Why do parts shrink during 3D printing?

The fused filament fabrication (FFF) process relies on the rapid melting and cooling of plastic filament to consolidate the designed geometry. Like most materials, polymers expand when exposed to heat and the degree of expansion is an intrinsic material property known as the coefficient of thermal expansion (CTE).

If you increased the temperature of polycarbonate and nylon by 50°C, you would observe a greater expansion in nylon due to its higher CTE. Conversely, if polycarbonate and nylon are cooled down by 50°C, you will see greater shrinkage in nylon compared to polycarbonate, as shown in the figure below.

Graph representing idealized thermal expansion and shrinkage of polymers with different coefficients of thermal expansion (CTE). Higher CTE materials will shrink at a greater rate when cooled. Source: AON3D

We can see that some degree of shrinkage is unavoidable since polymers need to cool from the extrusion temperature to room temperature. This effect is more pronounced for high-performance polymers such as PEEK, PEKK, and ULTEM™. When these polymers cool below their glass transition temperatures, the rigidity of the polymers increases. Due to this increased stiffness, thermal contractions are inhibited, which results in residual stress.

Shrinkage in additive manufacturing can also occur due to crystallization in semi-crystalline polymers such as PLA, PA (nylon), PEEK, and PEKK. The graph below illustrates this shrinkage, showing a decrease in the specific volume (volume per unit mass) as the polymer is cooled below its melting temperature.

Semi-crystalline polymers experience a volume decrease when cooled under the melting temperature Tm due to crystallization. Amorphous polymers don’t crystallize and show a linear decrease in volume under and above the glass transition Tg. Source: AON3D

In contrast, amorphous polymers do not crystallize and instead show a gradual decrease due to the aforementioned thermal contraction. The formation of crystals results in a sharp specific volume decrease which, in turn, shrinks the printed polymer.

Shrinkage is a direct contributor to residual stress in 3D printed parts. But what exactly is residual stress and how does it affect the performance of printed parts? In the following sections, we explore how residual stress occurs during 3D printing and provide guidelines on how to reduce it.

What is residual stress and why does it occur?

Heat is brought into 3D printed parts continuously from the top surface due to the layer-by-layer process. Layers below the newly deposited track will cool relative to the extrusion temperature and heat will transfer from the top layer in three ways:

  • Conductive heat transfer to the cooled layer below
  • Convective heat transfer to the air around the part
  • Radiative heat transfer to the rest of the part and the build chamber

As heat is lost rapidly from the top layer, the layer begins to shrink. But since the top layer is constrained by the previous layers beneath, the layer is not able to shrink or deform completely. Residual stress occurs when a material has the propensity to shrink but is physically restricted.

Residual stress is developed internally due to the thermal gradients during the 3D printing process. When parts are removed from the build sheet, the residual stress relaxes which manifests physically as warped geometries. Source: AON3D

Now if you were to remove this part from the build plate, you would immediately notice that the edges curl. This happens because the part is no longer constrained to the build plate and is free to deform, relaxing the residual stress. Higher residual stress equals greater warping in printed parts as well as poor interlayer strength.

Before we suggest improvements for dimensional accuracy and mechanical properties of printed parts, let’s first understand why poor layer weld strength occurs in 3D printing.

What causes poor layer weld strength?

If you’re familiar with 3D printing, you may have heard that FFF 3D printed parts are weakest along the Z-direction. The way parts are sliced and the hardware used to print the part both have a considerable effect on the Z-axis properties. Here we look at why poor layer weld strength occurs and how we can improve properties to match injection molded polymers.

The mechanical properties of all polymers, whether amorphous or semi-crystalline, are affected by the selected layer height when slicing the model. The example below shows tracks printed with a 0.2mm and a 0.4mm layer height with a 0.4mm nozzle diameter.

Porosity vs layer height. Source: AON3D

When the chosen layer height is smaller than the nozzle diameter, the deposited bead is squished. This allows the polymer to spread, creating a larger surface area for the printed layers to bond. A larger surface area allows for more polymer chains to interact between layers, increasing the interlayer weld strength.

In addition, the squished beads are more rectangular in shape which decreases the pore size between deposited tracks. Pores are detrimental to the mechanical properties of printed polymers since they are sites where delamination failure begins. The extrusion width setting on slicers should also be modified to ensure these pore sizes are minimized.

To further increase the interlayer weld strength, the right hardware is required. When a part is created in layers, the polymer chains on the surface of each layer are initially disconnected. To promote interaction between layers, heat is required to mobilize the polymer chains.

Fortunately, the 3D printing process delivers heat as a new layer is deposited. Heat is transferred to the previous layers which improves polymer chain migration and interpenetration. Here, the chosen extrusion speed and width play an important role; tuning the speed to ensure enough heat is transferred and the extrusion width to minimize pores is essential. But often, especially for high-performance polymers like PEEK, PEKK, and ULTEM, relying on the heat from extrusion is not enough. Therefore, a heated build volume is required to promote the movement of polymer chains between layers and to maximize the mechanical properties of parts in the vertical axis.

How to reduce shrinkage, residual stress, and poor layer adhesion?

The key is to establish a uniform temperature distribution during the printing process. Since all three of these phenomena are thermally induced, lower variations in temperature across the part will decrease warping and residual stress.

To minimize warping and ensure properties are maximized for high-performance polymers, a heated build chamber is required. Heated build chambers that can exceed 120°C are optimum for maximizing properties on high-temperature thermoplastics. Printers such as the AON M2+ allow homogenous temperature distribution within the build volume and will allow layers to stay warm enough for polymer chain migration to increase the layer weld strength.

Tested This product has been reviewed by our team.
Product data validated by the manufacturer.
Approximate starting prices based on supplier-provided information and public data. Prices may vary by region, over time and do not include additional products or services (taxes, shipping, accessories, training, installation, …).
$ 59,400

A high temperature industrial 3D printer for manufacturing full-scale, strong, high-performance parts. Combined with ungated access to the world’s most advance materials, including PEEK, PEKK, and ULTEM™, the AON M2+ unlocks additive manufacturing applications.

In addition, carefully engineered convective air circulation ensures that thermal stability is maintained and distributed across the volume of the part to prevent warping and localized residual stress.