Slic3r PE: Please reduce default values for "max volumetric speed" from 15 to 11.5 mm³/s
 

Slic3r PE: Please reduce default values for "max volumetric speed" from 15 to 11.5 mm³/s  

Page 4 / 4
  RSS
Tim
(@tim-m30)
Illustrious Member

Okay - now for a boring test to check a more continuous extrusion rate.

Test was a simple square with rectilinear infill, with Slic3r PE maximum volume rate set to 15 and 18 ... 

15 mm3/s  = all is well.

18 mm3/s  = adhesion suffering and the occasional extruder click.

Also tried 25 but it was a definite no go.

It is always wise to get more than one opinion......
Posted : 16/05/2019 1:10 am
Tim
(@tim-m30)
Illustrious Member
Posted by: Tim

Also tried 25 but it was a definite no go.

Let me qualify that : 25 mm³/s worked the first couple passes ... then died an inglorious death of clicking, skipping, and general mayhem.

It is always wise to get more than one opinion......
Posted : 16/05/2019 1:34 am
bobstro
(@bobstro)
Illustrious Member
Posted by: Tim

Bob - changing layers from 0.25 to 0.35 is a major part of the V6 gain (30% faster, right there), probably as much as the increased diameter.   

Definitely agreed. My point was that an $8 nozzle yields a significant amount of improvement without going the full route with an extruder upgrade. The V6 can handle up to 0.80mm nozzles pretty well.

My goal here is to better understand why people are saying 15 mm3/s is unrealistic for a V6 on my Mk3 ... lol.  So far, I haven't convinced myself Prusa choose a wrong number.  

For me, the reason I'm saying it's overly optimistic is for the same reason I feel the 200mm/s speed for infill is overly optimistic. When I slice a print large enough to actually hit the limits (as verified in slicer preview), I get extruder clicking, skips and jams. Dropping MVS modestly takes care of most such situations. Similarly, with other viscous materials, lower settings take care of many problems attributable to insufficient heating time and nozzle back pressure, notably flexibles.

My notes and disclaimers on 3D printing and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan...
Posted : 16/05/2019 3:53 am
bobstro
(@bobstro)
Illustrious Member
Posted by: Tim

Let me qualify that : 25 mm³/s worked the first couple passes ... then died an inglorious death of clicking, skipping, and general mayhem.

I believe in NASA parlance that would be a "successful partial launch".

My notes and disclaimers on 3D printing and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan...
Posted : 16/05/2019 3:54 am
Tim and Sembazuru liked
Tim
(@tim-m30)
Illustrious Member
Posted by: bobstro

For me, the reason I'm saying it's overly optimistic is for the same reason I feel the 200mm/s speed for infill is overly optimistic. When I slice a print large enough to actually hit the limits (as verified in slicer preview), I get extruder clicking, skips and jams. Dropping MVS modestly takes care of most such situations. Similarly, with other viscous materials, lower settings take care of many problems attributable to insufficient heating time and nozzle back pressure, notably flexibles.

I had issues with 200 mm/s infill - but usually with rectilinear, and then only until I figured out Slic3r was only providing half the requisite material for the pattern.  Rectilinear is every other layer, hence it needs a lot more than a planned 0.45 width. It really needs twice that for the ideal, but you can get by with 150% and get a reasonable interface.  Grid is bad because it requires the nozzle to remelt the prior path when it completes the crosshatch.  Grid wears nozzles.

I do lots of prints that are full size, and presses the printer to its limits ... I have never run into the 15 mm³/s limitation in a bad way.  At least that I can trace back to that one factor as root cause.

Seriously, printing has been rather dull and boring since I replaced my heat break.

This post was modified 1 year ago by Tim
It is always wise to get more than one opinion......
Posted : 16/05/2019 5:33 am
Tim
(@tim-m30)
Illustrious Member
Posted by: bobstro
Posted by: Tim

Let me qualify that : 25 mm³/s worked the first couple passes ... then died an inglorious death of clicking, skipping, and general mayhem.

I believe in NASA parlance that would be a "successful partial launch".

Actually, it'd be described "A successful launch with a mid-flight anomaly necessitating flight termination."

It is always wise to get more than one opinion......
Posted : 16/05/2019 5:37 am
Vojtěch
(@vojtech-p6)
Honorable Member
Posted by: bobstro
Posted by: Vojtěch

Btw, as another datapoint, I just tested Prusa (Filament-PM, Plasty Mladeč) PETG Black, the standard one which is used for printing MK3's, and I've achieved F510 at standard 230 °C without clicking on my printer. That translates to 20.44 mm³/s. But then, the BMG is a rather powerful extruder.

This is going to be one of those things where there are safe numbers to recommend to the average new user that will (hopefully) keep them out of trouble (I'm sticking with 11.5 for now), and then there are the tuning numbers for those of us trying to extract every bit of performance out of our printers (15-20 depending on hardware configuration with the v6?).

Even though I have a rather beefed up extruder I still think the 10 or 11 or 11.5 mm³/s number is a good PLA default, even for me. Because I know that while the printer can extrude more than 20 mm³/s of both PLA and PETG, the prints would look horrible and would probably not hold together well. 

I'm like the yellow speed sign showing a maximum speed for that big turn ahead of 40. You know you can do 80.

As expected, that printing temperature makes a difference, and a significant one. Do you normally print that filament at 230C? I do wonder how much the extruder matters. Yet another factor to consider!

I thought so. I checked the presets now and it's actually 240 °C. But well, that'd probably get me to similar numbers like I had with PLA.

The Prusa Bondtech Extruder Upgrade I've got on my printer has one unexpected benefit. I can tighten the gear spring as much as I like to prevent skipping. Because the motor doesn't drive the gears directly, even if the motor would get hot (it never does anyway), there is no thermal path from the motor to the filament, so no thermal induced skipping will happen. With this extruder it's no longer a balancing act to tighten that screw. That I assume is also why I'm getting higher maximum extrusion rates. But that extruded filament is very cold and rigid the moment it leaves the nozzle, so it's not a realistic printing speed.

Posted : 16/05/2019 8:14 am
Vojtěch
(@vojtech-p6)
Honorable Member
Posted by: Tim

So - thinking I'll do a slow speed test, moving kilograms of filament at limiting rates directly on to the hot bed. Separate strips 1.5 expected extrusion width, printing at 25 mm/s or whatever the standard layer 1 rate is. If I modify a standard 100 mm square with 20% rectilinear infill I should be able to punch up the extrusion to make an acceptable test. It will look messy but the infill is a continuous pour, suitable for continuous extrusion rate testing..

I still think a cylinder in vase mode is a better test. There would be no acceleration affecting the E axis at all, whereas for infill, the extruder will be stopping at the ends. Yes, it'd require going through a slicer, but the total time to print would verify whether the slicer did any speed limiting shenanigans.

Posted : 16/05/2019 8:23 am
Vojtěch
(@vojtech-p6)
Honorable Member

I have sketched out my idea for a nozzle that would have a bigger inner surface, allowing to melt more plastic, without impeding the flow of the plastic at the same time and be manufacturable on a common CNC machine. The design would need tuning based on real world performance, the surface would probably need to be made even bigger. STL file HERE.

 

Nozzle design

Posted : 17/05/2019 1:02 pm
bobstro
(@bobstro)
Illustrious Member
Posted by: Vojtěch

I have sketched out my idea for a nozzle that would have a bigger inner surface, allowing to melt more plastic, without impeding the flow of the plastic at the same time [...]

Hang on there, buster. You just made me order a set of 3D Solex nozzles and now you're trying to sell me something else.

But seriously... I did order a set of 3D Solex in 0.40, 0.60 and 0.80mm sizes. I'm very curious to see the results. This could be very useful for large functional PETG prints (e.g. bird feeders).

Your design does look like it might be less prone to filament hardening and sticking. Any plans to get some prototypes made?

My notes and disclaimers on 3D printing and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan...
Posted : 17/05/2019 4:58 pm
Vojtěch
(@vojtech-p6)
Honorable Member
Posted by: bobstro

Hang on there, buster. You just made me order a set of 3D Solex nozzles and now you're trying to sell me something else.

😎

But seriously... I did order a set of 3D Solex in 0.40, 0.60 and 0.80mm sizes. I'm very curious to see the results. This could be very useful for large functional PETG prints (e.g. bird feeders).

I'm very curious about your results with those, too.

Your design does look like it might be less prone to filament hardening and sticking.

And will still allow for cold pulls, hopefully.

Any plans to get some prototypes made?

I first need to do a bit of math to figure the best configuration out, the model is a mere 30 minute sketch of the idea. Standard E3D nozzles require custom made drills to manufacture, so I'll need to figure out whether and where I can get those made and then possibly have the nozzle made. (Another option is an EDM machine, but that also requires a custom form tool.) And the prototype will need to be polished smooth internally, because a rough surface could impede the flow a lot and make it not work. So yes, I'd love to have it eventually manufactured, but it won't happen tomorrow.

This post was modified 1 year ago by Vojtěch
Posted : 17/05/2019 6:05 pm
Tim
(@tim-m30)
Illustrious Member

Has anyone done the math (aka simulation) of flow characteristics of PLA melt in small channels?  Fluids have some very weird characteristics when laminar flow is no longer possible.   There may be all sorts of gotchas that make a multiple small channel design a crock. 

It is always wise to get more than one opinion......
Posted : 17/05/2019 6:42 pm
Vojtěch liked
Vojtěch
(@vojtech-p6)
Honorable Member

Fluid dynamics of high viscosity sticky fluids are damn hard to simulate accurately. Particularly when you have to model the heat flow and changing characteristics of the fluid based on temperature in addition. I'm afraid it's hard enough that it makes sense to make some good guesses and then experiment. Yes, it may turn out to be useless.

Posted : 17/05/2019 7:41 pm
bobstro
(@bobstro)
Illustrious Member

The 3D Solex nozzles have two narrow channels for feeding filament, which is great for increasing the heated surface area, but I wonder how it'll affect retractions given the non-Newtonian nature of molten filament. Only one way to find out. I'm glad you guaranteed these would work!

I did do a bit of pre-order research, and apparently you can still do cold pulls with them. They look odd, but it works. I do expect them to clog more easily, so trying the larger sizes with filled filaments may be a challenge. I'd love to be able to print Colorfabb XT-CF20 at something more than 1mm^3/s... although 2 won't be that much more exciting.

My notes and disclaimers on 3D printing and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan...
Posted : 17/05/2019 8:36 pm
timo.m
(@timo-m)
Estimable Member

@bobstro

Did you have the chance to use your 3D Solex nozzles by now? What they are promising sounds very interesting but I find close to nothing regarding actual user experience.

Posted : 23/04/2020 1:59 pm
bobstro
(@bobstro)
Illustrious Member
Posted by: @timo-m

Did you have the chance to use your 3D Solex nozzles by now? What they are promising sounds very interesting but I find close to nothing regarding actual user experience.

Yes! I was in the midst of updating my notes pages when I diverted all my attention to COVID-19 response printing. I've got some raw data in the PDF linked to on my notes page that you can check out.

In short: Both 0.4 and 0.6 Solex nozzles significantly increased throughout. Surprisingly, PLA restricted rates more than PETG. With a 0.6mm nozzle, I could not extrude fast enough to cause extruder clicking when printing PETG at 250C, perfect for face Shields.

Surprisingly, the 0.8mm version was unremarkable. I suspect it might be clogged. Sadly, my focus is on the immediate crises, but I hope to return to more fun testing soon.

In short though, they won't guarantee good finish, but the Matchless nozzles can increase throughput with a V6 hotend to near-Volcano rates. Maintenance may be an issue, but the 0.6mm has done well with a half-dozen spools of PETG.

 

My notes and disclaimers on 3D printing and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan...
Posted : 23/04/2020 10:05 pm
ron
 ron
(@ron)
Estimable Member

@bobstro,

I find here many insterresting methods to evaluate the maximum volumetric speed. I also view some information about the Solex nozzle I didn't know. Your values for MVS and PETG are quite suprisingly high for a 0.40 brass nozzle.

I found another interesting approach by Stefan ( https://www.cnckitchen.com/blog/flow-rate-benchmarking-of-a-hotend ) you may already know. I think I will go its way to test some of my filaments and share the data too.

 

This post was modified 3 months ago by ron
...
Posted : 08/06/2020 8:34 pm
Page 4 / 4
Share:

Please Login or Register