www.structural.de > 3D-Printing

Structural assessment of 3d printed functional parts


3D printing in FFF / FDM technology with plastic material is now affordable for everybody. The question arises, weather those 3d printed parts may be used as parts of load-bearing structures, especially for small structures in fair trade stalls, event construction, membrane constructions, in larger artwork, decorations, or for museum furnuture. Here quite often is a need for small unique joints and connectors, which are not available in the market. Forces are mostly low, and life time is often restricted to some weeks or month, which allows to ignore most considerations about environmental exposition, creeping and aging of the plastic material.
structural.de works since January 2018 at FEA based and methodically sound design methods for such parts. In litarature and research, not many results are published yet. Serios filament manufacturers like Ultimaker, Polymaker or Formfutura do issue data sheets with tensile strength values, but all those values are tested in XY plane, means in printing layer directions. Data tested perpenducular to the layer planes, i.e in Z-direction, are not available. Only few companies, e.g. Stratasys and Polymaker, have published its strength values for Z-direction as well.

Standing printed specimen for tension test of ultimate Z-stresses

It is widely known, that the tensile strenght in Z-direction perpendicular to the printing layer plane is often much less than the one in XY-direction. 3d-printed parts in FFF / FDM plastic technology show a very clear orthotropic behaviour. The problem is, that many functional parts have stresses in many directions, and it is only in few situations possible to turn the part so that major stresses run in the plane of the 3d printing process. So, we must live with the fact that major stresses will most likely occur also in Z direction, even if we try to optimize and even if we do the design stress check in FEA with different stress limits according to the direction.So, the limit stress in Z-direction will finally determine the usability of a material for load-bearing functional parts. 

Test device for small tension tests up to 2 kN

Therefore, hpl structural has done quite a lot of tensile strength tests for many different filament types in both XY direction and in Z direction. From the 5% quantils, we derived limit stresses for FEA based analysis methodes, based on typical safety factors used for plastic material. The tests focused on materials which are easy printable with common 3d printers with normal nozzle temperatures up to 280°C. "Rocket-science" high-temperature materials like PEKK and PEEK have been left out. Those materials have, by the way, its advantages especially under high temperature conditions, but are not significantly stronger. Also the promising metal-based laser-melting and laser-sinter methods are still too expensive for our field of activities and had (yet) to be left out. We do test ULTEM 1010 currently, but without further hope of a widespread successfull application of this expensive material in FFF printers. It just prints too ugly, warps too much, and strenght values are just average.
With the ultimate tensile stress obtained from the tensions tests, we are able to design the functional part as STL file, do the structural analysis directly in our Strand7 FEA software, validate the limit design stresses for the particular printer and material with the test specimen, and finally produce the functional part directly on the desktop, at least for test purposes.In this process, those materials turned out to be most promising, which are easy to print, have a good melting behaviour, stick nicely to the printing bed, have a good layer-to-layer adhesion, show no warping - and - most important - show a ductile behavour (plasticity) before breaking.

In terms of tensile stresses and ductility, we were very pleased with 

Polymaker PolyMAX PLA (Base: PLA) (as all PLA restricted to room temperature conditions)

Polymaker PolyMide (Copolymer PA 6 / PA 6.6 „Nylon“)(indoor and outdoor use) 

materials are very ductile at least in its XY printing plane and have a relatively low difference between the tensile stress limits in XY directions vs. Z-direction. Interestingly enough,good printability turned out as a sign for good layer adhesion and good Z-.strength.

But also the following fiber reinforced materials have impressed us, since they are very stiff (high Elastic Modulus) and have a very high tensile strenght in XY-direction:

3DXTech CarbonX
Nylon CF, rigid.ink Carbonyte and DSM novamid 1030 CF (Carbon fiber filled Polyamid)

Owens Corning / XStrand  GF30PA6 (Glass fiber filled Polyamid)

fiberthree F3 PA CF pro (Carbon fiber filled Polyamid) - with also a very good Z-strenght

Those materials did not only show high stiffness and a very hight tensile strength in XY-direchtion, but also surprised with a very low warping even with 100% infill, which favourites them for high-performance parts acting like a plate.  

But even with those high-quality materials, and considering the common safety factors for plastic given by literature, the magnitude of the usable stresses is only about the same as of timber (5..16 MPa; SLS). This is admittingly not very much, but consider - you have a material without limits in geometry, and you can produce the part right now on the desktop!

Small 20mm² test specimen as STL-File

Test results  of 20mm² small specimen (Filament tensile strength results)

Filament manufacturers or users of filaments suitable for load-bearing functional parts: Please feel free to send printed specimen (use STL file linked above) of your filament (minium 5 specimen in XY plane printed, minium 5 specimen in Z-direction printed) to hpl structural. Infill: 100% at +- 45 Degrees.;  Bottom and Top Layer: 1 Layer. Edges: 1 Line. Nozzle diameter 0,4 mm. We will test the specimen and add the results to the table.

Small Specimen for strength tests

A very interesting alternative to plastic is the idea of 3d printed parts used as models for metal casting. Even with common cast aluminum alloys, the usable tensile stresses raise up to 10 times the value of the plastic part. And - for aluminum, the Eurocode 9 (DIN EN 1999) gives the engineer an officialy acknowledged code of practice for all the design work, also for large load bearing structures. All the known problems with plastics like aging, temperature and UV light exposition and  creeping are simultaneously vanished. For this reason, hpl structural tests a handy technology chain in cooperation with a local foundry. 

Our present design method results for the plastics in a quite low usable stress level, due to the required safety factors for plastic and the limits of the material. The distance of the usable stresses to the breaking stresses is quite high. Therefore, we expect only a very low influence of creeping and aging for our purposes, as long as the ambient temperature is significantly lower than the vicat temperature of the plastic. To back this theses, we started middle term loading tests with typical stress levels of 5 MPa for periods of one to three month, the typical load duration times for fairground and event structures.

A second test device allowing tension forces up to 20 kN will help to verify the theory at test part which are similiar in shape and size to the real structural part

Second test device for tension stress tests up to 20 kN

structural.de uses the following printers for the test specimen and functional test part:

- Malyan M200 / Primacreator P120 with BuildTak printbed for small PLA and PETG parts
- FelixTec4 Dual Extruder (
www.felixprinters.com ) with BuildTak printbed for PLA; PETG an non-problematic PA
- Anycubic i3 Mega with Micro-Swiss wear-resistant nozzles and Ahltec Carbon Fiber printbed for fiber-reinforced polyamide
- Anycubic i3 Mega with Zortrax M200 Perforated Printbed for all strongly-warping, adhesion-problematic materials
- Anycubic Chiron for large-scale demonstration models in PLA
- CreatBot F160 (Peek version) with perforated GFK board + Polysmooth, see below) and chamber heating for high-temperature materials  

By the way - how we print ULTEM:
ULTEM seems to adhere to nothing. ULTEM does not even stick to PEI, though it is PEI. Here one possible way to succeed:  Get a perforated glass-fiber (GR4) board from
www.digikey.com . The type you need is Twin Industries 7100-1010. The boards have the size of 254 mm x 254 mm x 1,76 mm and can be lasercut to fit to your printers printbed.  Sand the boards with fine sanding paper and clean with alcohol afterwards. Then, put them into your printer as printbed and level nicely.  Then, print a thin layer (e.g. 0,2 mm) with Polymaker´s "Polysmooth" full area as a binder layer. Choose extrusion multiplier in your slicer larger than 1 (1,25..1,5) in order to extrude enough material which will fill the perfroation holes. The perforation filled with filament will greatly improve the adhesion to the perforated board. Then, level the printbed again very carefully and rather sharp for the surface of the Polysmooth layer. Your printer is now ready to print ULTEM (if he is able to reach 375°C at the nozzle). We experienced superior adhesion for ULTEM 1010 by 3DXtech with 375°C nozzle temperature and 90° bed temperature with 20 mm/s speed in the  CreatBotF160 printer, which we improved a bit: The parts cooler (not needed for Nylon and PEI) was redirected to cool the extruder stepper motor instead of the part.  Works great at least for our small specimen objects!  The finished part can be cut off the bed with a good knife. After some prints, the perforated board must be cleaned totally with alcohol to get rid of the Polysmooth and can be prepared again.

- 2019 -


  zurück zu: www.structural.de