06 – Test Instruction 9.1.5 Seal Washout (closed loop)
Source: OCP UQD Specification v2.0.0 §9.1.5 (Figure 14), Table 6 (note 1), §10, §15.3.3
Related: 02-Parameters.md · 03-Tests.md (test A7) · 04-Backlog.md · 05-Pressure_Drop_9.1.4.md
Reading guide: Text without marking is a requirement from the OCP specification. Rows marked (internal proposal) are our own recommendations where the specification lacks data. They must be approved before the test is run.
1. Purpose
Verify that the sealing geometry of the coupling withstands repeated connection and disconnection at high pressure and high flow without the plug or socket valve seal being pushed out of its groove ("seal washout") and causing leakage.
The test recreates the real scenario of connecting and disconnecting a liquid-filled server to a liquid-filled TCS loop during operation.
Out of scope: open circuit, for example filling and draining a server. There, extreme pressure drop and high velocity can cause seal washout. The specification refers to the QD supplier's guidance and to building up flow and pressure gradually within ratings (OCP Liquid Cooling Integration and Logistics).
2. Scope
| Product |
Configuration |
Mated position |
| UQD02 / 04 / 06 / 08 |
UQD plug + UQD socket |
Fully mated. The latch mechanism may be removed or bypassed. |
| UQDB02 / 04 / 06 / 08 |
UQDB plug + UQDB socket |
Fully mated, hard-stop (internal proposal, see MD-023) |
| Hybrid 02–08 (if BS-002 = Yes) |
UQD plug + UQDB socket |
Fully mated, hard-stop |
| Test type |
Samples per product |
Source |
| Own verification |
5 pairs |
§9 |
| Cross-verification |
3 pairs per combination and supplier (our plug + their socket, and vice versa) |
§10 |
The same pairs are run in both flow directions: first plug→socket, then socket→plug.
3. Test parameters
| Parameter |
Value |
Source |
| Number of cycles |
20 per flow direction, 40 in total per pair |
§9.1.5 |
| Flow directions |
P→S (plug to socket) and S→P (socket to plug) |
§9.1.5 |
| Connection speed |
40 ± 10 mm/s |
§9.1.5 |
| Test flow |
2 × rated flow |
§9.1.5, Table 6 note 1 |
| Test temperature |
60 °C. Tolerance missing; internal proposal ± 3 °C (MD-009). |
§9.1.5 |
| Test pressure |
6.9 bar(g) / 100 psi(g) operating pressure, excluding transient pressure spike |
§9.1.5 |
| Max pressure spike at disconnect |
≤ 30 psi (2.07 bar) above the pressure before disconnect, i.e. ≤ 130 psi(g) / 8.96 bar(g) at C |
§9.1.5 |
| Test medium |
Water or PG25 (BS-004) |
§9.1.5 |
| Pipe length WHA ↔ QD |
L < 5 × ID on both sides |
Figure 14 |
| Acceptance |
No observed leakage and no seal washout |
§9.1.5 |
3.1 Test flow per size
| Size |
Rated flow (L/min) |
Test flow 2× (L/min) |
Test flow (GPM) |
| 02 |
2.27 |
4.54 |
1.20 |
| 04 |
6.44 |
12.88 |
3.40 |
| 06 |
11.36 |
22.72 |
6.00 |
| 08 |
17.79 |
35.58 |
9.40 |
Allowed deviation (internal proposal): + 5 % / − 0 % of the test flow. A flow lower than 2× rated is not allowed.
3.2 Sizing basis (internal proposal)
| Size |
Hose ID (as 9.1.4) |
Max L WHA↔QD (< 5·ID) |
Velocity in hose at 2× |
Max ΔP across QD at 2× (Min Cv) |
| 02 |
6.35 mm (1/4") |
< 31.7 mm |
≈ 2.4 m/s |
≈ 16 psi (1.10 bar) |
| 04 |
9.525 mm (3/8") |
< 47.6 mm |
≈ 3.0 m/s |
≈ 9.6 psi (0.66 bar) |
| 06 |
12.70 mm (1/2") |
< 63.5 mm |
≈ 3.0 m/s |
≈ 7.4 psi (0.51 bar) |
| 08 |
15.875 mm (5/8") |
< 79.3 mm |
≈ 3.0 m/s |
≈ 6.8 psi (0.47 bar) |
ΔP across the QD is calculated as 4 × ΔP at rated flow with Min Cv. The pump must handle this plus the circuit losses at a system pressure of 6.9 bar(g).
The specification does not state hose ID for this test. The same ID as in 9.1.4 is proposed (MD-023). A short distance to the WHA is difficult to achieve for the small sizes, so plan the fixture with the WHA directly on the adapter.
4. Equipment
A (valve, pressurization)
│
┌─┴─┐
│Air│ Accumulator (air cushion)
└─┬─┘
│ B (valve, flow adjustment / bypass)
├────────⊗────────┐
│ │
┌─────────────┴──( P )──────────┴───[C]──┬──────▶◀──────┬──[D]────(LPM)──┐
│ Circulation pump psig │ QD │ psig flow │
│ WHA WHA │
│ ├─ L ─┤├─ L ─┤ │
└────────────────────────────────── return ◀──────────────────────────────┘
L < 5 × ID
| Pos |
Component |
Function |
| A |
Valve to accumulator |
Pressurizes the circuit to operating pressure |
| Air |
Accumulator with air cushion |
Holds the system pressure and absorbs volume changes during connect and disconnect |
| Pump |
Circulation pump with speed control |
Provides the test flow |
| B |
Valve in parallel branch (bypass) |
Adjusts the flow. Lets the pump run when the QD is disconnected. |
| C |
Pressure transducer/gauge upstream of QD |
Operating pressure and pressure spike |
| D |
Pressure transducer/gauge downstream of QD |
Pressure on the downstream side |
| WHA |
Water hammer arrestor, one on each side |
Limits the pressure spike at disconnect |
| LPM |
Flow meter |
Test flow |
| QD |
Pair under test, mounted in the actuator |
– |
4.2 Instruments and equipment (internal proposal where there is no requirement)
| Equipment |
Proposal |
| Pressure transducers C and D |
0–16 bar(g), ≤ ± 0.25 % FS, sampling frequency ≥ 1 kHz to capture the pressure spike (MD-024) |
| Flow meter |
Range up to ≥ 40 L/min, ± 1 % of reading, withstands 60 °C and 10 bar |
| Temperature sensor |
Pt100 in the flow upstream of the QD, ± 0.5 °C |
| Heater / temperature control |
Holds 60 °C throughout the circuit; insulated lines |
| Accumulator |
Pre-charge about 4–5 bar, volume ≥ 1 L, so that the pressure is held stable during volume change |
| WHA |
Sized for hose size and 2× rated flow, mounted as close to the QD as possible |
| Linear actuator |
Speed 40 ± 10 mm/s, stroke ≥ engagement length + 10 mm, force ≥ 2 × max coupling force at 6.9 bar (UQDB08 ≈ 300 N, choose ≥ 600 N), with position sensor |
| Fixture |
Coaxial guidance of plug and socket; hard-stop for UQDB |
| Data logger |
Pressure C and D, flow, temperature, actuator position; ≥ 1 kHz for pressure |
| Camera (internal proposal) |
Film the coupling surfaces during cycling to document leakage |
Instrument accuracy per ISO 18869:2017 §6.3 applies generally (§9). Fill in the requirements from the standard.
5. Safety
- The fluid is at 60 °C and pressurized to 6.9 bar(g). A leak or seal washout gives a hot jet. Test behind a protective screen or in an enclosed rig.
- The accumulator contains pressurized gas. Depressurize the circuit before changing samples.
- The actuator delivers significant force. Use an emergency stop and a pinch-point guard.
- Wear eye protection and heat-resistant gloves when handling hot parts.
6. Preparations
| # |
Step |
| F1 |
Check sample ID, marking and visual status (03-Tests §3). Photograph the plug and socket valve seals before the test (internal proposal). |
| F2 |
UQD: remove or bypass the latch mechanism so that the actuator can connect and disconnect. Document how it was done. |
| F3 |
Mount plug and socket in the fixture so that the flow direction is P→S: inlet in the plug, outlet in the socket. Tighten terminations to the torque in Table 5. |
| F4 |
Check that the distance WHA ↔ QD is < 5 × ID on both sides. |
| F5 |
Fill the circuit with test medium and bleed. |
| F6 |
Heat to 60 °C and check that the temperature is stable. |
| F7 |
Set the actuator speed to 40 mm/s and verify with the position sensor (pass 30–50 mm/s). |
| F8 |
Define dwell times (internal proposal, MD-022): mated until flow is stable at the test flow, at least 5 s; unmated until pressure at C is stable, at least 5 s. |
7. Execution
7.1 Operating-state set-up (Test Setup)
| # |
Step (specification) |
| S1 |
With the QD fully mated: start the pump and adjust pump speed and/or valve B until the test flow (2× rated) is reached. |
| S2 |
Stop the pump and disconnect the QD. |
| S3 |
Start the pump. |
| S4 |
Pressurize the circuit via valve A (or otherwise) until 6.9 bar(g) is shown at pressure transducer C. |
| S5 |
Stop the pump. |
| S6 |
Connect the QD. |
7.2 Test process (Test Process)
| # |
Step (specification) |
| T1 |
With the QD connected and the pump running: check that the flow has not changed. If adjustment is needed it may be necessary to start over from S2. |
| T2 |
With the pump running: connect and disconnect the QD 20 times at 40 ± 10 mm/s. |
| T3 |
The pressure spike at disconnect must be ≤ 30 psi above the pressure before disconnect. |
| T4 |
Observe leakage and/or seal dislodged from the valve during all 20 cycles. |
| T5 |
If no failure occurred in T1–T4: reverse the flow direction to S→P and repeat T1–T4 (internal proposal: start over from S1 after reversing). |
7.3 One cycle in detail (internal proposal)
1. Mated, flow stable at 2× rated → log P_C, P_D, Q, T
2. Disconnect at 40 mm/s → log P_C max (spike)
3. Unmated, pressure stable at C → check for leakage at the end faces of both valves
4. Connect at 40 mm/s → check that flow returns to 2× rated
5. Leak check at the coupling boundary → next cycle
7.4 If the pressure spike exceeds 30 psi
The specification gives the limit as a requirement in the test process, not as an acceptance criterion for the coupling (MD-024). (Internal proposal): a cycle with a spike > 30 psi counts as invalid, not failed. Adjust the WHA, accumulator pre-charge or flow and rerun the cycle. Note the event in the protocol.
8. Assessment
8.1 Failure criteria
| Failure |
Definition (internal proposal, MD-025) |
| Leakage |
Liquid dripping or seeping from a closed valve when unmated, or from the coupling boundary when mated, beyond the normal fluid loss at disconnect (cf. 9.1.2). |
| Seal washout |
The seal has wholly or partly left its groove, is visible outside the groove or is found in the circuit, or the valve does not seal after disconnect. |
| Functional failure (noted) |
The coupling cannot be connected, the valve does not open so that flow is absent, or the coupling jams. |
8.2 After the test (internal proposal)
| # |
Check |
| E1 |
Visual inspection and photo of both valves' seals, compare with F1. |
| E2 |
Optional leak check of unmated plug and socket at 6.9 bar(g) for 3 min (as 9.2.4), as support in borderline cases. |
8.3 Acceptance
- Per pair and direction: Pass (P) if no leakage and no seal washout were observed during 20 cycles.
- Per product: Pass if all 5 pairs pass in both directions.
- A failure in P→S means S→P is not run for that pair, since the test process only continues if no failure occurred. The pair fails.
9. Protocol
| Field |
Value |
| Product / size / configuration |
|
| Sample ID pairs 1–5 (plug / socket) |
|
| Test type (own verification / cross-verification against ___) |
|
| Date, operator |
|
| Test medium (water / PG25) |
|
| Test flow set value (L/min) |
|
| Temperature (°C) |
|
| Operating pressure at C (bar(g)) |
|
| Hose ID, distance WHA↔QD (mm) |
|
| Latch mechanism (UQD): removed / bypassed, how |
|
| Actuator speed measured (mm/s) |
|
| Dwell times mated / unmated (s) |
|
| Instruments (ID, calibration date) |
|
| Deviations |
|
9.2 Cycle log (one table per pair and direction)
| Cycle |
Q before disconnect (L/min) |
P_C before (psi(g)) |
P_C spike (psi(g)) |
Spike Δ (psi) ≤ 30? |
T (°C) |
Leakage (Y/N) |
Washout (Y/N) |
Remarks |
| 1 |
|
|
|
|
|
|
|
|
| 2 |
|
|
|
|
|
|
|
|
| … |
|
|
|
|
|
|
|
|
| 20 |
|
|
|
|
|
|
|
|
9.3 Summary per §15.3.3
| Flow direction |
Pair 1 |
Pair 2 |
Pair 3 |
Pair 4 |
Pair 5 |
| S-P |
P/F |
P/F |
P/F |
P/F |
P/F |
| P-S |
P/F |
P/F |
P/F |
P/F |
P/F |
10. Sources of error and tips
- Flow drops after a number of cycles: valves or seals may have begun to deform, or temperature, and thereby viscosity, has changed. Log the flow every cycle.
- Air in the circuit: makes pressure and flow unstable and dampens the spike. Bleed carefully at 60 °C, since gas is released on heating.
- Misalignment: the fixture must give coaxial connection. Side forces can cause damage that is mistaken for washout.
- PG25: higher viscosity gives higher ΔP and greater stress on the seals. Always report which medium was used.
- Normal fluid loss at disconnect is not leakage. Assess leakage only after the surfaces have been wiped.
11. Open points affecting the test
| ID |
Short description |
Impact here |
| MD-009 |
Tolerance for 60 °C missing |
Internal proposal ± 3 °C |
| MD-022 |
Dwell times per cycle not stated |
Internal proposal in F8 |
| MD-023 |
Hose ID, UQDB alignment (SA) and which half moves not stated |
Internal proposal: ID as in 9.1.4, coaxial, hard-stop |
| MD-024 |
Status of the spike limit and measurement method unclear |
Spike > 30 psi = invalid cycle |
| MD-025 |
Definition of leakage and method for washout check missing |
Definition in 8.1 |
| BS-002 |
Hybrid or not |
Scope |
| BS-004 |
Water or PG25 |
Test medium |
| BS-006 |
Partners for cross-verification |
Samples for §10 |