Bottom Surface Flowrate

Thoughts about Raise3D, 3D printing and making in general.
User avatar
Vagulus
Posts: 189
Joined: Mon Mar 25, 2019 5:27 am

Bottom Surface Flowrate

Postby Vagulus » Mon May 04, 2020 7:21 am

Using the latest (to the best of my knowledge) ideaMaker Manual https://support.raise3d.com/ideaMaker/5-1-3-4-solid-fill-15-1007.html in section 5.1.3.4 Solid Fill I cannot work out the difference between

(1) Bottom Solid Fill Flowrate refers to the flow rate of printing the interior of the solid part at the bottom of the model. and
(4) Bottom Surface Solid Fill Flowrate refers to the flow rate of printing the interior of the surface solid part at the bottom of the model.

My object is to have my Pro2 print a smooth, clean bottom surface on the Upper Disc of this model
200426-PancakeModel.png
200426-PancakeModel.png (14.36 KiB) Viewed 1632 times

which means obtaining that result on the top of support material.

This goal is being infuriatingly evasive. I will much appreciate a solution to the problem.

Thanks
Attachments
Pancake.stl
(215.12 KiB) Downloaded 30 times
--------------------------------------------
God so loved the World that
He did not send a Committee

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Mon May 04, 2020 1:18 pm

#1 you area searching for the most elusive result on the planet. Printing a perfect flat layer on top of support.
To understand this, we have to look at what the nozzle and 3D printer is doing at a physical level and then go back and find how the slicer settings and features could or would produce these motions and actions.

You ask why? How shall we count the reasons? First, the support layer top surface isn't flat. Unless we broke into some seriously advanced gcode generation of EXTRA movements like surface ironing with the nozzle, the support layer is not a flat surface. Ironing is an advance featured created by the open source community that was an extra addition of gcode commands to have the nozzle skim without extruding and kind of melt and shave the top surface of a layer like machining or ironing it. https://www.youtube.com/watch?v=gh5wC4Ti95s A feature other slicers may have implemented (Cura) but to my knowledge often requests as an addition to Ideamaker but not yet added and sure as heck, not this advanced to do it on a support layer top surface.

Next, how a breakaway layer is created. Basically, a layer bond is created when we lay a hot new layer and squish the round threads coming out of the nozzle to oval to maximize contact to both adjacent threads side by side, and then squish into the previous layer to flow and mold into and make contact with the previous layer structure. Conversely, to make a breakaway layer- we modify 2 parameters primarily. Since a layer takes time to print, and thus often the starting point and direction of the new layer is different than the previous layer, the now hot material is put down over colder now cooled plastic of the previous layer. Layer size comes into play, which is why small objects with either raft or support material- tend to have issues with what is supposed to be a breakaway bond because the period of time is short between how much the previous layer was cooled before the new hot layer was placed. Higher retained heat, the hot layer tends to melt and flow and bond better- often too good. The second factor is, again, related to the squish factor of nozzle height to the previous layer compared to nozzle diameter. Is the nozzle is higher than the nozzle diameter, then we drool the new hot layer onto the previous layer without any squish. That limited contact of just laying a round noodle of hot plastic and some time factor meaning the interface between the 2 layers is cooler limits the bonding action- something desired intentionally on a breakaway contact layer typical in support material or say a raft, or any other similar reasoning. The flip side is, this increased gap THEN affects that bottom first layer of the actual object with that weak round threads, typically just loose, not bonded to each other side by side and a generally poor shredded wheat appearance.

Back to your question- you got focussed on Flowrate.
Stop, let's takes a few minutes and understand flow rate.
As stated before, 3D printing depends on a couple of assumptions or facts or however you want to call or feel about them.
The nozzle natively puts out the same diameter of filament as the size of the bore. When the slicer creates gcode, it uses that fact when in 3d making the gcode segment AKA one line of gcode. In other words, in the slicer, a line has a height (AKA layer height), a width (extrusion width), and finally that straight line segment has a length. Those combine to make a volume in 3D space. The incoming 1.75mm diameter (nominal) filament represents a cylinder of infinite length. The slicer takes each segment,takes the input diameter of filament you entered into the profile, takes that layer height and extrusion width combined with segment length, and does the reverse math of volume into the nozzle = volume out of the nozzle. So while the filament is a larger diameter going in, the tradeoff becomes less linear length is pushed in as a ratio to the linear length of the segment of gcode length. The key is flow rate is a change from this one to one ratio. In other words, we have an ideal as calculated logical representation of the nozzle moving through 3D space, squirting this 0.4mm diameter hot noddle of plastic, and how much linear length of 1.75mm filament is required to make that length. Changing flowrate higher than 100% increases the length pushed in making an often fatter hot noodle out than the nozzle diameter as the plastic is forced to squish sideways and kind of flow out in front of the nozzle because it's coming out of the nozzle faster than the nozzle is moving in XY. This generally results in poor control of the extra plastic flow, poor surface finish, but can and does switch that plastic everywhere making sometimes a slightly stronger part when done right and reasonable proportions.

Effectively, what YOU are attempting to do or compensate, is increasing flow rate on the interface layers, to fill the gap in the interface layer where the slicer automatically makes the nozzle height higher than normal (height than nozzle diameter) compared to the top of the previous layer. And obviously, if we even begin to understand what this looks like at the gcode and functional nozzle motion and squirting plastic level- we'd know that you cannot compensate reliably or effectively and even hope for a reasonable surface finish, let alone just not make a fused mess. Again, 2 things happen with increased flow rate. You are squiring more plastic and more plastic faster than the nozzle is moving causing that plastic to blob up and kind of be plowed and pushed around my the nozzle in a largely uncontrolled way, and then also adding more hot plastic in any given space is more hot plastic that must cool and solidify. In a big layer, you just make this literal "hot mess" of plastic.

Again, if you want to do this right:
#1 understand the physical variables- what the nozzle is doing on a given layer or function. Specifically it's height relationship to the previous layer, and it's flowrate compared to the XY movement of the nozzle.
#2 Understand how settings and features in a slicer create the gcode, that commands the nozzle to move in these paths. So what feature, what setting, when and where that happens or kicks in, what setting X does and so forth.
#3 Using the wrong slicer or expecting a free slicer created by the company that sold you the printer to be perfect, when they aren't slicing experts.

In short, here's an analogy. You are driving down the road in the car. You say it's too hot in here. You change the radio station turning the tuner knob. You sit and wonder why you are still not getting the results and why changing the tuner knob doesn't change your body temperature feel and you are still uncomfortable.

You change flow rate WHEN you understand what flow rate does, how it will affect or improve the situation and as a rule, a plus or minus 25% flow rate less or greater than 100%, you probably have something else very, very wrong.

Again, we have a number of issues.
You want an ideal gcode print file- that to my knowledge, either only other advanced slicers might be able to create by regionalizing custom settings and actions to create this much more complex gcode and actions.
You don't understand the physical level of 3D printing, why we do things, logical actions and effects and relationships.
You don't have the experience and knowledge of years of reading MULTIPLE 3d printer groups and brands, seeing what each brand and group comes up with innovation wise to solve problems, have experience with multiple slicers and know the features sets of those.
You expect the free slicer given away with a Raise printer, developed by a team who has not yet proven themselves to really understand 3D printing, they simply keep adding features by copying (badly) what users tell them or copy a feature from another slicer without actually understanding the why and how.
You expect a rudimentary manual for a software, to teach you 10 years of printing experience, when again, the software developers pretty much prove over and over they don't understand it. So how are they going to write an in-depth guide n settings and worse, when to use them, and finally, how far is a sane adjustment.
Last edited by Jetguy on Mon May 04, 2020 1:29 pm, edited 1 time in total.

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Mon May 04, 2020 1:26 pm

In a nutshell, this is both an expectations problem, and flat out adjusting the wrong thing and being disappointed in the results.

You probably are never going to get anywhere near what you want with this slicer.
You need to be looking at changing the right variable- the layer height or distance at the interface layer to about the exact layer height and better yet, if really using proper alternate support material (washable or dissolvable), maybe closer than normal layer height.
You really need ironing, and that's not an option in this slicer.
And, finally, if doing this with just same material support (the support is the same plastic as the print) you will NEVER be able to have a perfect flat bottom. Back to the expectations issue. Why? Because with same material support, in order to create that breakable interface bond layer, you do that by minimizing contact surface for bonding. If the top of the support is too smooth and flat, you'll never be able to break the bond and just get one fused mass, and conversely, if you compensate with gap, then the bottom of the first layer of the part is terrible, because you are just drooling lines of plastic over a surface, and then hoping they somehow bond straight and even and flat?

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Mon May 04, 2020 1:42 pm

Again, if we take this moment and learn nothing else.
#1 Support material printing in all known current states (same support or even alternate support), generally will leave the bottom surface of the object printed on top less than perfect. This is due to so many limitations, but understanding what makes a strong layer bond, and what makes a weak breakaway one, you see why this is a challenge.
If you want that flat bottom on BOTH objects, the both object need printed on the bed side by side, not vertically.
Support and raft are used when there is no other possible way to print. They are tools of last resort and come with tempered expectations.

#2 Flowrate seems to be the most misunderstood and most abused slicing setting. I see over and over using flow rate to compensate for what amounts to a a nozzle to previous layer distance issue. On first layer, often bed leveling or Z gap limit switch adjustment, warped bed, or other limitations but at the physical level, over and over, the nozzle isn't the right height, so the user increases flow rate, and then plows the nozzle pushing along an ugly blob of molten plastic just kind of slopping it around. And the results look sloppy.

#3 Understanding the nozzle diameter to nozzle height relationship is key in 99.9 percent of all discussions. This affects so much of how the plastic will bond and flow and shape. It's the difference between a good layer and a bad layer in so many ways.

#4 Understanding slicer settings. Sorry but a manual is not going to teach you what you need to know. No post I can make could cover every setting in the depth to actually know how and when to use it. And, in doing so, you might have to start thinking about the nozzle to layer function, understanding input VS output at the nozzle, and last, yes, understanding and reading and writing gcode. If you don't have the foundation, how can you build a house? If you don't understand how the nozzle should move, how can you expect a slicer and what setting to make it move in the way you want?

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Mon May 04, 2020 1:53 pm

Again, let me make it clear.
Even if I made super custom gcode. Even if I used 2 material support, AKA washable support, and I went to fine tuning and testing ironing the top that support, and adjusted and tuned the first layer of the new object on top of the support. Because support is a 3D printed layer and less flat and hard than a bed surface, I'll never be able to squish and mold the first layer of the printed object on top of support have the same finish and quality as a part printed on the bed directly.
NEVER. Not going to happen, no magical setting.
That needs to be the understanding and expectation.

OK, so you still want to do it anyway.
Again best result is complex, requires a good proven alternate support material print, requires fining tuning of that top surface, requires features Ideamaker doesn't have. In short, being honest, probably out of your current skillset. The physical printer might be able to pull it off, your software, experience, knowledge, and the required use of an alternate slicer is what keeps you from doing this.

Worse, again a lot of folks are doing this stacking as a method of rapidly producing multiple parts in a print. In the past, this method was for printers with limited bed area, and so this was a method to beat that limitation of bed size. Here, we are possibly using it for some of these larger COVID-19 face guard frame prints, and again, because multiple don't fit side by side, a way to do it, accepting the tradeoffs in surface finish on a part that we just need more mechanical aspect, and rapid copies of lesser quality is acceptable. But again, it's understanding WHEN to use something, what the tradeoff is, WHY we use it in the first place.

User avatar
Vagulus
Posts: 189
Joined: Mon Mar 25, 2019 5:27 am

Re: Bottom Surface Flowrate

Postby Vagulus » Tue May 05, 2020 6:00 am

Thanks Jetguy for taking over an hour working on this answer. I greatly appreciate your effort.

Sadly, it is not much use reminding me that a brief simplistic manual (produced by a 3D Printer company which is not renowned as a Slicer Company) is not going to make up for years of experience. It's true, and I am aware of that. That is why I ask questions in this august forum - to tap into those years of experience per persona your good self.

First, let me be clear that the model I posted is experimental. It was devised to provide me with a test of the horizontal surface finishes I get with a given template. I do use stacking, for example to keep individual items in a set or when an item is made of multiple parts such as
200505-MultiplePartModel.png
200505-MultiplePartModel.png (43.48 KiB) Viewed 1589 times

I am also aware that stacking will not decrease production times. My printer still has to lay down a given amount of material. Also, increasing model height generates an disproportionate increase in printing time. Printers seem to be happier printing wide rather than printing high.

Second, I am not printing ornaments. I am doing Mechanical Engineering Prototyping. My models are regularly bulky and use most of my Pro2's printing volume, 305x305x305 mm. Consequently, printing time is important so I am using 0.8 mm nozzles and I am happy to accept the resulting reduction in general surface quality. Getting a reasonable finish on parts printed on support material is a means to reduce post-print processing (cleaning up).

Jetguy wrote:#1 Support material printing in all known current states (same support or even alternate support), generally will leave the bottom surface of the object printed on top less than perfect.

This statement goes to the heart of the problem. Only slightly elaborated (see #2) it would have saved you a considerable amount of typing ;) . Perhaps you should just have said, "Yuh Can't - and that's that!, though it would be nice if you softened the blow a little. :roll:

Jetguy wrote:#2 Flowrate seems to be the most misunderstood and most abused slicing setting. I see over and over using flow rate to compensate for what amounts to a a nozzle to previous layer distance issue. ... And the results look sloppy.


Jetguy wrote:#3 Understanding the nozzle diameter to nozzle height relationship is key in 99.9 percent of all discussions. This affects so much of how the plastic will bond and flow and shape. It's the difference between a good layer and a bad layer in so many ways.

I would like to gain some understanding of this. I know I don't understand the relationship between Layer Height, Flowrate, and Extrusion Width Percentage. I have yet to find that altering Extrusion Width Percentage in ideaMaker has any effect at all. If you can point me to some documentation on this subject (kindergarten level, please :? ) I'd be delighted.

Jetguy wrote:#4 ... understanding and reading and writing gcode. If you don't have the foundation, how can you build a house?

I have extensive programming experience in programming and should have little difficulty in adapting to gCode. To start working with gCode I would expect access to a lexicon, a grammar and the syntax of a version of the language applicable to coding for a Pro2. gCode appears to be Open Source and various versions can be found. I am reluctant to choose one until assured that it is not going to crash my machine. My problem is that, despite several times of asking in this forum, nobody has directed me to this information. Again, if you can point me to some documentation on this subject (kindergarten level, please :? ) I'd be delighted.

In any case, this is the best I have done so far:
200505-Obverse-480x234.png

200505-Reverse-480x239.png

Is this (to quote an old Lady in the show 'Oklahoma') "... about as far as she can go"?

Finally, what is the difference between those two statements in the manual?

Thanks
--------------------------------------------
God so loved the World that
He did not send a Committee

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Tue May 05, 2020 12:10 pm

#1 In general, Raise 3D uses the Marlin flavor of Gcode. Gcode was developed for CNC machines mainly (subtractive machining) and then adopted by the open source community for 3D printing. In doing so, the 3D community uses existing codes in a way very different from the original context, so yes, you are correct and want to be pointed to a 3D printing specific flavor of gcode, and then even better, the exact firmware version. The N series motion control used Marlin firmware, and so Ideamaker makes Marlin formatted gcode, and then the Pro2 series came out and uses basically the exact same gcode, so even for purposes of discussion that even thought Raise3D claims a custom motion firmware "all new" on the Pro2, for gcode, it's still Marlin. https://marlinfw.org/meta/gcode/

Gcode, can be this simple. A line of typical printing gcode is
Command number, Axis applied to, Value in mm, Feed Rate in millimeters per minute
G1 X100 Y200 E4.8 F6000
First thing is, where you are now in the 3D world is the previous line of gcode (not shown). For the purpose of this discussion we'll assumed we started at the home parking position at X0 Y0 Z0, and via the starting gcode sequence that homed the printer, we also primed the nozzle so it's ready to extrude. Again, this technically means that all logical "I am here" positions both physical and logical in the controller are X0 Y0 Z0 E0. She further detail just to make life hard, Z, w home to a fixed gap between the bed and the nozzle. So Z0 logical, is actually Z0+gap.
Anyway, G1 is a coordinated move. What that means is that all movements START at the same time and END at the same time even though they have radically different distances and there is only one feed rate given. So that means as a move, each axis will be going a different speed to achieve same start time, same end time, different distance. Put another way, the move is a diagonal move in a straight line. The line of gcode is the new target position, AKA where we will be when the move is finished. So again, we start at position X0 Y0, move diagonally to X100 Y200, and during that move, we push 4.8mm of raw solid wire of 1.75mm diameter filament into the extruder and obviously, a much longer because the nozzle is smaller than 1.75mm of the nozzle sized hot noodle of filament comes out and squirted down as the nozzle moves.

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Tue May 05, 2020 1:05 pm

Give me a bit today, this is going to be multiple replies to answer your questions.
Where I am try to get you is to a point you can analyze the gcode and actually see the changes.

Nutshell of where this is going.
Extrusion width changes at the same layer height should result in wider or narrower spacing between adjacent nozzle paths, and changes in extrusion volume (E values) when comparing 2 identical lines of gcode from different slices.

Here's where I think the disconnect is now that you gave more information:
#1 you are using a larger nozzle than stock, specifically 0.8mm
#2 You probably are running stock hotends
#3 you likely may not have properly slowed your feedrates down given the massive volume difference when increasing nozzle size, so you are possibly hitting the extruder feed rate limit at which it begins to skip steps feeding filament or simply slips and grinds resulting in less extrusion than commanded AKA under extrusion. In doing so, changing flowrates won't fix it by commanding more when the extruder is already at it's limit.

Simple proposed test.
Start the print file you've already printed. Easy, it's already on the printer. When it starts, change the feed rate percentage on the screen to 50%. Let it finish the print basically at 1/2 speed. Is there is a difference? (Probably yes- thus exposing what I said, you exceeded the rate limit of the extruder without understanding this change when you double nozzle size).
Note, the ideal value may not be half. I don't know where you started, I don't know your extruders limits. It could be less or more.

Again, what I'm seeing is, #1 changing the wrong variable and not getting results, #2 Raise sure as heck isn't going to post limitations of the printer so what I'm sharing is data and examples I've gathered with over 10 years of experience. In other words, the software and the firmware limits don't prevent a user from exceeding the limits of the hardware, and in the case of the extruder, you get bad unreliable less than consistent extrusion or massive under extrusion above a given volume. And more confusing, there is a long term max rate, where as short term after an extruder sits there hot not extruding, you can feed faster than that for just a few seconds and then the extruder kind of hits the wall.

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Tue May 05, 2020 3:11 pm

Lesson 2, slicing as it relates to gcode changes.
You asked about extrusion width.
Extrusion width is used for 2 calculations.
The first, is how far 2 parallel adjacent nozzle paths will be placed that are supposed to bond together. Say you want to use 0.8 extrusion width on a 0.8 nozzle. On one hand, this is what Raise literally just told someone to do and I'm guilty myself of using it, but technically that's incorrect and we will get into that. Point being, for this, we'll use 0.8 because raise says to.
OK, so the nozzle squirts the plastic and resulting width is 0.8mm, or centerline of the nozzle, is half diameter, so, say from the edge of the outline of STL file dimension on that layer, the wall, we will place the shell path of the nozzle 0.4mm inside the outline path of the STL because that will give us that outline. Then on infill, the nozzle path spacing is overlap and thee extrusion width. So extrusion width is 0.8mm, spacing is 0.8mm (one half on one side, one half on the other side) minus the percentage overlap. Overlap being a percentage, is a percentage of extrusion width.

And say if we sliced a 20mm square cube, and first layer infill lines are parallel to X or Y, we could easily examine the resulting gcode and see how the resulting spacing between the nozzle paths changes as we change overlap, or we change extrusion width.

Then also, if these are all the same layer height, and we begin comparing files and we adjust extrusion width, then for the same linear length of a line of extrusion, the resulting filament E length would be different because a wider stripe of the same height and length is more volume, and E represents the volume as a length of incoming 1.75mm filament.

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Tue May 05, 2020 3:15 pm

Going to your example pictures, either you under exuded first layer or Z gap or layer high resulted in the nozzle being to high, even on the bottom puck. The top puck exhibits exactly what I'm saying, a layer over support material cannot be squished well, so it is not well bonded adjacent strands of extrusion all kind of loose and tending to deform and drag around before cooling solid.

Because I feel you are not getting first layer right as a baseline for comparison, you then don't have a proper frame of reference how the upper should compare. In other words- both are not great IMO.

User avatar
Vagulus
Posts: 189
Joined: Mon Mar 25, 2019 5:27 am

Re: Bottom Surface Flowrate

Postby Vagulus » Wed May 06, 2020 3:14 am

Jetguy wrote:Simple proposed test.
Start the print file you've already printed. Easy, it's already on the printer. When it starts, change the feed rate percentage on the screen to 50%. Let it finish the print basically at 1/2 speed. Is there is a difference? (Probably yes- thus exposing what I said, you exceeded the rate limit of the extruder without understanding this change when you double nozzle size).
Note, the ideal value may not be half. I don't know where you started, I don't know your extruders limits. It could be less or more.

I am all for doing a test!

There is a problem with this one in that I do not know what you want me to do. This, I think, is a terminology problem in that I can find no reference to 'feed rate' leave alone how to 'change the feed rate' in the Pro2 Manual - the only references to feed relate to changing the filament. Then you talk about changing the feed rate percentage on the screen when the printer has started. :?:
(1) Are you saying there is some adjustment that can be made on the fly through the Touch Screen on the printer (I am using a Pro2)?
(2) Is there some change you want me to make in ideaMaker?

There is no point in doing a test if I am not doing the right test. We have to be talking about the same thing, so please step me through what you want me to change (and how to change it).

Please be patient. I may be a slow learner, but when I have learnt it I've got it. Hopefully I'll also know why I did what I did. :D
--------------------------------------------
God so loved the World that
He did not send a Committee

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Wed May 06, 2020 4:21 am

1, Yes, on the LCD screen of the printer, under the tuning page, there is "Feed Rate". Normally defaults to 100% meaning the printer reads the feed rates in the print file and executes them as written in the file. Changing the percentage changes the printer to go faster or slower than the rates in the print file. We want slower because you changed to a bigger nozzle and likely, failed to correctly adjust the ideamaker profile for the larger nozzle slower. By doing it at the printer, we change the real time execution.
During a print, touch on the word TUNE at the bottom of the screen to bring up the TUNE page and menus.
home screen.png
home screen.png (44.47 KiB) Viewed 1539 times

Then, on the TUNE page, in the middle of the screen you see "Feed rate" and to the right 100 over 100.
you touch on the numbers and it will allow you to enter a nw number- 50 for 50% AKA slowing down all speeds by 50%
tune screen.png

2, If the setting works and the print is better, THEN we change the rates in the ideamaker profile so that all future print files for this larger nozzle are made with the corrected speeds and thus print correctly.

Refer to this manual https://support.raise3d.com/RaiseTouch/ ... 7-729.html
And, semi related topic discussion- about exceeding a printers feed rate limitations and causing a jam- or just under extrusion in your case https://forum.simplify3d.com/viewtopic.php?t=4877

User avatar
Vagulus
Posts: 189
Joined: Mon Mar 25, 2019 5:27 am

Re: Bottom Surface Flowrate

Postby Vagulus » Fri May 08, 2020 7:45 am

Short hiatus here. I am having trouble getting a new nozzle extension/retraction servo motor from the agent. I'll be back when I have my machine running. :roll:
--------------------------------------------
God so loved the World that
He did not send a Committee

User avatar
Vagulus
Posts: 189
Joined: Mon Mar 25, 2019 5:27 am

Re: Bottom Surface Flowrate

Postby Vagulus » Sun May 10, 2020 5:50 am

Hi Jetguy

Back in action with a single-extruder Pro2. Raise3D's Australian National Agent does not have a $20 spare part for a $8K machine! :shock: Even worse, they'll have trouble getting them. It appears they have been out of production for a while. :cry:

Download https://1drv.ms/u/s!ArMxdUYEh9fKkCARMwPRAZTvikQ0?e=Nbfyip
Printing the Pancake thingy at 100% rate then 50% rate gives the results shown in the images Obverse and Reverse.
There's not much difference, is there. The only identifiable change was the printing took twice as long at 50% rate. That's hardly a surprise, but it removes any advantage I was trying to gain in production times. :(

I have included in the download some screenshots of the current print settings (with the speeds for 0.4mm nozzles for comparison). I had used an on-line print speed calculator to try to get some ides of nozzle speeds.
https://www.printitindustries.com/pages/print-speed-calculator
There is also a copy of the template I used to print these models.

Hope this helps.

BTW: 50% print on right.
--------------------------------------------
God so loved the World that
He did not send a Committee

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Sun May 10, 2020 1:30 pm

I feel like we just went to comparing apples to oranges.
We started this topic, printing on top of support, and then wondering why the bottom surface was bad.
We find out in addition, using a bigger nozzle and probably not fully tuning feed rate given the larger nozzle size.
The test was to reprint this object stack while reducing speed to ensure we didn't max out the melt rate of the extruder.
Again, original object pictures you posted earlier in the topic.
200505-Reverse-480x239.png

Only showing the bottom because the topic was primarily about bottom issues.

The latest prints you shared so others can see what we are talking about.
coins top red.png
coins bottom red.png

Then you print these at different speeds on the bed directly. While "a" test, we've shifted gears here to a different topic- how to tune 3D printing. Where to even begin? The shiny surface of the prints in general on the top side (super glossy looking print) is an indicator of how hot you printed and how much heat is retained in a layer before the next layer is printed. Hot layers (too much latent heat in the plastic) is very good for part strength. This is because the plastic tends to bind super well to adjacent threads below, beside, and above any given line of extrusion. They all flow and melt together as kind of one big mass. The downside to extra heat is VERY poor overhang, dropping, movement after placement of threads. In other words, it's a tradeoff of brute strength of the part with a massive degradation in dimensional quality, support, overhangs, and other details. Heaven forbid we print a cone or tower or some other tall and thin object, this could be a disaster.

OK, so something to know tuning. A dry dull looking finish is indicative of a cooler printing temperature. This is where the plastic is cooling extremely shortly after leaving the nozzle and solidifying. This is extremely good for dimensional accuracy, overhangs, and other tricky to print shapes. The downside is reduced part strength since the layer to layer bond or adjacent line bonds are weaker due to not melting into each other as much.

You have to find the UPPER and LOWER ranges of the temperature to find the ideal middle ground range. In other words, you print one test piece on the low end of the temperature range of that filament, trying to achieve the point where the extruder doesn't fail to extrude, but we still get a finished print with reasonable quality and a duller dry looking finish. Then the high range is where we see this shiny surface, drooping and moving threads that shifted before hardening. Given the upper and the lower- now we can make smart decisions on tradeoffs again for that nozzle and that plastic.

So- the lesson is every filament and then nozzle size has a RANGE of temps of printing. I say printing because two totally different concepts- extruding melted plastic is NOT 3D printing. 3D printing is BOTH extruding hot melted plastic AND placing it down AND it bonding to adjacent layers and last cooling to solidify BEFORE gravity and other forces allow it to move out of place from where the nozzle places and attempted to shape it. The larger the nozzle, the bigger the hot noodle of plastic, the MORE CRITICAL getting this printing temperature right to tune to the specific plastic, but also nozzle size. The larger the diameter of the hot noodle, the more heat it retains and the longer it takes to solidify.
Fan airflow and other details cannot always force faster cooling trying to remove more latent heat as we increase nozzle size.

So yes, IMO, based on only a picture, I think you could stand to reduce the extrusion temperature setting by 10-15C from it's current value for that filament and that nozzle. Your latest pictures show slightly deformations and flowing of adjacent threads and just a shiny big well melted together surface. If going for super strong and final dimensions and or warping are not the most important factor, then sure, we can leave the same temp.

I'm just saying, I have no idea where you are going. you complain of finish, the coins have lettering, part of this lettering deforming is heat related during printing, also nozzle size. We start adding support material and worrying about bottom finish, but then we print the latest examples directly on the bed and get given nozzle size, relatively good results (IMO) on the bottom.

User avatar
Vagulus
Posts: 189
Joined: Mon Mar 25, 2019 5:27 am

Re: Bottom Surface Flowrate

Postby Vagulus » Mon May 11, 2020 3:18 am

Misunderstanding here, I think. (Let's stay on the one subject for now - Feed rate. We can come back to temperature ;) )

These images are rotated 90 degrees anticlockwise for some reason. As you showed them there are images of two pairs of 'coins'. In both images the coins in the top row were printed at 50% rate (set in the printer) the ones in the bottom row at 100%. Your top picture shows the tops of the coins, the bottom picture the bottoms. In both case the LOWER coins were printed on the BuildTak (print platform unheated which would explain the duller appearance), the UPPER coins were printed on the support material. All I sought to show here was that reducing the feed rate made little difference.
--------------------------------------------
God so loved the World that
He did not send a Committee

User avatar
Vagulus
Posts: 189
Joined: Mon Mar 25, 2019 5:27 am

Re: Bottom Surface Flowrate

Postby Vagulus » Wed May 27, 2020 7:32 am

Hi there Jetguy

When I wrote,
Vagulus wrote:All I sought to show here was that reducing the feed rate made little difference.

I was thinking that we had made progress with our experiment. I am not being frivolous when I say that we had shown feed rate to NOT be the cause of the issue. In Scientific terms that is some success. Of course, we have not solved the problem, but we know what it ISN'T! :geek:

FTR: I have pursued the issue of what it is. Two weeks of skull-tearing frustration have driven me to the conclusion that the major culprit was Layer Thickness. With 0.8 mm bore nozzles I had been working with a Layer Thickness of 0.55 mm. With the printing head 'homed' the clearance between the tip of the nozzles and the BuildTak was set to 0.2 mm which is standard. While printing that would give a distance of 0.75 mm between the tip of the nozzle and the last layer of material. I came to the conclusion that that was not a good idea when I was laying down a 0.8 mm dia. 'sausage' of material.

I changed to using a setting of 0.64 mm Layer Thickness and got an immediate improvement. I also adjusted Extrusion Width and changed the support material pattern from Lines to Grid (which seems to give a much more rigid support column). Outcomes aren't perfect but at least I can print a serviceable part for my prototype. My major issue at the moment is surface finish. Disregard the text. It was just there in the testpiece. My conclusion is that IdeaMaker does not provide for text of sufficient size to be clearly rendered using 0.8 mm bore nozzles.
200527_3 Disc Top 640x661.png

You can see that the individual lines of filament are not squishing together to form a somewhere-near-flat surface. Extrusion width is currently set at 1.04 mm with is 130% of Nozzle Bore - a little above the general recommendation of 100-120%.

Do you have any thoughts on possible fixes?

Thanks
--------------------------------------------
God so loved the World that
He did not send a Committee

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Wed May 27, 2020 1:24 pm

See the gaps between your infill lines on the top of the part AKA screen door effect?
There are basically 2 functions that affect that.
Extrusion width sets the distance between adjacent paths based on an assumed squished width the plastic will achieve- obviously that's not as wide as expected.
OR
Extrusion volume is not as expected, and the mount and rate of plastic coming out of the nozzle is being stretched thin.

So here's the problem I'm facing. I know how to setup a slicer correct for a given nozzle. I know how to properly measure filament distance extruded, figure out a rough estimate of an extruders reliable fastest rate, and a few other technical details.
Getting you on that level of understanding with a couple of forum posts- not something I'm sure I can do.

But more than that, a problem you and I are facing, they have screwed up ideamaker so badly in the last couple of releases, I have no idea the bugs and problems that exist. Raise3D is not a group of slicing experts. They took a basic slicer, that mostly worked, took a bunch of user input and requests from people who weren't qualified to operate a toaster, let alone a 3D printer, and then ruined the slicer for all involved.
Between putting settings in so many places and duplicate and conflicting
Last edited by Jetguy on Wed May 27, 2020 1:31 pm, edited 1 time in total.

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Wed May 27, 2020 1:25 pm

Here's a thought: if you install a larger nozzle- first layer nozzle homed gap might need to be higher.....

Next, I reread the post and saw this:
You can see that the individual lines of filament are not squishing together to form a somewhere-near-flat surface. Extrusion width is currently set at 1.04 mm with is 130% of Nozzle Bore - a little above the general recommendation of 100-120%.

1.04 mm with is 130% of Nozzle Bore - a little above the general recommendation of 100-120%........ and you wonder why you get the screen door effect?

So here is now what I'm seeing.
You failed to also calibrate first layer nozzle gap when homed when you changes nozzles. Yes, a MECHANICAL adjustment.
Because then you attempted in ways I would never dream of to compensate for first layer problems, you've now created rest of the print problems.

Jetguy
Posts: 3107
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: Bottom Surface Flowrate

Postby Jetguy » Wed May 27, 2020 1:44 pm

Right here is where you've gone wrong. You aren't following the "I changed something major (nozzle size), and gone back to basics tuning my slicer and printer for the new nozzle".

The first thing to do when you have a suspected calibration problem is go back to the basics- extrusion volume testing of a KNOWN test.
Example, this topic https://forum.raise3d.com/viewtopic.php ... ish#p11695
The instructions to follow https://forum.raise3d.com/viewtopic.php ... fish#p2620

That test is valid for every printer, every nozzle size, every slicer. It is the how to to get extrusion volume right, and that's one of the most important variables to get right.
so
#1 Go back to a reasonable baseline state of settings in your ideamaker profile after changing the nozzle. In other words, let Ideamaker auto set extrusion width based on nozzle size and layer height rather than specifying.
#2 THEN ensure you set the nozzle first layer mechanical homed gap corresponding to your expected first layer height- now larger with the lager nozzle.
#3 Then perform the above extrusion volume test- except make one minor change- scale or double the size of the 20mm test cube to 40mm square. Because we are doubling the standard nozzle size, doubling the physical test makes it easier and better for accurate tuning https://forum.raise3d.com/viewtopic.php ... fish#p2620
#4 Then save that flow rate and other tuning details back into the print profile and attempt to fine tune your actual part you are trying to print.


Return to “General”

Who is online

Users browsing this forum: No registered users and 2 guests