horst.w
Posts: 795
Joined: Fri Oct 17, 2014 5:00 pm

Re: Horizontal lines in test tower

Fri Nov 17, 2017 7:00 am

... and here the results:
20171117_123623.jpg
20171117_123617.jpg
That are the beginning of print of your tower with your settings, only adjusted for the diameter of filament. Compare the difference to a "normal" print of mine.


It takes some time to comment your settings, I want to do it later.
But the issues are clearly caused by your settings, its not S3D.

H.

InspGadgt
Posts: 34
Joined: Wed Oct 18, 2017 7:59 pm

Re: Horizontal lines in test tower

Fri Nov 17, 2017 12:49 pm

That's what I'm trying to figure out...I figured it was my settings in S3D but I'm not sure what. I wouldn't think it was my printer settings otherwise my Cura prints would be affected as well. I'm not sure what settings in S3D to adjust as I am still really new to the software. So far I like it a lot...I just need help figuring it out.

horst.w
Posts: 795
Joined: Fri Oct 17, 2014 5:00 pm

Re: Horizontal lines in test tower

Fri Nov 17, 2017 2:37 pm

Dear friend!

To compare the print results of Cura and S3D is not fair!

Cura is a 1/2 or even a 3/4 automatical program that will print very well, when all parameter of the printer and the filament-material are in a certain range of standards.
S3D does it not so, S3D musst be turned from the first point to the last point. There are only basics but that is not sufficient. Therefore you can feel free to use even really exotical materials.

Your settings

Extruder:
Your ExtrusionWidth is only 0.40 mm while the nozzle is also 0.40 mm. That is ok when you adjust the perimeters to a thin wall so as you do it when you only use it with 2 perimeters. BUT your LayerHigh is 0.2 mm, will say it is 50% of the extrusion width.
If 0.4 mm is not needed, print the ExtrusionWidth with the standard from S3D, that automatically 120% of the nozzle-value = 0.48 mm

Layer:
LayerHigh is 0.20 as written above 50% of the ExtrusionWidth. The nozzle presses the material very hard and therefore the sidewalls of the extrusion become more and more round and bulge outside.

Temperature:
May be ok, BUT ...
you are printing very, very slow -see below-

Speeds:
the setting of 60 mm/s seems ok but there are also 50% for Outline-Underspeed. When only printing 2 perimeters, 50% of material is printed with a speed of only 30 mm/s AND additionally you have set Speed Overrides below 15 sec one layer down to 20% (of normal speed of 60). Because your model is very smal the printer is therefore running only with 12 mm/s all the time.

What makes that?
The tower has 20 x 20 mm and is the only one object to print. That cause that the hotend of the printer is moving only in this area all the time while the coolling - even with 100% - has no success because there is no place for the fan wind so that the top of the print has no efficient cooling. The hotend heats with its heat-emission the upper layers additionally what is staying smooth.
I could study how the extrusion has pressed the second layer out of the walls vertical line and that are your lines in the surface.

The same lines are inside the tower.
If you want to test something, don't use a such small object. a cube or a tower must be minimum 50 x 50 mm and the best results you can get needs a second object 10 cm beside the first one to win time for cooling. Without harden the upper layer you will have anytime such lines caused be the pressure of the nozzle.

There is another curiosity of your settings.
Scipts - Starting script - has M - commands, M 201 and M205. That concerns the acceleration and the jerk and set them very deep. Using your proflie my printer was very, very slow and also very noicy and has written these values to the eeprom. Unluckily I dont remember the true values of mine but I set them now to M201 > X1000 Y1000 and M205 to X20 *), may be to much but is running.
I dont know your printer and its firmware but the M205 commands parameter are X4 Y4 . This Y-parameter seems to be not existing. Have a look to this link:
http://reprap.org/wiki/Gcode#M201:_Set_ ... celeration *) M205 X20 is NOT the X-axis, its a parameter for both, X and Y axis!

H.

InspGadgt
Posts: 34
Joined: Wed Oct 18, 2017 7:59 pm

Re: Horizontal lines in test tower

Fri Nov 17, 2017 4:09 pm

I think you are misunderstanding me...I'm not complaining that my Cura prints are better than my S3D prints. I only compare because I want to learn how to get as good prints out of S3D as I do with Cura which should be possible. I see people all the time using S3D that get great prints. I already prefer S3d over Cura for a couple of reasons...first, with S3D I am finally able to print multiple pieces at a time of my 3D LabPrint airplanes. No matter what I did I could not get Cura to print multiple thin wall pieces without having some major holes. S3D did it the first try! Second I also want to print miniatures and S3D's customizable supports are so much better than anything else out there. So please understand I'm not trying to complain about S3D...I just want to understand how to get better with it.

As for my settings...thank you for that evaluation. .4 line width is what I am using in Cura as well. Do you think Cura is altering this in the Gcode? As for layer height...from my understanding the smaller the layer height the better quality the print is...is this not correct? I used to use .15 even to get a better print but that takes even longer. The factory file for the 3D LabPrint planes I believe uses .25 layer height but I lowered it slightly to get rid of some small holes I was getting and increase layer adhesion which is critical for the airplane prints. As for print speed...I don't think the Anet A8 in it's stock frame is capable of much faster than this...it has a lot of flex in the acrylic frame. Until S3D I couldn't even print at that speed without hugely sacrificing quality...Previously I was printing at 40mm/s. Remember this is a direct feed printer with the extruder stepper on the X axis carriage, not a nice, light Bowden's setup. I do plan to build an aluminum frame for it soon. I'm just waiting for some things to sell on e-bay so I can afford the aluminum extrusions. Eventually I'd like to build a Bowden's setup...maybe for my 2nd printer. I can try a bigger tower and multiple towers...but if it were a heating and cooling issue then the Cura print would suffer the same issue as I used the same values for that in both prints. The M commands in the start of the script I was using to try different acceleration and jerk settings without having to modify my printer. The results on the test print even without those command was exactly the same. As for the commands themselves I was just using what someone had posted earlier in this thread. M205 is the jerk setting for the current version of Marlin according to the earlier post.

Thank you for your very thought out and detailed post. I'll try some of these this weekend and see that happens.

horst.w
Posts: 795
Joined: Fri Oct 17, 2014 5:00 pm

Re: Horizontal lines in test tower

Fri Nov 17, 2017 6:47 pm

InspGadgt wrote:
...I just want to understand how to get better with it.

... As for layer height...from my understanding the smaller the layer height the better quality the print is...is this not correct?
...
3D-printing is the result of the accordance of a lot of wishes, technical conditions and also the knowledge of interaction of these parts.
You have some mayor requirements and all other conditions are the consequence of them.

Temperature e.g. Filament has a band of temperature, say 190 up to 215 ° (~PLA)
below the 190° the material is to viscose for printing, above 215° the material will perhaps burning to coal inside the nozzle.
All material are very individully, even from spool to spool may exists differences.
What to do? in case of minor temperatures you must print slowly, in case of high temperatures you need high speed. The reason is that the material needs a consistent melting process inside the nozzle with a small band of temperature for becomming an optimal viscosity and at once enough flow that the temperature can not rise up inside (that blocs the extrusion); near the collar of entrance into the melting chamber it must be not more than smooth but not liquid.

You can reverse the arguments: if you want to print quick you must increase the temperature. The dependencies are the same.

Cooling e.g. Cooling is a indispensable condition. PLA needs stringently cooling. With 1,25 gr./cm³ it has an high spec.weight and it seems that it is a bad heat conductor, will say it stores the heat very good. But also it shrinks 3 - 5 % when it cools down (>>> warping). For building layer by layer an object, it needs to cool down the top of print, so that the upper layer ist harden bevor the nozzle is extruding the next. If not the pressure of the nozzle and the new material will change the position and the form of the former layer. And this pressure is needed for the better bonding together of the layers.

Filament Diameter and Extrusion Multiplier:
The diameter is a given value, the best to measure with a micrometer screw. The Multiplier is only the corrective for it. And all together is the data base for the processor of the printer to calculate the feed rate.
I'm wondering you are using 1.0 for the EM, I'm find out the best for my filament 0.92 by accurate diameter measurement

You have ask to the proportion between Layer High and Extrusion Width.

The LH has its optimum round about 30 % of the EW, it must be always wider than high; a very universal EW is 0.48 mm (120% of a 0.40 nozzle) and 0.15 mm LH. The molten material has a high drag to flow out of the nozzle but the sidewall of this rope is nearly vertical. More LH becomes rounder side walls and needs to be print slower, better colling ...

H.
Last edited by horst.w on Sat Nov 18, 2017 5:36 am, edited 1 time in total.

InspGadgt
Posts: 34
Joined: Wed Oct 18, 2017 7:59 pm

Re: Horizontal lines in test tower

Fri Nov 17, 2017 8:28 pm

Interesting...I hadn't thought of any corolation between speed and temp but now that you mention it, it makes perfect sense. And I wonder if that could be one of the reasons for the differences in print in the 2 test towers...while the speeds were set the same for both slicers when physically watching the prints I noticed the Cura print was moving faster than the S3D print. If I can get a good print at faster speeds I would really like that as most of the prints for my planes take 14-16 hours. So I have to start them right when I get home from work or they won't be finished by the time I wake up in the morning.

Well with the planes 3D LabPrint recommends a small amount of over extrusion by using a multiplier of 104% for better layer adhesion. I haven't measured my PLA but figured the closer it is to 1.75mm the closer to 1.0 you would need on the multiplier. Previously I've used 1.0 assuming my PLA was close to 1.75mm with good results. I guess that's why S3D defaults the multiplier to .96 or .98.

What you are saying is quite interesting. I'd be interested to get your take on the factory file settings that 3D LabPrint provides for it's planes. Since their STL files cost money I think they would be somewhat adverse to uploading a factory file so I'll just have to post the settings:
Attachments
Settings.zip
(681.77 KiB) Downloaded 46 times

horst.w
Posts: 795
Joined: Fri Oct 17, 2014 5:00 pm

Re: Horizontal lines in test tower

Sat Nov 18, 2017 3:16 am

The diameter of a high quality filament is very constant but 1.75 or 2.85 mm is only the nominal value. Mostly they vary +/- 0.05 mm , will say you have bought a 1.75mm filament between 1.80 and 1.70 mm. And this is a great difference. It needs to measure out the real diameter which is mostly constant +/- 0.01

The multiplier is as written the corrective to the diameter. You can change the diameter OR the multi, changing both makes no sense.
The both values (and others) define the quantity of extrusion, depending to the LayerHigh, Layers ExtrusionWidth and at the top also the rate of heat-expansion, because the rate of shrinking define the result of exact measures of the object.
For a fine quality (= fine bonding, accurate angles, a.s.o.) a moderate over extrsuion can be possitive, thats also a question of the model to be print, but I would say it is a +/- value of 1 or 2 % of the flow. Find out your basics data of the actually spool of filament and vary the flow with displays settings of the printer.

Comments to your settings-zip-file will follow later ...

H.

Here we are ...

The settings may be run on its special printer and with its special filament which is not named.
But the temperatur of 230 ° is to low for ABS / ASA (~245° and also no cooling but the best for a inhoused printer with more than 50° C inside the box). For PLa or its derivate (e.g. carbon) it seems to hot or the speed must be strongly higher than 3600 mm/min = 60 mm/sec. *)
Printing with only one perimeter and no Infill will say it prints always a outline with 70% = 42 mm / s and when a layer needs only 15 sec or below for one layer it reduce the speed down to 30% of 60 mm/s = 18 mm/sec.
A LayerHigh of 0.25 mm is always critical for my printer how ever it may be caused. To find the best balance between LayerHigh, LayerWidth, Speed, Temperature, Cooling and the models conditions is very hard and it needs a lot of experiences. To know the causes for a bad result and how to improve it is more than the half print ;-)

I think, that is enough of theory, print your models and show pictures when you are believing better results are possible. And always attache your settings and if possible the complete factory-file or STL-files.

*) go to menue Tools > Options there you can switch the dimensions to mm/ s which is better manageable

H.

horst.w
Posts: 795
Joined: Fri Oct 17, 2014 5:00 pm

Re: Horizontal lines in test tower

Sun Nov 19, 2017 3:35 am

Yesterday evening I have upload a spool of HD-Class (~PLA HT) to save it from waste.
Diameter is not round, it has 2.84 to 2.92, also I calculated 2.88

Settings:
Diameter 2.88
Multiplier 0.92
ExtrusionWidth 0.48
LayerHigh process 1 = 0.20 (small base socket); process 2 (overhang area) 0.12; process 3 0.20 (wide socket)
Temperature: (manufacture's recommendation 195° - 225°) 195°
Cooling 4 steps from 10 to 90% (layer 10)
Speed 65 mm/s; Outline 80%; NO overrides

and here the result while the print is running there were no improvements needed
20171118_233448.jpg
20171118_234231.jpg
H.

InspGadgt
Posts: 34
Joined: Wed Oct 18, 2017 7:59 pm

Re: Horizontal lines in test tower

Mon Nov 20, 2017 3:28 am

Well I've applied some of what has been discussed including upping the speed...I increased it to 80mm/s and am printing out the first 3 pieces of the fuselage for my plane and it is looking much better than my last attempt. It took a few tries but I'm really starting to zero in the settings at least for this PETG print. I still get some zits, far less than the last plane and I'm not getting the lines I was getting before. I'm only about 4mm into the print but so far so good. Hopefully it stays like this for the whole print.

horst.w
Posts: 795
Joined: Fri Oct 17, 2014 5:00 pm

Re: Horizontal lines in test tower

Mon Nov 20, 2017 9:54 am

Happy printing!

If you are interested, have a look to the german section of reprap.org

http://forums.reprap.org/read.php?238,776574

a flying ellipse.

The autor has also published a number of videos on YouTube e.g. https://www.youtube.com/watch?v=L9iDHNIykRQ

H.

Return to “General Discussion and Tips”