I would like to build a standing shelf where the supports which hold each successive plank are 3D printed (to obtain special shapes).

However, I read that PLA flows under constant stress/pressure. Still, this doesn't stop the author of the article from using PLA for a hanging shelf, which obviously is subjected to constant negative pressure.

Which material suffers the least from plasticity/non elastic deformation under stress?

Answers with data for multiple materials are welcome.

I found that the phenomenon is called "creep" and is related to ISO 899, but I couldn't find any data for the common filament plastics and I don't know the theory behind it, so I'm not sure whether it's unavoidable or it appears after a threshold stress is reached.

Information: it's a living room shelf which will hold books and other stuff and is supposed to last a decade. I will of course add a safety factor and I could even fix the planks to the wall (in hidden places), but ideally the 3D printed material should have NO creep.

  • 1
    $\begingroup$ Would you be willing to spend money on an abrasion-resistant nozzle so that you could use carbon fibre-reinforced filament (which also costs more)? $\endgroup$ Commented Jan 2, 2020 at 19:07
  • $\begingroup$ I was thinking about non-reinforced filaments, but if you can find creep data and/or how to calculate a threshold to completely avoid creep for reinforced materials, I think the answer would be fine. I don't know how creep works (is there a threshold? is it unavoidable for plastics?), but if a reinforced filament is the only way, or if a pure filament requires twice as much material to avoid creep, a reinforced nozzle is definitely not an issue. $\endgroup$
    – FarO
    Commented Jan 2, 2020 at 22:55
  • $\begingroup$ Potentially this question could be moved to the Chemistry section of SE. $\endgroup$
    – FarO
    Commented Jan 3, 2020 at 13:55
  • $\begingroup$ I don't think it would be a good fit for Chemistry.SE as it is about the mechanical properties of the material. As I don't have much experience of 3D-printed items under stress, I can't really give an answer. Maybe you could print one support which will take the maximum forces and keep it loaded at three times that for some time. If it fails in a month, you know the material wasn't suitable. Or just print new parts every time the old ones are looking suspicious. Or there's lost-PLA casting for metals. $\endgroup$ Commented Jan 3, 2020 at 14:10
  • 2
    $\begingroup$ @AndrewMorton you are right, Engineering.SE would be better. Related question (unanswered): engineering.stackexchange.com/questions/19111/… $\endgroup$
    – FarO
    Commented Jan 3, 2020 at 15:55

1 Answer 1


Your question can not be answered theoretically -- only empirically.

You need to print some trial brackets with materials of interest and measure them. The question is: under what conditions should you test them?

The problem is that creep involves both compression and tension, and the behavior may be different. It is also impossible to fully translate material specifications into component behavior without a really good model that includes the details of the infill, adhesion to peripheries, and all the microscopic detail of a real 3D-printed part.

The problem with typical PLA may be the low temperature. Raising the temperature of a normal PLA printed object to 160°F (70°C) softens it to the point of nearly being limp. I have used this for force-fitting PLA parts by warming a pot of water and placing parts in it for a few moments. That temperature is hotter than your room, but a hot summer day in the sunshine could soften the part to the point of failure.

For anything load bearing, I would want a material with a higher plastic temperature.

There are PLA formulations which are annealed after printing. This is claimed to allow the PLA to slowly recrystalize and become both stronger and usable at higher temperatures. I don't have experience with this.

Depending on your printer, you can also consider using a higher temperature filament, such as ABS or PC (polycarbonate). PET-G is a little better than PLA, but it softens at a lower temperature than ABS.

As important as the material itself is the anisotrophy of the printer parts. Be sure to print the parts so that the major stresses are along the strongest axes, typically X and Y, and not along the weaker Z asis. Choose your infill to contribute to the strength, and use plenty of it, or design the part so that the infill is not intended to contribute to the strength.

  • $\begingroup$ mind using °C too, as printers are operated using °C usually? °F is not used in most countries worldwide. $\endgroup$
    – Trish
    Commented Jan 15, 2020 at 17:46
  • $\begingroup$ I know, and usually C is used, but in this case, because it represented the thermometer I used for the purpose, and I was using temperature below the boiling point of water, I thought the F would be appropriate and stick out as being an unusual temperature for 3D printing. I can edit in the C values. $\endgroup$
    – cmm
    Commented Jan 15, 2020 at 17:51
  • $\begingroup$ Thanks! the main benefit Celsius has, it is easily transferred to Kelvin. $\endgroup$
    – Trish
    Commented Jan 15, 2020 at 17:57
  • 1
    $\begingroup$ I had one of those Kelvin thermometers, but it stopped at zero. I figured it must have been a bad design. :) $\endgroup$
    – cmm
    Commented Jan 15, 2020 at 17:59

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .