Sealing a mini mill enclosure for flood coolant

Short answer: seal the door with a full-perimeter EPDM cord compressed 20-30%, keep every fastener outside the splash line, slope the floor to a 0.5 in [13 mm] drain, and stand the whole box on a catch tray. Flood coolant is low pressure, but it runs for the whole cycle, so it wicks through any gap a door leaves open. If you are still weighing flood against mist and air blast, settle the coolant choice for a small CNC first; the sealing job below assumes flood, because flood finds every seam you skip.

Where splash finds a way in

A cutter spinning at hobby-mill speeds of 5,000-20,000 rpm throws emulsion at the door, the hinge and the latches for as long as the cycle runs. One splash bounces off. The fiftieth creeps into a joint you could feed a feeler gauge into, and gravity drags it down to the floor seam. Five paths do nearly all the leaking on a hobby box, and each has a fix that costs minutes now instead of a spilled sump later.

This is the hardware a Type 4 washdown enclosure leans on. A compressed full-perimeter gasket, latches that pull the door square, drainage that never breaks the seal. What a NEMA 4 rating asks of gaskets, doors and hinges walks through that hardware for electrical boxes. Enclosure type ratings are manufacturer ratings; a catalog box is not UL listed, and your DIY build answers to the water test alone.

What you need

• 1/4 in [6 mm] EPDM cord, about 48 in [1.2 m] for a 12 x 10 in door, plus splice glue made for the cord
• A door stop or retainer that sets the closed gap; the next section works it out
• 4 corner latches and a piano hinge with a stainless pin
• A 0.5 in [13 mm] bulkhead drain fitting with a chip guard over its mouth
• A catch tray with a 1 in [25 mm] lip; a full-size sheet pan, 18 x 26 in [460 x 660 mm], clears a 16 x 14 in base on every side
• Bulkhead fittings with o-rings for the coolant lines, and a drip loop for the power cord
• Modelling clay and calipers, to set the squeeze before coolant ever runs

A sealed wet box also traps motor heat, so size for it with enclosure size and heat for a hobby CNC before you cut sheet.

Steps to seal the box

1. Form the floor slope first. Wedge the frame so the drain corner sits low while the epoxy base cures: 1/8 to 1/4 in per foot of fall, the slope drain piping runs at. A 14 in [355 mm] deep floor then drops 0.15-0.3 in [4-7 mm] from the high wall, enough that coolant runs to the drain and chips ride along instead of damming.
2. Fit the drain at the low corner. Drill for the 0.5 in [13 mm] bulkhead, seat it with an o-ring each side, and keep its mouth out of the chip pile. The drain ends over the tray, and the tray is the last line before your bench.
3. Retain the gasket and glue the one splice. A groove in the frame edge holds the cord square; a screwed retainer strip does the same job on a flat frame. Pick the section with the gasket hardness and thickness guide, and use EPDM for a water-based emulsion: it is the hot-water elastomer, good from about -50 to 150 C. Make one splice, at the top centre, glued with cord adhesive; a dry butt joint wicks like a straw. Foam tape is the other common door seal, and the foam or rubber door seal comparison shows why it loses under running water.
4. Set the door stop, not the latch force. The stop fixes the closed gap; the latch only holds the door there. For a 12 x 10 in door on 1/4 in cord, the next section lands on 0.19 in [4.8 mm].
5. Keep the hardware dry. Hang the door so the hinge pin sits outside the gasket line, run a soft vinyl fin across the hinge gap, and fit a latch at each corner so the door pulls square at four points. Retainers through polycarbonate take a specific torque; the screw torque guide for a plastic door seal gives the numbers.
6. Bring every line through the top. Coolant supply and return pass through bulkheads with an o-ring each side; power drops in and loops below its fitting before it runs down the wall. It is the same sealing job as a probe through an enclosure wall.
7. Prove it on water. Close the door, run the pump on plain water for 10 min, and wipe the seams with a dry paper towel. Water finds the leaks for free; emulsion leaves a film that takes an evening to scrape off the bench.

Setting the door stop on a 12 x 10 in door

The inputs: a 12 x 10 in [305 x 254 mm] door, typical for a 16 x 14 in hobby mill box — scale everything to your opening; 1/4 in [6 mm] EPDM cord, the common cord stock; a 25% squeeze target, the middle of the 20-30% band door gaskets are designed to.

Two rules do all the work. Gasket length is twice the width plus twice the height. Closed height is the free height times (1 − squeeze).

The arithmetic:

• Length = 2 × (12 in + 10 in) = 44 in [1.1 m]; order 48 in [1.2 m] so the splice and trim are covered.
• Closed height = 0.25 in × (1 − 0.25) = 0.1875 in, call it 0.19 in [4.8 mm]; the stop crushes the cord by 0.06 in [1.6 mm].
• Band ends: at 20% squeeze the cord closes to 0.25 in × 0.80 = 0.20 in [5.1 mm]; at 30% it closes to 0.175 in [4.4 mm].

Decision: shim the stop until the closed gap reads 0.175-0.20 in [4.4-5.1 mm] at every point, aiming at 0.19 in. Measure with clay: press 1/4 in balls onto the frame at the four corners and four mid-spans, close the door to its stops, open it, and caliper the squashed discs. Treat 0.20 in as the weep line — under 20% squeeze, the latch corners pass splash — and 0.175 in as the bow line, where a 1/8 in [3 mm] polycarbonate door flexes enough to lift a corner off the seal.

On a flat plastic door you can seat an o-ring in a groove instead of a face cord; standard o-ring practice caps gland fill near 85% so the ring keeps room to spread. O-ring versus flat gasket sealing for acrylic panels weighs the ring against the strip on plastic, and raised lip or flat face gasket in a sealed enclosure sorts out the face geometry.

Checks that prove the seal

• Clay discs read 0.175-0.20 in at all eight points, and again a week later; a reading that stays in band means the stop holds the squeeze, not the latch.
• A paper strip pulled from the closed seal drags with firm resistance at every point; a strip that slides free marks a dead spot.
• A 10 min plain-water run leaves the seams dry when you wipe them with a white paper towel.
• 1 cup [240 ml] poured on the high corner reaches the drain and lands in the tray; water that pools means the slope never formed.
• After the first real session, the bench under the tray is dry; a puddle there points at the floor seam or the drain fitting, not the door.

Mistakes that put emulsion on the bench

• Filler over a field of screw holes. Filler cracks with door flex and each hole wicks anyway; bond retainers on gasketed bosses and leave the wall with 5 fasteners instead of 50.
• Compression by latch force. Latches crushing a fat cord work for a day, then bow the door; the stop sets 0.19 in and the latch only holds it there.
• A dry splice. One unglued butt joint leaks more than the whole perimeter.
• Over 30% squeeze. Thin polycarbonate doors bow and lift a corner off the seal.
• Lines through the bottom. A fitting sitting in the puddle leaks every single session.
• A drain with no tray. The drain alone empties onto your bench; the 1 in lip is what catches whatever the drain passes and whatever the seams weep.

If the retrofit turns into filler and hope, start clean instead. A custom stainless enclosure carries the slope, the 0.5 in drain and the gasketed door in the drawing from the first cut, or you set them yourself in the 3D enclosure builder; either way it is delivered in 10 business days from your approval of the drawing.

Catalog dimensions to compare

Compare these catalog dimensions with your component layout. Open each enclosure for its rating, material and usable space, then check those specifications against your application.

Questions