Undersized holes

[Leeway]

Depending on the part and material, drilling is not always an option in plastics. Case in point I have one production part that uses 2 1.5" by 5/32 steel roll pins to add strength. It's a fairly thin part, so requires the backbone. I seal it in with the same Pet G filament in a 3D pen. Now I get steel strong good looking 3D printed parts in a production setting for very little cost really. I had to modify the drawing, because a drill bit destroyed the part with heat. Uncontrollable really. Change the size. Then I was faced with S3D not giving enough material around the cavity. Well fiddle some more with the settings and we have a good part. That is where S3D is for me. A good manipulative tool that needs work. But they all do. Some are better than others now at particular things. None are the ultimate printing experience. For paid software, S3D needs to continue to update and include new features no matter how meager to keep enjoying the benefits of paid software and being the best at what they do right now. Just a users opinion, but I would not have the product I have today without listening to my customers wants and needs. It kind of cascades downstream. If S3D makes a better update, I will be able to provide my customers with a better quality product.
Right now it's all about what can you make the current tech do to meet your needs. So, the more work arounds that you are aware of, the better print you might get.

[upsm]

@Leeway, well there will always be a corner case where "nothing works" .. you can't fight it.. sometimes you can design around it, sometimes you have to lie and cheat to get what you want, edit g-code manually... I doubt there will ever be a slicer that will solve all cases with all materials.

What I do get very upset about is issues like this one with undersized holes that's here "forever", that is solved on professional machines. The same UP Plus2 I'm using with Simplify3D, if I use it with their original (closed source, works only with their machine) slicer the holes are perfectly sized! So it can be done. How come TT have this figured out I don't know; I assume it has to do with them making the $500,000 printers too so they have some experience in controlling them. I printed some stuff almost 20 years ago on stratasys, didn't need to drill the holes, they were perfect. The stratasys from 20 years ago uses hotend that's even less precise then what modern hotends we use on our printers do, so again it is possible. The question is when will we crack it not "will we" but still, reprap project is entering second decade and we still battle with hole sizes

Second thing that gets on my ganglias big time is "hiding parameters" all slicers do. Look at the comment in g-code Simplify3D generates. The slicer "knows" it's starting overhang, or base, or perimeter, or whatever. Why can't we set "any" parameter based on that too? Why not allow user to set "any parameter" for "any" section. I want for e.g. to change temp for support. Or I want to print only outer layer with extruderX and everything else with extruderY or ... like who cares, if you use/calculate parameter - make it changeable trough config but in order to hide "too much options" from users they all decide to make only the basic options available and then "tweaking" and "working around it" to solve "corner cases" gets difficult

One other thing on the original question. It's decently hard to figure out something is a "hole" from STL file alone. It's just a bunch of triangles and one would have to add some "pattern recognition" into code to properly solve it. Reading the "proper" 3D solid files would be, in my opinion, much better solution that could produce much better results.

[Leeway]

Agreed. This is the kind of thing I would be listening to if I were the boss. Some software does give the option to use newbie, normal and advanced settings. Something like that might work well here in S3D. Support is certainly an issue for them, but as forums go, this one is okay. Better than most. Lots and lots of answered questions to common issues. If anyone would jump on the issue, I would expect it to be S3D. People don't like paying for software that is just like free stuff. The premium cost can't just be the slick user interface which is what it seems like now. I have tried a few slicers and I do like S3D. Just hope they do listen to customers and do something about issues like this one. Parts will always have holes. Holes generally require a tolerance. Pro grade software should achieve inner and outer tolerances in one punch. As mentioned, it's not the hardware.

[dkightley]

Please don't accuse me of jumping to the defence of S3D's producer.......but I have to pass a comment here....

Parts will always have holes. Holes generally require a tolerance.

You're quite correct that holes need to have tolerances.....but surely isn't this a function of the design of the part? If I were designing a part that needed to have a hole in which a shaft needs to freely rotate....then I would design the internal diameter accordingly. A 5mm diameter shaft will not rotate freely in a 5mm hole!...unless you're talking high tolerances, when the manufacturing process will demand the hole be finished using say a reamer!

Similarly, if I were designing the hole to allow a shaft to be an interference fit, then I would also determine the internal diameter at the design phase....and dimension the hole to suit - taking into consideration the material and manufacturing process used!

I suppose what I'm really saying is that whatever component part is being designed, then you have to take into consideration the limits and restrictions of the manufacturing process being used. If that means designing a hole 1mm bigger so it ends up the correct size, then so be it!

I'm sure many think nothing of rotating a part on the printer bed so it prints okay.....or splitting the part into two, etc This business of hole diameters....to me....is just another thing that you take into consideration in the design phase.

[Jeff_Birt]

I think your expectations on precision have to match the limitations of the process. Even when doing a design for injection molding plastic to have to keep the process I mind. When the mold is make you just don't say, 'the plastic will shrink by 3% so we'll make the mold 3% larger'. Depending on the mold design and part design areas will cool at different rates and might produce undesired affects.

When 3D printing with the FFF process you have plastic cooling at different rates too, i.e. two areas with the same volume of plastic might have different surface areas and will cool at different rates. A hole in one area surrounded by a large mass of plastic might wind up a slightly different ID than the same sized hole elsewhere. Then there are machine differences, filament differences, changes in room temperature, voltage fluctuations, etc. There are innumerable variables to try and account for to make the print 'perfect'.

Part of the design process is accounting for shortcomings in the manufacturing process (no matter what manufacturing process.)

[Leeway]

Right. When everything else is working okay. tough it out on the holes. That is what we do. No software adjustment. It's called a second OP. An undesirable option if possible.'

[upsm]

Please don't accuse me of jumping to the defence of S3D's producer

I didn't notice anyone was accusing Simplify3D producer? The undersized holes is a general problem all generally available slicers have, Simplify3D is not better nor worse in that regard so no accusations needed. Maybe this discussion evolves into solution, that would be nice, but no accusations for sure.

You're quite correct that holes need to have tolerances.....but surely isn't this a function of the design of the part? If I were designing a part that needed to have a hole in which a shaft needs to freely rotate....then I would design the internal diameter accordingly. A 5mm diameter shaft will not rotate freely in a 5mm hole!...unless you're talking high tolerances, when the manufacturing process will demand the hole be finished using say a reamer!

If the result of the holes would be +-0.5 or +-1mm or something like that then I'd agree with you, but it's not. We are not talking about "tolerances" here, we are talking about miss-generated G-Code that produces holes that are seriously out of spec. 3mm hole will often be under 1mm, 8mm hole will be under 5mm etc.. and different extruder properties with different plastic properties will make this different.

Part of the design process is accounting for shortcomings in the manufacturing process

Yes, only in this case, the manufacturing process is not limiting the size of the hole. Example - TearTime UP Plus 2, original cpu with original slicer (closed source), you give it part - part comes out dimensionally perfect. There's a tolerance of course, that's the limit of manufacturing process, but we are talking ~0.2mm here so both inside openings (holes) or outside dimensions will be as they are in STL with some tolerance (under 0.2mm) and it will not depend on the hole size, so 1mm hole will be 1mm +- 0.2mm, 3mm hole will be 3mm+-0.2mm and will, for sure, not be 1mm or 0.7mm instead of 3mm. How can they achieve that? No clue! But they obviously can! The exactly same machine, only cpu replaced to run open source firmware with G-Code generated by Simplify3D/Slic3r/Cura/CraftWare and 1mm hole is not a hole at all and 3mm hole is printed as 1mm. The manufacturing process is 100% same, the only difference is the software.

So these "undersized holes" are not "manufacturing process limitation" but a "software limitation". Yes one can work around it, it's not even hard to work around it; and to be honest there's 10 pages of feature requests and at least 3 bugs I'd like implemented/fixed before I'd even suggest Simplify3D devs try to tackle the undersized holes problem. So, yes, in my opinion, this is super low priority issue, but one should not just write it off as "limitation of manufacturing process", as the solution exist, we just need to find it.

[brian442]

We are not talking about "tolerances" here, we are talking about miss-generated G-Code that produces holes that are seriously out of spec.

I have checked several gcode files myself, and S3D generates the EXACT coordinates that I would expect for these internal holes. So I haven't found any reason to think that the gcode is actually incorrect. If you find otherwise, please feel free to post your factory file and the exact line in the gcode file where you are seeing the problem. With every file I have looked through, the coordinates are EXACTLY what they should be.

That being said, you have to keep in mind there's lots of other factors that can screw up your tolerances. Things like over-extrusion, first layer being too close to the bed, even poor STL file exporting from your CAD package. So make sure you check these things as well.

[Scott_M]

I have checked several gcode files myself, and S3D generates the EXACT coordinates that I would expect for these internal holes. So I haven't found any reason to think that the gcode is actually incorrect. If you find otherwise, please feel free to post your factory file and the exact line in the gcode file where you are seeing the problem. With every file I have looked through, the coordinates are EXACTLY what they should be.

That being said, you have to keep in mind there's lots of other factors that can screw up your tolerances. Things like over-extrusion, first layer being too close to the bed, even poor STL file exporting from your CAD package. So make sure you check these things as well.

Checking the G Code was the first thing I did as well, and the toolpath IS where one would expect it IF you were using an endmill, but we are not, we are extruding a flexible string of plastic that does not do what an endmill does. So the path needs to be adjusted to take this into account. And it does not do this. That is the problem.

I just can't understand why so many people are defending a known issue that is a problem. Yes I will work around it until it is fixed, But it still needs fixed ! And the longer people accept it the longer it will take to get fixed.

Scott

[upsm]

Checking the G Code was the first thing I did as well, and the toolpath IS where one would expect it IF you were using an endmill, but we are not, we are extruding a flexible string of plastic that does not do what an endmill does. So the path needs to be adjusted to take this into account. And it does not do this. That is the problem.

amen to that