This topic contains 31 replies, has 6 voices, and was last updated by jhr 1 year, 3 months ago.
-
Posts
-
As mentioned, just asked in the Cura forum if any Ultimaker owner ever has spotted the rims. Thrilling. 🙂
It could be a CURA issue and not showing up on other printers. To get the effect you are getting is in the fine setting. (20 micron) There are other printers that can do 20 micron but have bigger nozzles. (.3 vs .4 mm) Small changes can make noticeable differences when dealing with these small dimensions. We are looking at a 1, 2, maybe 3 micron difference. You may not see this with other printers if for no other reason than a different nozzle size or different extruder. Using one of your printers where the effect is intermittent, try printing the same sample, several times, but with a 2 percent increase/decrease in size. If it is a CURA rounding issue it should fade in and out with changes in size.
So far just one reply in the Cura forum, but that guy claims it is unknown to Ultimaker users.
Note that the picture of the traffic pylon is not by me, but by CEL themselves, which is an indication that the rims appear on some?/many?/all? Roboxes. I could imagine that some users haven’t noticed the rims yet, either
- because they don’t have the required eagle eye,
- it didn’t jump into the eye because of the shape of their printed objects or
- because their Robox hides the rims under the described mechanical jitter.
I just wanna make sure that CEL becomes aware of the fact that there’s a problem here – although a literally small one – which negatively impacts the print quality. So far it seems they haven’t realized this: neither the discussion I had with Simon Coleman in my related Freshdesk support ticket indicates that they did, nor the fact that they posted that picture of the – otherwise really excellent – dual-color pylon.
This looks to me like a very slightly bent z axis screw. The robox has a 0.5 mm pitch screw. Which equates to 5 layers on fine. Basically the z axis threaded rod is pushing and pulling the gantry just a little.
OK. I finally got chance to print out your little house and here’s the result:
It is done on green Kickstarter edition of Robox, AM 2.00.03, fine profile with ColorFabb’s black PLA/PHA filament. It seems not to have those 5th layer anomaly.
-
This reply was modified 1 year, 4 months ago by
click.
@click that looks like the normal setting 200micron. The screw pitch is 500micron so thr pattern would be 10 layers high on that.
A good test might be to set the layer height to 125micron. That would change the repeat to 4 layers instead of 5 if the problem is caused by the z screw.
I bet you are right. I was sure I’ve picked Fine, but probably picked Normal automatically :(. Started printing it again!
I have found several artifacts in printing parts.
- There is a vertical pattern of approx 2mm in vertical plane and inline with the X & Y axis’s. This may be due to the belt pitch, belt tension or infill pitch. It also seems to be affected by the peripheral speed. I still have to determine exactly the cause of this. The exact amplitude has still to be determined but not large but visible, will need to investigate further.
- When printing a taper which is slightly of the X and Y axis but rotates around one of them (approx 1.6degrees), in the vertical plane get a circular pattern that can be seen. It also is not large but occurs at roughly 0.1mm intervals. I believe this to be a CAD to STL issue as the CAD I am using (Geomagic Design) gives resolution options when exporting to STL file. The resolution options seem quit coarse before the file size exceeds the the maximum that AM can accept. Again this is a low level and I estimate <10micron level.
- It is also possible to see the individual stepper moves. Be it X-Y this is extremely minor estimate it at <1micron.
- The main point. With a default fine profile there is a 5 layer artifact. It needs further investigation. dr-woo suggestion of a cube is good. This should show if the gantry moves due to Z screw concentricty and give a fixed offset. If its a another effect then the offending layer will be different e.g. be larger/smaller around the 4 sides of the cube. Need to look at this in detail. Normal profile seems to give much more random layer effects.
Sorry if the above seems long winded, I am currently working on producing casting patterns. Casting houses have a bad feeling about 3D printed plastic for moulds however with the Robox I have achieved a result in that the one I am currently using tell me the patterns I have mode by 3D printing are the best they have seen. The initial patterns were relatively simply the next are very complex and I need to minimize printing artifacts where I can.
Secondly I have had some professional SLS parts made through company I am working for and the surface finish is inferior to the Robox. I think its a great machine but need to iron out these blemishes.
I have not supplied any photos due to difficulty of getting ones that clearly show the issues. I am working on this as well
@hagster A bent, not perfectly centered/not perfectly perpendicular Z axis thread leads - at least to my understanding - to the jittered layers (as in the left house of the following image). The 5th-layer-rim (as in the right house) goes ALL AROUND the print, so it is not a shifted layer, but rather a squeezed layer. The 5th-layer-rim can’t be explained with mechanical problems IMO, because of the repetitive, mathematical accuracy with which it reappears.
@john Let’s please stick to the repetitive 5th-layer-rim issue, which is what this thread is about.
And don’t compare the Robox to SLS or SLA/DLP printing. Those are completely different techniques, which do have different, often better surface finishes, but comes with other downsides, such as limited choice materials, and much higher operating costs, not to speak of much higher price points of the printer itself.
You can and should compare the Robox only to other FDM printers, or as such, the Robox makes a pretty good job. Except for the jitter I’ve seen on many machine and except for the 5th layer rim in Fine mode.
OK. Here are pics of little house done in fine settings:
I am not sure what to make of it.
-
This reply was modified 1 year, 4 months ago by click.
@click Your print sort of clearly shows the 5th-layer-rim issue. Look closely at the roof. The black glossy filament is not the best choice to visualize the effect, so is my house model. I’ll test print and upload a simple cylinder later on…
You must be logged in to reply to this topic.
