What happens to the temperature inside your nozzle?
-
Ok everyone @deckingman is looking for some diamond tipped miniature drill bits, for science!
-
drilling one brass and one "regular hardened steel" nozzle (not vanadium) might give enough data for conclusions?
-
@T3P3Tony said in What happens to the temperature inside your nozzle?:
Ok everyone @deckingman is looking for some diamond tipped miniature drill bits, for science!
Or EDM
-
...........Or alternatively, one of the multitude of nozzle\hot end\silicone sock manufacturers who happily relieve us of our hard earned cash, could pick up the batten and do the tests. I'm just a pensioner scratching about in his garage with very limited resources and getting bugger all in return.
-
@deckingman said in What happens to the temperature inside your nozzle?:
Thinking about it some more, I don't like idea of packing a nozzle with thermal grease of any sort. The main reason being that in general, filaments are poor thermal conductors. So if the grease/compound did make any difference, it still wouldn't be a fair representation of the true nozzle temperature when filament is loaded. But I'm sure I can find a way to block off the 0.5mm hole in the end.
I absolutely agree with this. In fact, the thermocouple or thermistor itself has different thermal characteristics than any filament... but there's simply no other way to get a reading.
As an alternative to filament, what about a LOW temperature silicone? It might be the closest you'd be able to get to the thermal characteristics of molten plastic, yet still be easy to work with (and clean up) once it's cool.
Of course, even those results would be missing the variable of cold filament moving into the melt chamber, absorbing heat from the hot block and melting, and then the molten filament extruded (along with a significant amount of heat energy.) (Typing that last line, I'm reminded of how a HVAC heat exchanger works...)
Perhaps it'd be better to test the temperature of the filament actually being extruded instead of the nozzle temp? Afterall, the temperature of the block and even the nozzle is less important than the temperature of the molten filament. Would it be physically possible to test the filament as it's exiting the nozzle, but before the ambient air has had a chance to cool it? Is there any instrument that can measure a 0.5mm area without obstructing it?
To be fair, I don't think there's a really viable way to test this fully. As mentioned very early in the thread, your tests offer clues, but we might never really understand what's going on.
Personally, I love these kinds of threads, because they remind me of everything I don't know and make me question everything I think I do.
-
@garyd9 said in What happens to the temperature inside your nozzle?:
Perhaps it'd be better to test the temperature of the filament actually being extruded instead of the nozzle temp? Afterall, the temperature of the block and even the nozzle is less important than the temperature of the molten filament.
I'd go as far as to say that the temperatures of the block and nozzle are completely irrelevant and that the only thing that matters is the temperature of the filament. In reality, until we find some other way to melt filament (and that comment could lead to whole new discussion), filament is heated by thermal transfer from the hot surfaces to which it is contact, so the hot block and nozzle temperatures will be close to that of the filament, but it's only the filament temperature that is important.
Would it be physically possible to test the filament as it's exiting the nozzle, but before the ambient air has had a chance to cool it? Is there any instrument that can measure a 0.5mm area without obstructing it?
I don't see why not. One would need a very small thermocouple. But a thermocouple is just formed by fusing the juntion of two different metals - I don't think there is any limit on the minimum size of those two metals. "Micro" temperature sensors probably exist (I don't have one laying around in my tool box though).
And of course to make life a bit easier, we could use a bigger nozzle - say 1.0 mm instead of just 0.5mm. But then having mentioned that fact, I can already see a whole new dimension to this testing - the effect of filament temperature as a function of nozzle diameter - which then leads to an investigation into the distribution of heat from the outer surface of the filament through to the core - and how this might be affected by different nozzle materials - all of which will be affected by filament flow rate through the nozzle - etc. There is enough there to form the basis of a higher education thesis I'd have thought
To be fair, I don't think there's a really viable way to test this fully. As mentioned very early in the thread, your tests offer clues, but we might never really understand what's going on.
Agreed. As I said in one of the opening statements of my blog. " In my opinion, it is not wise to assume that the temperature seen by the hot end thermistor is a true reflection of the temperature at the nozzle tip when there is some airflow passing over the nozzle". I think I have proven that statement to be true. I think also that I have shown that various nozzle materials react differently when subjected to some (inevitably) deflected part cooling air. But I haven't been able to demonstrate if this translates to a change in the temperature of filament, either inside the nozzle, or at the point of exit. We can still only guess (but maybe our guesses might be closer what truly happens, than they were before I did the tests).
-
Great work. Since I joined the Duet community at the start of the year I have learned so much. But I always felt like there were chunks of information missing.
The point of "we know what makes a print work, and we use it, but we might not know why it works" comment is so true. Seeing unexpected results comes from these chunks of missing information.
I have been using A2 machine steel nozzles exclusively for over a year. Not because I print with abrasive materials, but because I figured they were more durable, and I could replace them less. I never really put temperature into the equation. I knew that hardened steel is less heat conductive, and I have tended to use higher temperatures because of this. But seeing your results as certainly opened my eyes wide, and I am going to go back to using high quality brass nozzles.
But on top of that, it makes me think greatly about the impact of cooling near the nozzle. Maybe the cooling ducts that are popular are thought to be great, but over doing their job and causing heat loss in unexpected ways.
It opens up a new direction of thought when thinking of some of the consistency issues I have been having.
CR10S, August 2018
Anycubic Photon S December 2019
Ender 5 Feb 2020
Ender 5 Plus May 2020
Anycubic Photon Mono X Nov 2020
~
https://3dimensiongames.com/ -
These nozzles allows to measure and/or stabilize the temperature closer to the tip.
-
@RyanP said in What happens to the temperature inside your nozzle?:
................ But seeing your results as certainly opened my eyes wide, and I am going to go back to using high quality brass nozzles.
It was never my intention to put people off using hardened steel nozzles. They still have their place - with certain filament types, I'd say they are essential. But just be aware of what part cooling air might do to the filament temperature. It may be that you can still print that part which needs to be made from an abrasive material, without any part cooling?
But on top of that, it makes me think greatly about the impact of cooling near the nozzle. Maybe the cooling ducts that are popular are thought to be great, but over doing their job and causing heat loss in unexpected ways.
Exactly so. In my opinion, it is impossible to blow cooling air over a part that is being printed a fraction of a mm below the nozzle, without some of that air passing over the nozzle itself - either directly from the fan ducts, or indirectly by being deflected back from the part being printed. What might come close would be a sort of reverse Berd air system - a small annular tube around and close to the nozzle, but with holes pointing away from the nozzle.
It opens up a new direction of thought when thinking of some of the consistency issues I have been having.
That's great. I wonder too how many people have been suffering with what appear to be extruder issues (grinding filament, missed steps etc) which happen to coincide with switching to a metal nozzle?
-
@ All. For info - I'll shortly be running those tests again with a home made silicone "sock" around the nozzle. And I'll be detailing a very cheap and quick method to make such a sock.
I'm calling it a "nozzle trainer sock". A sock (as in item of clothing) covers the foot and ankle, but here in the UK, one can buy "trainer" socks, which cover just the foot but not the ankle. This sock covers the nozzle but not the hot block - hence "nozzle trainer sock".
I'll do another blog post and put a link on this forum in the next couple of days or so.
-
Certainly I never took your test or your post as a slag on metal/steel nozzles, simply realizing that I am not necessarily using the best nozzle for my current printing needs.
When I first got into 3D printing with my Creality CR10S, I frequented its reddit sub. Over time the answer being shared around bothered me because every problem can't be "under extrusion".
Since I converted my now three printers two Duet board and started participating in this forum, I have learned so much, especially volumetric flow rates and their place in printing. Focusing on nozzle pressure made so much sense to me.
With this information that you have provided compliments that info. I still don't know exactly how the correlate, but if a certain amount of volume pressure in the nozzle is needed to maintain the quality of print, then exact nozzle temperature would need to be a highlighted aspect of that.
@zapta These nozzles look great. I will certainly look into these.
-
@RyanP said in What happens to the temperature inside your nozzle?:
But on top of that, it makes me think greatly about the impact of cooling near the nozzle. Maybe the cooling ducts that are popular are thought to be great, but over doing their job and causing heat loss in unexpected ways.
They may be.. or they may not be. Or perhaps that are, but we are compensating for it somehow. If the fan was really cooling the filament within the nozzle to a significant degree, wouldn't it be reasonable that many people would notice print issues on their 2nd or 3rd layer (or at the lowest layer they turn on their cooling fan?)
I can tell you that often times I have failures when I think I know "the why" of what I'm doing - and more often have successes when I just figure things out with trial and error. That tells me that either I don't really know "the why", or that I only know enough to make me dangerous.
As an example, I have the perfect part cooling (in theory.) It's a berd-air ring that blows down at a ~45 degree from a ring at the nozzle. It's high pressure and low volume, so it shouldn't "over cool" the part. It's arranged in a way that it can't blow directly on the nozzle (and the volume of air is low enough that defection should mostly be muffled inside the infill of a partially printed part.) Yet, if I don't carefully balance filament temperature and cooling air pressure, my PLA prints either curl madly on overhangs or I get insane amounts of stringing. Oh, and because the universe loves to kick me in the face when I think I understand "the why", I get MORE stringing at lower filament temperatures (which is exactly the opposite of popular opinion. )
(This isn't meant to derail the thread - it's only an example of how knowledge isn't always helpful, and sometimes can be harmful.)
Anyway, the kind of stuff that Ian is doing is actually helpful to me in many ways. I can't say that I've ever been able to DIRECTLY take something on his blog and use it, but many times I've read things and said to myself.. "aaah! That might explain..."
Oh, and up until recently, I've almost exclusively used E3D hardened steel nozzles. They let me switch between filament types without having to change nozzles as often. Recently, however, I ended up switching from 0.8mm steel nozzle to brass and found an improvement in flow with that size nozzle. (Ironically, I had my part cooling fan turned off for those prints.) With 0.4mm nozzles, I've never noticed any effective difference in printing with brass vs hardened steel. I also never had a need to vary temperatures for different nozzle types (though I do let my hot end temperature "settle" for a couple minutes before starting a print - so perhaps that gives the hardened steel time to get up to the proper temp.)
"I'm not saying that you are wrong - I'm just trying to fit it into my real world simulated experience."
-
Every single word you typed sounds like my experiences lately. Problems....solutions....why?....don't know. Same problem, different solution, why?....don't know.
This forum has proven itself invaluable when it comes to finding the solutions. The Whys are sometimes there, but other times it can be more like "Try this, it may fix it".
We all have to take out individual machines, and filter through all the experience of each other to find what actually works on our own machines.
-
@zapta Those nozzles look interesting. I'd be a bit concerned about the sensor falling out. How is it retained?
Also, as I mentioned in my blog, if you do have localised cooling of the nozzle (as opposed to the hot block) then potentially you could end up heating the filament inside the hot block to 30 Deg C or more higher than the normal print temperature. If that happened with PET-G, it would carbonise and you'll never clean it out, so you could end up with a wrecked hot end. Therefore, use with caution and be especially careful not to have part cooling air blowing over the nozzle. -
First, I think most people blow way too much air on their parts. When using PLA, you almost don't need to blow air at all, except in specific areas, to preserve small details. I think slicers should take that into account. For example, why blowing air during infill? It just reduces the layer adhesion!
Second, what about blowing hot air? It needs to be below the glass transition temperature. This is what Stratasys does: the chamber is kept at 100°C (ABS), and the fans blow air at that temperature. And they blow a lot of air (2 big fans on each side).
Frédéric
-
@garyd9 said in What happens to the temperature inside your nozzle?:
They may be.. or they may not be. Or perhaps that are, but we are compensating for it somehow. If the fan was really cooling the filament within the nozzle to a significant degree, wouldn't it be reasonable that many people would notice print issues on their 2nd or 3rd layer (or at the lowest layer they turn on their cooling fan?)
If you look at any decent nozzle seller you will see that with steel nozzles they say "you need to increase your print temp by 10 degrees"
-
@fma Regarding the temperature of the cooling air, I think I covered all that in my blog. I was especially careful to note the ambient temperature and the fact that the bed was unheated, so any deflected air might be cooler that if the bed was heated.
I would be hesitant to say that "most people blow way too much air on their parts". I haven't seen the results of any surveys asking people what volume flow rate they use, have you? Is that statement based on any factual data or are just guessing?
Also this statement quote - "When using PLA, you almost don't need to blow air at all, except in specific areas, to preserve small details"
...... is only be true for certain nozzle sizes and slowish speeds. When using larger nozzles and/or printing at high speed, one can lay down a lot of plastic in a very short space of time. If cooling isn't used, then the previous layer will still be at or above the Tg temperature when the next layer is added resulting in the printed part sagging.
-
@deckingman said in What happens to the temperature inside your nozzle?:
@fma Regarding the temperature of the cooling air, I think I covered all that in my blog. I was especially careful to note the ambient temperature and the fact that the bed was unheated, so any deflected air might be cooler that if the bed was heated.
I would be hesitant to say that "most people blow way too much air on their parts". I haven't seen the results of any surveys asking people what volume flow rate they use, have you? Is that statement based on any factual data or are just guessing?
Facts: I meet a lot of people doing 3D printing in my hackerspace and around, and I see what they are doing.
...... is only be true for certain nozzle sizes and slowish speeds. When using larger nozzles and/or printing at high speed, one can lay down a lot of plastic in a very short space of time. If cooling isn't used, then the previous layer will still be at or above the Tg temperature when the next layer is added resulting in the printed part sagging.
Yes, true. That's why I do think that an heated chamber keeping the part at the correct temperature¹ would be a huge improvement, even for PLA. This way, you don't have to bother that much with air flow.
¹ finding that correct temperature would need some experiment, though, and probably change for each filament brand/color/...
-
@fma said in What happens to the temperature inside your nozzle?:
keeping the part at the correct temperature¹ would be a huge improvement, even for PLA
I started my 3d printing "experience" with "difficult to print materials" PP and HDPE (and HDPE800 which is some even higher density polyethylene than normal HDPE and warps even worse), so I'm experimenting with heated chambers for almost a decade. WRT PLA 25-30C is kinda best printing temperature, and temp of the air you blow is best around 25-30C. Blowing hotter air (e.g. 40C) produces very bad prints. Try it yourself. There was a huge issue UP printers had with PLA where the designer made the air that cools the heatbreak work as part cooling too and no matter how much air they pushed the results were terrible. That air was 40-45C. Bringing colder, room temp air, to cool the part solved all the issues. You can find measurements and discussions in lenght on the pp3dp forum.
ABS works great in 70C chamber with 70C air being circulated in the chamber (no need to blow directly under the nozzle, enough to circulate air. I made it usually in the way that first layer of air going from the heaters creates laminar flow over the top layer of the print (never actually visualised the flow but..) and it works ok, problem is hotend's heatbreak, I experimented with a lot of stuff (peltie's, external air trough tubes etc...) and so far water is the only thing that works ok, but those are all hotends I made myself and are not "very good" (single piece aluminium turned on lathe, no multimaterial heatbreak, no interchangeable nozzle ..)... today there are some interesting solutions available that one should try but one needs time
PC is similar only chamber temp 130C works better than 70C.
PP is on the other hand super nasty to print, it's Tg is 0C but I had best print quality with 40-50C chamber and very slow air movement (208C printing temp and slow 8-12mm/sec print speed with pauses on small layers)
cooler nozzle than the block is expected, it is important that low conductivity nozzles are not too long so that temp differential to the tip is not too big. You compensate for the tip temperature by increasing the temp of the heater block. If the nozzle is too long you have to get your block "too hot" and that can damage the filament, but as long as the required heat increase is in the safe zone for the filament you don't have a problem, you just need to temp-tower find the best temp for certain print settings (fan speed, nozzle size, print speed, layer height, extrusion width combo).
-
Thanks for the feedback!
