Pesquisar no resumo do documento
Preventing warping and improving adhesion of high temperature
PLA: A practical guide
By Razzak Al-Gurnawi ([email protected])
In the 3d printing community, PLA has traditionally been regarded as a material that allows
easy and trouble free printing. Its properties allow it to adhere to a variety of surfaces at a
relatively low temperature. While this is generally true, it is not always the case. Many users
in the 3d printing community have experienced problems with first layer adhesion and
warping, which are issues usually associated with ABS.
While this guide attempts to address some of the issues encountered with bed adhesion and
warping, it is by no means a complete guide. It is worthwhile noting at the outset that there is
no substitute for proper calibration of the printer, especially ensuring the level of the printer
bed relative to the nozzle is appropriate and consistent.
It is important to note at this stage that PLA comes in several variants. For the purpose of
this discussion I will be focusing on the NatureWorks brand PLA. Most popular in the 3d
printing world seem to be the following variants:
Natureworks Ingeo 4 Series 4043D:
Natureworks Ingeo 4 Series 4032D:
The two products have some very important differences in properties which must be
considered carefully before printing. The major differences between the two materials lie in
their thermal properties. The 4043D plastic has a melting point of 145°C-160°C, while the
4032D plastic has a melting point of 155°C-170°C, making the latter more suitable for higher
temperature applications which traditionally have been the realm of ABS.
This increased temperature resistance of the 4032D plastic raises several problems when it
comes to 3d printing. The two major ones are:
• Poor bed adhesion
• Warping of printed parts
These require higher extruder and heated bed temperatures, which in themselves cause
Poor bed adhesion
The 4032D plastic will not adhere to the printing surface as easily as the 4043D plastic. A
plain glass sheet may not be sufficient as the surface is too smooth for the plastic to bond to.
A part printed on a glass sheet typically seems to detach itself with minimal lateral loads
(poking with the finger).
There are many documented ways to overcome this poor bonding of the first layer:
• Sandblasting the glass bed
• Using painters tape on the printing surface
• Using glue sticks
• Applying hairspray to the printed bed
• Using watered down PVA glue
• Lemon juice (I assume this has to do with keeping the surface free from oil?!)
Your mileage may vary with any of these techniques and many of them will work better than
others. They are all generally inexpensive and experimenting with what works is
encouraged. I personally have found the PVA glue technique to be excellent. Adhesion has
been so good, I accidentally broke a slender piece while removing it.
To apply this technique, mix a quantity of PVA with water. 2 parts water to 1 part PVA works
very well. Apply this to the printing surface using a brush. The heated bed will dry this layer
very quickly as it heats up. To decrease adhesion, simply add more water.
I have found that a heated bed at a temperature of 90°C works very well. It is possible to go
lower, but I have suffered poor results at anything below 70°C. Nozzle temperature can vary,
but 220°C +/- 10°C seems to be the optimal range. This avoids the filament rolling over the
surface and sticking to the nozzle.
Once good first layer adhesion was achieved, catastrophic warping of the plastic part is
usually experienced, an example of which can be found in Fig. 1.
Fig. 1 – Catastrophic warping
After an extensive review of online material, consultation of the material data sheets and the
material vendor, the reason for warping is believed to be due to the steeper temperature
gradient experienced by the part in the Z direction. This temperature gradient crosses the
glass transition temperature of PLA. Crossing this critical temperature threshold is believed
to cause the warping problems experienced with high temperature PLA (4032D). This does
not usually occur with regular 4043A PLA which is usually printed at bed temperatures
around 50 degrees.
Glass Transition Temperature
Wikipedia defines the glass transition temperature as the reversible transition of an
amorphous material from a hard and relatively brittle state into a molten or rubber like state.
When undergoing this phase transition, the material experiences a large change in material
properties. This change in material properties also affects the expansion coefficient of the
material as shown in Fig. 2. While the material experiences one rate of expansion between
points A and B, once it goes through the transition (the region between points B and C) its
rate of expansion changes to the rate between points C and D.
Fig. 2 – Thermal expansion vs. temperature
Temperature gradient across the printing Z axis
A 3d printed part can typically be assumed to experience a temperature gradient as shown
in Fig. 3. If we heat the print bed to approximately 90°C, the lower layers (left hand side of
the graph) will be above the transition temperature of the material, while the upper layers
drop below it (right hand side of the graph). In practise, this means the part expands at
different rates across its Z axis. This causes the layers to curl on one another.
Fig. 3 – Hypothetical temperature distribution across the Z axis of a printed part
Based on the analysis above, it would be prudent to print the part in question using a
temperature which is below the material’s glass transition. This however is in direct conflict
with the adhesion problem faced earlier, where it is established that it is important to print
the heat resistant PLA at higher temperatures to improve adhesion.
The solution to this is to use a high temperature to apply the first layer of plastic and get a
good level of adhesion. After the first layer is printed, the temperature is dropped below the
glass transition point (around 50°C -55°C). This allows the print to experience the same
lower rate of expansion (Fig. 2 region A to B) across its entire z depth. Curling is not
experienced at the first layer due to this temperature as the material layer is thin enough to
accommodate expansion and contraction.
This has indeed been tested successfully several times using a Printrbot LC v2 with a
heated bed and a glass sheet. The results are very consistent. An example of this is shown
in Fig 4.
Fig. 4 – Successful printing run