Edmund Boland ANCA CNC: The Cutting Tool Plan Explained
Edmund Boland, director general of ANCA CNC, has revealed a plan for the cutting tool industry that covers grinding automation, tool geometry data and shop-floor connectivity. This page explains the mechanism behind each shift, where it applies, and where it does not. Read it if you specify, grind or buy cutting tools.
What the Edmund Boland ANCA CNC plan changes on the floor
A plan for the cutting tool industry is not a press release about the future. At the machine level it is a list of decisions: which grinding spindle runs which wheel, how the tool geometry is measured, and where compensation for wheel wear gets applied. Boland's message lands on the same three levers any tool room already pulls.
The first lever is the grinding cycle itself. A fluting pass on a 6 mm end mill removes a fixed volume of carbide, and the wheel edge wears by a predictable amount per pass. If that wear is corrected only at the start of a batch, the last tools in the run drift out of tolerance. Correcting it every 8 to 12 tools keeps flute rake angle inside ±0.5° across the batch.
The second lever is measurement. Tool geometry is only as good as the instrument that verifies it. Optical presetters resolve edge radius to about 2 μm, while a touch probe on the grinder resolves position to about 1 μm but says nothing about edge sharpness. The two answer different questions and neither replaces the other.
The third lever is data flow. A grinder that logs wheel dressing intervals, spindle load and coolant temperature gives the tool room a maintenance signal before scrap appears. That is the practical core of the plan: connect the measurements that already exist instead of adding sensors that nobody reads.
- 1Wheel wear compensationApply every 8–12 tools, not once per batch.
- 2Two measurement systemsOptical for edge radius, probe for position.
- 3Logged spindle loadEarly signal for dressing and coolant issues.
Why grinding parameters set the limit on tool life
Carbide removal in fluting is a brittle fracture process, not a cutting one. Each grit particle loaded with a few newtons of force either micro-fractures the carbide or rubs and generates heat. The difference between those outcomes is depth of cut per grit, which the feed rate and wheel specification control together.
Push feed too high and the tool edge develops micro-chipping that only shows up after the tool has run 200 parts. Push it too low and the wheel glazes, spindle load climbs, and the wheel needs dressing sooner than the cycle plan allows. The usable window is narrower than most published tables suggest.
Coolant matters more here than in milling. Straight oil at 20 to 30 L/min per grinding zone carries heat away and keeps the wheel pores clear. Water-soluble coolant at the same flow rate can work, but the concentration has to be checked weekly; a drop from 8% to 5% changes the lubrication film and the edge quality follows.
For tools below Ø3 mm, the mechanics change again. The wheel cannot be dressed coarse enough to keep grit load low without weakening the tool edge. That is why small-diameter tool grinding leans on finer wheels, lower feed per pass, and more frequent in-process checks.
- 1Feed too highMicro-chipping appears after a few hundred parts.
- 2Feed too lowWheel glazing, rising spindle load, early dressing.
- 3Coolant checkVerify concentration weekly; 8% to 5% changes the film.
Tool geometry data is the part engineers underuse
Most tool rooms store geometry as a drawing and a nominal tolerance. The useful version stores the measured value per tool: actual rake angle, actual relief angle, actual edge radius. Once that data exists, a batch that cuts 15% fewer parts can be traced to a specific geometry deviation instead of blamed on the workpiece material.
The mechanism is straightforward. Cutting forces scale with rake angle and edge radius. A 5 μm increase in edge radius raises the ploughing component of the force, which raises the temperature at the cutting edge, which accelerates flank wear. You cannot see that chain from a pass/fail inspection.
This is where the plan's emphasis on connectivity earns its keep. A grinder that exports measured geometry per serial number lets the tool room build a wear curve instead of a scrap count. Over a few months the curve predicts when a wheel or a dressing cycle needs attention.
The limit is data hygiene. If the presetter is calibrated once a quarter and the grinder probe once a year, the exported numbers drift apart and the curve is noise. Calibration frequency has to match the tolerance you are trying to control, not the maintenance calendar.
- 1Store measured valuesPer tool, not just nominal plus tolerance.
- 2Force chainEdge radius to ploughing force to edge temperature.
- 3Calibration frequencyMatch it to the tolerance you control.
Where automation pays off and where it does not
Automated loading on a tool grinder pays back when the batch is repeatable and the cycle is long. A 12 mm end mill with a 4-minute grinding cycle and a 200-piece batch is a good candidate: the operator is free for most of the shift and the cycle stays consistent. A mixed order of thirty different geometries is not.
The reason is setup time, not machine time. Every geometry change needs a new wheel form, a new program, and a first-article check. Automation does not shorten those steps, and a robot arm waiting through a 25-minute setup is capital sitting idle.
The same test applies to in-process measurement. Probing every tool makes sense when the tolerance is tight and the wheel wears fast, for example a Ø2 mm tool at ±0.005 mm on diameter. Probing every tool on a Ø16 mm roughing cutter at ±0.05 mm wastes cycle time for data that will never trigger a correction.
A practical rule: automate the operations that repeat more than fifty times per setup, and leave the rest manual. The plan's direction is right, but the payback threshold is a batch-size question, not a technology question.
- 1Good candidateLong cycle, 200-piece batch, one geometry.
- 2Poor candidateThirty geometries, 25-minute setups.
- 3Probing thresholdTight tolerance plus fast wheel wear only.
Matching the plan's shifts to your batch profile
Use the batch profile on the left to decide which shift applies first.
| Shop profile | First shift to apply | Why | Skip if |
|---|---|---|---|
| Long runs, one geometry | Automated loading | Setup cost amortized over 200+ parts | Batch under 50 parts |
| Tight diameter tolerance | In-process probing | Wheel wear drifts within a batch | Tolerance wider than ±0.05 mm |
| Mixed small orders | Geometry data logging | Traces variation without new hardware | No presetter on site |
| Fine edge requirement | Coolant and wheel control | Edge radius drives tool life | Roughing tools only |
| High scrap rate | Calibration audit | Drifting instruments mimic process faults | Instruments calibrated monthly |
The verdict on the plan
If your batch repeats more than fifty times per setup, invest in automated loading and probing first. If it does not, spend the same money on geometry data logging and calibration instead, because that is where the scrap actually comes from.
Questions engineers ask about the plan
Does the plan apply to shops that only regrind tools?
Yes, and the priorities differ. A regrinding shop already knows the incoming geometry, so the useful shift is measuring what comes back after grinding rather than what arrives.
Log the removed material volume per tool. That number predicts wheel life better than a pass count, because two tools with the same diameter can need very different stock removal.
How much does in-process probing actually improve tolerance?
It removes the drift between the first and last tool in a batch, which is usually the dominant error source. On a Ø2 mm tool at ±0.005 mm, that drift can be most of the tolerance band.
It does not improve the machine's positioning accuracy. If the grinder repeats to ±2 μm, probing will not make it repeat to ±0.5 μm.
Is straight oil always better than water-soluble coolant for tool grinding?
For fine edges on small tools, usually yes. Oil carries heat away faster and keeps wheel pores clear without the concentration drift that water-soluble fluids show.
For larger roughing geometries with generous coolant access, water-soluble fluid at stable concentration is often enough, and it avoids the filtration and disposal cost of oil.
What is the smallest batch where automated loading pays back?
There is no fixed number, but the break-even sits where setup time per batch is small compared with total grinding time. In practice that is often around 50 to 100 parts for a simple geometry.
Below that, the robot spends more time waiting through setup than it spends loading. Measure your own setup time before buying.
How often should presetters and grinder probes be calibrated?
Match the interval to the tolerance you control. If you hold ±0.005 mm on diameter, a quarterly presetter calibration is too slow; monthly is a reasonable starting point.
Cross-check the two instruments against a known master at the same time. If they disagree by more than a third of your tolerance band, one of them has drifted.
Need machined parts to the tolerances this plan describes?
We machine prototypes and production runs in aluminium, stainless, titanium and engineering plastics, with 100% inspection before shipment and reports on request.
12-hour quote100% inspectionNo minimum order
More machining notes
We publish setup notes, tooling trials and inspection data from the factory floor.
