Best filament for a battery enclosure?

Bambu Filament and Accesories
41

As it happens, Ultimaker sells their own label of PET-CF filament, and they provide a fair amount of information on how to anneal it, including how much it will shrink along the XY dimensions and a different amount for Z

Their free Ultimaker Cura slicer software can even automatically compensate for shrinkage. I downloaded it and gave it a quick look. It is compatible with what looks like a pretty long list of printers, but, alas, Bambulab printers are not among them.

Their recommended temperature profile for doing the annealing looks a lot more fancy than anythiing else I have yet seen.

Yesterday I ordered a Kg of the Bambulab PET-CF, so I’ll give it a try and see how it goes.

1 Like
42

This is the first I’ve heard about it. Thanks for the head’s up.

43

Interesting, I was just about to say that in order to pre-compensate for the shrinkage you could probably do a standard material shrinkage test such as the CaliCross (Printables). I’ve personally used it before to calibrate material shrinkage for anything that I print with that I care about dimensional accuracy on.

But having the manufacture specify that information certainly takes a lot of the guess work out of it.

However, I am still a little concerned about warping, mainly from this CNC Kitchen annealing in salt/sand video which I watched a while back, and I believe is the one you were referencing previously:

P.S. In my opinion, material shrinkage doesn’t get nearly enough attention in the discussion of producing dimensionally accurate regular 3D printed parts (let alone annealed parts), despite it being pretty critical to producing any kind of functional part that is larger than say 150mm.

And this is not a problem that solely affects notoriously high shrinkage materials such as ABS. Even for Bambu’s own PLA and PETG, I measured an average shrinkage of around 99.7% for PLA and 99.69% for PETG. So for a part that is say 200mm long, it would end up coming out a whole 0.6mm short just from shrinkage if not compensated for…

Furthermore, many slicers don’t even have an option to set the material shrinkage rate in the slicer’s material profile (including Bambu Studio!), and the very few that do have the option (Orca Slicer, Super Slicer, etc.), only have it set to 100% by default.

I always found it perplexing that a company such as Bambu that markets itself as completely beginner-friendly/plug and play, and says you can produce dimensionally accurate parts, does not even have a field in their slicer to compensate for shrinkage, let alone a pre-tuned value for their own filaments!

Again my opinion, but I think if Bambu simply included a pre-tuned shrinkage value for these materials in the material profile, as well as a pre-tuned X-Y hole compensation value, then this would probably solve a lot of an dimensional accuracy issues people experience with their printers, especially if you are a new user and don’t know any better, and would go a long way to bringing the printers within the “standard” FDM accuracy goal of ±0.2mm.

I at least know from my own tests with the A1, calibrating material shrinkage as well as XY hole and contour compensation (I ended up leaving the latter as 0 in my case) certainly helped improve the dimensional accuracy of the parts compared to the defaults of no compensation.

44

Thanks for the heads up - would you agree with bignazpwns’s warning about the stock activated carbon filter on the P1S/X1C not being sufficient for printing Bambu’s PET-CF (or any other Bambu filaments) and that an additional external activated carbon filter would be required?

I’m mainly asking out of curiosity, I probably won’t be using PET-CF as my material of choice mainly due to the lower impact resistance than say ABS/ASA, as well as it not being compatible with the AMS (so no chance of using secondary material as support interface for perfect overhangs).

45

Yes you are correct that there are certifications for things such as general lithium ion batteries. For example, the UN 38.3 test. Unfortunately this kind of test not only requires a lot of money to have officially undertaken and for your product to be certified (around $5000-7000), but there are 2 further key issues with the test itself:

  1. You need to submit at least 16 copies of the battery pack you have designed for testing. For a one-off battery pack that is already costing me around £450 in battery cells and parts (let alone testing equipment), this is really not financially feasible, especially in addition to the $5000 minimum testing fee.
  2. Even if you were to not have your pack officially tested, and decide to try and carry out the tests at home, some of the tests require lab grade testing machinery, and many of the tests are destructive in nature, which would mean that even if the pack passes the test and does not undergo a major malfunction such as thermal runaway, the cells would likely be rendered permanently inoperable, which amounts to essentially throwing away the money you spent on them.

As an example, here’s an overview of the UN 38.3 tests:

  • T1 – Altitude Simulation (Simulation to 50,000 ft altitude atmosphere)
  • T2 – Thermal Test (12-hour dwell times at -40°C and 72°C, repeated 10 times)
  • T3 – Vibration (1G from 7Hz to 18Hz, 2g from >18Hz to 200Hz, 3-hour test in each axis)
  • T4 – Shock (34.6g shock pulses in positive and negative direction)
  • T5 – External Short Circuit (Pack conditioned at 57C, short maintained for 1-hour or more)
  • T6 – Impact (9.1kg mass dropped from 61cm)
  • T7 – Overcharge (Charge at twice manufacturer’s recommended current)
  • T8 – Forced Discharge (12 V power supply connected in series at maximum recommended current)

Tests 1-4 I do not have the relevant equipment in order to perform. For the remaining tests, I’m assuming that the BMS is supposed to be bypassed (otherwise for example short circuit protection would kick in immediately and would remain active for the full hour), in which case test 5 would likely end with the cells being left permanently damaged and unable to hold a charge, and test 7 & 8 would at a minimum shorten the lifespan of the cells.

As for guidelines for the manufacturing of safe lithium ion batteries or other ebike systems, I have found several papers, presentations and articles online from a variety of sources (such as the NASA presentation I linked in an earlier post), but I have yet to find specific UK government regulations which dictate exactly how a battery must be constructed in order to pass some relevant certification.

But if anyone knows such a specification for the UK that is relevant for li-ion batteries or ebike systems, then I would be happy to read it and consider it when finishing the design.

46

So far I haven’t yet found anything which implicates PET-CF, unless you were to perhaps light it on fire and breath the fumes it gave off during combustion. I did a quick read of the Ultimaker MSDS for PET-CF, and if anything, it makes PET-CF sound almost benign

There was an interestng thread here:

which mentions someone who thought he became sick while printing PETG-CF. There was also further talk in that thread about how a number of people thought the carbon filter built into the X1 was not effective.

Also, another guy from the bambulab community said he got seriously sick from printing ABS and ASA an had to go to the hospital:

I have more links to chaase down, but if there’s a smoking gun regarding PET-jCF, I haven’t yet found it, at least not yet.

What I can agree with is that if you want to completely scrub the air in one pass, then I can certainly believe the tiny filter in the X1 will be inadequate for that. On the other hand, if you were in no hurry and could wait for it to cycle through the air say, 1000 times, then maybe by the end of that it will have reached past the point of diminishing returns,and doing more after that might likewise not have much of an effect.

Regardless, there are certainly chemicals that even activated carbon is unable to pull out of the air, so for that reason I’m inclined to think that venting to the outside is just obviously going to be the better way.

2 Likes
47

The activated carbon filter does kind of work when the chamber fan is running, but the issue is that the air passing through the filter is only partially scrubbed before exiting. Plus it only filters VOC’s and not small particles which would be one of the issues with CF filaments. The most important issue, which you would really want it for in the first place, is that you typically leave the chamber fan off for the filaments that put out VOCs (ABS/ASA) to prevent warping. This completely negates the point of having it.

I personally use a bento box, which is a HEPA and activated carbon filter air scrubber that lives inside the printer chamber. Its purpose is to recirculate the air inside the printer, filtering out VOCs and small particles with each pass. I wouldn’t call it the best solution, but it does very well. I also have an air purifier which also has a HEPA and activiated carbon filter. This combination is probably good enough for most situations, but I honestly found that venting outside is the best solution. My vent solution has an inline duct fan that pushes the air outside. It maintains negative pressure in the print chamber, which prevents all VOCs and small particles from escaping.

1 Like
48

I agree, venting outside is the best answer, but it’s also my current struggle. While my venting solution has been working quite well to keep VOCs and small particles from escaping my print chamber, I’m finding that it also slightly lowers my chamber temperature. I need to find or design a solution to keep my chamber temperatures up to assist with printing materials that are prone to warping while keeping the vent fan running.

49

Can you maybe defer your venting until after the printing is completely over? You could have one or more big air scrubbers in the same room but outside your printer to hopefully continually mop up anything that might leak out of the printer during the printing. Maybe there would be other ways also, but at least this would be easy and seemingly straight forward.

Edit: Oh, I just now read your post just above that, and I see that you’re already doing it.

50

If you got a big enough air filter (one or maybe two of them) with an h13 or h14 filter in it, then I’m guessing it could clean the air fast enough that it could keep up in real-time with whatever is leaking out and thereby stay ahead of the curve.

51

That’s basically how I handle it right now. For prints that I’m not worried will warp, I will just keep the vent fan on since it’ll only typically drop by a degree or two. For large or warp prone prints, I just leave it off and let the bento box and air purifier do their job.

Since the vent fan does a good job at evacuating the air without causing a big hit to temperature, I’d really like to figure out a method that will let me maintain or go a bit higher in chamber temperatures while leaving the fan on. Right now I’m looking into outside insulation for the printer and/or an internal heat source.

52

The bambulab pet-cf seems to sell very quickly. Yesterday the 1kg rolls came into stock. I ordered one then, and now the 1kg rolls seem to already be on backorder. Suposedly the 500g rolls are cooming into stock tomorrow but can be ordered now, so I did.

53

I have also tried a printed attachment, different to the one you have planned, but the important thing to learn from this is that shocks should not be underestimated during normal riding. The total weight of the battery results in high forces acting on the connection. Sooner or later these connections will break, in my test it only took a few kilometres. Basically, you would have to use metal connections. But then it could happen that your housing breaks out. What I can recommend are Velcro straps. Thread them all the way round or through the housing! You need about 3cm of fastening surface to ensure that they hold securely, but more is better. The advantage is that the straps do not tear, they absorb shocks and hold the battery on the bike. Even if the battery has a bit of play, it wouldn’t matter. It won’t fly off.

Unbenannt
Unbenannt1

This is the housing made of PLA. You can clearly see signs of use. But these will not disappear over time. These short Velcro straps, which are threaded through the housing, hold the battery securely on the carrier even in the event of potholes.
Nothing has ever broken into the housing either, there are also electronics built in, soldered cables and so on. You just have to make sure that everything in the housing is immobile (I used foam to fill gaps).

As far as the use of materials is concerned, I can’t say anything against PLA. It is inexpensive, stable in the required respects and can be printed relatively easily and quickly. However, as I have already explained, I would consider PETG for a few other reasons. More expensive materials are not always better just because they theoretically achieve certain measured values under laboratory conditions.
Or why does ABS break faster than PLA? Or why does PAHT-CF break faster than PLA? Why can a PLA withstand an extreme bending test and ABS cannot?

I will list a few factors again and then it should be good (everyone has to gain experience themselves and everyone is welcome to burn as much money as they want). In the pictures you can see a housing made of PLA:

  1. it has been in use for about a year.
  2. it has been used in summer up to 40°C and shows no damage due to temperature.
  3. it survived a fall from the bike while travelling. Without sustaining any significant damage.
  4. it is not waterproof as it is not intended to be submerged in water. It can withstand any downpour, even heavy rain. It does not soften or crumble just because it has been exposed to moisture. Due to the ventilation grille, the inside of the housing should have been under water after being exposed to heavy rain. Since it was opened a short time later for other reasons, it turned out that there was no water inside. Electronics etc. all remained dry, no corrosion.

What is most important is the construction of such things, it makes or breaks a lot of things, no matter what material you use. The colour white or lighter colours always contribute to the fact that heat is not absorbed as much as by black (I think we all took part in physics lessons). Such a housing can be protected against adverse environmental influences by coating it. Inside and out.

Have fun building! :slight_smile:

54

That is a good objection. To counteract this, it may make sense to paint the housing. But at least against water, I painted my PLA housing with matt clear lacquer on the outside and inside.

55

Thanks for the info on a practical solution to this - I’ll look into the bento box and nevermore filter when I have some time. I’m guessing they will require some printed parts to create, but as a temporary solution I do at least have a small air filter with an H13 HEPA filter and activated carbon filter the I mentioned previously, which I suppose I could just put right next to the printer and have it running continuously whenever the printer is on.

For the bento box/nevermore, how exactly can you hook them up to the printer such that they are turned on automatically when the printer starts, and turn off a while after the printer ends, so they don’t run 24/7 even when the printer is not being used? For an open source printer such as a klipper machine, I can see it being easy to hook up some kind of control circuit such as mosfet to control the power via a GPIO pin on the raspberry pi, but for a Bambu printer such as the P1S or X1C, that seems like it would be much more difficult?

I may also consider getting an air quality monitor that can detect particles (such as PM2.5) and VOCs to place near the printer for piece of mind. One example of such a monitor is one sold by IKEA, if anyone is interested:

But regardless, it is very dissapointing that there is this kind of “false marketing” surrounding the P1S and X1C saying that they are safe to use with ABS and carbon filled filaments, when the activated carbon filter won’t filter carbon fibre particles as you said, it won’t be used when printing ABS to prevent warping, and even when it is in use, it’s not 100% effective. It’s rather scary that they can sell such printers as being safe to use in a normal environment (as opposed to an industrial setting with special air filtering in the room), and yet clearly the filtration isn’t effective, as mentioned by yourself, @NeverDie and the posts he linked where people got sick from using their bambu printers…

If anything, it does make me reconsider my upcoming decision to get the P1S, since it was supposed to be a competent and safe entry into printing materials such as ABS that need enclosures to prevent warping, and need filtration to prevent health hazards. If there are going to be all these important safety issues when using even non-filled ABS/ASA with the P1S, then it certainly makes me wonder whether I should just go with P1P and stick to PETG…

56

You can can print your own bento box or you can purchase a printed and partially assembled one from Voxel. I didn’t feel like sourcing my own hardware and printing one, so I purchased mine from Voxel for $35. Figured that was worth saving me from the hassle of finding my own fans, magnets, activated carbon, HEPA filter, and power supply.

For controlling the power of the bento box I just utilize a smart power outlet, which is controlled via my Home Assistant server. I have my printer, bento box, vent fan, and air purifier all controlled by a smart strip. This gives me the ability to individually control them however I need. I don’t run my bento box unless I’m printing filaments that need the filtration. I found that at least PLA does not appear to emit small particles or VOCs while I print with my door closed, so it’s not worth running the bento box. The great part about controlling these through Home Assistant, I can create automations for tasks that I want. For example, I have my vent fan turn on any time the chamber fan of the printer is on. For another example, I turn on my air purifier anytime certain metrics for VOCs or small particles are met. So if I wanted to turn on the bento box when a print starts and stop when the print is finished, I’d create an automation that looks for the start and finish steps of the printer that the HA server is capturing.

I picked up one of those IKEA AQ monitor recently. It’s pretty accurate from what I’ve seen, but it requires extra hardware in order to allow Home Assistant to read its data. Plus it doesn’t provide a lot of data; the VOC sensor output is just an arrow that points up, down, or to the right. Up for when VOCs are going above normal readings, down for when they are dropping below normal, and to the right when they at normal levels.

I ended up researching and finding my AQ monitor from AirGradient. It is fully open source and the various sensors can be replaced and/or potentially upgraded. It captures quite a bit of data: Temp, humidity, small particles (PM0.3, PM1.0, PM2.5, PM10.0), VOC levels, NOx levels, and CO2. Not as cheap as the IKEA one or what you can find on Amazon, but honestly worth it in my opinion.

All this talk about toxic fumes would surely scare just about anyone. In reality, the potential for being poisoned or getting sick very much depends on the environment of where you print. In my case, my printer literally sits in the same office as me, so I kind of went overkill on researching and implementing different levels of mitigation. What’s important to understand is how much of the air is saturated with the fumes and how much if it are you breathing. If you print in a separate room in your house that you don’t work in like I do, then you really don’t need to worry. You also have to remember that air in your house is typically being moved around and being replaced quite often, greatly reducing the saturation of fumes. A simple thing of just having good ventilation in a room that can replace the air often is enough to mitigate the fumes; the basic act of opening a window does that. Lastly, just having an enclosed printer contains a surprising amount of the fumes, that’s why a bento box and an air purifier would pretty much take care of anything that might be harmful.

1 Like
57

Forgot to mention … turning on my air purifier almost instantly drops small particle readings to zero and VOCs not too long after. Since it’s moving air in the room, it’s reducing the saturation of fumes that are clumping together around the printer. While it runs, small particles stay zero and VOCs drop below baseline levels.

1 Like
58

Does it do a true PM0.3 measurement, or is that figure some kind of software estimation? I’ve tried looking into this in the past, and this is where I had trouble getting a clear answer. For instance, does it even matter which it is? if it’s not a true measurement, is it just as accurate? If not, how big is the possible error?

59

What are you meaning by a “true” measurement? The monitor uses a Plantower PMS5003 pm sensor. The sensor uses laser light scattering to detect particle sizes. From everything I’ve read, it does okay ish at less than PM1.0, but relatively accurate at PM1.0 and larger. So the PM0.3 readings aren’t exactly accurate, but regardless of its accuracy, it does show when levels are rising or lowering and that’s all I’m really looking for right now.

Sensors like the Plantower PMS5003 are designed to be a low cost option. These perfectly work for using captured data to see trending and provide relatively accurate data on whats currently going on the air around the sensor. Someone like myself can’t reasonably justify spending the amount of money needed to get lab grade results. My level of research is just “Did it go up or down?” and “Roughly how much did it change?”.

1 Like
60

I don’t have a good answer to that, because my thoughts on the matter are muddled. I think the source of my confusion is that quite often when I see these sensors referred to on various websites, they are referred to as PM2.5/PM10 sensors. Here’s an example of that: Sensing the Air Quality: Research on Air Quality Sensors
Yet the sensor itself reports more than that. So…iwhy don’t they seem to be getting credit for that? Why aren’t they referred to as PM0.3/PM2.5/PM10 sensors, or whatever their capability actually is? OK, so maybe it’s a stupid question. I’m just not sure. But most often things get badged the way they do for some sort of reason.