Summary

Drawing lines, die marks and galling are routinely described with the same word, but only one of them gets worse on its own. Drawing lines come from material flow and grain, die marks are the tool surface copied onto the part, and galling is adhesion wear that transfers metal, accelerates, and scratches every following part. Die material, coating and lubricant decide which of the three you get, and the finish requirement on the drawing decides whether it matters.

Three defects that look alike, and only one that gets worse

A rejected deep drawn housing usually arrives described in one word: marks. The word says little, because three unrelated mechanisms produce similar-looking surfaces. Drawing lines come from material flow and grain structure. Die marks are the tool surface copied onto the part. Galling is metal welded and torn from one surface onto the other.

The distinction decides who owns the problem. A drawing line present since first article is an early material or tooling decision. A die mark appearing after a long run is a maintenance event. Galling that starts mid-run and worsens is a process failure that will not settle by itself. Sorting the three by appearance, cause and behaviour is the first useful step in any surface complaint; a wider catalogue is collected under metal working defects.

Observation	Drawing line / orange peel	Die mark	Galling / pick-up
Where it sits	Follows material flow; shifts with coil and blank position	Fixed relative to the tool, the station and the stroke	Starts at a high-pressure zone: draw radius, ironing land, die entry
Under low-angle light	Faint banding, dull matt patch, pebbled orange-peel texture	Parallel lines matching the polish direction on the tool	Raised smears and torn metal, then a bright lump followed by a deep score
Behaviour over the run	Stable; present on part one and part one hundred thousand	Stable until the tool is polished, then it moves or disappears	Accelerates; each stroke adds material to the die and deepens the score
Root cause	Coarse grain, yield-point elongation, work-hardening, flow pattern	Tool polish direction and grit, a soft spot, a burr on the tool	Adhesion wear: cold welding at the interface under pressure and heat
First response	Check grain size, temper condition and draw ratio	Inspect and re-polish the station that owns the mark	Stop, remove the built-up material, then change coating or lubricant

Drawing lines: material flow and grain, usually cosmetic

A drawing line is a track left by the material rather than by the tool. The blank carries a grain structure inherited from rolling and annealing, and during the draw each grain deforms a little differently. When the grain is coarse, usually because an annealing cycle ran too hot or too long, that uneven deformation shows as orange peel: a pebbled matt texture on a wall that was smooth before forming. Polishing the die will not remove it, because the roughness is in the sheet.

A second mechanism is yield-point elongation. Low-carbon steels such as SPCC, DC01 and DC04 can develop discrete bands when stretched after the coil has aged, showing as diagonal stretcher strains on the first draw. Temper rolling suppresses the effect, and a coil stored long enough for it to return will band the first parts drawn from it. Both mechanisms are repeatable rather than progressive, and both follow the flow pattern the tool imposes, which is why a drawing line tracks the material and not the station. Grain behaviour is related in the materials guide.

Drawing lines run with the material flow, so they follow the wall and the draw radius instead of crossing a sealing band.
Orange peel is a grain-size problem; no die polish will remove it because the roughness sits in the sheet.
Stretcher strains are a yield-point effect in low-carbon steel and often appear after a coil has aged in storage.
A drawing line is normally cosmetic: it is repeatable, it does not deepen, and it deposits nothing on the tool.
It stops being cosmetic where it crosses a seal band, coincides with a wall-thinning limit, or is deep enough to raise a stress concentration.

Die marks: the tool surface, printed onto every part

A die mark is the tool's own surface copied onto the part. Every punch and die carries a texture: the direction and grit of the final polish, the waviness left by a dressing stone, a soft spot, a small burr. That texture is embossed into the workpiece under draw pressure, and because it belongs to the tool, the mark sits at a fixed position. That is the practical test: if a mark stays in the same place while the coil, the lubricant and the material lot all change, it comes from the tool.

Polish direction matters as much as grit, because a texture running across the flow direction scrapes while the same roughness running along it leaves the surface alone. A mark on a high-pressure land is also where the lubricant film breaks first, which is how an appearance complaint becomes a galling complaint. On a fresh die, toolmaker's polish marks fade as the tool beds in.

Progressive die tooling with polished punches and inserts for drawn parts — The polish direction and grit on a punch or die insert is printed onto the part. A texture running across the flow direction scrapes; the same roughness running along the flow is far more forgiving.

Galling: the defect that feeds itself

Galling is adhesion wear, and it changes the process rather than only the appearance. At sufficient local pressure and temperature the oxide films on the two surfaces break down, clean metal meets clean metal, and microscopic junctions form across the interface. When the surfaces slide apart, those junctions fracture on the weaker side, leaving workpiece material on the die or die material dragged along the part.

That is why galling belongs in a different class of problem. Transferred workpiece material is work-hardened at the point of transfer and can be harder than the die it sits on, so the lump behaves like a cutting tool: it scores the next part, then the one after that, and each pass raises friction, heat and pressure at the same spot. There is no stable state. A drawing line measuring 5 µm on the first part still measures 5 µm on the ten-thousandth; a pick-up can go from a smear to a torn wall in one shift.

A dull matt patch at a high-pressure zone, where the lubricant film has broken down.
Fine bright smears of workpiece material on the die, visible on the tool rather than on the part.
A raised lump on the die and a matching score on the part, running along the direction of flow.
Transfer on both surfaces, local heating and discolouration, and a clearance that has effectively closed at that point.
Deep scoring or a torn wall, which in stainless can turn a surface complaint into a wall-thinning or cracking failure.

Die material and coating: what each step actually buys

The die surface is the more reliable lever against galling, because it changes the interface itself rather than the film on top of it. Die treatments buy three different things, and no single treatment buys all three: hardness, which resists abrasive wear; chemical inertness and low surface energy, which resist cold welding; and topography, which decides whether lubricant is retained at the contact or squeezed out of it.

Die surface	Mechanism at the interface	Pick-up resistance	Effect on interval between polishes	Where it fits
Uncoated polished cold-work steel (D2-type)	Hard matrix, but steel-on-steel contact once the lubricant film breaks	Low, and poor on stainless or aluminium	Baseline; thousands of strokes on mild steel, far fewer on stainless	Mild steel, low volume, simple geometry
Through-hardened steel, 58–62 HRC	Higher hardness delays abrasive wear, but does not stop adhesion	Only slightly better than uncoated	Modest gain, mainly against abrasion	Abrasive blanks, coated steel
Nitrided surface layer	Hard compound layer over a diffusion zone; good against abrasion and erosion	Moderate; the brittle compound layer is removed before final polishing	Better than uncoated, but re-polishing removes the benefit locally	High-cycle tools that are polished infrequently
Hard chrome plating, typically 5–25 µm	Low-friction, chemically less reactive surface that resists cold welding	Good; the standard upgrade for stainless and aluminium	Several times the uncoated interval, and it can be stripped and re-plated	Draw radii, ironing lands, other high-pressure zones
PVD TiN, 2–4 µm	Hard ceramic layer applied at a temperature that does not soften the tool	Good on mild and coated steel	Longer than hard chrome on abrasive work	Progressive dies, sharp edges, coated blanks
PVD TiCN	Harder than TiN and more wear-resistant in abrasive contact	Good, with better wear life	Longest of the PVD family on steel	High-volume steel and stainless
DLC-type carbon coating	Very low friction against aluminium; limits adhesion and smearing	Best available against aluminium transfer	Large gain on aluminium, less on stainless at high interface temperature	Aluminium at moderate interface temperature
CVD coating	Thicker, denser layer with good coverage on complex geometry; deposited hot	Good	Long, where the tool grade and heat treatment suit the process	Complex inserts, high-volume work

Two cautions follow. A coating applied to a draw radius is often polished away during tryout, leaving the zone that needed it most uncoated. And treating the whole tool is rarely necessary: marks form where pressure, sliding distance and temperature coincide, which in a drawn part means the draw radius, the ironing land and the die entry rather than the flat faces. Treating two or three zones is often the most economical route, and keeping them consistent is a tooling design question as much as a surface one — see stamping die design.

Lubricant: chosen for formability, or for finish

Lubricant and die surface do the same job by different means, and are often chosen by different people for different reasons. Formability drives the usual choice: whether the draw tears, whether the wall thins past its limit. Finish pushes the opposite way, because what keeps a surface clean is a film that survives the contact pressure and holds the two metals apart. A high-film-strength lubricant leaves fewer die marks and less pick-up, and more residue; a thin one cleans easily but breaks down early and lets the die texture print onto the part.

The chlorinated question sits on top of that. Chlorinated additives are the traditional answer to severe stainless drawing: at the interface temperatures a stainless draw generates, the additive reacts with freshly exposed metal to form a boundary film that carries load where a plain oil film cannot. The cost appears downstream. That residue resists a simple alkaline wash, can be carried into an annealing furnace where it leaves carbon, and if trapped in a drawn recess can later appear as corrosion under a plated finish. Non-chlorinated lubricants clean more easily but may not carry the load, so the trade is between forming risk and cleaning risk. Pre-treatment is described under surface treatment.

Match the lubricant to the material and the reduction actually used, not to a general preference.
Test cleaning on a crevice sample rather than a flat panel, because that is where residue survives.
Treat a lubricant change as a process change: it can move the drawing-line pattern, the visibility of die marks and the polishing interval at the same time.
Do not use lubricant to hide galling that a die surface treatment should be solving; a film that holds at low stroke rate may not hold at full rate.

Why stainless and aluminium pick up more than mild steel

Mild steel is forgiving for two reasons. Its flow stress rises gently with strain, so the pressure at the die radius stays moderate, and its thermal conductivity, around 50 W/(m·K), lets friction heat disperse into the sheet and the tool instead of concentrating at the contact.

Austenitic stainless removes both advantages. SUS304 and 316L work-harden steeply, so the pressure the material exerts on the die radius climbs as the wall is drawn, and the surface layer itself partly transforms to martensite as it passes. Thermal conductivity is roughly a third that of mild steel, around 16 W/(m·K), so friction heat stays at the interface and raises the temperature exactly where the boundary film is trying to survive. A freshly exposed, hardened stainless surface at that temperature is close to the worst case for cold welding to a steel tool.

Aluminium fails differently. The bulk metal is soft, but it carries a hard oxide whose fragments act as an abrasive third body between die and part, scoring both. Where the oxide breaks, fresh aluminium smears onto the tool, and because the smear is soft it fills the small pits in the die surface, so the die's own topography is then transferred to every following part.

Material	What it does at the interface	Pick-up tendency	Practical consequence
Mild steel: SPCC, DC01, DC04	Moderate flow stress; conductivity near 50 W/(m·K) disperses friction heat	Low to moderate	Uncoated tool steel is often adequate; drawing lines are the more likely complaint
Austenitic stainless: SUS304, 316L	Steep work-hardening; conductivity near 16 W/(m·K) traps heat; the passive layer renews	High	Plated or coated die surfaces and a boundary-additive lubricant are usually needed
Aluminium: 5052, 5754	Soft matrix under a hard abrasive oxide; fresh metal smears readily	High, and different in character	Coatings that limit smearing, and die pits that fill with aluminium to watch
Copper and brass	Soft, high friction, and transfer with no oxide to slow it down	Moderate to high	Coatings help; blank cleanliness matters as much
Tin-plated steel, SPTE	The soft tin layer transfers before the steel does	Moderate	Build-up is soft and can be wiped off, but it changes clearance
Zinc-coated steel	Zinc is soft and the build-up can bond to the coating	Moderate to high	Watch the die entry and the ironing land; the build-up is not always visible on the part

The reading is that a die surface which holds on mild steel will not automatically hold on stainless, and a coating that is excellent against aluminium transfer is not necessarily right against stainless at high interface temperature. Die surface, lubricant and material have to be selected as one set. The more difficult grades are covered in the stainless steel deep drawn parts guide.

When the mark matters, and when it does not

Surface finish becomes a functional requirement only when the surface has a job to do. A mark matters on a sealing face, because a line crossing the contact band is a leak path. It matters on a sliding or bearing surface, because raised metal and roughness change friction and wear. It matters when the surface is visible after assembly, an appearance requirement but still a requirement. And it matters when the part is subsequently plated, because a deposit is conformal: it follows the shape it sits on, including a scratch, rather than filling it.

Where the surface is	Does the mark matter?	Why	What to specify
Sealing face: ball seat, O-ring land, gasket face	Yes, critical	A line across the contact band is a leak path from inside to outside	No tool marks at 10×; maximum Ra on the band; leak test; boundary sample
Sliding or bearing surface: spool bore, plunger bore	Yes	Roughness and raised metal raise friction, wear and stick-slip	Maximum Ra, roundness, no raised metal; marks along the sliding direction are the worst
Visible after assembly: trim, bezel, outer shell	Often	Nothing functions differently, but it is the surface the end user sees	Boundary sample and a stated lighting condition rather than a number alone
Subsequently plated or painted	Yes, unless the coating is thick	Plating is conformal and copies a mark instead of filling it, so a shallow score stays visible under a thin deposit	Finish controlled before plating; mechanical finishing or a thicker deposit where the mark must vanish
Internal, non-functional, not visible	Usually not	Nothing touches it, nothing seals against it and nobody sees it	No requirement, or a coarse Ra limit — but say so explicitly
Surface that is formed again later	Sometimes	A sharp-bottomed score acts as a stress raiser and can start a crack at a later necking or flanging step	No sharp-bottomed scores at 10×; depth and radius limits rather than Ra

The 10× criterion is a written acceptance rule, not a judgement, and a usable version names four things: magnification, illumination, viewing direction and the exact area. Ten times magnification under diffuse illumination of roughly 500 lux is a reasonable starting point, with a low-angle beam added to catch raised metal. The view should be normal to the surface, and the inspected area should be the sealing band rather than the whole part. Where two people can reach different verdicts on the same part, the remedy is a boundary sample. How surface and tolerance interact is covered under ironing tolerances and surface finish.

A drawing note that survives production: Sealing land: closed region, airtightness required. No drawing lines, die marks or raised metal visible at 10× magnification under diffuse illumination of about 500 lux, with a low-angle beam check for raised metal, viewed normal to the land. Maximum Ra 0.4 µm measured across the lay with a 0.8 mm cutoff. Boundary samples approved at first article and retained. 100 percent leak test after assembly.

Writing a finish requirement that can be measured

Ra measured on a formed surface is not the same quantity as Ra measured on a machined surface. A turned surface has a periodic lay set by the feed, so one number describes it well. A drawn wall carries the die surface plus the grain-scale deformation of the sheet, and it differs between the bottom, the radius and the wall of the same part. Readings along the lay and across it can differ by a factor of two or more, and changing the cutoff changes the reported value on an unchanged surface. A visual standard is therefore often the more useful half of the specification, because a single deep score in an otherwise smooth wall can sit inside any Ra limit and still fail the function. Note practice is set out in the engineering drawing guide.

Name the surface and the area; a finish value with no area attached is not a specification.
State the direction of measurement relative to the lay, the cutoff and the instrument class.
Approve a boundary sample of one passing and one failing part at first article and keep it with the drawing.
Where finish is critical, make the die polish condition a controlled process step: on an as-drawn surface the finish of the part is largely the finish of the die.

Deep drawn parts with formed walls, radii and bottoms inspected for surface finish — The bottom, the radius and the wall of the same drawn part are three different surfaces. A single Ra value taken on one of them describes the measurement as much as the part.

Maintenance interval and the trade you are actually making

The cost of a surface requirement rarely appears on the part. It appears in the tool room, in press downtime, in the handling risk each time a coated tool is removed for polishing, and in the capacity lost while that happens. The number that matters is the interval between polishes, and it moves with each of the following:

Material: stainless and aluminium are the least forgiving, mild steel the most.
Die surface: plating or coating the high-pressure zones commonly extends the interval several times over on stainless.
Lubricant: whether the film survives the pressure and temperature the operation actually reaches.
Blank condition: burrs, grit, rust and a scored coil edge all carry material into the die.
Reduction per pass and ironing: heavier reductions raise interface pressure and strengthen the case for a coating.
Press speed: faster strokes raise interface temperature and shorten the interval.
Workpiece coating: zinc and tin transfer to the die and change the friction condition.

Those drivers explain why two tools making the same part can need polishing on very different schedules, and why an interval quoted for one job rarely transfers to another. The response is not linear in any single factor: moving from uncoated steel to a plated or coated surface commonly multiplies the interval on stainless, while the same change on mild steel may show only a modest gain. A documented interval, with the last good part inspected before each polish, is more informative than any table.

Finishing cost and finish quality sit on opposite sides of one decision. A more expensive die treatment, a better lubricant, a shorter polishing interval and a lower reduction per pass all buy surface quality, and each costs something in tool preparation, cleaning, press time or operation count. Because those levers are additive, the cheapest real improvement is usually not a new coating but a shorter requirement: if an internal surface touches nothing, seals nothing and is seen by no one, a finish callout on it buys nothing, and deleting it is free. Relaxing it is still a specification change, so record it against a boundary sample and let the functional surfaces carry the budget. Where money is needed, spend it on the two or three high-pressure zones rather than uniformly.

We deep draw and progressive-die stamp housings in SUS304 and 316L, SPCC, DC01 and DC04, copper and brass, aluminium, SPTE and DT4E on presses from 25 to 350 t, with in-house tool design and build, so die surface decisions are made alongside the tool rather than after it. Where a surface requirement cannot be met inside the draw — a mirror finish on a wall that must also be ironed, or a sub-micron Ra on a deep internal surface — it has to come from a later operation, and the drawing should say so. Our limits are set out under deep draw metal stamping capability.

Drawing lines, die marks and galling on a deep drawn part: where they come from and how die material changes them