Cylindrical Grinding Cost: Quoting OD, ID and Centerless Work

Grinding is where a part stops being "machined" and starts being "finished" — and where a lot of shops quietly lose money. By the time a shaft or bore reaches the grinder it already carries the full cost of turning, heat treat, and handling, so a grinding miss isn't a cheap mistake. Cylindrical grinding cost is driven by completely different things than turning or milling: not how much metal you remove, but how little, how precisely, and how many times you touch the wheel to get there.

This article builds a grinding unit price the disciplined way — setup, cycle, wheel and dressing, in-process gauging, secondary handling, overhead, margin — but with the details that separate OD (outside diameter), ID (internal/bore) and centerless grinding so the number you quote survives the first article and the production run that follows.

Why grinding doesn't price like turning

On a lathe you remove millimetres of stock and the cost tracks material and cycle time across that removal. On a grinder you remove hundredths of a millimetre, and the cost tracks something else entirely: the tolerance, the surface finish, and the time the part spends not cutting — dressing the wheel, spark-out passes, gauging, and load/unload of a part that is usually hard and often heavy.

Three things make grinding its own cost category. First, the stock removed is tiny, so material is almost irrelevant and process time dominates. Second, the wheel is a consumable that wears and must be dressed, so a share of wheel cost and dressing time lands on every part. Third, grinding is almost always the last operation, which means scrap here destroys all the upstream value already in the part — so the estimate has to carry inspection and risk, not just cutting. The disciplined build-up is the same skeleton as any CNC machined part; the line items underneath are what change.

The three jobs: OD, ID and centerless

The first question on any grinding quote is which configuration the geometry forces, because each has a different cost profile.

Type	How the part is held	Best for	Cost driver
OD (cylindrical)	Between centres or in chuck	Shafts, journals, outside dia	Load/unload + pass count
ID (internal)	Chucked, small inside wheel	Bores, bushings, bearing seats	Small wheel wear, frequent dressing
Centerless	Supported on a work blade	High-volume pins, rollers, dowels	Setup-heavy, then very low per part

OD grinding between centres is the workhorse — accurate and flexible, but every part is loaded, dialled and unloaded by hand, so labour per part stays high. ID grinding fights physics: a small internal wheel can't take much load, wears fast, and needs frequent dressing, so cycle and consumable cost climb even though the surface is small. Centerless grinding flips the economics — the setup is fussy and slow, but once the wheels, regulating wheel and work blade are dialled in, parts flow through continuously at a tiny cost each, which is why it dominates high-volume pin and roller work. Matching the configuration to the batch is half the quote, the same way machine class drives a CNC turning quote.

Setup: the cost that punishes small batches

Grinding setup is heavy and it is fixed per job: mount and balance the wheel, dress it true, set the workhead and tailstock or the centerless blade and regulating wheel angle, set the wheel-to-work offset, then grind and gauge a first-off. On a centerless job, dialling the blade height and regulating-wheel tilt for size and roundness can eat an hour before a single good part exists.

Because that cost is divided by the batch, the per-part number swings violently with quantity:

Batch size	Setup cost (€150 total)	Setup per part
1	€150	€150.00
10	€150	€15.00
50	€150	€3.00
250	€150	€0.60

This is why a one-off reground shaft and a 250-piece centerless run are not the same part priced twice. Quote the centerless setup against a batch of five and you lose the job; quote it against a batch of 500 and forget it scales — both are setup-allocation errors. The same batch-division logic underpins every job; see the machine shop hourly rate calculation for building the rate that setup time is billed against.

Cycle time: passes, spark-out and gauging

Grinding cycle time is not "diameter divided by feed". It is the sum of rough passes, finish passes, spark-out, and the dead time around them. Walk the operation and add a time for each piece:

1. Rough passes — remove the bulk of the grinding stock (typically 0.1–0.4 mm on diameter left after turning) at higher infeed.
2. Finish passes — light infeed to hit size and finish without burning the part.
3. Spark-out — passes at zero infeed that let the wheel and part relax to true size; skipping it is how you miss roundness and Ra.
4. Dressing share — the wheel is dressed every N parts; that dressing time, divided across those parts, is real cycle.
5. Gauge and load/unload — in-process or post-process gauging plus handling a hard, often heavy part.

The trap is pricing only the rough cut, where the metal actually comes off, and treating spark-out and gauging as free. On a precision journal, spark-out and gauging can be the majority of the cycle. Always cost the full cycle against the fully burdened rate of the specific grinder — grinders often carry a higher hourly rate than a lathe because of the machine, the coolant filtration, and the skill the operation demands.

Wheel, dressing and coolant: the consumable line nobody itemizes

A grinding wheel is a consumable, and every part it touches owes a share of it. The wheel wears, gets dressed back to true (which throws away wheel material), and is eventually scrapped. To get the per-part consumable cost honestly:

Wheel cost per part = (wheel price ÷ parts per wheel life) + (dressing-stick share + dressing time × rate, spread per part)

Add coolant and filtration cost on precision work, where clean, well-filtered coolant is the difference between meeting Ra and burning the part. None of this is large on any single part — but on a 5,000-piece centerless run, an unaccounted wheel and dressing cost of even €0.15 per part is €750 walking out the door. List it as a line so it sits in cost, not in margin, the same discipline a complete manufacturing quote template enforces on every job.

Tolerance and finish: the whole reason the part is on the grinder

A part is ground because turning couldn't hold the tolerance or the finish — so the tightness of those callouts isn't a multiplier, it's the job. A diameter at IT6 with a fine Ra needs more finish passes, full spark-out, and gauged verification; tighten to IT4 or add a roundness and cylindricity callout and you add passes, slow the infeed, and raise the scrap risk on a part that already has all its upstream cost baked in. Reading those callouts correctly is the estimate; the GD&T-for-estimators guide covers how tolerance classes and geometric callouts translate into grinding time and scrap allowance.

This is also where grinding risk has to enter the price. Because it's the last operation, a part burned or undersized at the grinder loses everything spent before it. A precision grinding quote should carry a small, explicit scrap/rework allowance sized to the tolerance — not buried hope that the first-off holds for 250 parts.

A worked example: a hardened OD journal, batch of 50

Take a hardened 4140 shaft, one 30 mm journal to grind to IT6 with a 0.4 µm Ra, between centres, after heat treat. Quantity 50.

1. Setup — mount/balance/dress wheel, set centres, first-off gauged: €150 ÷ 50 = €3.00.
2. Cycle — rough + finish + spark-out + gauge + load/unload ≈ 3.0 min at a €85/h burdened grinder rate = €4.25.
3. Wheel + dressing + coolant — share per part ≈ €0.45.
4. Scrap/rework allowance — ~3% on a part carrying upstream value ≈ €0.55.
5. Cost per part = 3.00 + 4.25 + 0.45 + 0.55 = €8.25.
6. Price at 35% margin — using price = cost ÷ (1 − margin), 8.25 ÷ 0.65 = €12.69.

Push the batch to 250 and setup falls to €0.60, cost drops to €5.85, and the same 35% margin gives €9.00 — the quantity break a buyer expects, now defensible line by line rather than discounted by feel. The margin step uses the same arithmetic as any job; if the difference between dividing by 0.65 and multiplying by 1.35 ever trips you up, the margin-versus-markup guide settles it.

From a drawing to a grinding price

The slow, error-prone part of cylindrical grinding cost estimation is not the arithmetic — it's reading the drawing completely under deadline: which diameters carry a ground tolerance, the IT class and Ra on each, the roundness and cylindricity callouts, whether the geometry calls for OD, ID or centerless work, and the heat-treat state that sets the wheel and the scrap risk. Miss any of those and the unit price is wrong before a margin ever touches it.

QuoteBuddy reads the technical drawing and surfaces those features — ground diameters, tolerance classes, surface-finish and geometric callouts, and the operations the geometry implies — so the estimator builds the price from a complete picture instead of a hurried scan. The cost engine then assembles setup, cycle, wheel and dressing, scrap allowance, and your target margin the same way every time, on every ground part.

Start a 30-day trial and run a few real precision-part drawings through it — from upload to a complete, itemized quote PDF — and see whether the grinding price it builds matches what you'd have quoted by feel, with the margin landing exactly where you set it.