Case Study: Cutting Die Lubricant Use on a Hot Forging Press
A hot forging press moved from a manual spray gun to timed air atomizing die lubrication: 45 mL per stroke, 50 percent less lubricant, fewer laps.
A hot forging press applies die lubricant once per stroke. That action sets the die temperature, the release of the part, and whether the next forging fills its cavity. Flood the die and the face chills, the billet skin freezes, and laps open at the corners. Starve it and the forging sticks and the flash land wears.
This case describes a European closed-die forging plant that lubricated its dies by hand with an air-assisted spray gun. The change was to external mix air atomizing nozzles on timed injection, and every figure is a working value for a press of this class.
Case Snapshot
The reference configuration below is the generic shape of the installation.
| Item | Reference configuration |
|---|---|
| Industry | Closed-die hot forging, carbon and alloy steel |
| Duty | Die face lubrication, one application per forging stroke |
| Equipment | Forging press on a 10.0 s cycle, billet at 1,150 to 1,250 °C, die face 250 to 400 °C, 300 x 250 mm face per die half |
| Original setup | Manual air-assisted spray gun, 3.0 s of trigger time per stroke, weekly brushing of the open pipe |
| Nozzle chosen | External mix air atomizing flat fan, 65 degree, six nozzles on a manifold, timed injection |
| Operating point | 45 mL per stroke, 0.8 s spray window, 2 bar atomising air at 7 Nm³/h per nozzle |
| Material requirement | Hardened stainless liquid tip and air cap, 316L body, tungsten carbide tip available |
| Compliance | EN 10204 3.1 certificates, IP65 coil and connection, documented spray settings |
The press class matters more than the exact forging. A closed-die press on an 8 to 12 s cycle is the common shape in hot forging. The same sizing transfers to screw, hydraulic, and mechanical presses.
The Challenge
The press forged a steel component of about 1.5 kg from an induction-heated billet. It ran a 10.0 s cycle, or 360 strokes an hour and about 1,080,000 strokes a year across two shifts. The die face ran at 250 to 400 °C.
Lubrication was manual. An operator loaded the billet and sprayed both die halves with an air-assisted gun, holding the trigger 3.0 s of each 10.0 s cycle. That 30 percent duty and a flow of about 108 L/h while open put about 90 mL of diluted lubricant on the faces.
The drop size was the first problem. A hand gun at that flow makes drops of 150 to 300 µm, which land where they are thrown. The flash land soaked while the deeper pockets stayed nearly dry.
Volume was the second. The gun had no meter, so two operators ran different die temperatures from one setting sheet. The spread around the 90 mL target was about plus or minus 30 percent.
Water was the third. At 90 mL over about 0.24 m² of active die surface the spray laid down roughly 375 mL/m². That is twice what a metered film needs. Every extra millilitre has to boil off, and boiling takes heat out of the die. The flood pulled the face from about 320 °C down to about 230 °C each cycle.
Rejects were the fourth problem. Laps, underfill, and sticking at ejection ran at 4.2 percent of output, which is 3,780 rejected forgings a month. The fifth was cost. The press used 97,200 L of mixed lubricant a year at about $38,880, the open pipe was brushed weekly, and tips were replaced every 6 to 8 weeks. Die life averaged 12,000 strokes, or 90 die changes a year.
The Solution
The plant stopped treating lubrication as a hose and started treating it as a metered film. The target was a thin, even graphite film across the active surface: the flash land, the draft walls, and the pocket floors. About 5 µm is the depth a water-based graphite lubricant leaves once the carrier flashes off.
That depth sets the liquid volume. A typical water-based graphite die lubricant carries about 10 percent solids by volume and holds about 25 percent of what lands. Leaving 5 µm of graphite therefore needs about 200 µm of liquid. Over 0.24 m² of active surface that is about 48 mL per stroke, and the target was set at 45 mL.
The second step was the nozzle. The duty needs droplets below about 100 µm, so the film follows the die contour instead of pooling. A hydraulic nozzle at 34 L/h makes drops of 150 to 300 µm. An external mix air atomizing cap takes the same flow to 25 to 60 µm.
External mix rather than internal mix was the third step. A graphite suspension is a solids load, and an internal mixing chamber holds it in a confined volume. That volume sits static for 8.6 s of every 9.4 s cycle, so it silts. External mix passes the liquid through a short passage to a tip with no chamber, so only a thin film waits between strokes. The internal vs external mix guide works the same decision.
Layout was the fourth step. Six nozzles sit on the manifold: two per die half for the pockets and the face, and two more for the flash land rim and the ejector area. Each is a 65 degree flat fan mounted 250 mm from the die face and aimed within 10 degrees of normal. It covers 318 mm against a 300 mm face.
Timing was the fifth. A solenoid valve replaced the operator’s finger, opening 0.4 s after the part clears the cavity and closing 0.5 s before the die shuts. The spray window is 0.8 s inside the 9.4 s cycle, an 8.5 percent duty.
Supply discipline was the sixth. A 200 µm basket strainer sits on the pump discharge, and a 100 µm element sits at the manifold inlet. The atomising air passes a 5 µm coalescing filter and a dryer. The setting sheet now carries numbers: air pressure, liquid pressure, spray window in seconds, and volume per stroke in millilitres.
The Engineering Detail
The film has to be thin, even, and early
Lubrication does three jobs in closed-die forging. It releases the forging, it lays a graphite barrier that keeps hot steel from welding to the tool, and it draws heat out of the die. All three depend on a film that is thin and even. A thick film wastes lubricant and over-cools the die, and a patchy one leaves metal-to-metal contact at the flash land.
Of the 3.4 s die-open window in a 9.4 s cycle, the spray uses 0.8 s. The carrier then takes 0.5 s to flash off. Sprayed too late, the face is still wet when the billet lands. Sprayed too early, the graphite oxidises before it does any work.
Water is the carrier, and the die runs in film boiling
A die face at 250 to 400 °C sits above the film boiling band for water. The Leidenfrost point for water on steel falls at roughly 180 to 220 °C. A drop that lands on the face rests on a vapour cushion instead of wetting the metal. Drop velocity therefore matters as much as drop size, and air atomizing supplies both.
The cooling limit follows from the same physics. Every millilitre of water has to boil off, and the energy comes out of the die. A 45 mL film over 0.24 m² is 188 mL/m², a controlled draw of heat. The old 90 mL flood was 375 mL/m², and its swing from 320 °C to 230 °C on every cycle drives heat checking at the tool surface. Keep the face above about 200 °C at the end of the spray: too cold and the billet skin chills into laps, too hot and the graphite burns back.
Internal mix or external mix for a graphite suspension
An internal mix nozzle shears air and liquid together inside the cap and makes the finest drops for a given air volume. The fluid is the problem. A water-based graphite die lubricant carries 10 to 30 percent solids between roughly 1 and 15 µm. A closed chamber gives that solids load nowhere to go between strokes.
Stokes’ law gives the settling rate. A 10 µm graphite particle at a density of 2.2 g/cm³ falls through still water at roughly 4 mm per minute. A 20 mm deep chamber is covered in about five minutes of true stillness. Dispersants slow that, and the coarse fraction still drops out and re-enters the tip as lumps. External mix avoids the chamber: the liquid runs a short passage to a tip and meets the air at the cap face.
Timing, duty cycle, and the worked arithmetic
The valve duty cycle is the fraction of the press cycle during which the valve is open. It sets the ratio between a catalogue flow figure and the flow the pump actually sees:
duty cycle = spray window / cycle time
At 0.8 s in a 9.4 s cycle the duty is 8.5 percent, so the manifold passes 11.8 times its average flow while open. Size the pump, the header, and the strainer on that instantaneous flow. Valve life follows cycles rather than hours. At 1,080,000 strokes a year a solenoid rated for 10 million operations would last nearly nine years, and a seal kit service sits at 4,000,000 cycles.
| Step | Value | Arithmetic |
|---|---|---|
| Cycle time | 9.4 s | set by the press after the change |
| Strokes per hour | 383 | 3,600 / 9.4 |
| Lubricant per stroke | 45 mL | target from the 5 µm film calculation |
| Lubricant per hour | 17.2 L/h | 383 x 45 mL = 17,235 mL/h |
| Spray window | 0.8 s | set by the die-open window |
| Valve duty cycle | 8.5 percent | 0.8 / 9.4 |
| Instantaneous manifold flow | 202 L/h | 17.2 / 0.085 |
| Flow per nozzle, six nozzles | 33.7 L/h | 202 / 6 |
| Atomising air, six nozzles | 42 Nm³/h | 6 x 7 Nm³/h |
| Air per hour, average | 3.6 Nm³/h | 42 x 0.085 |
| Air per stroke | 9.3 normal litres | 42 x 0.8 / 3,600 |
Run the chain backwards and it holds. At 383 strokes an hour, 45 mL per stroke is 17.2 L/h. Over 1,080,000 strokes a year that is 48,600 L, and dividing the annual volume by the annual stroke count returns 45 mL per stroke. The same test on the old setup: 108 L/h held 3.0 s of a 10.0 s cycle is a 30 percent duty and an average of 32.4 L/h. At 360 strokes an hour that is exactly 90 mL per stroke.
Standoff, angle, and the covered die area
Coverage width for a flat fan follows W = 2 x h x tan(theta / 2), where h is the standoff and theta is the spray angle. At 250 mm with a 65 degree fan, that gives 318 mm. It contains the 300 mm die face with about 9 mm of margin per side. Standoff is a volume decision wearing the costume of a mounting decision, because the deposit per unit area falls as the fan widens.
| Standoff | Covered width | Overlap between neighbours | Relative deposit | Note |
|---|---|---|---|---|
| 150 mm | 191 mm | none, a 9 mm gap | 1.67 | bare stripe at mid band |
| 200 mm | 255 mm | 22 percent | 1.25 | tight, sensitive to misalignment |
| 250 mm | 318 mm | 37 percent | 1.00 | chosen: even coverage, mask kept clear |
| 300 mm | 382 mm | 48 percent | 0.83 | overlap past target, some waste |
| 350 mm | 446 mm | 55 percent | 0.71 | 146 mm of pattern lands past the die |
The 30 to 40 percent overlap band is the working target for an even film, and only the 250 mm row sits inside it. At 350 mm the pattern still covers the die, but the deposit per unit area falls about 29 percent. Aim within 10 degrees of the face normal: past 20 degrees the fan footprint distorts and the far edge thins.
Air consumption and the cost of atomisation
Atomising air is a running cost that a hand gun hides. The chosen cap uses about 7 Nm³/h at 2 bar, so six of them draw 42 Nm³/h while the valve is open. At an 8.5 percent duty the average is 3.6 Nm³/h, or 9.3 normal litres of air on every stroke.
At 1,080,000 strokes a year the manifold uses 10,080 Nm³ of compressed air. This plant delivers compressed air at $0.03 per Nm³, so the atomising air costs about $302 a year. The air-assisted gun it replaced spent about $270, on 9,000 Nm³, so atomisation added roughly $32 to the air bill.
That is the honest figure for a timed injection duty, because these nozzles fire for 8.5 percent of the cycle. Where the air cost really bites is a continuous line, as the air cost guide sets out.
Solids, strainers, and the changeout interval
Graphite is a soft abrasive and a settling one, and a header that runs 8.5 percent of the time offers little motion to keep it mixed, so strainer mesh is the first control. The liquid tip has a 1.5 mm orifice, and general practice filters to about one third of the smallest orifice. That puts the strainer at 500 µm or finer, but a graphite line wants finer still, because graphite agglomerates into lumps once it settles. The plant runs a 200 µm basket strainer (80 mesh) on the pump discharge and a 100 µm element (150 mesh) at the manifold inlet. A 45 µm screen (325 mesh) blinds within a shift.
Line velocity is the second control. At the 202 L/h instantaneous flow, an 8 mm bore header runs at about 1.1 m/s. That is fast enough to hold the graphite in suspension and scour the pipe on every injection. A 20 mm header would run at about 0.18 m/s, and the graphite would settle along the bottom of the run.
Changeout follows the same facts. Graphite has a Mohs hardness of 1 to 2 and the tip is hardened stainless, so the plant measured a flow drift under 2 percent after 1,000,000 strokes. Wear is not the limiting item. The deposit that dries on the cap and tip between injections is, so the schedule is written in stroke counts.
| Task | Interval | Trigger for doing it sooner |
|---|---|---|
| Pattern check against the reference sheet | 250,000 strokes, about 12 weeks | streak or ragged edge in the fan |
| Air cap and liquid tip clean | 500,000 strokes, about 6 months | dry deposit visible on the cap face |
| Strainer wash out | 250,000 strokes, about 12 weeks | rising differential pressure |
| Liquid tip replacement | 3,000,000 strokes, about 2.8 years | flow drift over 5 percent at fixed pressure |
The Results
The numbers below compare the twelve months before the change with the twelve months after, at the same annual output of 1,080,000 strokes. They are working figures for this class of press.
| Metric | Before | After | Unit |
|---|---|---|---|
| Lubricant per stroke | 90 | 45 | mL |
| Lubricant per year | 97,200 | 48,600 | L |
| Lubricant spend per year | 38,880 | 19,440 | $ at 0.40 per L |
| Atomising air, average | 3.0 | 3.6 | Nm³/h |
| Compressed air spend per year | 270 | 302 | $ at 0.03 per Nm³ |
| Cycle time | 10.0 | 9.4 | s |
| Rejected forgings, laps and sticking | 4.2 | 0.9 | percent |
| Die change interval | 12,000 | 16,500 | strokes per die |
| Die changes per year | 90 | 65 | count |
| Operator spray time | 144 | 8 | minutes per shift |
Lubricant was the largest saving, halved from 90 to 45 mL per stroke. That 50 percent cut sits inside the 30 to 60 percent band published for the change from flooding to air atomisation. The volume fell without a loss of release because the drop size fell at the same time.
The reject rate fell from 4.2 to 0.9 percent, about 2,970 fewer forgings a month. Those rejects were laps, underfill at the corners, and stickers at ejection, the symptoms of a film too wet in one place and too thin in another. Die life rose from 12,000 to 16,500 strokes, removing 25 die changes a year, roughly 20 press hours.
The operator line matters as much as the money. One person went from about 144 minutes of an eight hour shift holding a trigger, 30 percent of the shift on one action. That fell to about 8 minutes on a manifold check. Only one number rose: compressed air, by about $32 a year against a lubricant saving of $19,440.
Why This Matters
Most die lubrication problems are answered by buying a different lubricant. This case shows why volume, drop size, timing, and coverage have to be fixed together, because the fluid was the same before and after.
The method transfers in four steps. Work out the film depth the die needs, then convert it to millilitres per stroke from the die area and the solids fraction. Check that figure against the strokes per hour. Choose a mix geometry that survives the fluid, size the standoff from the coverage width, and set the spray window inside the die-open time.
Three thresholds are worth carrying into any hot forging press. Keep the die face above about 200 °C at the end of the spray, or laps and cold shuts follow. Keep the fan overlap between 30 and 40 percent, or the film goes patchy. Hold the atomising air at the cap data sheet pressure, because consumption climbs with absolute pressure while the drops barely change.
The maintenance logic is the other half. On a graphite suspension the nozzle does not wear out, it silts up. The changeout interval therefore comes from the solids in the fluid and the stroke count, not from a calendar.
The same arithmetic runs on any press with an 8 to 12 s cycle, and a pattern test on the actual die settles the last of it. If your press is flooding its dies to get release, send the die face drawing, the cycle time, and the volume per stroke to the BoreJet application team through the contact page, and browse the nozzle range used here on the air atomizing nozzle page.
Your Duty May Differ
Send the Vessel Drawing and the Pump Curve.
We size the head class, the count and the operating point for your tank, not a catalogue guess. A coverage test on the actual vessel beats any estimate.