Process

Anisotropy in metal-printed parts: why orientation is part of the plane

· 4 min read · Samylabs

A part manufactured by laser powder bed fusion does not have the same properties in all directions. It is less resistant in the direction in which it was grown than in the plane of each layer, and the difference is not a mere laboratory nuance: in terms of elongation and fatigue, it can be significant.

This leads to the consequence that is hardest for someone coming from machining to grasp: the manufacturing orientation is part of the part specification. Changing it changes the product, even if the drawing hasn't been touched.

Where does the difference come from?

Three things happen at the same time and they all push in the same direction.

The grains grow upwards. Upon solidification, heat is primarily dissipated to the preceding layer, which is solid, cold metal. The grain grows in the opposite direction of heat flow, thus elongating in the direction of build and often spanning several layers. A microstructure with grains elongated along an axis responds differently depending on the loading direction.

The defects are flat and horizontal. The lack of fusion occurs between strands and between layers, so it tends to form elongated defects that are perpendicular to the growth direction. A flat defect is much more damaging when the tensile force pulls perpendicular to it, which is exactly what happens under vertical loading.

The surface is worse on inclined faces. Faces that look downwards are rougher, and roughness is the point where a fatigue crack starts.

How much does it weigh

Without giving numbers that depend on each material and each parameterization, the pattern is constant:

Property How it behaves
Elastic limit Moderate difference between directions
Tensile strength Moderate difference
Elongation Considerably lower in the growth direction
Fatigue life This is where it is most noticeable, by far

The practical takeaway: if your part operates under static load and has sufficient clearance, anisotropy probably won't cause any problems. If it operates under fatigue, that's the first thing to check.

What reduces it

Heat treatment. This is the most effective method. Stress relief helps; recrystallization treatments can significantly equalize the directions, at the cost of time and furnace requirements.

Hot isostatic pressing. Closes internal porosity, which is one of the three causes. Expensive, and only closes pores that do not connect to the surface.

Sweeping strategy. Rotating the fill direction between layers—typically a few degrees on each—prevents defects and grains from all aligning on the same axis.

Orient the part so the load is distributed evenly. This is free and the most effective method. If you know where the stress will be directed, position the part so that this direction aligns with the layer plane, not the growth axis.

None of them eliminate it. They reduce it to a manageable level, which is the goal.

What to write in the specification

Here's the practical application of all the above. A complete LPBF part specification states, at a minimum:

  • The alloy and its supplier or purchasing criteria.
  • The manufacturing orientation, indicated on the model, not described in words.
  • The complete heat treatment, including its cycle.
  • The direction in which the required properties are measured.
  • The test specimens, manufactured using the same build and orientation as the part.

That last point is the one most often overlooked and the one that invalidates the most tests. A test specimen manufactured separately, in a different build or orientation, does not represent your part: it represents another part.

The three most costly mistakes

Reorienting to save time without warning. Tilting a part reduces the number of layers and lowers build costs. It also changes its properties. If the change isn't returned to the person who signed the plan, the product has been altered without anyone's knowledge.

Moving the part to a different position on the build plate. This doesn't change its orientation, but it does affect the thermal environment and gas flow. For demanding parts, the position may also freeze.

Give a single property number. A card that says "elastic limit X" without stating the direction is giving half the information, and usually the favorable half.

How would we approach it

Decide the orientation based on load, not build time, and fix it in the model from the beginning. If it needs to be changed later for production, make it a conscious decision that goes back to the design stage, not a shop floor adjustment.

Always make test specimens from the part, even if no one has asked for them yet. They cost little while build is underway and are impossible to obtain afterward.

And if the part is subject to fatigue, assume the discussion will be about the largest defect and its location, not the average density. This is the rule that prevents the most problems.

—

How all this is demonstrated and what is frozen at each level, in how to qualify an LPBF part. Where the defects that orientation exacerbates come from, in porosity and density. And why heat treatment is performed with the part still on the platform, in deformation and stress relief.

And if you want to see what comes out of your specific geometry —volume, build height and an order of magnitude of cost— without sending us the file, there is the quote calculator: the STL is read by your browser and does not leave your computer.

→Keep reading