UQD Bypass Jumper Hoses

PARTS ON THIS PAGE: RH-CSK-JMP-06 · $360 RH-CSK-JMP-06-10PK · $2,250 RH-CSK-RACK-04 · $2,200 RH-CSK-RACK-12 · $4,700 RH-CSK-CAP-06 · $440 RH-CSK-SPL-KIT · $400 RH-FC-300 · $26,600

What a bypass jumper is for

An ORV3 rack manifold presents a supply and return socket at every server position. Until a server blind-mates into those sockets, the circuit through that position is open — the socket valves hold coolant back, but nothing flows. That is fine for storage and useless for commissioning. Flushing debris out of the manifold, filling and venting, running a pressure-decay test, and exercising the CDU all require a closed loop with real flow paths through it.

The jumper is that closed loop. Its plug halves mate the manifold sockets the same way a server's couplings would, open the valves, and route coolant from supply back to return through 3/8-inch hose. Teams that skip them end up improvising with NPT adapter stacks and shop hose, which puts worm clamps, mixed metals, and untested seals into a loop everyone is about to certify clean.

What to look for in a jumper

Any hose with two couplings on it will technically close a circuit. Whether you want it in a loop that feeds cold plates is another question. The checklist we build to:

• Dry-break at both ends. UQD plug halves valve shut on disconnect, so pulling a jumper off a live manifold leaves a film on the coupling faces, not a puddle in the rack. Our ends are machined to the OCP UQD spec.
• 316L on wetted metal. The jumper sits in the same coolant as your cold plates. These use 316L barbs with crimped 304 stainless collars — no worm-drive clamps to back off under vibration.
• EPDM hose and seals. Compatible with PG25 and treated DI water across 34 F to 140 F, which is why EPDM is the standard seal material in glycol loops.
• Sensible length. 24 inches spans a supply/return pair on an ORV3 manifold (48 mm OU pitch) with enough slack to dress the hose away from rail hardware. Other lengths and end configurations are built to order.
• Individual testing. Each jumper is decay-tested at 100 psi before bagging. Design working point is 125 psi at 104 F.
• Flow capacity. Passes flush flows to 20 LPM without becoming the restriction in the branch.

SKUs and pricing

The 10-pack works out to $225 per jumper against $360 for singles — about 38% off, or $1,350 saved per pack. Ten is the practical unit: enough to close out a partially populated rack for flush and decay testing, or to spread two or three per rack across a row during staged bring-up.

If you also need caps, plugs, and coolant sampling on site, the rack kits bundle the same jumpers with the rest of the commissioning hardware in a foam-cut case, priced roughly 11% under buying the pieces separately.

Bypass jumpers and related commissioning hardware. Prices in USD; in-stock items ship in 5-7 business days.

Placement: flushing vs. decay testing

The two jobs want jumpers in different places, and running the wrong setup wastes a day.

Flushing is a velocity problem. Jumper every position and the flush flow divides across all of them — per-branch velocity drops and debris settles instead of carrying out. Work in banks instead: jumper two or three positions, cap the rest with a 6+6 set, and run the flush cart (ours is the RH-FC-300; any rig with 25-micron absolute filtration works) until the effluent runs clean at the filter. Move the bank and repeat down the manifold. Each jumper passes 20 LPM, so a bank of three gives the cart something to push against without deadheading it.

A decay test is the opposite. You want the whole volume closed and every interface that will see coolant included in it. Jumper every position you can and cap the remainder. Testing against bare socket valves only proves the shutoffs; a jumpered position proves the mated interface a server will actually blind-mate into. Two practical notes. First, EPDM hose relaxes slightly on first pressurization — let the loop sit 15 to 30 minutes at test pressure before opening the logging window, or the creep reads as a phantom leak. Second, keep field test pressure at or below 125 psi with jumpers inline; the manifold itself carries a serialized 225 psi factory hydro cert, which makes the jumper the limiting component in the field loop.

When the test passes, pull the jumpers, cap the open sockets, and pull a baseline coolant sample with the sampling port kit (RH-CSK-SPL-KIT) so the chemistry record starts at day zero.

Common questions