nozzle size

[Tinchus2009]

Hi. I have been experimenting with the excel sheet. Then I went back to something more "practical" (I like to print) that also helps me to better understand and learn the use of full control.
I was looking at the intetrface here: https://colab.research.google.com/gist/fullcontrol-xyz/589c78de0093698a07ec724af6428f09/fullcontrol-lampshade.ipynb

I created a lamp shade, but I still dont understand how to set up nozzle size. In the scrip I can see how to set up wxtrusion width, but the extrusion amount calculation depends on the nozzle size, so how so I set this value or how is fullcontrol doing that calculation in order to create extrusion values on the gcode?
Here it is the ipynb I created just in case you want to chek it out.
Thanks in advance
fullcontrol_lampshade4_chica_redonda_suave.zip

[fullcontrol-xyz]

The nozzle size isn't used actually, since you can print all different widths and heights from a single nozzle size. FullControl basically takes the width and the height that you set and multiplies that by the length of each segment in the print path to calculate the volume of that segment. Depending on the printer that volume is then converted to a number which is that e number in G code. I've printed the exact same g code with nozzles ranging from 0.4 to 0.8 mm before and the parts coming out were basically identical. That works if you're printing wider than the nozzle diameter that you probably wouldn't be able to print a very nice line that's 0.4 mm wide from a 0.8 mm nozzle.

[Tinchus2009]

I understand, you are talking about using volumentric flow printing. But my problem here is: I did simple test prints, they came out "ok", but I measured the width of the paths and they were off by too much: using a 0.6 mm nozzle, and having set an extrusion width of 0.8 in fullcontrol, I got reauslt like paths having 0.7 in some prints and 0.85/0.9 ikn other nd the difference between them was exactly hwn changin nozzle from 0.6 to 0.8 mm. and that is what caused my original question here.

Volumentric printing is a way og generating a gcvode independant of material and nozzle size, I use that in many of my profiles, but there is still something missing for me here:

.- when you say full control calculates de volume and then depending on the printer the volumen is converted to a numer: the "printer" is set somewhere on fullcontrol configuration, because to get that number you still need the nozzle size, there is no way to get extrusion number without knowing nozzle size. If I want to do that in my usual slicer, I activate volumetric printing, the printer is set on volumentric mode, the firmware need to know the nozzle size and with that the firmware does what you do following your exaplanation..
.- Pehaps Im missing somethinf on the full controil configuration? Do I need to set up my printer on volumetric mode? I guess no because I can see on the gcode extrusion lenghts and not volumen extrusion

So probably Im missing some configuration where I set my kind of printer? I saw that there are some defaults, the one I shluld use a "generic" , but how I set the nozzle size for the "generic type" printer?

[Tinchus2009]

Ok, I went back to the excel sheet. I can see there that you are asumening on most printer you are using 0.4 nozzle, I can see that value set on some of the printer presets, but just as a comment and not as a value that user can set.
can also see tthat user on the excel he can choose between mm3 or mm as extrusion unit.
Si please correct me if Im wrong: following your previous explanation, if you calculate the volume of a path you still need the nozzle diaim to get the E value if the printer usus mm as extrusion unit. You wouldnt need it if the printer is set to volumentric mode.
And here it comes where this aprouch is failing for me: the material im using is pmma, it gets solid REALLY FAST. So fast that is problemtiv when you try to extrude more than 20% above nozzle diam (you get almost an inmediat jam).
The aprouch fullcontrol has works (actually worked very well on a ABS print I did) works because if you calculate the E number asuming a 0.4 nozzle to get that number and later you use a 0.8 nozzle, will work going to higher nozzle sizes, but not in reverse (I did a test using a 0.2 mm nozzle and filament was grinded since too much material was trying to be extruded through it and back pressure was clearly too big)
Also another problem is the ability to have precise control over path width wich also I have confirmed, because it is posible to extrude a 1.2 mm width path with a 0.4 mm? following the usual maths for extrusion the answer is no, but in practice it is possible but we all know width is not going to be real precise.
Can you confirm my guesses? If Im right, is is possible to set in the code the noozle size it it is just hardcoded?
I could make a suggestion here: in orfer to get rid of all these and probably simplify also the code of fullcontrol, you could just calculate volumen os the path and leave that on the gcode. The user then would just need to set on the start gcode the command to set the firmwware on volumentric mode, and the firmware will take care of this. But in my opinion, it would be just more simple to create a nozzle size param.

I did more test on the gcode generation (with the excel sheet):

This code generate volumettric extrusion code:
G0 X50.0 Y50.0 Z0.2
G1 X100.0 E0.8
G1 Y100.0 E0.8
G1 X50.0 E0.8
G1 Y50.0 E0.8
G0 Z0.4

If I change the extrusion units to mm ofcourse I get the mm extruded code:
G0 X50.0 Y50.0 Z0.2
G1 X100.0 E0.332601
G1 Y100.0 E0.332601
G1 X50.0 E0.332601
G1 Y50.0 E0.332601
G0 Z0.4

So by doing these 2 gcode geenration tests I think I can confirm the extrusion calculation is done asuming a 0.4 nozzle.
so for me, I can stay on the 0.4 and 0.6 mm just to be on my safe confort zone (im very picky regarding extrusion widht jajajaj) but I really think it would be an improvement to add nozzle size parameter buecause by doing so yo could achieve beter precision on extrusion widths.
Regading printer type: what I have just done is to create a new type of printer on the excel, that uses mm3 as extrusion units and in the start gcode I set the gcode command to instruct firmware to work in volumetric mode. Still calculation done based on 0.4 mm nozzle would not be real ok if you are using a 1mm nozzle, and the problem would be important I guess if using a 2mm nozzle

[fullcontrol-xyz]

Hey,

Si please correct me if Im wrong: following your previous explanation, if you calculate the volume of a path you still need the nozzle diaim to get the E value if the printer usus mm as extrusion unit. You wouldnt need it if the printer is set to volumentric mode.

No, nozzle diameter is not used to calculate E. As far as I'm aware the is the case for all forms of gcode and all slicers. The key benefit of volumetric units is you don't need to know the diameter of the filament reel because the printer firmware is aware of it and uses it to translate E into motor steps.

When you're finding the extruded lines to not be exactly the correct width, that is due to fundamental experimental factors that are not really related to FullControl or slicers. As extrusion pressure increases (e.g. extruding wide from a narrow nozzle) you will potentially get under-extrusion, but it hugely depends on your extruder tightness level (e.g. the screw on the extruder), extrusion temperature, material rigidity (to transfer force from the extruder drive gears to the melt chamber), layer height, and lots of other things.
FullControl is outputting the exact correct data for how much material is required to be extruded (ignoring any mismatches in the assumed cross sectional shape like rectangle or stadium, etc., which is also not hugely relevant here). The problem is the gcode is a list of instructions of what the printer should do in terms of motors. If you tell the motor to feed in the same amount of material to the hot end, the same amount must come out of the nozzle, unless there are uncontrolled losses (leaks, slipping, etc.) and then it's the case that really complex experimental factors are coming into play, which gcode is not designed to consider).

I can see there that you are asumening on most printer you are using 0.4 nozzle

This isn't the case. I might refer to a 0.4 nozzle for the purpose of trying to keep line widths the same as or within reasonable bounds of the nozzle size, but this data is used in the FullControl design (i.e. my set value of EW) not the the FullControl algorithms that convert the design to gcode.

So in summary, users just need to design EW that is reasonable for your nozzle diameter (I've found that to be between 75% and 250% of the nozzle for an E3D nozzle (maybe only up to 200% for more 'pointy' nozzles. It doesn't matter if you're using volumetric or not, as long as your printer and/or FullControl is set up with the correct feedstock filament diameter. If you are finding that you're printer is under-extruding for particular printing conditions and you want to compensate for that, that's is logical, but not really the sort of thing that would go into the back end of fullcontrol because it's not repeatable across systems/materials, etc. You could either optimise you steps/mm for E in your firmware to achieve the correct correlation between gcode E and physically printed material. Or you could deliberately over-extrude by using M221 gcode command in the design or start-gcode or by setting EW to be wider by the desired over-extrusion factor.

Most people don't even know this kind of consideration exists. We have previously done the under-extrusion compensation for mechanical properties research, where we printing under lots of different conditions, weighed the specimens and then reprinted them with adjusted EW values to ensure that all actually weighed the same. That allows us to assess the effect of printing parameters without the normal unintentional variation in material flow which would obviously hugely affect performance - see attachment. That allowed us to find that all these printing parameters that are conventionally thought to hugely affect mechanical performance actually have basically no effect. You just need to set up your printing to consider the effect of those parameters directly and not all the indirect effects, especially uncontrolled variation in extrusion rate.

2021 - Bulk-Material Bond Strength Exists in Extrusion Additive Manufacturing for a Wide Range of Temperatures, Speeds, and Layer Times.pdf

[fullcontrol-xyz]

Hope that message doesn't come across to arrogant or dismissive! 👆

Always happy to have conversations about this stuff. I am certainly an expert in this topic, but am also fully aware that my understand won't translate to all situations and conditions, so it's valuable to be presented with situations that don't fit my understanding 🙏

[Tinchus2009]

NOT AT ALL! I really preciate you take the time to answer me, and I will also do a clarification regarding this: nothing I said above or nothing I could say ion future is a or will be a negative comment or a bad critic. Today my thinking is "This thing I found (talking about full control) is GREAT. Because I see the potential of doing things I can not do using regular slicer or without not doing a huge amount of gcode hacks, tweaking etc...

Going back to the topic: Im maybe not understandfing because Im too used to think on the normal way on how slicing is done. The maths I have in head is:

So basically the nozzle size is required to calculate extrusion amount.
On my above pst i did a mistake, I read it again and I notice I used "nozzle size" in many places where I should have used "filament diameter" Filament diameter is the only value a printer on volumetric mode needs to know, not the nozzle size.

So after seeing my mistake and reading your answer, what I think you are doing is doing the job normally the firmware on volumentric mode would do: translate extrusion volumen to E movements? Is that the way fullcontriol is working? (Im very sorry to answer these kind of questions, I usually like to know how things works, I hate just "clicking" lol).

[fullcontrol-xyz]

That image is one way to calculate extrusion area yep. You can also have an even simpler option if you juts ignore the round corners and calculate area as width*h. FullControl uses that assumption by default, but also has an option for this stadium shape cross section in case people want to use that.

Regardless, you still don't need nozzle dia either way. You equation in the image does not include nozzle diameter, it only used w, h and pi.

Once area is calculated, it's then combined with the length of the current straight-line section to get volume, which is then converted to either volumetric or non-volumetric E.

[Tinchus2009]

Thanks for the answer. I think I will leave this topic for now because I think you are using too much time trying to explain me and I dont understand jajajajaja. Still is like I dont know how you canlculate that E number: how can you calculate a command to make the extruder move certain amount of mms of filament if the software dont even know filament diameter... if you move 1mm of 1.75 mm diameter filament you get a volume but 1mm of 2.85 mm filamente diameter will give you way more, so how the software know... it is a real mistrey to me.

Another question: I read the paers, all of them, SUPER NICE! have 1 model I usually print, a vibration dumpener, itr is just a square, and I use a giroid infill that has good shock absortion features. And I use TPU. But retractions are crazy, I have almost 35% of time wated on retractions and travel moves.

You apaper about the insole ios a great example to me and I would like to redesign my model and use full control. Can you share the mth formula for that infill you used on the isole paper or can you share the excel sheet so I can see how you work in such situations? I have also seen all tutorials.

[fullcontrol-xyz]

Sorry I never replied to this. I haven't got the design easily accessible, but it basically a combination of mathematical line equations for sine waves, where amplitude is fluctuating as a function of Z. Some are defined along the x direction as wavy lines, and others are defined in polar coordinates to be wavey arcs. However, I'd recommend going to the python version of fullcontrol for that sort of geometry... you're pushing towards the limit of Excel with it, whereas the python version could go far more complicated. Wiht some help from chat gpt, you could gradually build up complexity. Get it to create functions to generate liner or arc wavey lines, then start adding them line by line.