Coolant for CNC Machines: 7 Proven Selection Checks

• 1
  Key takeaways
• 2
  Coolant families compared for CNC operations
• 3
  Coolant problems and what they usually mean
• 4
  Match the fluid to the alloy, not the machine brand
• 5
  Coolant for CNC machines: concentration control decides tool life
• 6
  Sump volume and turnover rate set the maintenance load
• 7
  Water quality and mix procedure are part of the coolant spec
• 8
  When MQL or near-dry machining is the wrong answer
• 9
  Coolant cost sits in maintenance, not in the drum
• 10
  Step by step: choosing and commissioning a coolant

Start with the workpiece. Aluminium 6061 and 7075 cut cleanly with synthetic fluids because the fluid rinses chips instead of clinging to them. Cast iron is different: fine graphite swarf absorbs oil and turns the sump into sludge, so many shops run cast iron dry or with a light synthetic and a strong chip conveyor.

Stainless 304 and 17-4PH work-harden under the tool edge. They need a fluid that carries heat away fast, so higher concentration and higher flow beat exotic additives. Titanium TC4 and Inconel sit at the other end. Heat stays in the cut, and straight oil or high-pressure emulsion at 70–100 bar keeps the edge alive.

Magnesium AZ31B and AZ91D are a special case. Water-based fluid reacts with fine magnesium fines and creates hydrogen. If you machine magnesium, that is a separate risk review, not a coolant swap.

• 1
  AluminiumSynthetic or semi-synthetic, 5–8%, good chip evacuation.
• 2
  Stainless and tool steelEmulsion at 8–10%, high flow, clean sump.
• 3
  Titanium and nickel alloysStraight oil or high-pressure emulsion, never a lean mix.

A refractometer reads the dissolved solids in the mix. On a freshly charged sump, a 7% emulsion might read 7.5% because of the way the fluid refracts. That offset is normal. What is not normal is drifting to 3% after two weeks of top-ups with straight water.

Low concentration causes rust on fixtures, sticky residue on the table, and a fast drop in tool life. High concentration causes foam, skin irritation, and on aluminium it can leave a gummy film that ruins anodising. On aluminium, keep the mix in the 5–8% band. On stainless and steel, 7–10% is common.

The check itself takes a minute. Pull a sample from the return line, not the surface. Note the reading, the date and who topped up the sump. That log is the difference between a stable process and a weekly fire drill.

A small machine with a 60 L sump and a single shift may sit idle for 16 hours a day. Bacteria grow in still, warm fluid. A large central system with 3,000 L and continuous circulation stays cleaner because the fluid keeps moving and gets filtered.

If you run several machines, a central system usually wins on labour. One charge, one filtration loop, one concentration reading per day. A single-machine sump is cheaper to install but costs more in labour per litre.

Add a skimmer if your machines see hydraulic leakage or way-lube carryover. Tramp oil floats, seals the fluid from air, and gives anaerobic bacteria a place to grow. The smell usually shows up two weeks after the first slick.

• 1
  Sump volumeSize the charge so the fluid turns over at least twice per shift.
• 2
  Filtration20–50 μm on central systems; a chip conveyor alone is not filtration.
• 3
  Tramp oilSkim daily on lathes; weekly may be too slow.

Hard water reacts with the emulsifier and forms scum. Above roughly 200 ppm as CaCO3, expect more foam, shorter sump life and cloudy mix. If your tap water is hard, a softener or a reverse-osmosis blend pays back quickly on a central system.

Mix order matters. Always add concentrate to water, never water to concentrate. Adding water to a drum of concentrate creates an unstable invert emulsion that separates within hours. Use a venturi mixer or a paddle at low speed, and never pour concentrate straight into a sump.

Check pH weekly. Most emulsions run between 8.5 and 9.5. Below 8.0, bacteria are winning and rust risk climbs. Above 9.8, you risk skin irritation and attack on aluminium. A simple dip strip is enough for daily checks.

Minimum quantity lubrication uses 10–50 mL/h of oil carried in air. It works well on aluminium and titanium finishing where the cut is open and chips clear easily. It also avoids the whole sump maintenance problem, which matters if you run one shift and hate weekends.

It fails in deep pockets, deep-hole drilling and high-pressure through-spindle work. There is no fluid mass to carry heat away, and chips pack into pockets. If your part has a 6× diameter bore, MQL will likely burn the drill.

MQL also changes the shop environment. Oil mist needs extraction, and some facilities treat it as a combustible risk. That is a facility decision as much as a machining one.

The purchase price of a drum is a small part of the total. The bigger costs are labour for sump cleaning, disposal of spent fluid, tool wear from a lean mix, and scrap from rust or poor finish.

A shop running 20 machines can spend more on sump cleaning hours than on fluid. That is why central systems and skimmers look expensive on paper but pay back over a year.

If you outsource machining, ask how the supplier controls concentration and how often the sump is charged. A shop that cannot answer that question is likely topping up with water and hoping.

• 1
  Charge frequencyCentral systems: 6–12 months. Small sumps: 2–4 months.
• 2
  DisposalSpent emulsion is regulated waste in most EU and US states.
• 3
  Tool costA lean mix can double insert consumption on stainless.