Every other filament on this site has one density. LW-PLA has two: the density of the material on the spool, which is ordinary PLA at around 1.24 g/cm³, and the density of the part that comes out, which can be as low as 0.4. The difference is a chemical foaming agent blended into the polymer, and it makes this the only common filament whose printed properties you set at the machine rather than choose at the shop.
How the expansion works
The foaming agent is inert below roughly 230 °C. Above that it decomposes and releases gas, and the extrudate leaving the nozzle puffs up — an extrusion that would have been 0.45 mm wide can emerge two or three times that. The polymer solidifies around the bubbles, producing a closed-cell foam.
You exploit that by cutting the flow rate. Set the slicer to extrude something like 40% to 50% of the normal amount and let the expansion make up the difference. Get it right and the part has full-width walls made from a third of the plastic. Get it wrong and you have either a gappy, under-extruded shell or a dense, blobby part that foamed less than you assumed.
Expansion varies with temperature, so it has to be calibrated on your machine. Print a test tower with the temperature stepping up every few millimetres at fixed flow, measure the wall thickness at each step, and read off the setting that gives the width you designed for. It is half an hour of work and there is no way around it — hotend geometry, flow rate and residence time all affect how much the agent decomposes, so another person's numbers will not transfer.
What the weight saving actually looks like
Take a wing section that would consume 120 g in plain PLA. At a foamed density of 0.4 g/cm³ the same geometry is 67.7% lighter. In filament terms the relationship inverts: 120 g of material at 1.24 g/cm³ is 40.23 m of 1.75 mm filament, while 120 g at an effective 0.4 corresponds to 124.73 m of printed volume — you are laying down three times the volume per gram. Three times the volume per gram is also three times the distance the head has to travel to lay that gram down, which is why foamed PLA prints slowly for its mass. The length and weight converter will keep the two figures apart for you: the spool's own density describes what you buy, the foamed density describes what you end up holding.
For a radio-controlled aircraft, where every gram costs battery endurance, that is not a marginal improvement. It is the difference between a design that flies well and one that does not.
The number your slicer gives you is wrong
This is worth stating flatly because it costs people money. A slicer estimates mass by multiplying extruded volume by the density in the filament profile. With LW-PLA, the extruded volume is not the printed volume — the material expands after it leaves the nozzle — and the profile density is the spool's, not the part's.
So the gram figure and the cost figure on your slicer's summary screen are both meaningless for a foamed print until you tell it what you are actually doing. Weigh the first successful part, back-calculate the real printed density, and put that in the profile. This material is the reason a length-and-weight tool has to ask which density you mean.
Four settings that make or break it
The nozzle window runs from 200 all the way to 260 °C, and its upper half is an expansion control rather than a melt-quality setting — which is why it looks so much wider than any other PLA's. The bed sits between 50 and 60 °C and no enclosure is needed.
- Print slowly. Foaming takes time in the melt zone; pushing material through quickly gives the agent less opportunity to decompose and produces inconsistent expansion.
- Keep cooling low. Aggressive part cooling quenches the foam before it has fully risen.
- Dry it. Moisture interferes directly with the foaming reaction, so a damp spool gives erratic expansion — which reads as a calibration problem and is not. Four to six hours at 45 °C.
- Retraction needs rethinking. Pressure in the melt zone behaves differently when the material is expanding, and settings carried over from plain PLA usually leave blobs.
What LW-PLA is genuinely bad at
- Dimensional accuracy. Wall widths depend on an expansion ratio that depends on temperature, flow and speed. Precision fits are not what this material is for.
- Surface finish. Foamed surfaces are matte and slightly textured, and they never come out glossy.
- Strength. Tensile between 10 and 35 MPa, and the number depends on how much you foamed it. Strength-to-weight is excellent; absolute strength is not.
- Small, detailed parts, where expansion swamps the feature size.
- Being trusted straight out of the box. Uncalibrated, it prints badly enough that people conclude the spool is faulty.
- Cost. $40 to $70 per kilogram, though a kilogram makes three times as many parts, which changes the arithmetic considerably.
Where it wins
Radio-controlled aircraft wings, fuselages and control surfaces — the application it was designed around and still the dominant one. Lightweight props and cosplay armour that has to be worn for hours. Drone bodies. Large display pieces where postage is charged by weight. Insulating or sound-damping parts, since a closed-cell foam does both.
For anything where the mass of the part is a design constraint rather than an afterthought, nothing else on a desktop machine comes close. For everything else, plain PLA is cheaper, easier and more accurate — which is what the comparison between the two works through in detail.