The Extrusion Myth We All Believed

this is brute force my extrusion force testing machine it works by measuring the force needed to push plastic through a 3d printer hotend in some recent testing of the e3d revo which is this one fitted we observed some ripples in the force data and then calculated that these must have come from the interaction of the two main drive gears in the clone dual drive extruder that we were using for this testing now though we need to find out if this is just because it was a poorly made clone or of the original bondtech design for dual drive extrusion is a little bit flawed [Music] while many 3d printers now use an extruder where the filament is driven from two sides this is a relatively new thing the way we used to do it back in the good old days was a hobbed bolt which was literally a bolt and an idler bearing when the innovation of dual drive extrusion came along it was thought to be an improvement as it increased the contact area with the filament which in theory means stronger grip more force and more reliable printing right well i have some questions about that like does dual drive extrusion actually provide more force is more falls even needed and if it is needed are we doing it the best way possible i'm going to try and answer these questions today by running some tests on different extruders to see what effect dual drive gears has on the outcome my setup obviously involves brute forsythe and he'll be fitted with three different extruders the clone dual drive extruder a genuine e3d titan and a genuine bontec bmgm the hot end will always be the revo 6 the filament will just always be pla and the temperature 220 degrees celsius the flow rates that we're going to run through are 2 6 10 14 and 18 millimeters cubed per second in addition to the normal methods of analysis like taking averages and standard deviation i'm also going to be using fourier analysis which is quite a massive operation but all you really need to know is that it will highlight frequencies that are involved in impacting extrusion force if you want to know more about it there is a link in the description to a fantastic three blue one brown video that explains it much better than i can to make sure that the tests we're going to do on the new extruders will show ripples we need to run the same tests on the clone drive system that we know will definitely make them just to test the test if you like as you can see we can confirm that the ripples still appear their frequency is what we expect and the other random drops as well as the three kilogram limit also still show up so that's a green light to go we're good to run more tests using this setup time for some real testing in my mind if we want to verify that the dual drive gears are causing the problem we need to remove them from the situation without changing anything else and see if the problem persists using an e3d titan which is this one right here would kind of do that but it does change quite a few other things too what i decided to do was just file the teeth off the second drive gear so that it becomes an idler our test is now basically identical in every single way to our control but it's no longer dual drive looking at the results it does look like the ripples have gone for the most part but with some present still at the lowest flow rate using very analysis to identify any dominant frequencies we find that the previous ripple frequency has basically gone but others still exist and we seem to have created a new one in the process comparing the mean force this didn't really change much at all between the two tests which means that the second gear didn't really add any force as such but it also means that the hot end provided equal resistance with and without the second driving gear interestingly i think i might have actually improved this extruder slightly with my modification because the standard deviation and the range which are like measures of how far the force varies are reduced meaning the extruder was able to provide a more consistent force without the second drive gear let's take a look at the single drive gear titan now to see if that reveals anything more interesting uh what this is not what i expected at all the main things that jump out at me straight away are the high forces the massive waves and how different this is to the low forces that look very smooth let's utilize our friend for rear analysis on test number 15 to see what's going on straight away we see peaks at 0.371 hertz and 0.742 hertz exactly double these correspond to around 20.2 and 10.1 millimeter pitch on the filament the closest i can see for a match on the extruder is around 23.8 millimeters which is the circumference of the drive gear it seems reasonable that this could be the cause but it's two millimeters difference which is quite a lot analysis of the 14 and 10 millimeter cube per second flow rates shows pitches of 21.3 which is a bit closer to the ideal 23.8 but still seems a bit far away maybe there are measurement areas i'm not aware of yet definitely something to investigate later looking at the lower flow rates overall we can see they're much smoother than we found on the clone extruder but that doesn't mean we can't use fury analysis to highlight important features that are impacting the force for example we find the effects of the large gear teeth at 0.34 millimeter pitch as well as the effects of the hobbed teeth biting the filament at 0.75 millimeter pitch both matching well for the two and six millimeter cube per second tests going back to the basics for a minute and comparing to the unmodified clone extruder the control at the start the average forces are similar up to 10 millimeter cube per second but increase significantly for the titan at 14 and is more than double at 18 millimeter cube per second it's a different story for the range and standard deviation though with the exception of again 18 millimeter cube per second as the titan far outperforms the clone with less than half the standard deviation in nearly every case and a half to a third of the range which really is quite a big improvement so what about the high flow rates and average forces then they don't seem to make any sense how can they possibly be larger on the titan than the dual drive clone especially when they were both using the same hot end i think we can explain this by two points firstly the e3d titan must grip the filament better than the clone this is the only way it would be able to apply more force secondly given the first conclusion and the known slipping of the clone there must be more filament being extruded by the titan compared to the clone if we apply the same amount of thermal energy then to a larger amount of filament the temperature will rise less meaning it's more viscous meaning more force is required for that extrusion and i think that does explain why the forces could creep up it seems like a small cause having a very large effect but i do think this is the most likely scenario maybe testing the bontec bmgm will give us more insight this looks to be following patterns much more similar to the titan than the clone extruder which i guess is a good thing let's have a deeper look starting in with the high forces much like the titan the forces on the bontec really saw at 18 millimeters cube per second and also experienced strong frequency patterns our investigation suggested it could have been the drive gear on the titan so what about the bond tech looking at the fourier analysis for test 15 there are two main areas with peaks around 0.410 hertz and 0.742 hertz which translates to 18.3 and 10.1 millimeters respectively 18.3 is approaching but still quite a few millimeters away from the drive gear circumference at around 24 and even more strangely the 0.742 hertz is exactly the same frequency of wave that we saw with the e3d titan a totally different extruder with totally different geometry while it's possible that this is just a coincidence perhaps it's actually a component of the system that didn't change maybe something about this stepper motor rotation hot end melt rate or spring stiffness of the load cell maybe even a combination of all of these that's going to be a question for another time though because i don't have all the information now for now let's concentrate on lower flow rates and forces fourier analysis indicates disturbances from the teeth on the large gear and the hob teeth on the drive gear on multiple tests but looking at the graphs we can see that while these are present these effects are actually quite small in fact for flow rates up to 14 millimeter cube per second the standard deviation range and average all quite comparable to the titan 18 millimeter cube per second is a bit of a different story though increasing to nearly 7.3 kilograms of force on the bontek compared to 6.3 on the titan with range and standard deviation also significantly larger so having run through all that data what does it really mean and what answers have we found mainly of course the extrusion force is just the most awesome thing ever jokes aside though i think we need to be careful in how we interpret the results for this and remember what we're really measuring the load cell is measuring how much the hot end resists the flow this is equal to the force provided but it's not the same thing i believe what we're looking for in an ideal extruder is very consistent forces which leads to consistent flow but this in turn seems to lead to higher resistance which demands higher force measuring high forces does not mean it's a stronger extruder it means there is more resistance but of course provided the extruder can sustain that force constantly it's actually not really a problem based on that i think we can conclude the clone extruder is pretty terrible compared to the real thing the forces fluctuated a lot more the filament skipped even at low forces and the result of that was poor extrusion even filing back the teeth on the dual drive gear system seemed to improve the design upon closer inspection i did actually find the idler only contained one of the little needle bearings installed instead of two which probably could have contributed quite significantly to the issues that we saw and explains some of the random behavior but that's what was provided with that clone extruder the titan's performance was definitely a vast improvement compared to the clone the level of consistency we saw at the low flow rates was pretty impressive and the amount of grip it was able to provide up to 6.3 kilos was well also fairly remarkable when looking at the bmgm we saw very similar patterns that we did with the titan but with even higher forces which with my current understanding is not really a good thing by my recollection it didn't slip on these tests though proving a strong grip on the filament the whole time so let me try and answer the questions proposed at the start of the video does dual drive extrusion actually provide more force is more force if needed and if it is needed are we doing it the best way possible firstly yes it does provide more thoughts but i don't think it's a good thing and i also don't think it matters the force exerted by the extruder is not really determined by the extruder the force will be applied exactly as necessary to overcome the restrictions of the hot-ended nozzle an extruder that provides more force to do the same job must surely be negatively impacting the nozzle's resistance by some means potentially for example like inconsistent extrusion velocity however the reason why none of that matters is that more force does not mean more performance provided your whole system is operating correctly if filament starts to grind on the extruder teeth this can only mean that the force needed to extrude has increased and this can only come from the hot end during normal operation this would only happen if your hotend is unable to transfer any more thermal energy into the filament in other words the melting capabilities of the hot end have been reached attempting to grip the filament more strongly and push it harder will only lead to poor extrusion quality what i'm saying is that once the threshold for good extrusion has been met by the extruder exceeding this will not provide any additional benefit from testing here today this seems to be entirely possible on a single drive titan extruder does this mean that all extruder development is pointless then no absolutely not getting the extruder into a smaller and lighter package that can provide more consistent linear motion of filament seems like it would be the key to the best print performance possible thank you very much for listening and i'll see you in the next one

Channel: Vector 3D

Views: 73,028

Keywords: 3D Printer, 3D Printing, bondtech bmg, bondtech vs e3d titan, e3d titan, single drive extruder, extrusion force tested, bondtech extrusion for test, dual drive flawed, is dual drive broken, 3d printed, extruder inconsistent, extruder inconsistent extrusion, inconsistent extrusion, prusa inconsistent extrusion, issue 602, voron issue 6, 3d printing salmon skin, mk3s inconsistent extrusion, inconsistent extrusion velocity, extrusion myth, 3d printer myth

Id: moSZ05FkLQU

Length: 13min 30sec (810 seconds)

Published: Wed Mar 16 2022