What Grinding Burn Actually Is
Grinding concentrates a lot of energy in a small contact zone. A typical cylindrical grinding contact zone is 1–10 mm2 and the heat flux into the part can exceed 10 kW/mm2. Most of that energy goes into the part as heat. If the heat input is higher than the coolant can carry away, the surface temperature of the part rises past the metallurgical transformation point — and the surface changes.
On hardened steel, the visible symptoms are temper colors: yellow at ~200°C, brown at ~240°C, purple at ~290°C, blue at ~320°C. Each color corresponds to a specific tempering temperature, and each one means the surface has been softened. The harder the original steel, the more serious the damage: a 60 HRC part tempered to 350°C surface temperature drops to ~55 HRC on the surface and has residual tensile stress that initiates cracks.
The invisible damage is worse. Burned surfaces have:
Reduced fatigue strength — fatigue life drops 50–90% on burned surfaces vs ground surfaces.
Residual tensile stress — the surface is in tension, not compression, and cracks propagate from it.
Microstructural change — on through-hardened steel, the martensite tempers; on case-hardened steel, the retained austenite can transform.
Hardness drop — the surface is softer than the bulk.
For bearing rings, gears, and shafts under cyclic load, burned surfaces fail in service even though the part measured correctly at final inspection. Burn is the most expensive defect in cylindrical grinding because it is invisible until the part fails.
Three Mechanisms of Thermal Damage
The contact zone heats the part in three overlapping ways:
Friction at the grit–workpiece interface. Each grit sliding against the part generates heat. The total is the product of the sliding distance (high), the normal force (low), and the coefficient of friction (moderate).
Plastic deformation of the chip. Most of the energy goes into the chip, not into friction. The chip is plastically deformed and heated; some of that heat conducts back into the part.
Re-grinding of adhered chips on the wheel face. Chips that weld to the wheel face (called "loading") create a rubbing interface that generates concentrated heat.
All three are temperature-limited. If the heat can be carried away by the chip, the coolant, and conduction into the bulk of the part, the surface stays below the metallurgical limit. If the heat cannot escape, the surface temperature spikes and burn happens.
How to Recognize Burn
Five detection methods, in increasing order of difficulty and accuracy:
Visual temper colors. Look at the ground surface. Yellow/brown/purple/blue coloration means burn. Free, instant, but only catches severe burn.
Acid etch / nital etch. A 2–5% nitric acid etch reveals the heat-affected zone as a dark band on the surface. Common in production shops. Catches moderate burn.
Surface roughness anomaly. Burned surfaces tend to be rougher than non-burned surfaces on the same part. Use a surface roughness comparator or a portable profilometer.
Microhardness traverse. Cut a cross-section, polish, and measure Knoop or Vickers hardness from the surface into the bulk. Burn shows up as a soft layer at the surface. Most accurate destructive test.
Residual stress measurement (X-ray diffraction). The lab standard for assessing burn. Burned surfaces have tensile residual stress; non-burned surfaces have compressive residual stress (which is actually desirable). Expensive and slow.
For shop floor, the practical approach is visual + acid etch + microhardness on first article and after process changes. X-ray is reserved for certification, aerospace, and high-value parts.
The Causes of Burn, in Order of Frequency
Across a thousand production lines, the same five causes account for most burn incidents. Knowing the order helps you diagnose:
Wheel too hard for the material. A hard wheel holds its dull grains. The grains rub instead of cutting. Heat builds. Burn follows. This is the single most common cause of burn on hardened steel.
Insufficient coolant flow at the contact zone. Coolant carries away most of the heat. A clogged nozzle, a misaligned nozzle, or a low-flow pump can drop coolant delivery by 50% with no obvious symptoms.
Infeed too aggressive. A 0.01 mm infeed per pass generates roughly twice as much heat per unit time as a 0.005 mm infeed. The contact zone has more energy to dissipate. On a thermally limited process, this shows up as burn on the second or third pass.
Wheel glazed or loaded. A wheel that has not been dressed often enough wears smooth on the face. The contact area grows, the sliding distance per grit grows, and the heat per grit grows. Burn follows.
Spark-out too short. Spark-out is the pass with no infeed that lets the wheel run out the last of the cut. A short spark-out leaves residual heat in the surface that can re-temper the part.
The Cures, in Order of Effort
The right fix depends on the cause. Apply the cheapest fix first:
Re-check coolant flow. Open the nozzle, confirm the pump is delivering rated flow, and check the filter. If the flow at the contact zone is below 15 L/min for a small
grinder or below 40 L/min for a large one, you have a coolant problem. Fix it before changing anything else.
Switch to a softer wheel grade. Move one letter down (J to I, K to J, etc.). The wheel will release dull grains faster, generate less heat per grit, and reduce burn. Cost: minimal. Side effect: slightly faster wheel wear.
Switch to a coarser grit. Move from 100 to 80, or from 80 to 60. Each grit cuts a thicker chip, the chip carries away more heat, and the surface temperature drops. Cost: minimal. Side effect: rougher surface finish.
Reduce infeed. Cut infeed by 30–50% per pass. Take more passes to reach size. Cost: longer cycle time. Side effect: more passes, but each pass is cooler.
Dress more often. Add a dress cycle between parts or every 3–5 parts. The wheel face stays sharp, the contact area stays small, and the heat stays manageable. Cost: a small cycle-time hit. Side effect: more wheel wear (but better surface finish).
Switch to CBN. On hardened steel above 50 HRC, CBN generates 30–50% less heat per unit stock removed than aluminum oxide. The cycle time drops, the burn risk drops to near zero, and the wheel lasts 50× longer. Cost: 10× the wheel price. Pays back inside 12 months on high-volume hardened steel.
Add a creep-feed or high-efficiency grinding (HEDG) setup. Slow traverse with high depth of cut, using a coarser grit and high-pressure coolant. Different machine kinematics, but very low burn risk and short cycle times. Best for high-volume production of bearing rings and similar.
Use a high-pressure coolant system. 70+ bar coolant directed at the contact zone, instead of 5–10 bar from a standard nozzle. Reduces burn by 30–50%. Cost: pump + plumbing, $5K–$15K per machine.
Burn vs Other Surface Defects: Quick Diagnostic
Burn is one of several surface defects. Knowing which one you have speeds up the fix:
|
Defect |
Looks like |
Caused by |
Fix |
|
Grinding burn |
Yellow/blue/purple temper colors; soft surface layer; residual tensile stress |
Heat at contact zone; wheel too hard; insufficient coolant; aggressive infeed |
Soften grade; coarser grit; more coolant; lighter infeed; consider CBN |
|
Chatter marks |
Regular wave pattern on the surface; vibration noise during cutting |
Regenerative chatter; wheel too hard; machine rigidity insufficient |
Soften grade; coarser grit; reduce wheel speed; stiffen setup |
|
Wheel loading |
Glazed wheel face; shiny part surface; reduced cut rate |
Material softer than wheel; insufficient coolant; wrong abrasive for material |
Dress more often; switch to softer grade; coarser grit; check abrasive type |
|
Surface roughness too high |
Rough finish; passes visual but fails profilometer |
Grit too coarse; dress too rough; feed too fast |
Finer grit; finer dress; reduce feed rate |
|
Microcracking |
Fine cracks visible under microscope; appears after acid etch |
Excessive heat; aggressive infeed; wrong wheel for material |
Reduce heat input; softer grade; lighter infeed; consider CBN |
Coolant: The First Line of Defense
Most burn problems are really coolant problems. Three things to check first:
Flow rate at the nozzle. Measure with a flow meter. Typical target: 15–25 L/min for small cylindrical grinders, 30–60 L/min for larger machines. Below that, the coolant cannot carry the heat away.
Nozzle position and alignment. The nozzle must point at the contact zone, with the coolant stream hitting the wheel–workpiece interface. Misalignment by 5 mm can drop heat removal by 30%.
Coolant cleanliness. Dirty coolant (swarf above 50 μm) blocks the nozzle and reduces flow. Magnetic separator + 25 μm bag filter is the standard setup.
For severe burn problems, upgrade to high-pressure coolant (70–100 bar). This is the single biggest step-change available without changing the abrasive.
Choosing the Right Wheel for Low-Burn Operation
Two paths to a low-burn operation:
Aluminum oxide with optimized parameters. Softer grade (I–K for hardened steel), coarser grit (60–80), more frequent dressing, and ample coolant. Cycle time is the trade-off — you take lighter cuts to control heat.
CBN. Harder than aluminum oxide, runs cooler, holds profile longer, and dresses less often. Cost per part is lower on high-volume hardened steel. Capital cost is the trade-off — CBN requires a higher-rigidity machine and a higher spindle power.
The break-even between the two paths is around 5,000–10,000 parts per year of hardened steel. Below that, aluminum oxide with disciplined parameters wins on simplicity and capital cost. Above that, CBN wins on cycle time, surface finish, and per-part cost.
CBN for Burn-Prone Applications
CBN wheels cost 10× as much as aluminum oxide wheels but last 50–100× longer on hardened steel. The math on burn reduction:
Aluminum oxide on hardened steel: burn rate 2–5% of parts in production.
CBN on hardened steel: burn rate below 0.1% of parts in production.
On a $50 part with 2% burn rate, 100,000 parts/year burns 2,000 parts × $50 = $100K of scrap. CBN eliminates 95%+ of that. The CBN premium ($100K upfront over an aluminum wheel run) pays back inside the first production year.
CBN is not free, though. It requires:
Higher spindle power (typically 11+ kW for a 400 mm CBN wheel).
Higher rigidity machine (CBN does not forgive deflection the way aluminum oxide does).
Higher-pressure coolant delivery.
Higher operator skill (CBN dressing is different from aluminum oxide dressing).
On the right machine with the right setup, CBN is a step-change. On the wrong machine, CBN will burn parts faster than aluminum oxide because the operator does not know how to dress it.
Spark-Out: The Overlooked Burn Prevention
Spark-out is the last pass with no infeed, allowing the wheel to run out the residual material and the part to cool under the wheel. Most operators skip it because it looks unproductive.
What spark-out does:
Removes the last 0.001–0.005 mm of stock that the previous pass could not reach.
Burnishes the surface to a smoother finish.
Gives the part time to dissipate the heat from the previous pass before the next one (on multi-pass cycles).
A typical spark-out time is 2–10 seconds, depending on the part diameter and the previous pass depth. Skipping spark-out is one of the most common causes of burn at the end of a traverse cut, where the heat input is concentrated at the end of the cut.
Burn Detection on the Production Floor
For production environments, three practical methods work without lab equipment:
Visual inspection under bright light. Temper colors are visible to the naked eye on most ground surfaces. Train operators to recognize the colors. Stop the line if burn appears.
Nital etch test. Etch the ground surface with 3% nitric acid for 5–10 seconds, rinse, and dry. Burned surfaces show a dark gray to black band; non-burned surfaces show uniform light gray. Standard test in most bearing plants.
Portable microhardness tester. Modern Knoop/Vickers testers are handheld and quick. Measure Knoop hardness at the surface; a reading 10%+ below bulk hardness means burn.
Run one of these on the first part of every shift, on every new setup, and on the first part after any process change. If burn is detected, stop the run, fix the cause, and re-cut from a fresh stock.
Burn in Cylindrical vs Centerless vs Internal Grinding
The cause of burn is the same in all three processes (heat at the contact zone), but the relative risk differs:
Cylindrical: moderate burn risk. The contact zone is open; coolant reaches it easily. The main burn cause is the wheel-grinding parameters.
Centerless: moderate burn risk. The contact zone is small but the parts are small. Coolant reaches the contact zone easily.
Internal: high burn risk. The contact zone is inside a bore, with limited coolant access. The wheel is small and runs hot. Internal grinding burn is the most common defect on small bores.
For internal grinding, high-pressure coolant through the wheel center is often necessary to control burn. Standard flood coolant is usually insufficient.
Documentation and Process Control
Burn is the kind of problem that one operator fixes in the morning and the next operator reintroduces in the afternoon. The cure is process control, not operator heroics.
Document the following for every part that has a burn risk:
Wheel specification (full marking: A60L5V or equivalent)
Wheel speed (m/s)
Workpiece speed (RPM)
Infeed per pass (mm) and number of passes
Spark-out time (s)
Coolant flow rate (L/min) and pressure (bar)
Dress depth (mm) and frequency (parts per dress)
Wheel life (parts per wheel)
Pin the document next to the machine. When burn appears, you have a baseline to compare to. This is also the documentation a customer auditor will ask for when you ship parts to aerospace or medical customers.
FAQ
Q: What does grinding burn look like?
A: Temper colors — yellow, brown, purple, or blue discoloration on the ground surface. Each color corresponds to a specific surface temperature during grinding. Yellow at ~200°C, blue at ~320°C. Even mild burn (yellow) means the surface has been softened and residual tensile stress has been introduced.
Q: Is burn visible to the naked eye?
A: Severe burn (purple/blue) is visible to the naked eye on most surfaces. Mild burn (yellow/light brown) is often invisible until the part is acid-etched or measured for microhardness. A 10x magnifier helps but is not enough for definitive diagnosis.
Q: What is the difference between burn and overheating?
A: Burn is the metallurgical damage from overheating — the surface has been tempered or phase-changed. Overheating is the temperature rise itself, which may or may not have caused damage. A part can overheat briefly without burn; a part that burns has been overheated past the metallurgical limit.
Q: How do I prevent burn when grinding hardened steel?
A: Use a softer grade wheel (I–K), coarser grit (60–80), ample coolant flow (15+ L/min for small machines, 40+ L/min for larger), moderate infeed (0.005–0.010 mm per pass), and dress the wheel frequently. For high-volume production, switch to CBN.
Q: Can burned parts be re-ground to remove the damage?
A: Yes, if the damage is shallow (below 0.02 mm). Remove the burned layer with a light roughing pass, then re-finish. If the burn is deep, the part is scrap. Acid etch first to determine the burn depth.
Q: Does burn always reduce fatigue life?
A: Yes, sometimes by 50–90%. The exact reduction depends on the depth of the burn, the material, and the loading condition. Aerospace and bearing specifications require no burn precisely because of this fatigue reduction.
Q: What is high-efficiency deep grinding (HEDG)?
A: HEDG is a grinding mode using slow traverse with high depth of cut, coarse grit, and high-pressure coolant. It produces low heat input per unit volume, low burn risk, and short cycle times. HEDG requires a specially configured cylindrical grinder with high spindle power and high-pressure coolant.
Q: Is CBN always better than aluminum oxide for hardened steel?
A: For high-volume hardened steel above 50 HRC, yes. CBN produces less burn, holds tolerance better, and is cheaper per part above ~5,000–10,000 parts per year. For one-off or low-volume hardened steel, the wheel cost does not amortize and aluminum oxide with disciplined parameters is the right choice.
Bottom Line
Grinding burn is metallurgical damage from heat at the contact zone. The visible temper colors are the symptom; the invisible damage is residual tensile stress and reduced fatigue life. Burn is preventable but common, and it is the most common cause of in-service failure on ground parts.
Three levers solve most burn problems, in order of effort:
Check the coolant (flow, nozzle alignment, cleanliness).
Switch to a softer wheel grade.
Switch to CBN if production volume justifies the wheel cost.
For deeper reading on related grinding problems, see Grinding Wheel Selection and Chatter in Cylindrical Grinding. For machine selection, see the full Yelin CNC cylindrical grinder line or contact our engineering team.
English
