Reducing waste from Bambu Lab AMS (lite) prints

bamboo lab printers are great at multimaterial printing apart from the excessive filament waste today we investigate methods for saving filament and [Music] [Music] time currently every 3D printer sold by bamboo lab has a capacity for multi material 3D printing that includes the AMS add-on for the X1 and P1 series as well as the AMS light for the A1 Mini which I'm using for today's video the multicolor prints do look great but the amount of time and filament needed is definitely a weakness so let's see what we can do to improve that starting by understanding how the system works most 3D printers that do multimaterial with a single nozzle have to back the filament out to vacate the hot end so the new filament can then be inserted but if the sequence isn't spot on this tends to leave a Blog which can cause jams and make the prints fail bamboo lab printers take a different and more reliable approach to this the filament path comes the whole way through the middle like on other machines and we're representing this here with the green line the difference here is that bamboo lab printers have a sharp blade built into the tool head activated with this lever that gives the end of the filament a very clean brake and makes it easy to withdraw it without any jams but now when the new color or material is fed in we have the tradeoff there's still a significant amount of the old filament that needs to go somewhere and where does it go well it gets extruded out as waist purges the printer needs to keep on pushing filament through until not only the old filament is gone but the output from the nozzle is completely made up of the new filament without any hint of the old one and as you can see here all of this uses up filament but also takes a considerable amount of time so why is it so important that the old filament is purged completely what we're trying trying to avoid here is bleed where there's bits of the old filament contaminating the new filament and in a multicolor print we can see this as a discoloration the glove on the left has it very slightly with a dark band near the wrist and the glove on the right has bleed the whole way through because for the majority of the Red glove it's discolored darker where the black before it hasn't been completely purged and we can see the same thing on this gnome where orange and a little bit of green has discolored the white area on the front because it hasn't been purged completely so bamboo lab made a trade-off they gained reliability in terms of the filament unloading and loading but they lost out in terms of the print duration as well as the amount of filament required so let's see what we can do to improve the situation to test out some improvements we need a suitable test print and I went for this Marin multimaterial Remix by CPUs firstly it's small so relatively fast to print and most importantly it has bands of black next to White which is the situation most likely to show up color bleed so when away this is kind of like a torture test I started by printing a version not in multicolor as a baseline I Shrunk the model down a little more and I lowered the infill density to only 5% as we can see it's mostly Hollow and the print time is just under half an hour the A1 Mini produced a nice and clean print although it does look a little bit plain in only white in this video we're going to test print time as well as the amount of filament used and the 5 G measured matches the 5 G predicted by the slicer which gives us the first data point on our graph obviously the waist filament column is empty because we're only printing in a single color so with the exact same size and slicing settings we now apply colors to the different parts of the model after we slice there's a couple of things to note the total filament has gone from 5 G to 104 and the print time from half an hour to over 7 and 1/2 hours but even the base model without any of the waist uses four times more filament and that's because if we look inside rather than and all of this still being hollow each color is treated as its own separate area with perimeters and solid infill for the top and bottom layers to help track the results when printing this version and all of the ones to follow I set up a tray to catch all of the filament wastage this allowed me to use scales to verify what the slicer was predicting and when I spread out and flatten all of the waist filament it acts like a visual companion to the graphing of the results which now looks like this and as we can see the system is very inefficient I printed a lot of these models and not one of them failed but this is the trade-off that we discussed earlier it's actually pretty Stark seeing those two side by side so let's get into our changes the first thing we're going to do is to come to others and then come down to flush options and we're going to tick flush into objects infill and flush into objects support now the support one isn't going to do anything here because I don't have support enabled but flush into infill will show a difference and be indicative of what both settings do this is what what the infill looked like beforehand and this is what it looks like after where this area previously only had orange filament infill the slicer is now attempting to do some of the purging inside which should mean a reduction in those blobs that get wasted but looking at the graph we can see the result is almost meaningless the print time is still exactly the same and we've only saved 02 of a gam from this setting and you're probably thinking the impact will be minimal because I only had 5% infill so I sliced two versions at 50% infill and we can see the infill is now a lot denser inside the model and when flush into infill is turned on there's a lot more potential for the purging to take place within but even so the results aren't really that different we only Sav 2 minutes of print time and under 2 G of filament I printed the 5% infill version anyway just to see if I could see the infill through the white sections and as far as I can tell it's not evident at all when we compare it to the Baseline on the left to my eyes these look absolutely identical so it's it's probably worth turning on this setting but don't expect any meaningful gains but don't worry we can do better thanks to a suggestion from my Patron David in bamboo studio and oras lcer there's a button called flushing volumes when we click it it displays a table that informs us how much filament will be purged when changing between various colors now interestingly every value here is 280 mm Cubed on the left we have the from color and across the top the twocc color which means if we're changing from Green to Black The Purge value is 280 but if we're going from Orange to Green the value is also 280 which is quite unusual but if we click Auto Cal it will make more intelligent decisions for instance if we're going from a light color like green to a dark color like black the black is a lot stronger and we won't need to purge as much before the black takes over which is why the value is lowered to 114 the opposite of that is when we switch from a dark color like black to a light color like white we need to purge for a lot longer to make sure all of the Black is gone which is why that value is now much higher at 557 so this version is more nuanced but when we click okay and reslice we actually use more filament overall the old value was just over 80 g and now it's ballooned up to 98 G and that's because overall the flush volumes are far too conservative in trying to get rid of color bed to illustrate the problem I printed out some physical versions of the volume of filament that would be purged up the top we have our Vault 280 and when using autocal we can see that some of them are smaller in safe filament but the bigger ones are so much bigger that over the balance of a print all of those gains are lost and we end up using more filament so let's click on flushing volumes again but this time come down to the global multiplier David has experimented and told me that we can have it and still be safe from bleed if we click okay and then reopen this dialogue we can see that all of the autoc calculated values are now hared and we finally start start to see some gains in efficiency by harving that Global multiplier for filament flushing we save over 28 G in filament and we even save 50 minutes from our print time so the question is is print quality affected looking at the final result I would say absolutely not I can't see any bleed here whatsoever and comparing our original multicolor print on the left to our new one with 50% flushing I still can't see any deterioration in the results we can see a difference when it comes to the filament that gets purged the original one on the left has completely transitioned through the white whereas the 50% reduced version on the right is white with a tinge of orange by the time it finishes however the last of the previous color is taken out in this purge block and that looks quite different to the original as well as we clicked auto calculation it's now concentrating on purging white rather than purging every color uniformly finally some improvements but in some situations we can do even better the next thing I wanted to look at was the GCO code sequence for when the material was changed that's the sequence that starts when the filament is severed with the cutter and then the tool head Parks over on the left where the old filament is withdrawn the new filament inserted and the purging takes place in the slicer we can get to this by going to the printer profile going to the machine G-Code Tab and then scrolling down until we see filament change G-Code in this format it's almost impossible to read so I copied it into a text editor and did some indenting to make the conditional statements a little clearer unfortunately this was more or less a dead end because bamboo lab have a lot of specific g-codes and as far as I can find there's no published references for them we can also see that a variable for flush length is used in this G-Code which means we don't need to make any changes here because the changes made in our flushing volumes will already translate over to this G-Code perhaps I could have played with some of the back and forth X movements that spit out the filament but I didn't really feel it was worth it looking for more efficiency I thought laterally for this model there are 229 filament changes and in situations where you need more than one printing five or more at a time will still have 229 filament changes and considering most of the time is spent changing filament and purging the amount of wasted filament stays the same and we don't really have that much increase in print time the graph demonstrates this really well the orange column for model filament understandably grows but the waist filament stays the same and the print duration only goes up a little bit as you would hope there's no reduction in quality from printing duplicates your only risk is a little bit of stringing between the models but that wasn't a problem for me and these models look as good as the original here's a real world graph showing all of our results we have our Baseline on the left second in is purging into infill not much change next the 50% reduction in flush multiplier which is very worthwhile and finally printing five at once which gives us five times the model output without increasing the waste hopefully there's something in this video that's helpful for you and if you have any further tips please leave them down in the comment section this type of system will always be reliable yet inefficient which is why I have a five head tool changer prusa XL on order and after years it's finally being shipped expect to see some videos about that in the near future as well as a repeat of this test to see how it Stacks up thank you so much for watching and until next time happy efficient multimaterial 3D printing good day it's m again if you like the video then please click like if you want to see more content like this in future click subscribe and make sure you click on the Bell to receive every notification if you really want to support the channel and see exclusive content become a patron visit my patreon page see you next time

Channel: Teaching Tech

Views: 235,028

Keywords: 3d printing, 3d printer, 3d printed, 3d print, ams, ams lite, bl, bambu lab, bamboo lab, a1 mini, x1 carbon, x1c, x1, p1p, p1s, waste, filament, flush, flushing, purge, purging, reduce, reducing, reduction, minimise, minimize, ratio, multiplier, infill, support, material, pla, efficient, effiviency, improve, how to, guide, tutorial, experiments

Id: p3ZIM3megIU

Length: 11min 48sec (708 seconds)

Published: Fri Oct 27 2023