I am not necessarily looking for a solution as swapping out the nozzle resolves the issue. I am just wondering what your experiences have been when using the hardened steel hot-end.
Your second paragraph states that your print profile remains the same for all the tests. Have you tried tuning in a new profile for when the hardened steel nozzle is in use by running the various built-in orca-slicer calibrations? i.e. does running the hardened nozzle with the print profiles tuned specifically to it result in the same or similar results produced by the stainless steel profile when using whatever print profile you’re presently using? It seems reasonable to expect that different nozzles might require different profiles to perform at their best.
I have two different types of nozzles here, don’t mean the material…
One has a SINGLE hole going through, the other has THREE holes/channel joining into the outlet.
The first has no issues with filaments, the second struggles badly with anything soft, like TPU or filled.
For the later it just likes to block up here and there.
Maybe the difference is not just in the material but also the nozzle design…
Both of my nozzles are from BL and the printer recognises both. There isn’t anything special about the as with your trip,e nozzle hot-end.
I can imagine why that struggles as it is very soft and any pressure required to split it into three parts will cause the material to resist and likely block the nozzle and/or cause inconsistent flow as tension increases to a point it it either pushes through at once or not at all.
The BL supplied nozzles are detected when installed and the printer knows which one it has to print with.
I will add that to my tests.
However, with the BL printers managing to detect things like plates, hot-ends and for some filaments to then automatically adjust things in the slicer, it’s strange the nozzle might be the only one left to the user to figure out.
I don’t mind the job, but, I can not find any information in the wiki, which is a great source of info, to suggest what changes to be made, let alone any indication that any should be changed.
It is as if they don’t expect swapping nozzles to then require manual adjustments.
I didn’t see you mention the size of the nozzle.
So I assume you are using 0.4 for both?
I have been printing more in TPU so I am quite interested in your findings. ’
For the record I gave up in printing TPU with 0.4 a long time ago. It constantly clogged on me. I only ever print TPU with 0.6 and 0.8 (hardened) and don’t have clogging issues anymore.
Is it possible that your TPU isn’t dry enough. If I only saw your bad print, that is what I would suspect is the cause. I dry all my filament until it stops losing weight and then vacuum seal it along with desiccant. So that may be why it works well for me
I asked Chat GPT the following question (after first explaining the problem and the fact that the stainless steel nozzle works perfectly)
If the TPU is not dry enough, are there differences in properties of the stainless steel vs hardened steel nozzle that would make the stainless steel work perfectly while the hardened steel be terrible?
This was it’s reply
Yes, the moisture content in TPU can definitely affect how it prints, and different nozzle materials could exhibit varying performance under these conditions. Here are a few ways how moisture in TPU and the type of nozzle used might interact:
Moisture Effects on TPU: When TPU absorbs moisture, it tends to bubble and pop during printing due to the steam created as the moisture evaporates. This can cause imperfections in the print, such as poor layer adhesion, stringing, and uneven extrusion.
Thermal Conductivity Differences: Stainless steel has higher thermal conductivity compared to hardened steel. This means stainless steel nozzles can more efficiently transfer heat to the filament, potentially helping to evaporate moisture more uniformly and quickly during the printing process. This might result in a smoother print despite the presence of some moisture.
Nozzle Temperature Fluctuations: The lower thermal conductivity of hardened steel nozzles might lead to less effective heat transfer. This could result in less consistent temperatures at the nozzle tip, potentially exacerbating issues caused by moisture in the filament.
Surface Characteristics of Nozzles: The internal surface finish of the nozzle can also affect how filament behaves during extrusion. Stainless steel nozzles typically have a smoother internal surface compared to hardened steel nozzles, which might help in mitigating the effects of moisture-laden TPU by providing a more consistent flow.
Given these factors, if the TPU filament isn’t dry enough, the difference in how each nozzle type handles the evaporating moisture could explain why the stainless steel nozzle still performs well while the hardened steel nozzle results in poor quality prints. To resolve this, drying the TPU filament thoroughly before printing can help improve results with either type of nozzle. Using a filament dryer or oven to pre-dry the TPU can be particularly effective in removing moisture and ensuring more consistent print quality.
Before that, when I initially explained the problem, this is what it said (in case moisture isn’t the problem)
It sounds like the issue might be related to the characteristics of the hardened steel nozzle compared to the stainless steel one. Here are a few possibilities and suggestions:
Nozzle Diameter and Precision: The hardened steel nozzle might have a slightly different internal diameter or less precision in its manufacture compared to the stainless steel nozzle. This can affect the flow of TPU, which is very flexible and can behave differently depending on slight changes in the extrusion path.
Heat Conductivity: Hardened steel nozzles generally have lower thermal conductivity compared to stainless steel. This could mean that the temperature at the tip of the nozzle is not as high as the set temperature, leading to under-extrusion and the gaps you’re seeing between the lines.
Nozzle Wear and Material Compatibility: While hardened steel is excellent for abrasive materials, it might not be the best choice for TPU if the specific characteristics of the nozzle (like surface finish inside the nozzle) affect the material flow.
Here’s what can be tried to improve the print quality with the hardened steel nozzle:
Increase the Print Temperature: Because hardened steel nozzles can lead to a cooler extrusion environment, increasing the temperature might help ensure the TPU is melty enough to form solid layers without gaps.
Adjust the Print Speed: Slowing down the print speed can give the TPU more time to heat up and extrude properly through the hardened steel nozzle.
Check for Partial Clogs or Inconsistent Diameter: Ensure the nozzle is clean and not partially clogged. Sometimes, even small inconsistencies can affect printing materials like TPU.
Experiment with Retraction Settings: TPU is very flexible, so retraction settings that work for other materials might cause issues like stringing or oozing with TPU.
If these adjustments don’t help, it might be best to stick with the stainless steel nozzle for TPU, given that it’s already producing good results.
Maybe they still print fine, but they are likely still wet. I have purchased quite a lot of filament, about 25 spools from Bambu and another 20 or so from other brands and only two were completely dry and lost no water after drying, one PLA Pro from Flashforge and one HIPS from Gizmo dorks.
Flashforge typically has between 1-2 grams of water. Most Bambu is between 4 and 6 grams with some as low as 2 and some as high as 9. I don’t remember what my TPU lost, but I think it was on the higher end.
Found some yellow TPU sample I forgot about the other day, just 10 or 12m…
As it was kept in a crappy zip bad for over 2 years I wasn’t really expecting anything good - using the standard Bambu 0.4 nozzle.
To be able to get the most of it I just printed tiny squares.
First and using standard settings recommended for TPU was a disaster…
The stuff was coming out more like foam.
Played with the speeds, temp, flow rate and extrusion factor but only got very marginal improvements.
Then I took the rest and placed in a bag with loose silica gel to go in a heated chamber at 45 degrees Celsius over night.
The TPU lost a little bit of weight…
The print came out MUCH better but still far from what you would get from PLA or PETG, rough, ugly, micro-bumps and -holes…
Back in the bag and placed in a warm spot for a few days while heating it up to around 50 once or twice a day.
I had no use for TPU for a very long time in my project but if you ask me one should get it bone dry and keep it that way during printing.
TPU will never come out as smooth as PLA, nature of being so flexible…
But it should come out acceptable and with good layer adhesion.
The one thing I learned while using TPU is that snail speed gets the best results…
Which was one of the reason I started to avoid using it for mixed material projects - in favour of things like silicone added for the soft and grippy feeling using a custom mold…
Really don’t have anything on my list that would require me to print it in TPU…
I appreciate your input, but, I’m not sure you read the initial post.
I am using HF TPU, that is high-flow, it is specifically designed to be capable of printing at high speed.
When I print at high speed as set by the Bambu Lab HF 95a TPU profile with a stainless steel novel I get excellent results, if anything, slightly better than what I get with TPU that isn’t HF.
My only issue is the quality when using the hardened steel nozzle.
There have been suggestions from several people that it has water in it. I am very doubtful as I have only experienced this once before and I because I screwed up.
I will run it through my spool heater though for two reasons:
My doubts aren’t facts and tests are good to rule things out.
I’ve only used the heater once before and it’s about time I have its current ratio of use-to-price.
That will take time obviously, plus, I am printing a lot of other things right now and some 0.2 nozzle projects. It will be tomorrow at the earliest before I publish the results.
If the poor results using TPU through a hardened steel nozzle were a known thing, I would have expected to see something written in the wiki about it.
I agree with Sim here.
A few weeks ago I purchased 3 spools of different color TPU (GiantArm brand) and they all came with desiccants in a mylar vacuum sealed bag.
They all printed terribly fresh out of the of bag: