There are many aspects that can affect the performance of an additive part, one of which is the surface roughness. It is a key factor in certain mechanical properties and is driven by additive modality and process parameters. 

Lyndsay Kibler, lead materials application engineer explains the different effects of leaving additive surfaces ‘as-printed’ versus machining, as well as some of the ways to characterize the surface finish of an additive part.

Lyndsay Kibler

What are the typical as-printed surface finish values for additive processes?

The most common surface roughness parameter is Ra, the arithmetic mean roughness value over a line. While many surface roughness parameters are available to characterize the profile of a surface, Ra is the most common parameter used across many industries today.

Generally, laser powder bed fusion (L-PBF) is going to result in a surface finish around 5-20 μm Ra (~200-780 μin Ra) on a vertical wall, depending on material and process parameters. The values for electron beam melting (EB-PBF) are a little higher than L-PBF and tend to be around 30-38 μm Ra (~1180-1500 μin Ra). Our binder jet technology splits the difference with an average vertical wall Ra of approximately 15 μm Ra (~590 μin Ra) after requisite sintering operations.

surface roughness
Visual examples of various surface roughness values. Units are in μin Ra. Source: GE Additive

 

Does surface finish affect mechanical properties? If so, which ones are impacted the most?

The surface finish of a part will affect some of its mechanical properties, but not necessarily all of them. The property most affected by surface finish is fatigue life. Most other properties, such as tensile, assess bulk material properties and are less affected by surface conditions. Often times, the higher the surface finish, the more peaks and valleys exist on the surface of the part. Fatigue life is driven by the valleys on the surface, since they behave like stress concentrators that could prematurely initiate cracks that propagate through the material. This reduces the overall fatigue life compared to machined surfaces where the surface is almost entirely smooth.

Another consideration with regard to mechanical properties is thin wall components. In general, thin walls (typically less than ~1 mm) often have similar mechanical properties to bulk material. However, surface finish can affect the minimum wall thickness, which is measured from valley to valley as shown in the image below. Thus, the deeper the valleys, the thinner the ligament when comparing to nominal part geometry. Ultimately, the thinnest ligament will limit the life of the component if placed in a critical location.

 

surface profile
Image shows how surface profile can affect the minimum ligament of a thin wall. Source: GE Additive

In the metal additive industry, is there a preferred method and equipment for measuring surface roughness?

Today, tactile profilometry and optical methods are most commonly used to characterize surface roughness. Tactile profilometry is typically used for applications with relatively low surface roughness values (smoother surface) and is less accurate for higher roughness values. Tactile profilometers measure roughness by dragging a tool, usually with a diamond-coated tip, in a straight line across the surface of the part. The machine will then measure the surface peaks and valleys across a given length of the part. For high roughness surfaces, the tool may not be able to reach the lowest valleys and thus may provide inaccurate readings. Benefits of tactile profilometry are the low cost of equipment, and compatibility with legacy characterization processes used for traditionally manufactured components.

Optical surface roughness measurements are usually taken using a chromatic light or laser microscope that maps the 3D surface of the part with high precision. This equipment is typically more expensive, but the measurements and results are very versatile. Optical measurements are especially useful for characterizing higher surface roughness values, as more detail can be captured over a wider area of the part. Where tactile measurements are reported as the arithmetic mean roughness over a line (Ra), optical measurements can be reported over an area (Sa). These values are not always comparable, so care must be taken when comparing surface roughness values taken from different measurement methods.

Both tactile and optical measurement methods play an important role in additive surface characterization, but both also require additional development to standardize measurement methods across the industry. 

Which process variables can influence surface finish in a build?

The beam-based modalities share a lot of process variables that impact surface finish including part location on the platform, both in relation to the laser (accounting for laser incidence angle) and gas flow; the angle of the surface in relation to the build plate; powder particle size distribution; layer thickness; and beam parameters such as power, speed, and spot size. For binder jet components, powder bed density, sinter cycle parameters, binder deposition rates, printhead resolution, and the properties of the binder material itself all impact surface finish.

While all of these parameters can be optimized for surface finish, they often come with trade-offs. For example, in L-PBF and EB-PBF applications, decreased layer thicknesses and slower beam settings often result in better surface finish, but this also increases build time. Similarly for binder jet, increasing the speed of head traverse reduces print times, but also increases surface roughness.

optical surface roughness
Example of optical surface roughness measurement profile. Source: GE Additive

 

How are these process variables controlled to optimize surface finish?

Again, many of the optimization processes are modality and process dependent. At GE Additive, we perform extensive parameter development studies across all modalities to optimize mechanical properties and surface finish at multiple build angles. Especially in L-PBF and EB-PBF, different parameters can be applied to different geometries, such that the parameter set could be different for a vertical wall than for an angled wall.

Another way to optimize surface finish is by adjusting the build orientation and layout to avoid a higher surface roughness in known high stress regions. For L-PBF and EB-PBF applications, parts can be oriented in such a way that any critical features are printed at more optimal angles to the build plate. Additionally, supports can be added to steep overhangs to improve surface quality.

Finally, there are opportunities to optimize the geometry of the part itself. Several features can be adjusted by designing for additive: flat overhangs can become self-supporting using chamfers and blends on downward facing surfaces, hole shapes can be adjusted in CAD to improve the top surface in the as-printed condition, and steep downward facing surfaces can be made more gradual.

If a part has been machined conventionally, can additive users consider leaving surfaces as-printed (without machining) in the final product?

Yes! There’s often a misconception that you always need to machine an additive part to achieve the most optimal properties for every application. One of the major benefits of additive is the design freedom that is no longer constrained by tool paths and machining access points. In many cases, features of additive components cannot be machined, which is why additive is chosen as a manufacturing method in the first place. Some examples of these features include shaped internal holes and channels, complex strut structures, and lattice geometries among many others.

Whether a component needs to be machined or can be left as-printed is dependent on the part requirements. Several reasons to machine surfaces include critical tolerance features, such as interfaces to other components, bolt holes, or threaded holes; and fatigue-limited areas where a machined surface would improve component life. Otherwise, if the rest of the component can be left as-printed, it’s often less expensive to do so.

By analyzing the component requirements and characterizing the material properties in the desired end state, users can address all these different considerations. (Click here for more information on generating material property curves.) This allows them to look at both the drawbacks and benefits of machining compared to leaving a surface as-printed, while balancing component requirements with the process-related mechanical properties and tolerance debits.

What other post-processing methods exist for improving surface finish besides traditional machining?

There are several methods available for improving surface finish, and several considerations need to be made when selecting the right method for a given application. Some methods, such as abrasive slurry flow, are useful for smoothing surfaces in holes and passageways to promote consistent fluid flow. Others, such as tumbling and grit blast, can improve the appearance of the part and overall surface finish by smoothing down the peaks on the surface. However, these methods will not address the valleys, and as previously mentioned, the valleys drive debits in fatigue life. So, if the goal is to improve fatigue life, an alternate method (for example, chemical milling or traditional machining) must be considered to remove enough material to remove the valleys altogether.

Binder jet offers a unique opportunity to improve the surface of a part while still in its “green” state (i.e., prior to sintering). A process as simple as manual sanding of surfaces—while not an industrialized solution—has shown to improve final part surface roughness. This may be scaled up to grit blasting or similar processes if the green part has enough strength to withstand this additional post-processing. The reason to perform the work on the part in its green state is that the level of effort is greatly reduced: it is much easier to remove some surface material from a green part than a final, fully sintered component.

 

surface roughness post processing
Examples of additive surfaces after various post-processing methods.  Source: GE Additive

Overall outlook

The surface finish of a part is highly dependent on the modality and processing parameters used, and many of these parameters can be adjusted to optimize surface finish. Several of these parameters come with trade-offs, whether in productivity or cost, and should be balanced with component requirements.

If as-printed surfaces will not satisfy component requirements even after process optimization, there are several methods that can improve surface quality. Selection of the post-processing steps should be governed by the requirements of your part in its intended environment.
 

If you are interested in finding out more about what levels of surface finish your parts need, then please contact us.