Laser Marking Glass Bottle Saves Man's Job
One of Jimani's regular job shop engraving customers showed up today with a problem. He had 2...
By: Jim Earman on 3/26/26 8:00 AM
Most marking problems don't start at the laser. They start at the drawing. A part arrives with a vague callout — "laser mark here" — no depth spec, no surface condition noted, no code size defined. The job shop does its best. The results are inconsistent. Parts get flagged at inspection. Everyone wonders what went wrong.
Design for laser marking (DFLM) is the practice of specifying marking requirements the same way you'd specify a tolerance or a surface finish: with enough precision that any qualified shop can execute and verify them. This guide covers the decisions that matter most — placement, part geometry, surface finish, coatings, and code layout — along with why getting them right before the first piece is marked saves significant time and money downstream.
Design for laser marking (DFLM) is the practice of specifying mark location, surface condition, mark type, and verification criteria on a drawing or work order before a part enters the marking process. When those requirements are left vague or undocumented, the marking step becomes a judgment call — and judgment calls produce inconsistent results, especially across production runs, multiple operators, or different job shops.
Laser marking is a permanent process. That's its primary value. The flip side of permanence is that a bad mark is expensive — rework is limited, and on many metal substrates there's no going back once the beam has run. A mislocated part number on a medical device or an unreadable Data Matrix code on an aerospace component can mean scrapped parts, failed audits, or missed shipments. The mark quality problem is almost always traceable to something that could have been addressed at the design stage.
The parallel to machining tolerances is useful here. A machinist can't hold a dimension that hasn't been called out. Similarly, a laser operator can't guarantee a specific contrast level, code grade, or depth unless those requirements are defined somewhere in the documentation. "Laser mark part number" tells the operator what characters to use. It says nothing about where, in what technique, on what surface condition, or to what depth — and those variables have a direct effect on whether the mark survives in service and passes verification.
In our job shop at Jimani, the questions we ask most often before running a new job are: Where exactly on the part should this go? What's the surface condition at time of marking? Does the mark need to meet a depth spec? Does the code need to pass a specific scanner or grade requirement? If none of that is documented, we ask before we mark — because permanent marks made on the wrong assumptions are nobody's idea of a good outcome.
A steered-beam fiber laser marks within a flat focal plane defined by the f-theta lens. Any surface geometry that pulls the marking area out of that plane — curvature, steps, tapers, sharp radii — reduces power density at the surface, which produces weaker marks, wider and less defined line widths, and degraded code readability. The impact compounds with distance from focus: a surface just a few thousandths of an inch out of focus may still produce an acceptable mark, but a surface off by .030" or more on a narrow lens will produce a noticeably softer result.
Flat surfaces are the easiest case. If the marking area is flat, within the marking field of the lens, and perpendicular to the laser beam, setup is straightforward and results are predictable. Problems arise when marking areas are close to edges, bosses, or pockets that interfere with either the beam path or the fixture that holds the part in position.
Curved surfaces require more planning. A general rule for cylindrical parts is that you can mark roughly 60° around a 1" diameter part without rotating the part — beyond that, the beam moves far enough from the focal point to degrade the mark visibly. This matters for code placement: a 2D Data Matrix code placed across a transition from flat to curved on a small medical instrument may read fine at the center and fail at the edges because the outer cells have moved out of focus. If circumferential marking is required, the part needs to be rotated, and that means either a rotary indexer in the fixture setup or a job shop equipped for tiling. Those are solvable problems, but they need to be scoped before the first piece runs — not discovered when the first batch fails a scan verification.
Stepped surfaces — where the mark needs to span two different height levels — require a z-axis repositioning between tiles or, in the case of very small height differences, a lens with enough depth of focus to bridge the step. In our experience at Jimani, a height difference of .010"–.015" is usually within the depth of focus of a standard lens on a 5" marking field, but beyond that, you're working outside the reliable focus window and mark quality will be inconsistent.
[Image: Side-by-side comparison of a well-focused mark vs. an out-of-focus mark on an aluminum substrate — shows the difference in line width and edge definition. Alt: Out-of-focus vs. focused laser mark comparison on anodized aluminum]
Laser marks should be placed on flat, accessible surfaces that fall within the lens's focal plane during a single setup, with the mark area clear of edges, holes, and radii by at least the width of the mark itself. Placement too close to a raised wall, boss, or edge shadow will cause uneven contrast or partial dropout because the beam clips adjacent geometry or the part blocks proper fixturing.
The access question is more significant than it might appear. A steered-beam galvo system delivers the laser from above, perpendicular to the part surface. If the marking location is inside a pocket or recessed below a shoulder, the angle of the beam as it moves toward the edges of the marking field may be obstructed by surrounding geometry. Pocket walls that are taller than the depth of the pocket plus the working distance of the lens will block the outer portions of the mark. This is a fixturing problem as much as a design problem, but it's easier to deal with when the marking pad is designed to be accessible from above without obstruction.
Near edges is where we see the most preventable problems in the job shop. A serial number or Data Matrix code placed .020" from a machined edge sounds fine until you realize that the ablation process creates a small heat-affected zone and that the edge itself reflects the laser in ways that can create artifact marks or reduced contrast right at the boundary. A minimum standoff of the mark height from any edge is a reasonable design guideline — more if the edge is a sharp burr rather than a clean chamfer.
Mark location also affects fixturing design. If you're running a high-volume production job, a repeatable fixture is essential for consistent placement. Marking that can only be done in one orientation, on a clearly defined flat pad, with a positive stop in the fixture, will produce far more consistent results run after run than a mark whose location is nominally specified but depends on an operator eyeballing placement each cycle.
Type II anodized aluminum with a dark dye gives the best contrast for laser ablation — the ablated white mark reads sharply against the dark background. Clear anodize, light-colored anodize, and chemical film coatings like Iridite or Alodine all produce readable marks, but the contrast is subtler and the visual result will be more dependent on lighting conditions during inspection and scanning. The surface condition at the time of marking is one of the most important variables in predicting final mark quality, and it should be specified explicitly on the drawing.
The mechanism behind contrast on anodized aluminum is worth understanding. Laser ablation removes the anodize coating by vaporizing it, exposing the bare aluminum beneath. The bare aluminum surface, once lightly textured by the laser, scatters reflected light and appears white. If the anodize being removed is a dark color — black, blue, dark gray — the contrast between the white ablated area and the dark background is excellent and readable under almost any lighting condition. If the anodize is clear or a light color, that same white mark on a light background is far more subtle, which creates real problems for automated scanning.
Hard anodize is a specific challenge. Type III (hard) anodize builds a much thicker, harder coating than Type II — the coating penetrates into the substrate as much as it builds up the surface, creating a ceramic-like layer. In our job shop, we find that hard anodize typically requires more passes, a 50 watt laser rather than a 20 watt, and higher fill density to achieve adequate ablation. Even then, the cleanup pass that produces the bright white appearance on Type II anodize rarely delivers the same result on hard anodize. If mark contrast is critical on a hard-anodized part, that should be discussed with your marking source before the parts are processed.
Bare aluminum — no coating — marks by ablation as well, though the contrast depends on the surface finish of the base material. A matte or bead-blasted surface produces a more visible mark than a polished surface, because the textured base provides a background against which the laser-modified area stands out. On stainless steel, the appropriate technique shifts from ablation to stain marking — where controlled heating of the surface creates a dark oxide layer without removing material. This is a non-penetrating process, which is why it's widely used for medical devices and surgical instruments where surface integrity requirements are strict.
[Image: Close-up of laser ablation mark on dark anodized aluminum vs. clear anodized aluminum, showing contrast difference. Alt: Laser ablation contrast comparison on dark vs. clear anodized aluminum]
A Data Matrix code requires a minimum cell size large enough for the production scanner to resolve at the scanning distance used in your process, a quiet zone of at least one cell width on all four sides, and placement on the flattest and most surface-uniform area of the part. Cell sizes below .010" on an engraved or ablated surface carry risk unless the mark has been validated against the actual scanner at production distance — code grading software will estimate readability, but a physical scanner reading a physical part is the only reliable verification.
Cell size is where most Data Matrix code failures originate, and it's almost always a documentation problem rather than a laser capability problem. The mark is generated correctly by the laser — the cells are present and dimensionally accurate — but they were specified at a size that the installed scanner cannot resolve. The fix is to verify the minimum cell size the scanner can read at production distance before the drawing is released, then specify that cell size on the drawing with a minimum, not just a nominal. Specifying a nominal of .010" with no tolerance means a shop could deliver .008" cells and be in spec — which may not scan.
Quiet zone is frequently omitted from drawings, and it matters. A Data Matrix code that runs to the edge of a machined face without adequate margin will fail to decode even if every cell is perfectly formed, because the reader algorithm depends on the finder pattern contrast. One cell width on all sides is the bare minimum; two cell widths is more reliable in production environments where parts may shift slightly in the fixture.
Code orientation also deserves a note. A Data Matrix placed on a surface that's subject to pooling fluids, grinding debris, or thermal discoloration in service may read fine off the machine and fail in the field. The MIL-STD-130 standard for UID marking on defense items is a useful reference even when the application isn't defense-related — it establishes minimum requirements for mark location, contrast, and scanner verification that represent genuinely good practice for any traceability application.
"Marking by hope" is what happens when an engineer notes a generic marking requirement — "laser mark per spec" or "mark P/N here" — without defining mark type, depth, minimum contrast, surface condition, character height, or verification method. Every downstream operator fills in those blanks differently, producing variable results that pass or fail inspection based on interpretation rather than a defined and measurable standard.
The phrase sounds a bit sharp, but it accurately describes something we see often in the job shop. A customer sends a part with a drawing that calls out a part number to be marked in a particular area. That's useful information. What's missing is everything needed to produce a consistent, verifiable mark: whether the part is marked before or after anodize; whether the mark needs to survive a specific surface treatment that comes later in the process; what the minimum character height is; whether the mark needs to be engraved to a specified depth because the surface will be abraded in service; and whether there's a pass/fail criterion for contrast.
The consequences show up differently depending on the application. For identification parts in a non-critical application, a variable mark might be acceptable. For a serialized medical device that will be tracked through sterilization cycles, a mark that degrades after repeated autoclave exposure — because it was ablated shallowly into hard anodize rather than engraved to a specified depth — is a traceability failure waiting to happen. For a UID-coded defense component, a Data Matrix that grades at a C when the contract requires a B is a non-conformance.
The fix isn't complicated. It's documentation discipline applied early, before the first piece is marked. A complete marking note on a drawing takes perhaps thirty minutes to write correctly the first time. The savings in rejected parts, re-mark costs, and audit responses over the life of that part number are substantial. We've reworked a lot of jobs over the decades at Jimani that could have been marked right the first time if the requirements had been defined before the parts arrived.
A complete laser marking callout should identify the marking location relative to a datum, specify the mark type (ablation, engraving, or stain marking), state the required depth or minimum contrast level, define minimum character and code cell size, note the surface condition at time of marking (before or after any coating process), and reference any applicable standard — such as MIL-STD-130 for UID items or AMS-2411 for anodized aluminum parts requiring marking depth requirements.
Location definition is the foundation. "Mark on face A" is better than nothing, but it leaves the exact position undefined. A datum-referenced center point or zone — "mark centered within a .500" x .500" window, .250" from datum B along the X axis" — is specific enough that a fixture can be built to it and an inspector can verify placement without ambiguity. On small parts with limited real estate, this level of specificity also forces the design team to confirm that the marking area is physically achievable before the drawing is released, rather than discovering the interference when the first batch arrives at the shop.
Mark type should always be called out because ablation, engraving, and stain marking produce different physical results — different depths, different surface profiles, different durability in service. An ablated mark on anodized aluminum has essentially no depth; an engraved mark may be .005"–.015" deep. Those two marks behave very differently in a wear environment or under chemical exposure. Calling out "laser mark" without specifying the technique leaves the decision to the operator, who may not know which one the design intent requires.
Depth specification matters most when the mark will be subjected to surface treatments after marking, or when it needs to survive abrasion or chemical exposure in service. Firearms markings, for example, have a federally specified minimum depth requirement for manufacturer and serial number markings. Medical device markings that must remain readable through repeated sterilization cycles may need to be engraved rather than stain-marked on certain substrates. If depth isn't critical, saying so explicitly — "surface mark, depth non-critical" — is also useful because it lets the shop optimize for cycle time and contrast rather than running extra passes to achieve depth that isn't needed.
Finally, if the part is marked before a coating process — before anodize, before powder coat, before plating — note that clearly and specify whether the mark is expected to survive the coating process intact. Some coatings will partially fill an engraved mark and reduce depth. Others will change the contrast of an ablated mark significantly. Specifying the sequence, and the expected condition after each step, is the kind of information that prevents a lot of rework.
[Image: Example engineering drawing detail showing a complete laser marking callout with datum reference, mark type, and character height — annotated. Alt: Engineering drawing with complete laser marking specification callout]
If you're working on a part that's headed for laser marking and the requirements aren't fully defined yet, that's exactly the kind of conversation we have every day in our job shop. Send us a drawing or a sample part and let's work through the marking spec together — before the first piece runs.

One of Jimani's regular job shop engraving customers showed up today with a problem. He had 2...
Industry: Consumer Goods
Industry: Consumer Goods
Material: Wood
Mark Data Type: Graphics, logo
Laser Mark Type: Engraved,...