Relationship between Filament, Extruder Size and Printing…

Hello All,
Thank you for taking the time to read this and help. The more I read and search the more confused I become.
I am new to 3d printing, but am experienced in traditional CnC (mills and lathes), SolidWorks, and Embedded Systems.

Here is the question, I am looking at machines/kits to buy and cannot seem to figure out what are the advantages/disadvantages of a smaller extruder nozzle? I see sizes ranging from .30, .35 up to .5 mm.
In my mind the size of the nozzle is a limit feature size, or at least line thickness, but where this is throwing me off is when I look at the software. In either the standard reprap software or in RepSnapper there does not appear to be a setting for nozzle size. What am I missing here? Does it not really matter in that what only matters, is how much filament is feed through the extruder?

How can it figure out how much is bring printed without knowing the nozzle size? For example. Take the exact same amount of filament being pushed through the extruder. A nozzle that is .30 will have much more squish (I don’t know what the proper term is. The amount that squeezes outward versus the amount that is rounded under the nozzle.) than a .50 nozzle. Also if your making a solid layer a .30 nozzle would need to make .30 passes assuming no squish, if a .50 nozzle made .30mm passes part of the line would roll up on to the previous line, instead of being laid exactly next to it.

Along this same line why use 1.5 over 3mm filament, is it that there is more control over how much is being extruded? Is this a limitation of the stepper resolution and gearing?
Isn’t there an ultimate smallest size limit, where eventually not enough is forced out to stick to the object? E.G. 1 step at X gearing equals so much plastic, but with 1.5mm filament that amount is half as much? Couldn’t someone increasing the gearing so much with a 3mm filament that it reaches the minimum size?

I hope this makes sense, thank you all for the help and time.

The nozzle size is input into the slicing program you use, repsnapper is not very good at slicing (creating Gcode from a .stl file) most people use something like Skeinforge or sl3cer, Pronterface I believe can have Skeinforge added to it as an extra component so you then get slicing and printing with one package. Skeinforge can output a n amount of filament used although I find it not very accurate.

Using a smaller nozzle than 0.5 means you can in theory have lower layer height.

3mm filament needs more force to extrude but is less prone to breaking. 1.5mm costs more when you factor in the rolls contain only half the amount for the same price well PLA does here in NZ.

Its very rare to print 100% fill, what I do is 3 solid layers then depending on how much strength the part needs either 25%, 50% or 80% fill ratio on the interior layers. 80% gives almost solid fill anyway. If you did go for 100% your parts would end up over size

---

__________________________________________________________________________
Experimenting in 3D in New Zealand

tjhj2 Wrote:
-------------------------------------------------------
>
> Here is the question, I am looking at
> machines/kits to buy and cannot seem to figure out
> what are the advantages/disadvantages of a smaller
> extruder nozzle? I see sizes ranging from .30, .35
> up to .5 mm.
> In my mind the size of the nozzle is a limit
> feature size, or at least line thickness, but
> where this is throwing me off is when I look at
> the software. In either the standard reprap
> software or in RepSnapper there does not appear to
> be a setting for nozzle size. What am I missing
> here? Does it not really matter in that what only
> matters, is how much filament is feed through the
> extruder?

Some slicers take the nozzle size and chose a layer height and width, others take the layer height desired and the filament width. In terms of the g-code generated the nozzle size makes no difference as it is the flow rate of the plastic relative to the feed rate of the axes that is important and the nozzle size does not change that. What the nozzle size affects is the maximum and minimum filament widths that you can extrude. The plastic needs to be stretched a little so you there is a limit to how wide the filament can be. The filament swells when it comes out the nozzle into fresh air (known as die swell) and you can't extrude wider than that without compressing it lengthways. That is like pushing a string and it tends to squirm and wont span gaps.

At the other extreme the is a limit to how much it can be stretched thinner than it comes out naturally. it starts to cut corners and eventually will snap when spanning gaps.

Since these limits depend on the die swell value (which depends on plastic, temperature, flow rate nozzle hole length), and not the nozzle diameter, it makes little sense for the slicer to be given nozzle diameter, IMHO.

>
> How can it figure out how much is bring printed
> without knowing the nozzle size? For example. Take
> the exact same amount of filament being pushed
> through the extruder. A nozzle that is .30 will
> have much more squish (I don’t know what the
> proper term is. The amount that squeezes outward
> versus the amount that is rounded under the
> nozzle.) than a .50 nozzle. Also if your making a
> solid layer a .30 nozzle would need to make .30
> passes assuming no squish, if a .50 nozzle made
> .30mm passes part of the line would roll up on to
> the previous line, instead of being laid exactly
> next to it.

Given the desired filament width and layer hight it can extrude the correct volume. Whether it builds correctly depends on the user choosing values within the range that the nozzle can achieve.

>
> Along this same line why use 1.5 over 3mm
> filament, is it that there is more control over
> how much is being extruded? Is this a limitation
> of the stepper resolution and gearing?
> Isn’t there an ultimate smallest size limit,
> where eventually not enough is forced out to stick
> to the object? E.G. 1 step at X gearing equals so
> much plastic, but with 1.5mm filament that amount
> is half as much? Couldn’t someone increasing the
> gearing so much with a 3mm filament that it
> reaches the minimum size?

Yes the filament size affects the optimum gear ratio, but by the square of the filament size as it is volume that is important. The common sizes are 3mm and 1.75, not 1.5, which give a 3:1 difference in speed and torque required. At one extreme the ratio has to give enough torque, depending on the filament and the motor, and also enough resolution. At the other end the speed is limited.

The other aspect that filament size affects is the ability to coil the filament and for it not to buckle when forced into the extruder. The force depends on the square of the diameter but stiffness depends on the fourth power. So soft plastics would buckle when at small diameters. Hard plastic would be too stiff to coil at large diameters.

---

[www.hydraraptor.blogspot.com]

Thank you both for the time and explanations,

That makes a lot of sense fundamentally.

Nophead
>Since these limits depend on the die swell value (which depends on plastic, temperature, flow rate nozzle hole length), and not the nozzle diameter,
> it makes little sense for the slicer to be given nozzle diameter, IMHO.

Yeah I didn't take into account that plastic is flexible and expands with heat. I had it in my mind that X amount pushed into the nozzle equals the same amount out. Opps. Thanks for clearing that up.

Now how does this translate into the real world for me or what should I do?

I am looking to make a mix mosh of parts so think in terms of flexibility. I would tend to go more towards precision over speed. On my personal CnC machines it doesn't matter to me if a job takes 8 hours. A full commercial machine can do it in 2, but I can turn it on and let it run, sure I can make compromises in quality and gain speed, but most of what I does requires precision and detail. Im not in a production rush. I am thinking .35 nozzle and 3 mm filament for cost. It gives me the precision of layer height Is this bad combo?

What size nozzle and filament would you recommend?

Thank you for your time and help,

tjhj2 Wrote:
-------------------------------------------------------
>
> Yeah I didn't take into account that plastic is
> flexible and expands with heat. I had it in my
> mind that X amount pushed into the nozzle equals
> the same amount out. Opps. Thanks for clearing
> that up.
>
It doesn't change in volume (it will expand with heat but contracts again when cooled) but it comes out bigger diameter than the nozzle hole and correspondingly shorter when extruded into free space.


> Now how does this translate into the real world
> for me or what should I do?

I measure the extruded filament diameter and that tells me what range of filament widths and layer heights I can use. I then input those to Skeinforge.


>
> I am looking to make a mix mosh of parts so think
> in terms of flexibility. I would tend to go more
> towards precision over speed. On my personal CnC
> machines it doesn't matter to me if a job takes 8
> hours. A full commercial machine can do it in 2,
> but I can turn it on and let it run, sure I can
> make compromises in quality and gain speed, but
> most of what I does requires precision and detail.
> Im not in a production rush. I am thinking .35
> nozzle and 3 mm filament for cost. It gives me the
> precision of layer height Is this bad combo?

Those are reasonably choices for quality over speed.

---

[www.hydraraptor.blogspot.com]

nophead Wrote:
>
> I measure the extruded filament diameter and that
> tells me what range of filament widths and layer
> heights I can use. I then input those to
> Skeinforge.

So.. say, I have a .4mm nozzle. When I test extrude, after cooling, the diameter of the plastic is generally .6mm.
Slic3r has nozzle diameter as an input. Do I input .6 or .4mm?
How do I figure out the min/max layer heights I can use? Is it dependent on the printing speed?

Thanks very much for any light you can shed on this.

I don't know, I have never used slic3r, and I don't know what it does with the nozzle size as it is not relevant to slicing in my opinion.

The max layer height is when you have a height over width ratio of 1.5. Any less and the object will be weak because the filaments are not squashed together enough. So you work out the cross sectional area of 0.6mm and work out a rectangle of the same area and 1.5 aspect ratio. That gives you the maximum layer height. There is no minimum, you can go as low as you want but you have to keep the width reasonable so the aspect ratio has to increase.

It does depend on speed because die swell increases with speed.

---

[www.hydraraptor.blogspot.com]

lincomatic Wrote:
-------------------------------------------------------
>
> So.. say, I have a .4mm nozzle. When I test
> extrude, after cooling, the diameter of the
> plastic is generally .6mm.
> Slic3r has nozzle diameter as an input. Do I
> input .6 or .4mm?
> How do I figure out the min/max layer heights I
> can use? Is it dependent on the printing speed?
>
> Thanks very much for any light you can shed on
> this.

For Slic3r, tell it your actual nozzle diameter, 0.40mm. It tries to do the math that nophead describes, and does a good job. It will not let you slice with layers taller than the nozzle size you tell it. It will adjust the extrusion rate to accommodate for different print speeds, or in other words, for the areas it is printing slowly it will not push as much plastic out as it will for the areas where you are printing quickly. For me at least, it gives best results when the layer height is roughly 2/3 of the nozzle diameter or lower.

But how can it do the maths I describe because nozzle diameter does not come into the equations, just the die swell diameter?

---

[www.hydraraptor.blogspot.com]

Nophead, I haven't dug into the software to see how Slic3r accomplishes this, I am guessing that it makes assumptions about the die swell ratio that from empirical experience printing at different layer heights appears to me to be to be fairly accurate. You should try it on one of your Prusas or Mendel90, I think you will be pleasantly surprised at how much easier it is to use than Skeinforge. It doesn't give you nearly the operator control that Skeinforge does, but that means the newbie learning curve is much much flatter.

It can't make a valid assumption because the die swell for HDPE is totally different from PLA. Also it depends greatly on extrusion speed and the length of the nozzle exit.

Skienforge has never been a problem for me as I only use a few modules.

---

[www.hydraraptor.blogspot.com]

Thanks for the responses.

I tried entering my actual nozzle diameter into Slic3r (.35), per MKouri's advice above, and had problems with my layers not sticking properly.
I switched to entering the diameter of the extruded filament (.65), and I'm getting perfect prints now, so YMMV.

Geeez. I dont know why I ever disagree with nophead. I switched from the 0.35mm nozzle I had been using since before converting to Slic3r to a 0.50mm nozzle, and the solid surfaces stopped looking as good,more than can be explained by the nozzle change. I did as lincomatic and changed what I told Slic3r was the nozzle size (now 0.60mm), and the quality went back to what I was expecting. I then dug in my plastic scrap bin for some older extruder priming threads, measured them, and got 0.38mm, not a larger thread like you would expect from die swell. I guess my nozzle had closed up some from use, and I was merely coincidentally telling Slic3r the right value for it to use.

Nophead, once again you've taught me to never doubt you.

I was thinking about this overnight (before seeing these last to posts) and was thinking that maybe the reason was that your nozzle wasn't actually the diameter you thought it was. In my opinion, they are very difficult to drill accurately and I would suspect that a slight knock would see a dramatic change in size.

The last assumption made based on me reading that some people make their nozzles by drilling a larger hole and then tapping it smaller (someone was experimenting with square holes).

On the nophead front. I studied Mechanical Engineering as a Research and Design degree, so while I don't practice in this field, I understand the ins and outs of doing research properly. When I first got into the RepRap scene 2 years ago I spent a few nights reading his entire blog (comments and all). What struck me is that when he "opens his mouth", whatever he says will be able to be backed up by sound experience, science and mathematics. So much so that if I see conflicting opinions about, the first thing I do is to see what nophead might have said on the topic.