Printing accuracy is 5x lower than the published specifications

[huguescmoi 1]

Hi,

I did the cube test to see what is the printing accuracy in X, Y, Z axis. According to the Ultimaker 2 specs, they should be 12,5 micron for X, Y axis and 5 micron for Z axis.

I printed a 10 mm cube in regular PLA at the highest quality possible in Cura. I then measure with a high precision digital caliper the distance between two faces of the cube (measured in the middle of the faces) in all three axis.

I got the measurements : 9.94, 10.02, 9.94

So that is a 60 micron difference with the baseline, or 5x worse than the published specs.

I knew manufacturers were often lying on their published performance specs but I did not imagine it was 5x !

Has anyone tried this test and compared it to specs ?

Edited April 16, 2016 by Guest

[neotko 1,362]

Plastic srinks as cold down.

Welcome to FDM

[IRobertI 432]

Like neotko says, the plastic will shrink as it cools. Different plastics will shrink a different amount, different geometries/temperatures/amount of infill/filament diameter/ambient temperature/etc etc etc will also shrink or affect the size in some other way. The numbers listed are for the physical movements of the machine, not what the plastic piece will be.

[avogra 68]

I'm not sure, how accuracy has to be read when looking at manufacturing (or other) machines. My feeling says, it means how accurately the printhead and buildplate can be positioned, not how accurate the product will be. Usually the latter isn't under control of the machine's manufacturer as material, settings etc. play a major role. I think alone the surface roughness is more than 12.5 micron in most cases.

But even then, I would be surprised, if Ultimaker could guarantee the mentioned accuracy for position of head and bed. Maybe they mixed it up with resolution?