Advanced nozzle design
-
@plasticfactory I love to see practical experimentation.
The length to diameter ratio of 10:1 reminds of laminar flow elements but it has been many years since I had any dealings with such devices. I can see how it would work in a continuous extrusion process but I'd have reservations about back pressure in a start/stop application such as we use in printing. On the other hand, achieving laminar flow would have it's advantages if the flow at the nozzle outlet is otherwise turbulent. I guess there is only one way to find out and that is to do some real life testing.
Please keep us posted on progress.
-
@adam-v3d Well then, lol. Honestly a bit of both. My feeling is that thermal properties may improve given the increase in remaining material, but honestly just don't know without testing. Any insight?
-
@deckingman If the goal is to feed pellets into this, how are air pockets avoided? Or am I misunderstanding and pellets are what's being produced?
-
@kb58 Pellet extrusion would use a very small screw/barrel, with nozzle attached at the end. Air pockets are avoided by melting in the barrel and pushing material forward with the screw. Because of the small scale and intermittent extrusion, pellet extrusion tends to have lots of fluctuation in extrudate size and drooling. As I said, this nozzle is primarily in consideration of known pellet extrusion issues, but I believe the benefits (if any) would translate to both methods of printing.
@deckingman Appreciate it! This is a bit different than the square orifice and you can fully expect additional updates posted here. Your concerns are definitely shared on this end. I'm hopeful that even with the start/stop, the longer die path will essentially help "form" the shape regardless. We will see!
-
@plasticfactory Increased thermal mass will improve temperature stability over time, but i don' think there were problems with that on standard nozzles, certainly won't be any worse. Your CAD clearly differs quite a lot to the real thing at the tip as nozzles have wide flat ends which aids in keeping the tip warm.
the conductivity is not changing.
I think the main change in taking down the diameter of the flow sooner in the path so you have less distance to melt the filament before demanding it to change form. If you just install it straight into a normal v6 you might end up with issues. This is why volcano and super volcano exist, it just increases the surface area to volume ratio to get more energy to the filament per unit of time.
The advantage you do have is that in the thin area you have significantly changed the surface area to volume ratio. You'll be able to control the temperature better, and will have greater chance of getting consistent temperature right to the core of the filament, but you may be too late in the flow path for that to make a difference.
Laminar flow may well produce a better extrude from the nozzle, but the higher mixing of turbulent flow makes it better for heat transfer. I think nozzle flow is quite laminar already though, apart from where the filament contracts into the orifice.
one think I did read in a paper a while back was a suggestion that a large amount of heat transfer into the material actually occurs at the point where the orifice starts, the turbulent flow there and increased pressure cause it. I don't recall the full conclusion of the paper but the smooth change may not necessarily be better.
To be honest though, all the knowledge in the world won't be able to predict if this nozzle will be better or not. Thermal flow and polymers are both super complex. I think you're approach to testing it to find out is 100% the way to go.
-
@adam-v3d said in Advanced nozzle design:
I think the main change in taking down the diameter of the flow sooner in the path so you have less distance to melt the filament before demanding it to change form. If you just install it straight into a normal v6 you might end up with issues. This is why volcano and super volcano exist, it just increases the surface area to volume ratio to get more energy to the filament per unit of time.
I'd try this with a maxiwatt circular heater. EDIT to add: I found maxiwatt heaters to be much more consistent in heating and highly resilient against outside disturbances such as a part cooling fan turning on or any other such effects. So it would be my choice for testing, to make sure heating before the nozzle is optimal.
-
Another thought that is skillfully grabbed out of thin air with absolutely no evidence to back it up:
While going to the accepted standards of extrusion is a good thing. I wonder if it will make that much of a difference. Remember that most commercial extrusion processes do their extruding into air. The flow has to be perfectly laminar because any sort of turbulence as the material exits the nozzle would result in dimensional issues with the end product.
In our case we extrude against a fixed surface, the previous layer, and all kinds of flow disturbances are bound to happen. Further, the fixed surface is a fractional nozzle diameter away so relatively close.I don't know but as I said before, I will be following this closely .... most intriguing!
And no, it certainly does NOT sound like you are soliciting so keep us informed !
-
@adam-v3d said in Advanced nozzle design:
@plasticfactory Increased thermal mass will improve temperature stability over time, but i don' think there were problems with that on standard nozzles, certainly won't be any worse. Your CAD clearly differs quite a lot to the real thing at the tip as nozzles have wide flat ends which aids in keeping the tip warm.
the conductivity is not changing.
I think the main change in taking down the diameter of the flow sooner in the path so you have less distance to melt the filament before demanding it to change form. If you just install it straight into a normal v6 you might end up with issues. This is why volcano and super volcano exist, it just increases the surface area to volume ratio to get more energy to the filament per unit of time.
The advantage you do have is that in the thin area you have significantly changed the surface area to volume ratio. You'll be able to control the temperature better, and will have greater chance of getting consistent temperature right to the core of the filament, but you may be too late in the flow path for that to make a difference.
Laminar flow may well produce a better extrude from the nozzle, but the higher mixing of turbulent flow makes it better for heat transfer. I think nozzle flow is quite laminar already though, apart from where the filament contracts into the orifice.
one think I did read in a paper a while back was a suggestion that a large amount of heat transfer into the material actually occurs at the point where the orifice starts, the turbulent flow there and increased pressure cause it. I don't recall the full conclusion of the paper but the smooth change may not necessarily be better.I really appreciate this response. I agree it's not much of a problem currently, so I'm hoping that this will maintain things, if not negligibly improve them (clearly I'm overly optimistic in general, lol). To explain, the angle used for the tip on the model is clearly not what we are using for blanks. It needs to be updated. The angle, and thus landing area for the flat, is still up for experimentation. I imagine the next step if we see improvements with the extrudate will be to experiment with the tip area.
Again, thank you for the thorough response. It gives me a lot to think about, in particular pull testing. I think layer bonding strength testing will absolutely be required here as a preliminary.
@oliof I have actually used the Maxiwatt pretty extensively. I love the concept, but consistently had issues with it maintaining temperature (260-300*C) and eventually had to move on. It sounds like my unit may have been faulty -- do you think it would be worth trying another? What temperatures are you running with yours?
As an aside, and I may regret posting too much here, but you may find this interesting. It's a nozzle/heater block combo (and 3d printed!). It will use the same flow path as the nozzle above. Now, before anyone says a word I know this may be very impractical in practice. But, I want to see how it performs!
@jens55 I welcome any and all thoughts, fresh perspective helps tremendously! Everything you said are valid concerns -- it's simply not the same process, and we may see no improvement. But, if the amount coming through the orifice is more consistent, I am confident it will improve the print (even if only by some negligibly but measurable amount). It definitely felt like using conventional standards would be a good baseline -- design can always be tweaked based off this is a foundation. And if my optimism bears no fruit... I will let everyone know.
-
@plasticfactory so you can even 3d print the nozzle? All that's left is the controller and wiring, then we can get a proper reprap!
In all seriousness, this is all very interesting and thank you for sharing! Do keep us posted
-
@plasticfactory I had no issues whatsoever with the Maxiwatt on a Hemera, but I only ever printed at 250ish max. With a proper PID tune to 270C it didn't buckle once...
-
@oliof Works for me. I'll try another and if issues persist, I will be thorough about config/hardware review. I'm very confident everything was exactly as it should have been, but at high temps it really struggled to keep up. I was also using mine on a Hemera (not that it particularly matters).
Good enough concept to try again, imo.
-
@plasticfactory Re your combined nozzle/heater block idea, this is effectively what a Diamond mixing hot end is and IMO, one of its major drawbacks. Changing a nozzle when it wears out or for other reasons, takes hours rather than seconds (and costs a relative fortune compared to simply changing the nozzle). It also makes it difficult if not impossible to achieve both high thermal conductivity needed to melt the filament with high abrasion resistance needed for some filaments.
-
Regarding the basic premise of following the design rules for an extrusion die - are you referring to dies used for things like aluminum? I'm not an expert, but my understanding is that, those metals are not melted in the process, but are cold-formed as they go through the die. Also, in the case of 3D printing, the plastic is heated into it's plastic region and is not liquid, so I is the laminar vs turbulent flow question really applicable? I guess I should pull out my old heat transfer and fluid mechanical textbooks and see what the equations say. Anyone know a resource for the mechanical properties of our plastics of interest at various temperatures?
-
@mikeabuilder I am using extrusion die standards typical for plastic extrusion; the same standards we start with when designing dies for our extrusion machinery, scaled down. For basic mechanical properties of material, (in my opinion) the best starting point is a) knowing the grade of resin specifically, and b) searching here: https://plastics.ulprospector.com/ (account required, its free) Though this will not provide properties at various temperatures, only its proper processing and formed states. Either way its an excellent resource for this type of thing.
-
Prepped for the HSM:
-
@plasticfactory A little late, but small update for today. The machining is finished on the new nozzle, and started another one for a proper cross section (quality check that will ultimately make a good photo). I feel confident previous issues are resolved after putting it on our inspection setup. Now that its in my hands I should be able to start doing actual prints! More updates and photos to come throughout the week.
If there are any tests in particular anyone would like me to do, please share!
As for the pictures, the "C" shaped steel is a very simple fixture made to screw a nozzle into and machine from. The black piece is the finishing electrode that reflects the shape of the machined area (the bulbous areas are fluid buildup).
-
@plasticfactory
Jingle tools !!!
-
@plasticfactory, what does it take to generate an electrode that fine that by all rights should disintegrate when you look at it ?
Thanks for that picture! Amazing! -
@jens55 We use a high speed mill designed for cutting graphite. It utilizes high cutting speeds, lower forces and a flushing system specifically for removing graphite dust. A proper graphite setup really is the difference between being able to do general EDM work, and to be able to do something like this (not to say it couldn't be done without).
It's definitely a difficult part to make; the last nozzle came out better but we are running one more through, but breaking the electrodes up into 2 -- one for the major ID, and one for the minor. This should help things out quite a bit. Just part of the development process! Once the process is validated it will be highly repeatable.
-
Quick update -- had some improvement on the last samples, but in doing so identified areas for further improvement. We are breaking the electrode up into 2 pieces and sampling again early next week. I'm optimistic this will be one worth sharing!
EDIT: rather than bumping this thread further I'll use this as an update. Currently, our EDM department is backlogged after receiving a wave of production demand. We have electrodes cut, but no EDM machining will take place for ~3 weeks. To anyone particularly intrigued by this project, do not worry -- it is line one for our AM prototype development.
