Pick the wrong coolant for magnet cutting and you’ll know within hours. On NdFeB, water-based coolant turns a freshly cut surface grey with Nd(OH)₃ corrosion. On ferrite, oil-based coolant creates a sludge that clogs your filtration system for no benefit — ferrite doesn’t corrode in water. And on SmCo, it barely matters which you use, because the material shrugs off both.
We’ve run thousands of magnet cutting jobs on our scies à fil diamanté sans fin across all three material families. The magnet cutting coolant choice is the single most common source of preventable quality failures we see from new customers — and it’s the easiest one to fix once y ou understand why each material reacts differently.
Why Does Coolant Matter During Magnet Cutting?
Coolant in diamond wire cutting serves four functions, and the priority ranking shifts depending on the material:
Lubrification. The diamond grits on the wire create friction against the workpiece. Coolant reduces this friction, which lowers cutting force, extends wire life, and improves surface finish. Oil-based coolants have inherently better lubricity than water — the viscosity keeps a thin film between wire and workpiece even under load.
Heat dissipation. Diamond wire cutting generates less heat than grinding or blade cutting, but it’s not zero. Localized temperature at each grit contact point can reach 100–200 °C momentarily. Coolant absorbs and carries away this heat. Water has roughly 2× the specific heat capacity of mineral oil (4.18 vs ~2.0 J/g·K), so it’s objectively better at cooling. For most magnet cutting, this advantage is academic — the heat generation is low enough that either type works. But on thick cross-sections (above 40 mm) at higher feed rates, water-based coolant keeps the cutting zone noticeably cooler.
Chip evacuation. The wire drags coolant through the cutting kerf, flushing out swarf particles. If swarf accumulates in the kerf, the loose particles get re-cut by the wire, scratching the cut surface and accelerating wire wear. Lower-viscosity coolants (water-based) flush more effectively through narrow kerfs. Higher-viscosity coolants (oil-based) carry particles better once they’re entrained but penetrate narrow kerfs less readily.
Surface protection. This is where the material chemistry takes over. On reactive materials like NdFeB, the coolant must prevent oxidation of the freshly exposed surface during cutting. On chemically stable materials like ferrite and SmCo, this function is irrelevant. This single factor — surface protection — is why coolant selection is material-dependent rather than universal.
Which Magnet Cutting Coolant for Which Material?
Here’s the decision matrix we use in our application lab:
| Matériau | Recommended Coolant | Pourquoi | Can You Use the Other? |
|---|---|---|---|
| NdFeB | À base d'huile (huile minérale blanche) | Nd-rich grain boundary phase reacts with water → corrosion | Water-based with corrosion inhibitor (≥3%) possible but risky |
| Ferrite | À base d'eau | Chemically inert to water; better cooling and chip flushing | Oil works but creates unnecessary sludge |
| SmCo | Water-based or oil-based | Corrosion-resistant; either works | True flexibility — choose based on other process constraints |
The table is simple, but the consequences of getting it wrong are not.
What Happens When You Use Water Coolant on NdFeB?
This is the mistake we see most often. A shop cuts la ferrite or silicon all day with water-based coolant, then switches to a NdFeB job without changing the coolant. The cuts look fine coming off the machine. Four hours later, the cut surfaces are grey.
The chemistry is straightforward. The Nd-rich phase (roughly Nd₇₀Cu₃₀ composition at the grain boundary) reacts with water:
Nd + 3H₂O → Nd(OH)₃ + 1.5H₂↑
This reaction starts immediately on freshly exposed surfaces. Within the first hour, a visible oxide/hydroxide layer forms. Within 24 hours, the corrosion penetrates along grain boundaries to a depth of 5–20 μm, weakening the inter-granular bonds. Parts that looked fine after cutting become chalky and fragile.
The hydrogen gas released by the reaction creates another problem that’s easy to miss. In an enclosed coolant system, hydrogen bubbles accumulate in the cutting zone, disrupting the coolant film and creating momentary dry-cutting conditions. We saw this on one of our test setups — intermittent surface roughness spikes that we couldn’t explain until we noticed tiny bubbles in the kerf. Switched to oil and the problem disappeared.
If you absolutely must use water-based coolant on NdFeB (some facilities require it for fire safety or environmental compliance), add a corrosion inhibitor at minimum 3% concentration. Sodium nitrite-based inhibitors work, but check compatibility with your downstream plating chemistry — some plating lines are sensitive to nitrite residue. And move cut parts to oil-coating or drying within 30 minutes. Any longer is gambling.
Why Oil-Based Coolant Is Standard for NdFeB
White mineral oil (sometimes called liquid paraffin or cutting oil) is the default for NdFeB diamond wire cutting because it solves the corrosion problem completely. The oil film excludes water and oxygen from the freshly cut surface, and the residual oil layer on the cut parts provides temporary corrosion protection during handling and storage.
Typical specs for NdFeB cutting oil:
| Property | Plage typique | Notes |
|---|---|---|
| Viscosity (40 °C) | 5–15 mm²/s | Lower viscosity for better kerf penetration |
| Flash point | >150 °C | Safety margin for continuous operation |
| Water content | <0.1% | Critical — even trace water causes problems |
| Additive package | Anti-rust, anti-foam | Standard cutting oil additives |
Flow rate matters more than viscosity. Sur nos SG20-R machines, we run oil flow at 2–4 L/min, directed at the wire entry point into the kerf. The goal is a continuous oil film on both sides of the wire throughout the cut. If you see dry spots on the cut surface after the wire passes — visible as matte patches versus the glossy oil-wetted areas — increase flow rate or adjust nozzle position.
One thing we learned the hard way: oil-based coolant degrades over time. As you cut, fine NdFeB particles accumulate in the oil. These particles are sub-10 μm and don’t settle easily. After 2–3 weeks of continuous use, the particle loading increases viscosity, reduces cooling efficiency, and creates a gritty film that accelerates wire wear. We now change oil every 2 weeks on high-utilization machines and filter continuously through a 5 μm cartridge filter during operation.
Can You Use Water-Based Coolant on Ferrite and SmCo?
Yes — and in most cases, you should.
Ferrite (SrFe₁₂O₁₉ / BaFe₁₂O₁₉) is a ceramic oxide. It’s already fully oxidized. Water does nothing to it chemically. Using oil on la ferrite is like putting sunscreen on a rock — technically harmless but pointless, and it creates cleanup problems you don’t need.
Water-based coolant advantages for ferrite cutting:
- Better heat dissipation (higher specific heat capacity)
- More effective chip flushing (lower viscosity gets into the narrow kerf better)
- Easier cleanup — ferrite swarf washes off with water; oil-based swarf sticks to everything
- Lower consumable cost — water-soluble cutting fluid concentrate costs roughly $15–30/gallon at typical 5–10% dilution ratios, versus $30–60/gallon for dedicated cutting oil
- Environmental and disposal advantages — water-based fluid waste is simpler to treat and dispose of per EPA guidelines
SmCo is similarly corrosion-resistant, so water-based coolant works fine. The only scenario where we use oil on SmCo is when the same machine is scheduled for NdFeB work immediately after — it’s faster to stay on oil than to flush and switch twice.
How to Manage Coolant When Cutting Multiple Magnet Materials
This is the real-world magnet cutting coolant challenge. Most shops cutting magnets process more than one material family. The question is how to handle coolant transitions.
Scenario 1: Dedicated machines per material. If you have separate machines for NdFeB and ferrite/SmCo, keep oil on the NdFeB machine and water on the others. Simple, no changeover needed.
Scenario 2: One machine, multiple materials. Here’s where it gets tricky.
The safe sequence is: ferrite/SmCo first (water-based) → flush → NdFeB (oil-based).
Going in this direction works because a small amount of residual oil in a water-based system has minimal effect on ferrite cutting quality. But a small amount of residual water in an oil system will attack NdFeB surfaces.
The flush procedure we recommend:
- Drain the water-based coolant from the tank
- Run 1–2 liters of oil through the system to displace water from lines and nozzles
- Wipe down the work area, guide wheels, and any surfaces where water droplets might linger
- Fill with fresh cutting oil
- Run the pump for 2 minutes before loading the NdFeB workpiece
Going the reverse direction (NdFeB oil → ferrite water) is less risky but still needs a flush. Oil residue in water-based coolant causes foaming, reduces cooling efficiency, and can create a film on the ferrite cut surface that interferes with downstream bonding processes.
Fair warning: we’ve had customers skip the flush “just this once” going from water to oil. Three out of five times it’s fine — the NdFeB surface picks up just enough oil from the first few seconds of cutting to protect itself. Two out of five times, the residual water in the lines contacts the cut surface before the oil film establishes, and you get corrosion spots. Not worth the gamble on production parts.
Does Magnet Cutting Coolant Temperature Matter?
A patent from TDK Corporation (US6386948) specifically addresses coolant temperature control for magnet cutting, recommending 20–35 °C for optimal results. In our experience, coolant temperature is a secondary factor for diamond wire cutting of magnets — less important than coolant type, flow rate, and cleanliness.
That said, there are scenarios where temperature matters:
Summer operations in non-climate-controlled shops. If ambient temperature pushes coolant above 35 °C, oil viscosity drops and the lubricating film thins. We’ve seen measurable increases in wire wear rate during summer months in shops without cooling. A simple tank chiller ($500–1000) solves this.
Cold startup in winter. Oil viscosity increases at low temperatures, reducing flow through narrow coolant nozzles. If your machine sits overnight in an unheated shop at 5 °C, the first 10 minutes of cutting may have inadequate coolant flow. Run the pump for 5 minutes before starting the cut to warm the oil through circulation.
High-volume production. Machines running 16+ hours per day with oil coolant see gradual temperature rise from cutting energy and pump heat. If unchecked, temperatures can reach 45–50 °C by end of shift. At that point, cooling efficiency is noticeably degraded. A recirculating chiller maintaining 25 ± 3 °C is standard for production environments.
For water-based coolant, temperature is less of an issue because water’s high heat capacity self-regulates better. But if you’re in a cold environment, check that your water-based concentrate doesn’t lose its corrosion inhibitor effectiveness at low temperatures — some formulations require minimum 15 °C to maintain full protection.
Coolant Filtration: More Important Than Most People Think
Magnet cutting coolant cleanliness directly affects three things: wire life, surface quality, and coolant longevity. Yet filtration is the most commonly under-specced component in magnet cutting setups.
Magnet cutting swarf is fine — typically 1–20 μm particles. These particles remain suspended in the coolant and get recirculated through the cutting zone. When they pass between the diamond wire and the workpiece, they act as loose abrasive, creating random scratches on the cut surface and accelerating diamond grit wear.
For oil-based systems (NdFeB):
NdFeB swarf is metallic and contains iron, so magnetic separation might seem logical. But remember — the blanks are unmagnetized during cutting, and the swarf is too fine for most magnetic separators to capture efficiently. We recommend a combination approach: coarse mesh pre-filter (50 μm) to catch large particles and wire fragments, followed by a cartridge filter (5–10 μm) for the fine swarf. Change the cartridge filter weekly on high-utilization machines.
One issue specific to NdFeB: the swarf reacts slowly with moisture in the oil, generating hydrogen gas. In a sealed coolant tank, this gas accumulates. We’ve had one case where a sealed tank pressurized enough to pop the lid — not dangerous, but surprising and messy. Use a vented tank or install a breather valve.
For water-based systems (ferrite/SmCo):
Ferrite swarf is ceramic and settles faster than NdFeB metallic swarf. A settling tank with baffles works well as a first stage. Follow with a paper belt filter or bag filter at 10–25 μm. Water-based systems need more frequent monitoring for biological growth — bacteria and algae thrive in warm, nutrient-rich cutting fluid. Add biocide per the fluid manufacturer’s recommendation, and check pH weekly (target 8.5–9.5 for most water-soluble cutting fluids).
Magnet Cutting Coolant Problems and Fixes
Grey or discolored NdFeB surfaces: Water contamination in oil coolant. Check for leaks in the coolant system, condensation in the tank (common in humid environments), and residual water from a previous ferrite run. Drain, flush, and refill with fresh oil.
Foaming in water-based coolant: Usually caused by oil contamination from a previous NdFeB run, or by using too high a concentration of water-soluble fluid. Reduce concentration to manufacturer’s recommended range (typically 5–8%). Add defoamer if needed.
Increasing surface roughness over time with no other parameter changes: Coolant contamination with swarf. Check filter condition and coolant particle loading. If the coolant looks darker than fresh fluid, it’s overdue for change.
Coolant smell (water-based): Bacterial growth. The fluid has gone rancid. Drain, clean the tank with disinfectant, and refill. This happens faster in warm environments and when machines sit idle over weekends with stagnant fluid. Running the pump for 15 minutes daily even on idle machines helps.
Wire life shorter than expected: Often coolant-related. Check flow rate (is coolant actually reaching the cut zone?), cleanliness (is swarf loading high?), and concentration (for water-based, is it too dilute?). We’ve seen customers blame the wire when the real issue was a clogged nozzle reducing coolant flow by 80%.
Practical Magnet Cutting Coolant Setup Recommendations
For a shop starting magnet cutting operations, here’s what we suggest:
NdFeB-only shop: Oil-based coolant system. White mineral oil, viscosity 5–15 cSt at 40 °C. 5 μm cartridge filtration. Vented tank. Change oil every 2 weeks or when particle loading visibly increases. Budget roughly $200–400/year in consumable oil for a single SG20-R machine running 8 hours/day.
Ferrite-only shop: Water-based coolant. Quality water-soluble cutting fluid at 5–8% concentration. Paper belt or bag filtration at 10–25 μm. Biocide as recommended. Monitor pH weekly. Budget roughly $100–200/year in fluid concentrate.
Mixed-material shop: Decide whether to run dedicated machines or manage changeovers. For volumes under 20 hours/week total, a single machine with changeover is usually economical. Above that, dedicate machines. Keep a written changeover checklist — the one time someone skips a step is the time NdFeB parts corrode.
Nous proposons découpe d'essai gratuite where we can demonstrate the coolant setup for your specific material — send us samples and we’ll cut them with your preferred coolant type so you can evaluate both cutting quality and post-cut surface condition.
