What Is Layer Thickness in DMLS?

Direct Metal Laser Sintering (DMLS) belongs to the powder bed fusion family of additive manufacturing processes. In DMLS a laser selectively fuses metal powder particles layer by layer to build a near‑net‑shape part. Layer thickness is the vertical height of each individual powder layer that is spread, melted, and solidified. It is one of the most influential process parameters because it directly controls build speed, resolution, and part integrity.

Commercial DMLS systems typically offer layer thicknesses in the range of 20 µm to 100 µm, with 30 µm and 60 µm being common defaults for many materials. Advanced machines can sometimes go as thin as 15 µm or as thick as 120 µm for specific alloys. The chosen value determines the number of layers needed to complete a part of a given height: a 50 mm part built at 30 µm requires roughly 1,667 layers, while the same part at 60 µm requires only about 833 layers. That difference directly affects build time and cost.

Layer thickness interacts with every other process variable—laser power, scan speed, hatch spacing, and powder particle size distribution. Optimizing the combination for a given application requires understanding how thickness influences the two most critical quality metrics: accuracy and mechanical strength.

The Impact of Layer Thickness on Part Accuracy

Accuracy in DMLS includes dimensional precision (how closely the final part matches the CAD model), surface finish (roughness and texture), and geometric fidelity (reproduction of fine features, sharp corners, and overhangs). Layer thickness plays a central role in each of these areas.

Surface Finish and Roughness

The most visible effect of layer thickness is on surface quality. Because DMLS constructs parts from discrete vertical slices, the build direction results in a characteristic stair‑step pattern on inclined surfaces. The severity of this stair‑stepping is directly proportional to the layer thickness. Thinner layers—20 µm to 30 µm—produce steps that are so small they often become indistinguishable from the natural powder‑induced roughness. The resulting surface finish can approach that of conventionally machined parts, with Ra values as low as 5–10 µm depending on orientation and post‑processing.

Thicker layers, 60 µm and above, produce more pronounced steps, especially on shallow‑angle surfaces. For a 45° slope, a 60 µm layer yields a step height of roughly 42 µm, while a 30 µm layer yields just 21 µm. This difference is easily visible and can compromise the aesthetic or functional surface of components such as turbines, implants, or molds. Consequently, parts that require minimal post‑processing or that will be used as‑built often demand thinner layers.

However, the relationship is not purely linear. Very thin layers (below 20 µm) may lead to poor powder spreading or excessive re‑melting of previous layers, potentially introducing surface irregularities. Therefore, the practical minimum is often set by powder flowability and recoater blade design.

Dimensional Tolerances

Layer thickness affects how accurately the laser can reproduce fine features. Thinner layers allow the laser to resolve smaller details because the heat‑affected zone is more contained in the vertical direction. For thin walls, small holes, and sharp edges, 30 µm layers generally achieve tolerances of ±0.05 mm to ±0.1 mm, whereas 60 µm layers might drift to ±0.15 mm or wider. The stair‑stepping also creates a systematic deviation: a sloped surface built with thick layers will be slightly offset from the intended CAD profile due to the discrete step geometry. This can be compensated by applying adaptive slicing algorithms, but the baseline error remains.

Internal features such as lattice structures or conformal cooling channels also benefit from thinner layers. A thin strut in a lattice may be only two or three layers tall if built with thick layers, leading to poorly defined geometry or incomplete fusion. With finer layers the strut can be built from many more voxels, improving both accuracy and structural integrity.

Dimensional accuracy is also influenced by thermal shrinkage. Thicker layers involve larger molten pools that can introduce greater residual stress and distortion. Thin layers, while building slower, often produce parts that distort less because the thermal gradients are smaller and more uniform. This is especially critical for large, flat geometries prone to curling.

Stair‑Stepping Effect (Cusp Height)

The stair‑stepping error is mathematically described by the cusp height: the distance between the ideal sloped surface and the actual stepped surface. Cusp height = layer thickness × cos(θ) where θ is the angle from vertical. Thinner layers reduce cusp height proportionally, improving the overall shape accuracy. For surfaces that are nearly horizontal, the cusp height becomes very small regardless of layer thickness, but for vertical or near‑vertical walls the effect is negligible. The worst case occurs at shallow angles (e.g., 10°–30°), where thick layers can create significant geometric deviations. Adaptive slicing—varying layer thickness across the build—can mitigate this, but most commercial systems still use a uniform thickness per build job.

Stair‑stepping also creates stress concentration points that can affect fatigue life, especially in parts subjected to cyclic loading. Therefore, controlling layer thickness is not only a surface quality issue but also a functional reliability concern.

Effect of Layer Thickness on Mechanical Strength

Mechanical properties in DMLS are governed by the metallurgical quality of the fusion between layers. The interface between successive layers is a potential weak point because it often contains oxide inclusions, lack‑of‑fusion porosity, or insufficient remelting. Layer thickness directly controls the amount of energy delivered to each interface and the depth of re‑melting.

Tensile and Yield Strength

Numerous studies have shown that reducing layer thickness typically increases the tensile strength and yield strength of DMLS parts. For example, Ti‑6Al‑4V built at 30 µm consistently exhibits ultimate tensile strengths (UTS) of 1,150–1,250 MPa and yield strengths of 1,050–1,150 MPa, while layers of 60 µm might drop to 1,050–1,150 MPa UTS and slightly lower yield. The difference arises because thinner layers allow deeper relative remelting: the laser penetrates into the previous layer, ensuring a stronger metallurgical bond. Thicker layers increase the risk of lack‑of‑fusion defects at the inter‑layer boundary, which act as crack initiation sites.

The effect is material‑dependent. For Inconel 718, a nickel‑based superalloy, thin layers (20 µm) can improve UTS by 5–10% compared to 40 µm layers, while maintaining ductility. In 17‑4 PH stainless steel, the reduction in strength from 30 µm to 60 µm is often smaller (≈5 %) because the alloy’s low crack sensitivity partially compensates. However, for aluminum alloys (AlSi10Mg), thicker layers (>50 µm) can dramatically increase porosity and reduce UTS by 15–20%.

Build orientation also interacts with layer thickness. Horizontal (flat) builds benefit less from thin layers because the layers are large in area, and thermal history dominates. Vertical (upright) builds, with many small layers, show a stronger correlation between thin layers and high strength. For highly stressed parts, the combination of thin layers + vertical orientation often yields the best mechanical performance.

Fatigue and Ductility

Fatigue life is extremely sensitive to surface irregularities and internal defects. Since stair‑stepping and porosity are both reduced with thinner layers, the fatigue performance improves substantially. A DMLS part built at 20–30 µm may exhibit a fatigue limit (e.g., 10⁷ cycles) 20–30% higher than the same part built at 60–80 µm. The improvement comes from fewer stress raisers at the surface and fewer sub‑surface lack‑of‑fusion voids. Ductility (elongation at break) also tends to increase with thinner layers because the material is more homogeneous. For Ti‑6Al‑4V, elongation can rise from 6–8% (60 µm) to 10–14% (30 µm).

However, there is a diminishing returns point. Extremely thin layers (below 20 µm) can sometimes reduce ductility if the heat input becomes excessive, leading to over‑tempering or coarsening of the microstructure. For each material there is an optimal window where strength and ductility both peak. For example, many EOS and Renishaw parameter sets recommend 30–40 µm for the best balance.

Porosity and Defect Formation

Layer thickness influences the porosity type and density. Thinner layers promote fully dense parts (>99.9 % relative density) because each layer is remelted more thoroughly. Thicker layers can introduce two types of porosity:

  • Lack‑of‑fusion porosity – Large, irregular voids between layers, often elongated in the build plane. These are the most detrimental to strength and fatigue.
  • Gas porosity – Small spherical pores from entrapped gas in the powder or process; less affected by layer thickness but can be exacerbated if thinner layers require higher energy density.

Optimization requires adjusting laser power and scan speed to maintain a stable melt pool for each layer thickness. A common guideline is to keep the energy density (J/mm³) constant, but this is only a starting point. In practice, thinner layers allow lower laser power (reducing thermal stress) while still achieving full melting. Thicker layers often require higher power to avoid lack‑of‑fusion, which increases residual stress and the chance of cracking.

Material‑Specific Considerations

Different alloy systems respond differently to layer thickness changes due to variations in thermal conductivity, melting point, and solidification behavior.

Titanium Alloys (Ti‑6Al‑4V)

Titanium is sensitive to oxygen pickup and has low thermal conductivity. Thinner layers (20–30 µm) are strongly recommended for Ti‑6Al‑4V DMLS because they minimize the heat‑affected zone and oxide formation. Thicker layers (≥50 µm) often result in porosity and alpha‑case contamination. The industry standard for aerospace and medical parts is 30 µm, balancing detail, strength, and build time.

Nickel‑Based Superalloys (Inconel 625, 718, Haynes 282)

These alloys have high hot‑strength and tend to crack if thermal gradients are too severe. Thinner layers (20–30 µm) help by keeping the melt pool small and reducing thermal stress. However, excessive thinness can lead to a lack of fusion if the powder does not flow properly. Many recommended parameter sets use 40 µm for a safe compromise.

Stainless Steels (316L, 17‑4 PH)

Stainless steels are more forgiving. Both 20 µm and 60 µm can produce >99.5 % dense parts with good mechanical properties if parameters are adjusted. For 316L, thicker layers (50–60 µm) are often used for non‑critical parts to reduce cost, while 30 µm is typical for functional prototypes or medical instruments.

Aluminum Alloys (AlSi10Mg, Al6061)

Aluminum reflects laser energy and has high thermal conductivity. Thinner layers (30 µm) improve surface finish and reduce porosity caused by keyholing. Thicker layers (≥60 µm) often lead to balling and poor density. Most aluminum DMLS processes are optimized for 30–40 µm.

Optimizing Layer Thickness for Specific Applications

Selecting the right layer thickness is a trade‑off between speed, quality, and cost. Engineers must align the choice with the functional requirements of the part.

High‑Precision Components (Medical Implants, Injection Mold Inserts, Turbine Blades)

For these applications, thin layers (20–30 µm) are nearly mandatory. The need for tight tolerances, excellent surface finish, and superior fatigue life outweighs the longer build time. Post‑processing costs are also lower because less machining or polishing is required.

Prototypes and Low‑Volume Functional Parts

When time and cost are critical, medium layers (40–50 µm) offer a good balance. Surface quality is acceptable for most non‑aesthetic uses, and mechanical properties are still high (>99.5 % density). For many structural parts, 40 µm is the sweet spot.

Large, Non‑Critical Tooling or Fixtures

For parts that do not require high strength or surface finish, thick layers (60–100 µm) can cut build time by 50–75% compared to 30 µm. This is often used for support‑free builds or one‑off tools. However, careful attention must be paid to support generation and thermal management to prevent warping.

Lattice and Lightweight Structures

Thin layers are essential for fine lattice struts (<0.5 mm diameter). A strut built with 60 µm layers may only be 3 layers thick, making it fragile or even incomplete. Using 20–30 µm ensures struts have at least 8–10 layers, improving strength and consistency.

Conclusion

Layer thickness is a foundational parameter in DMLS that interconnects with accuracy, strength, and process economics. Thinner layers (20–30 µm) deliver superior surface finish, tighter tolerances, and higher mechanical properties—especially fatigue resistance and ductility—at the expense of longer build times and higher cost. Thicker layers (60–100 µm) increase productivity but can compromise part quality, introducing stair‑stepping, porosity, and weaker interlayer bonds.

The optimal selection depends on material, geometry, and intended function. Material‑specific literature and machine manufacturer parameter guides provide starting points. For critical applications, performing a process validation build that compares two or three layer thicknesses is recommended. Advanced simulation tools can also predict distortion and microstructure, but empirical testing remains the gold standard.

By thoughtfully balancing layer thickness with laser parameters, engineers can achieve cost‑effective, high‑performance DMLS parts that meet the most demanding requirements in aerospace, medical, automotive, and tooling industries. For further reading, studies such as "Effect of Layer Thickness on the Microstructure and Mechanical Properties of Selective Laser Melted Ti6Al4V" and ASME guidelines on powder bed fusion offer deeper technical insight.