Minimum Wall Thickness 3D Printing: What Actually Sets It
Ask three suppliers what the minimum wall thickness 3D printing can hold on their machines and you get three different numbers. All three are correct, and none answers your question: they are machine limits. What you need before you quote is a decision about load, span and process.
The manufacturing minimum is the thinnest wall the process can build repeatably. It is a property of the equipment.
The functional minimum is the thinnest wall that will still carry the load, hold the dimension and stay put in service. It is a property of your geometry.
They are close for some parts and far apart for others, and noticing that the datasheet number is the first one, not the second, is where most useful conversations about minimum wall thickness 3D printing begin.
Why is there no single minimum wall thickness for 3D printing?
Because four things move it independently, and a datasheet quotes you the one it can measure. The process sets the floor, because resin, sintered nylon and molten metal fail below a minimum wall in different ways. The material within that process shifts it. The wall's own geometry shifts it further: a flat vertical wall, a wall inclined to the build plate and a wall at the top of a tall part are three different jobs. And where the wall is matters as much as how thick it is — an outside wall is a solid the machine lays down, an inside channel is a void it has to empty.
What is the difference between the manufacturing minimum and the functional minimum?
The manufacturing minimum is where repeatability stops. Below it the wall is not reliably formed, the layer does not bond, the melt does not fuse through its depth, or the feature collapses under its own weight before the build finishes. A wall under it is not a thin part, it is a scrapped build.
The functional minimum is softer, and three of its inputs are yours:
- Span. Stiffness scales with thickness cubed, so the same wall carrying load across a short distance and across a long one is not the same structure.
- Load path. A wall taking compression, shear or a bolt load is a different decision from one that only closes an envelope.
- Service environment. Heat, moisture, chemicals and cleaning cycles all push the safe value up.
Thickening is the obvious answer and the least efficient: it adds mass, build time, and in metal the most expensive material on the quote. Where a process page offers a recommended figure above the published minimum, that gap makes the distinction explicit. Our SLA process page, for example, publishes 0.4 mm as the minimum wall thickness and 0.6 mm as what it recommends for functional parts.
How do SLA, SLS and SLM compare on wall thickness?
They do not sit close, and the ranking is not the one most people expect. These are the figures published on our own process pages:
| Published figure | SLA (resin) | SLS (nylon) | SLM (metal) |
|---|---|---|---|
| Minimum wall thickness | 0.4 mm | 0.8 mm | 0.4 mm |
| Recommended, functional parts | 0.6 mm | 1.0 mm | not stated as one figure |
| Minimum feature size | 0.2 mm | 0.4 mm | not stated on the page |
| Minimum hole or internal channel | 0.5 mm hole diameter | 1.0 mm internal channel | 0.5 mm internal channel |
| Dimensional tolerance as standard | ±0.1 mm | ±0.3 mm | ±0.2 mm |
| Layer height range | 50 – 100 µm | 60 – 120 µm | 20 – 50 µm |
Source: the capability and tolerance tables on our SLA, SLS and SLM process pages.
SLS has the highest floor, and the reason is the material rather than the machine. Un-sintered nylon is held together between particles, so a wall that is too thin is held together by too little of that bond and comes out of the oven with a weak section even when the shape survived. Metal fails differently rather than more strictly: a thin wall in laser powder bed fusion risks not fusing through its full depth, which is a buildability problem more than a strength one. Resin cures as a continuous solid, so it tolerates thin walls well; its constraint shows up elsewhere, in fine features.
Swipe the diagram sideways to read it →
Read across the tolerance row and the table says more than the wall row. A wall nominally 0.4 mm wide on an SLS line carries ±0.3 mm of standard tolerance, so the ratio between the smallest feature you can hold and the variation you accept does much of the design work. That is the argument for treating minimum wall thickness 3D printing as a tolerance problem, not only a strength one.
Does the material change the number?
Within a process, yes, through properties you would not think of as wall thickness drivers. In metal, what varies between alloys is how the melt pool behaves: how fluid the material is at the beam conditions, how much heat it carries away from the wall, and how well it wets the previous layer. In polymer, the variation runs across the grade rather than the chemistry: a glass-filled nylon is stiffer and more dimensionally stable than an unfilled one at the same wall.
The larger point is that wall thickness is rarely the most efficient answer to a stiffness problem. A designer who doubles a wall has taken the first available fix. Restructuring is often cheaper and better: moving material into ribs, thickening only the section that actually spans between two supports, or opening the geometry into a stiffened shell. An illustrative example, not a specification: a flat panel drawn at six millimetres can often be restructured as three ribs at one and a half millimetres that are lighter and stiffer in the direction that matters.
Why do internal channels have their own minimum?
Because an internal channel is not a wall at all. It is an absence of material that has to stay absent, and the process has to fill it with something on the way in and get it back out on the way out. The published figures show the gap clearly: on SLS the minimum wall is 0.8 mm but the internal channel is 1.0 mm, and on SLM the wall is 0.4 mm while the channel is 0.5 mm.
The mechanism differs by process, which is why the numbers do not scale together. In SLS the channel is emptied of loose powder, and below the published size the powder does not reliably clear. In SLM it has to be designed with an evacuation path, because unsintered metal powder does not behave in a narrow passage the way nylon beads do — our SLM page states the channel minimum together with the requirement for a clear evacuation path, and those clauses belong together, since a channel at the right diameter with nowhere for the powder to go is still a trapped-powder defect. For resin the question becomes drainage and cleanliness. The constant across all three is worth putting on your drawing: a closed internal void is a decision to make early, because the clearing operation it demands is easy to design in and expensive to retrofit.
Does wall thickness depend on orientation and on where the wall sits?
On orientation, yes, and the effect is easy to underestimate. A wall is a stack of layers, so what matters is how many make up the wall and how each bonds below. At the fine end of the SLM layer range a 0.4 mm wall is twenty layers deep, plenty of opportunity for the melt to run through; at the coarse end it is fewer, and the wall does less work for you.
On position, the wall at the bottom of a build is the most stable to make and the wall on an overhang the least, because it is built over open space while it is still only partly bonded. That is why orientation and wall thickness are the same conversation: a wall that must be both thin and in a difficult angle usually needs a local thickness increase or a different design, and finding that out in CAD is far cheaper than in the build. Our article on why 3D printed parts warp covers the mechanism; the connection worth noting is that a thin, wide section is both a stiffness problem and a distortion problem, which is why ribbing solves two defects at once.
What should you ask a supplier before you commit a number?
Four questions, and what you are really looking for in each is whether the supplier is quoting a machine limit or engaging with your part.
| What to ask | Why it matters | If the answer is vague |
|---|---|---|
| What geometry does that minimum assume? | A flat vertical wall, an inclined wall and a top-surface wall are different features. | The number is a machine label, not a design rule. |
| Is that your manufacturing minimum or your recommended value? | The gap between the two is where the design decision lives. | The datasheet floor is being treated as a target. |
| How do you clear an internal channel? | The channel figure is meaningless without the evacuation or drainage route. | Powder or resin is not accounted for in the design. |
| What happens if a wall falls below it? | Some lines run it and let you decide. Some hold the build. | The failure mode is discovered with your material in it. |
The first row gets skipped most, and it is the cheapest to ask. Published minima are measured against a reference geometry the supplier chose, and a wall with a hole, a wall meeting a curve, or a wall at the end of a thin rib are less forgiving than a plain flat wall of the same thickness.
Where this leaves you
The published numbers are worth having, and are on our process pages for that reason. They tell you where the floor is for each process, and whether a design is worth sending at all. What they do not tell you is what your part should be — and that answer comes from span, load path and service environment, reached in the drawing, where changing it is free. Minimum wall thickness 3D printing is not a number to look up. It is the answer to a question about structure that happens to have a machine limit attached to it.
If you send a drawing with a wall you are unsure about, tell us what that wall carries and what it spans between. That is faster than a number, and it usually ends with the wall thinner than you expected.