Object dimension accuracy in printing

It seems that the prints tend to be slightly smaller in dimensions compared to the dimensions in the file. Is there a way to adjust the print scaling so that the print will be exactly the size the object is planned.

The problem is most serious.
There is not a simple problem of scale.
In my case I find that the outer dimensions are almost correct while inner dimensions are smaller by about 0.6 mm (as absolute value).

As mentioned, simply scaling the object may or may not help you. But on the first tab on the right hand panel in Repetier Host, object placement, there are two triangles one inside the other. Click this and you can scale your object. I usually find PLA shrinks by less than 1%, so scale up say .01.

The dimensional accuracy of the Vertex 3D printer is a problem.
Of course i understood that scaling may or may not help.
I am asking at Velleman if there is a real solution.

I was wondering if it really is a printer problem? Could it possibly be a slicer problem? I mean does it have something to do with the slicer settings like wall thickness or something? If the printer is told to print for example 20mm x 20mm box (outer dimensions), should the box corners to be created in G-code according to coordinates x 0,175 y 0,175 and x 1,825 y 1,825 due to the fact that the nozzle has 0,35mm diameter?

I have found the same. To be frank, i never expected a H tolerance fitting etc, and was pleasantly surprised by the accuracy of the machine. Inside diameters are the most inaccurate but that’s logical when you look at the process being used. Paint something round whole applying a lot of paint and you’ll see the same. I just take notice of it when I design things and if I need a tight fit I’ll machine it afterwards.

I have the same problems. I printed a small wrench for nut:

Drawing with Inventor

Measured nut

Printed too small

I have to scale the object to 1.06 (6%) and now is OK.

Any better sollution?

I have had to scale only from 1.006 to 1.010 like 1% or less 1.06 seems quite a lot (6%)?

Apart from the slicer issue, the flow rate is an important parameter as far as accuracy is concerned. Higher flow mean more material deposited for the same final layer height meaning increased width.

I just think that it is the wrong way to design 3d objects according to slicer limitations. I think the design should always have the object original measures and the slicer program should be able to figure out how to move the print head to get the result in the .stl file. If CNC router is told to carve piece it takes in to account what tool is used is it 2.5mm drill or 5.0mm. Then the code is created according to the used drill.

I think this is definitely a slicer problem…Printer does what it is told to do. But how you tell to the printer do something is the key in this case…

I think the shrinking of the PLA does not explain 6% class errors, perhaps 1% max…

I am new to 3d printing as I assume you are by your comments. If I contrast 3d printing to the process of removing material then when I ‘tell’ my mill or lathe to remove material then, assuming no backlash etc., the bed in the former and the cutter in the latter simply follow the directions of movement with no degrees of freedom. With an additive process; like 3D printing the extruder follows the path the slicer has defined in the g code in exactly the same way but there are more degrees o freedom. For example, the temperature and the flow rate will both influence the volume of material that is extruded. Both of these will influence the dimension of the object printed. The increased degrees of freedom are inherent in an additive process and not a fault of either the printer or the slicer. I suggest you experiment with the variables under your control until you find the parameters that produce the results you want.

With respect to the issue of shrinkage that you refer to then if you take into account the temperatures involved and the relative coefficients of expansion of PLA and ABS then you might expect shrinkage in general of .1,62% and 1.48% respectively. But if you reflect on the process of slicing then it will become apparent that exactly what shrinkage you get will depend upon whether the critical dimension is printed vertically or horizontally.

I just checked the g-code creation with Simplify3d. According to the preview it looks like it does it right even if the nozzle diameter are changed. I tested the calculation with a high resolution .stl washer that has 10mm outer diameter and 5mm inner diameter. I have to test print the object tomorrow and see how the diameters will become in reality. So maybe it is not a slicer problem? Hmm…

I also found this video:

https://www.youtube.com/watch?v=5mjR6R1wWmQ

I wonder does this equation work? Nozzle Diameter / Actual Wall Thickness = Extrusion Multiplier

Though I don’t see the Extrusion Multiplier field in Cura, only in Simplify3D???

I think it’s a hardware problem.

I use a Kossel mini and a Prusa i3 3D printer also, and they do not have the dimension accuracy problem. I usually slice in Cura just as the repetier host for Vertex do.

Then it could be fixed with new firmware perhaps?

It would be interesting to have some comment or advice by a Velleman support.

I think Velleman could release optimal parameter/software/configuration setup and a test object how to print dimension accurate objects when K8400 printer is used with Velleman standard PLA. Velleman should know the best parameters for standard setup.

-Slicer, all parameters?
-PLA colour etc?
-Firmware?
-Printer, setups?
-Test object, dimensions?

Is there an update on this issue?
I think everyone is struggling with this.
Has anyone checked if Simply3D has the same problem or is that software able to create the correct object?

Thanks
Jan

Inner and outer diameters mostly have the same error, so I don’t think it’s a scale problem.

I think playing around with the flowrate could resolve some issues?

Any reaction from velleman on this?

Apart from this issue the K8400 is a perfect printer.
This could be a showstopper for people who need accurate holes.

Jan