Ability to shift infill pattern in cura

[mburrows 0]

Hello Ultimaker Community, and especially the developer of Cura!

 

It is 1 week ago since i have ordered my Ultimaker V2 Printer, and I can't wait until it will be ready for me to pick it up.

 

Meanwhile I am heavily constructing some things in openscad and slicing the models in cura. My aim atm is to construct & print a specific rotor, which is modeled perfectly symmetric, and i want to rotate it at high velocity, probably with my dremel tool or a salvaged motor.

 

But the toolpath view in cura shows me that the infill is not alligned to the center of the buildplate, and so the symmetry in my object is gone, and it would probably wobble when rotated at high rpm's. I have a screenshot that illustrates this issue:

is it possible to "shift" the infill pattern, so that the crosses would go exactly through the center of the buildplate ?

 

thank you in advantage,

 

mburrows

edit: thank you Didier for posting my image first ;-)

 

[Dim3nsioneer 558]

Daid will certainly give you a more detailed answer.

AFAIK, shifting the infill is not possible. However, for your application, I recommend you use 100% infill, then you should not have any problem with inbalances as everything is filled equally.

 

[DidierKlein 729]

This is the picture you're trying to post (just click on My Media icon to select it).

I would go with dim3sionner's solution too. And probably not print it as showed on you screenshot, the object will probably move to much will being printed

 

[gr5 2,224]

I think you will need some support also to keep it from falling over. Maybe model the support in cad - maybe every 2 inches it could use 3 supports that you cut and sand off later.

Also try any infill > 25% - for example try 50% - I think you will find that is more balanced/radially-symmetrical.

 

[mburrows 0]

Thank you guys for all your useful tips, and quick answers. You are a relly helpful and supporting community :-)

I was already concerned about the instability of this particular model, so support may be a good idea, but i thought i might get away with just low % infill.

I have the idea that maybe a slight changing of the dimensions of the buildplate could solve my problem. I will try that.

Anyways i have added my issue on github, maybe its just a simple fix, and Daid can make me happy soon.

good night, and happy printing :cool:

 

[mburrows 0]

Add: I tried out changing the dimensions of the buildplate, and it did indeed change the infill. Unfornately the exact position is somehow difficult to predict and i suspect it would be different values for every % of infill.

 

[DidierKlein 729]

Not sure about this but if you change the buildplate size in Cura you might smash the head around the printer...

Just add some infill it will help you with that issue

 

[nallath 1,124]

Your best option would be to set your infil to 0% and increase your wall thickness to 1,2. This will created pretty strong prints that are quite light.

 

[mburrows 0]

@nallath

This is also worth a try, and maybe even better than shifted infill. Thank you. I will report back, what worked best, when my printer arrives.

@didier: the change of the buildplate size was just for testing the infill pattern created. It just showed me, that the support structure is bond to the buildplate size. Maybe i will dig into the source code of cura engine myself, since i will have some spare time in some days. It should be easy to shift it to the middle, once i found the pice of code that computes the grid.

Maybe Daid can point me into the right direction, to where the calculation is done in the code.

meanwhile, happy printing

 

[Daid 306]

Infill is calculated with the front left corner as "center" point. (as that is actually 0,0) So the grid is rotated around that point. If you want the grid lines to match up with exactly the center of the model, you most likely have to select a proper infill% or tweak the machine size a bit.

But I cannot tell you exactly the values you will need. As it's a lot of "emergent behaviour" which results in the infill.

 

[mburrows 0]

But I cannot tell you exactly the values you will need. As it's a lot of "emergent behaviour" which results in the infill.

 

I think I have located the source in CuraEngine in "infill.cpp" in function "generateLineInfill".

I find it a bit difficult to fully understand how exactly you generate the infill there, because i do not get your main idea behind it. If you would just put some comentarys in there, i guess i would be able to find a solution to my problem on my own, given, this it is even possible.

thanks a lot, and again: I really do apreciate your responsive and open community here.

 

[Daid 306]

It's a bit hard to describe in words what I do there. Some images could do wonders there.

But the basic steps are:

- Rotate the polygons so the infill direction is on the X axis.

- Calculate how many infill rows there will be (min/max X position)

- For each line in the polygons, calculate the crossings with the X rows, and store that crossing.

- Now that we have line crossings for each row, you can do "even-odd" to get the infill lines (after sorting the crossings on Y position)

- Finally rotate the result back by the inverse rotation you used in step 1