer, the document is a validation method set, not a pass/fail standard.

Key points

1. The specification covers design and prototype qualification only; production and factory acceptance testing is out of scope.
2. Hydrostatic test: maximum rated operating pressure for 5 minutes, then 3x maximum operating pressure for 2 minutes, checked under UV light with no deformation.
3. Gas leak detection uses hydrogen (5% H2 in 95% N2) or helium, with an example criterion below 0.6 SCCM for water-tight; pressure decay example limit is under 0.5% over a 5-30 minute dwell.
4. Burst test runs to rupture or three times minimum proof pressure at a controlled 1-5 bar/min, with an example minimum above 20 bar.
5. Thermal resistance is computed as R = (Ts - Tl)/P at a fluid temperature typically 45 °C +/- 2 °C, across several flow rates.
6. Pressure drop is measured at no fewer than six flow rates from 25% to 150% of rated flow, with a curve fit and minimum Cv reported.
7. Flow rate testing goes up to velocities below 1.5 m/s to ensure no erosion or leaks.
8. Operational temperature cycling uses 10 to 65 °C with 72-hour dwells; shipping cycling uses -40 to +70 °C with helium leak rate of at most 1x10^-5 mbar-L/s.
9. Salt spray is 5% NaCl at 35 °C for 8 hours; humidity is 40 °C at 93% RH and 10% RH for 16 hours each.
10. Cleanliness: ultrasonic flush with particulate below 50 micrometres; fluid compatibility follows ASTM D8040, D471, D543 and D7933.

Using it in design

• Reference the specification in a cold plate purchase requirement and state which tests, limits and sampling the customer requires, since many limits are examples.
• Set the maximum operating pressure of the TCS loop first, because the hydrostatic (3x) and burst thresholds derive from it.
• Request thermal resistance and pressure drop curves at the stated fluid temperature and at several flow rates, including 25% to 150% of rated flow.
• Require fluid compatibility results for the actual coolant and all wetted materials, including elastomers, plastics and brazed joints.
• Specify the cleanliness requirement (below 50 micrometres) and align it with the filter rating of the CDU.
• Decide separately what production acceptance testing will be done, since this specification does not provide it.

Who should read it

Procurement and thermal engineers specifying or qualifying direct-to-chip cold plates, and laboratories or vendors planning the qualification test programme.

Document contents

1. License
2. Compliance with OCP Tenets
3. Overview
4. Test Specification
5. Description
6. Goals and Objectives
7. Common Abbreviations
8. Test and Validation Requirements
9. Leakage & Pressure
10. Leak Detection
11. Gas Pressure Decay / Dwell
12. Hydrostatic Pressure / Dwell
13. Burst Pressure
14. Thermal & Flow
15. Thermal Resistance
16. Pressure Drop
17. Flow Rate
18. Mechanical & Structural
19. Retention Force
20. Flatness
21. Stiffness
22. Reliability & Environmental
23. Operational Temperature Cycling
24. Non-Operational Temperature Cycling (Shipping)
25. Shock
26. Vibration
27. Salt Spray (Corrosion)
28. Fluid Compatibility/Corrosion
29. Humidity
30. Cleanliness
31. Flush/Cleanliness
32. Compliance

Read the document

Publication details

Authors

Ben Sutton; Brad Zakaib; Gerald Singer; Ian Winfield; Tyler Loughran

Publisher

Open Compute Project Foundation

Year

2026

Document language

English

Source type

Industry body

Access

Open licence — downloadable

Licence

OWFa 0.9 — Open Compute open specification licence; redistribution permitted

Licence evidence

www.opencompute.org

Citation