Because I am not yet convinced evaporation is a real issue. The super lube silicone oil specs say less than 2% evaporation at 100c in 22 hrs (Data sheet )
As the room the enclosed printer (H2D) sits in is around 15c at all times, well below 100c, the evaporation rate is substantially below the specified one.
In the model train hobby where electric motors are also oiled the maintance interval is entirely base on 40 hrs of active running time. Not self time. It takes years of shelf time to justify cleaning out the old oil and relubricating. Similar for consumer electronics with oil like tape deck motors and spindles.
Has someone done a documented test on the evaporation of the oils at room temp?
Alright, I’ve seen someone pointed at Microlube GL261 for the linear rail.
Any other better alternatives for the grease and for the oil? My issue with all those greases is that there are so many of different recommendation from different manufacturers. For example, linear rail.
BL recommends lightweight oil, Voron recommends Mobilux EP2 (which is a heavy and thick), Prusa recommends GL261 (which is hard to find in the US btw).
Why those are all different recommendations for the same thing? Rail is rail. Do I get it wrong?
I mean I posted a link to a study, but I can dig up another one from the stone ages.
This was tested, retested, tested some more, then tested again for laughs. Of note is “typewriter oil” which is a similar light oil to what we use on these rails (of course chemistry matters in the larger scope) and it showed roughly 1% in 5-6 weeks, which is about a ~2% loss in 3 months.
The second you heat that chamber for a print however you are going to increase that rate by 5-20x, depending on the chamber temps, and suddenly your 10 hour print is another 1% loss on top of it. That study shows losses of ~2% in just 3.5 hours at 55C as well. Round that to 4 hours and your 40 hours is now 15-20% loss just from a 55C chamber.
More if it’s hotter.
The individual volatility of a given type or brand of oil may vary of course, and the manufacturer still needs to assume the worst because they have zero control over the environment the printer is kept in.
I found something interesting. I previously thought that people who were using Superlube 21030 on their H2D were either ignorant of Bambulab’s 92003 recommendation or else just wanting to save money, since it is much less expensive. However, it now looks as though they were the smart ones and I was the ignorant one: 21030 is apparently the better choice with respect to metal wear on the H2D.
Grok analysis
They are not the same grease with different part numbers. 21030 is the one to put on the H2D lead screws. 92003 is the one Bambu named, and it is the worse fit for that job.
Both are NLGI 2, NSF H1, PTFE-fortified, and safe on plastic. The base oil is the whole difference.
On the H2D, grease goes only on the three Z lead screws. Those are steel threads sliding in a nut. A PAO film at 69 cSt is the right weight for that: it stays in the thread, carries the sliding load, and the PTFE handles the metal contact. 92003’s silicone oil is roughly two orders of magnitude heavier at room temperature, and silicone is a poor oil for steel-on-brass. Super Lube adds PTFE to 92003 specifically because plain silicone grease is weak on metal. It will not hurt the nut, and it is what the Bambu wiki lists, but it lubricates the screw worse and costs more.
That viscosity gap is why H2 owners who have run both prefer 21030 over Bambu’s 92003 recommendation. Lucas 10533 sits in between at about 230 cSt at 40 °C; 92003 is the outlier.
92003 is the better tube only if the part is rubber or plastic: O-rings, idler bushings, a plastic gear that must not see hydrocarbon oil. It is also the more inert dielectric grease. None of that is the H2D lead-screw job.
Neither belongs on the X guide rail or the Y/Z linear rods. Those take oil. Super Lube 52004 is the matching oil; grease there packs the carriages and picks up dust.
Brass: neither has an active-sulfur EP package, so neither etches yellow metal. 21030 is still the better brass-nut grease because the PAO film actually stays in the contact and it passes a 100-hour salt-spray test. 92003 seals moisture well and does not attack brass, but it does not lubricate the thread as well.
Nothing the H2D actually greases is a plastic that 21030 is incompatible with.
I had also thought maybe Bambulab advised in favor of 92003 because it is more plastic compatible, but I haven’t found evidence that 21030 is incompatible with any plastic it might contact on the lead screws. As a check against Grok’s analysis, if anyone knows differently, please post.
The difference in viscosity perhaps explains why my subjective impression of 92003 was performing a bit “stiff”, which is what lead me to order the BX300/F I discussed above.
There had also been a long thread on the forum where people discussed their H2 brass nuts getting stripped out:
but in their case, the grease took on a brass color from the metal fragments, not the gray/black that I’m observing. For some reason they even thought they were better off replacing their H2 brass nuts with 3D printed plastic nuts, which on its face seems pretty radical.
FWIW: My H2D Pro that’s not allowed to talk to bambulab (or anywhere else on the Internet for that matter). NEVER gives any maintenance warning. All the progress bars remain stuck as if it just got maintained.
So I’d venture it’s not the printer’s firmware tracking maintenance at all, it’s bambulab’s cloud ‘spying’ on your printing and telling you when the printer needs cleaning and greasing. Sadly there’s nothing intelligent in the printer going on at all regarding maintenance needs.
As I’ve not looked into the variety of printers you are describing I can only talk to the adhesion issue.
It maters of the item to lubricants is horizontal or vertical because if it’s vertical you want to stick the lubricant to the surface so a grease is needed vs an oil.
My plan is in the USA is: Lucas Oil 10533 white lithium grease for the grease vertical screws Super Lube 52004/52008 ISO 68 light weight synthetic oil for the oiled rails (including the pain in the butt small one inside the print head that does the movable extruder )
The NIST study from 1913 on mineral oils isn’t really 92003 with modern lubricants and devices. Transformer oils are used as coolants in a transformer - not our use case.
So for the typewriter oil, after 3 months 98% of the oil is still in place.
For the heated chamber - that is printing with nylon or ASA , PC. Thats active use. And the 40hrs of model train maintenance obviously doesn’t apply in those conditions. My point with that 40 hrs is it’s operational hours. Of the train sits idle then there is no required re-oiling needed.
I’ve been printing almost exclusively PLA over the 3 months. Thats at the 16c in the room the printer sits in.
I think this may settle it: when the leadscrew sheers off the brass, if it’s large it remains brassy in color. But as it gets ground down into smaller particles, it oxidizes and becomes black. The whiteness of the grease makes it stand out even more. So, yeah, the blackness is from the nut.
But maybe that’s by design? The lead nut is the softer, sacrificial metal. If you accept that wear is unavoidable with even the best lubricants, then…maybe the black type is to be expected. The brassy type would be the red flag. That just leaves the one question: is wear truly unavoidable with even the best lubricants, and is the rapid blackening a sign of excessive wear, or is rapid visible blacking to be expected with even minimal wear?
[Edit: I subsequently had Grok evaluate this post to make sure I was being accurate, and it’s comments seem worth repeating:
refinement
The color mechanism is right. Coarse brass stays reflective. Once it is ground fine, the fresh surface oxidizes in the oil and the particle looks black. Grease does not stop that, because the particle is born in the contact, where the film is already broken. The black is nut metal.
On the H2D the rear nut is the one that goes, often around 1,500–3,000 hours, on machines that were greased on schedule. That rate is a parts or alignment problem, not the design working as intended.
Brassy is not the only red flag, and black is not the benign case. A gold chip means the thread tore off a piece: a burr, a misaligned nut, or a film that collapsed. A black paste means that tearing already happened and the chips were milled down. Both are wear. The difference is stage, not severity. A faint grey over a normal service interval is ordinary. Grease that is obviously black, especially with play at the nut, is the failure people in the stripped-nut thread actually reported, and a lot of those reports were black rather than gold.
The open question at the end has a direct answer. Some wear is unavoidable in a sliding brass nut. Rapid visible blackening is not what minimal wear looks like in a translucent PTFE grease. Minimal wear leaves a slight tint. A screw that turns the grease black between reminders is shedding enough brass to matter.
The TL;DR to my cliff-hanger question at the end: “A faint grey over a normal service interval is ordinary. …A screw that turns the grease black between reminders is shedding enough brass to matter.”
Naturally, just because Grok said it doesn’t by itself make it true, so if you know or think differently, please post a rebuttal as a check on Grok’s conclusion.
]
Ok so you personally dont feel the interval applies to you, so the company should change it for everyone despite mountains of evidence that calendar aging matters to oils?
If you’re so convinced that 3 months equals 2% and the laws of physics has changed since 1913 (it hasnt) then by all means press the button to ignore that as many times as you like. It doesnt mean bambu should ever change it.
I don’t consider a 113 year old study that covers a completely different type of lubricant under completely different conditions a “mountain of evidence”.
There are still folks that will change their motor oil on a car every 3000 miles or 3 months like it’s 1976. Fine. But as an engineer I need to be convinced with valid scientific evidence before I believe a default that just feels wrong. To each their own.
@Gerrit Since you were involved in that other lengthy thread (currently now 565 posts long), what’s your current conclusion in regards to all this, especially given the updates in this thread? After all, you are the designated “Guidance Counselor”, and this thread is calling out for guidance big time.
Guess you ignored the first link I posted, along with half of what I said. As an engineer you shouldn’t be having trouble understanding that hydrocarbons evaporate.
Well, according to the stats, you are a guided counselor also, as you have offered 10 solutions also.
The title doesn’t mean I am an expert on everything, and I also need to do research on certain topics and others come from experience.
I can also simply take the title down. There are 50 guided counselors on the community
But the stripped thread topic is about to pass 600 replies soon yes.
We all can add to that thread.
I am about to leave on a holiday, so I might be less active in the coming period.
@Gerrit Ah, OK. I hadn’t seen that title before, so I thought Bambulab had anointed you with an official capacity. You should leave the title up though. It suits you. (I mean that as a complement, as you generally do have very good insights and good guidance too).