Dual Spiral Lightbulb / Week of June 09, 2012

Initially I chose the half-size Spiral Lightbulb for this week's print. When printed at half-size the print was more of a test of physics rather than a test of a reprap style printer. The pressure from a thread being extruded was enough to compress the tendrils. I modified the print to include a second spiral in the opposite direction to add structural support and I also removed the base.

[www.thingiverse.com]

As typical for my first print of an model I printed this at 20mm/s and at 0.33mm layers and 0.65 thread width. I intended to turn the 40mm overhead fan on after the first layer but forgot until more than half way up the print. Estimated print time was 35 minutes and actual time was 50 minutes. Min layer time was 10s.

For judging this contest I am looking for quality vs actual print time.




Edited 2 time(s). Last edit at 06/09/2012 08:49PM by billyzelsnack.

Very nice, I have not printed the original yet, so I'm really looking forward to attempting this one.

It's just 53.1mm high! cool.

---

[richrap.blogspot.com]

i seem to be having problems slicing this with slic3r 0.8.2 any one else ?

i tried exporting in netfabb makes no difference

Update on that

Seems to not crash if you use the quick slice feature

Edited 2 time(s). Last edit at 06/09/2012 07:13PM by Enlightx.

I tried it in 0.8.1 with the default settings and it said it successfully sliced. A quick look at a visualization of the gcode and nothing looked crazy.

I couldn't slice it on my laptop. It kept running out of VM. It worked fine on my desktop.
I tried using my same settings from the tower challenge and they need tweaking.
It is so small that you need to slow down times per layer I think. Printing a couple at a time might help as well. Doubt I will be able to get the 50 min time even dividing by two.

---

Lee

Hi Guys,

I printed one on an Up! printer for reference.

Settings:
Z layer height 0.30mm
10deg overhang support
solid fill
"fine"

What was the print time on the Up!?

btw. How in the world did you get those supports out?

Edited 1 time(s). Last edit at 06/09/2012 08:56PM by billyzelsnack.

Here is my shot at it, pics are straight off of the machine :


I used the latest version of Slic3r, the settings are below:
; layer_height = 0.2
; perimeters = 2
; solid_layers = 4
; fill_density = 0.25
; nozzle_diameter = 0.3
; filament_diameter = 1.75
; extrusion_multiplier = 1
; perimeter_speed = 30
; infill_speed = 50
; travel_speed = 150
; scale = 1
; single wall width = 0.60mm

I had the min layer time set to 10 seconds, and a desk fan running next to the machine. Time taken was 1hr 10. I always find I get the nicest overhangs when you set the single wall width equal to the natural diameter the extrusion takes in free air. I find this can change depending on loads of things, plastic colour, type temperatures all sorts. For Makerbot green ABS and my 0.3 nozzle I measured it to be 0.6mm. I printed it in ABS as I find it generally prints nicer on a bowden setup.

The bridge layer came out a tad scruffy tho



Edited 2 time(s). Last edit at 06/10/2012 11:55AM by rhysjones.

Very nice. What's the machine?

I'm definitely going to try your overhang tip. Is that an ABS only tip or does it work with PLA too?

It was printed on an eMaker/RepRap Pro Huxley with one or two minor changes.

I find the overhang tip is a general rule, but the problem with having fat walls is it makes the solid fill not quite so nice, so if the print is fairly basic I tend to turn it down a bit

Edited 1 time(s). Last edit at 06/10/2012 01:17PM by rhysjones.

For fun I printed one at half size. 26.5mm. 0.17mm layers so the same numbers of layers in my other print. 0.85mm thread width! haha. I set wipe to zero so the nozzle would not be playing in the tiny tendrils so long so it's a bit more stringy. 10s min layers. 37 minutes actual build time. It looks nicer in real life than in the photo though that is probably due that it is so small I can't even focus on it if it's closer than 8 inches from my eye.



Edited 3 time(s). Last edit at 06/11/2012 12:07AM by billyzelsnack.

Hey Billy I have to ask. What does wipe do? I have searched the Kisslicer forum but came up with nothing.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

It took me awhile to notice, but if you hover (without clicking) over any of the entry boxes it'll show a tooltip. Wipe's tooltip is:

Before moving to a new region, back-tracks over the most recent path by this distance.

im having an interesting time with this model.

got a thread going in the general section.

bascially what ever speed settings i use i cannot alter the speed of the print.

the speed settings will alter the round base but as soon as you get to the main fiddly bits of the print i have no control at all over anything.

iv now completed around 5 prints from 30mm/sec to a crazy 300mm/sec all around 1hour 13 mins

rhysjones can i ask could you use the platter part to export the gcode in latest slicer or did you have to use the quick slice?

I could use either. There is something funny going on though, I sliced one and then had a change of plan and resliced with different settings and the slice time went from 15 minutes down to 90 seconds, I hardly changed anything. I am running the github version though rather than the official release.

Edited 1 time(s). Last edit at 06/11/2012 03:33AM by rhysjones.

I agree the challenge on this one is more than just the print. It really maxes out the software.
I tried Kisslicer for the first time as well. That will take several calibration cubes to get right before I jump into something like this with it.
I did get the start of a decent print with it though, but it fouled about 1/3rd way in.

With Slic3r, I have even had my desktop crash the program a few times.
It's a win7 64 bit, 2600K overclocked and liquid cooled with 16 GB ram and a 60 GB SSD. Nothing wrong with the hardware.
Just a challenging drawing. When it crashed though, I was trying to do a couple of these on a plater. It will only complete the processing if I use a single bulb.

---

Lee

Here is my best effort yet.

New Slic3r
Print time was 78 minutes.



bottom_layer_speed_ratio = 0.65
bridge_fan_speed = 100
bridge_flow_ratio = 1
bridge_speed = 40
complete_objects = 0
cooling = 0
disable_fan_first_layers = 1

duplicate_distance = 6


extrusion_multiplier = 1
extrusion_width_ratio = 0
fan_always_on = 0

filament_diameter = 2.85
fill_angle = 45
fill_density = 0.2
fill_pattern = rectilinear

first_layer_height_ratio = 1
first_layer_temperature = 185

gcode_flavor = reprap
infill_acceleration = 50
infill_every_layers = 1
infill_speed = 50

layer_height = 0.2

min_print_speed = 10

nozzle_diameter = 0.33

perimeter_acceleration = 25
perimeter_speed = 45
perimeters = 4

randomize_start = 1
retract_before_travel = 2
retract_length = 4
retract_lift = .2

retract_speed = 30

skirt_distance = 6
skirt_height = 1
skirts = 1
slowdown_below_layer_time = 15
small_perimeter_speed = 45
solid_fill_pattern = rectilinear
solid_infill_speed = 50
solid_layers = 3

support_material = 0
support_material_tool = 0
temperature = 185
threads = 2
travel_speed = 150



Still want to try it a few more times. My camera or photography skills aren't that great either.

Edited 1 time(s). Last edit at 06/11/2012 11:47AM by Leeway.

---

Lee

This is a tricky little blighter, I had a quick first test, printed much faster than I expected, maybe a little too fast as I had major issues keeping it cool enough and a few 'twangs' of the structure as the head crashed into it a few times, nothing got snapped off, but a few missing chunks and some strings formed.

0.2mm layers
0% infill
2 solid layers

Print time was just 39mins 30 seconds.





I'm going to try another attempt, with only a single outline (if Slic3r will let me?) 0% infill again - as that seems to stop it doinglots of tiny pointless fills and extra extruder moves.
And maybe a little cooler filament, and Lots more cooling, also a bit slower.

---

[richrap.blogspot.com]

richrap Wrote:
-------------------------------------------------------
> I'm going to try another attempt, with only a
> single outline (if Slic3r will let me?) 0% infill
> again - as that seems to stop it doinglots of tiny
> pointless fills and extra extruder moves.
> And maybe a little cooler filament, and Lots more
> cooling, also a bit slower.

To stop all the travel moves from the infill you can also set it to do 10+ perimeters and it will do each post one at a time.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Leeway Wrote:
-------------------------------------------------------
> I agree the challenge on this one is more than
> just the print. It really maxes out the software.
> I tried Kisslicer for the first time as well. That
> will take several calibration cubes to get right
> before I jump into something like this with it.
> I did get the start of a decent print with it
> though, but it fouled about 1/3rd way in.
>
> With Slic3r, I have even had my desktop crash the
> program a few times.
> It's a win7 64 bit, 2600K overclocked and liquid
> cooled with 16 GB ram and a 60 GB SSD. Nothing
> wrong with the hardware.
> Just a challenging drawing. When it crashed
> though, I was trying to do a couple of these on a
> plater. It will only complete the processing if I
> use a single bulb.

Having problems trying to slice two on the one Platter also. Slicer (0.8.2) crashes on two different machines, one PC has lots of power. Will slice two on platter if settings have far fewer layers and big nozzle (.5mm), but I don't have a big nozzle to get to those settings (settings chosen just to see if would slice). Only reason I'm was going to try print two up was because I was having trouble cooling one on it's own without suffering heat loss at the nozzle. We'll see how it goes!
NumberSix

---

[numbersixreprap.blogspot.com]

I printed one with two 3/8" cylinders sitting beside the bulb for that reason, but my settings were way off on it. It did slice fine though, just had some bridging issues.

---

Lee

Blobtastic!

Left. .33 and 50 minutes.
Middle. .33 and 23 minutes. 195C No wipe. 3s min layer.
Tiny. .17 and 37 minutes.
Right. .33 and 16 minutes. 180C No retract or wipe. 3s min layer.

Estimate for Middle and Right was 9m50s. I no longer think retractions are what causes the build time under-estimation. I think it might have to do acceleration. My acceleration settings are still set at defaults and I'm wondering if estimations just presume some sort of acceleration level and for my printer with its double pulley setup it's off by a factor of two. Maybe I should gander at the RepG source because it seems to be just as wrong as the slicer.

btw. I uploaded a reduced triangle version to help with the slicer troubles. I don't want this to be a slicing competition. [www.thingiverse.com]

Edited 5 time(s). Last edit at 06/11/2012 07:31PM by billyzelsnack.

Sublime Wrote:
-------------------------------------------------------

>
> To stop all the travel moves from the infill you
> can also set it to do 10+ perimeters and it will
> do each post one at a time.


Thanks for the Tip, I'll try that tonight, cheers, Rich.

---

[richrap.blogspot.com]

A tricky print to get right, this one. I've tried a few times now, going for as much speed as I can without sacrificing too much quality, but the strands are a little too thin for my setup to do cleanly I think.

My second attempt was a decent balance, but with quite a bit of stringing (I'm having a few issues with that at the moment ) Stringing seems to be from deposited plastic and not from the nozzle per se, so z-lift would help at the cost of a lot of speed!



Mendel, 0.5mm nozzle ABS

Slic3r - defaults plus:
0.28mm layers
1mm retraction, no z-lift
2 perimeters
0% infill
15sec min layer time
Upped perimeter speeds by a bit (to 38mm/sec IIRC)

41.10mins print time

Image gallery of attempts (3 so far)

benosteen Wrote:
-------------------------------------------------------
Stringing seems to be
> from deposited plastic and not from the nozzle per
> se, so z-lift would help at the cost of a lot of
> speed!

Interesting insight. Now that you say it I think that might be the cause of my fine strings as well.

Stringing seems to occur when the move from the end of a perimeter to the next one trails over a part of the hot, recently laid perimeter, especially an overhanging one. Subjectively, it looks like they are the most prone to swelling or curling up after being laid down.

I need to pick up some calipers to check the filament width too, in case it's wider than I think leading to the gcode extruding just slightly too much.

However, I'm now considering how to use it aesthetically If you can't beat them...

UP! print time was 1:16 4.7g material

In my opinion, the biggest drawback to the up is the closed-source software. It's a great little machine, but the company has it's own view of how things should be done, and that includes not allowing people to disable supports. However, that being said, the supports were pretty easy to remove using a pair of needle-nose pliers.

On my old Rep-Rap I printed the original spiral lightbulb out without supports no problem. I'm building a MendelMax 1.5 at the moment!

It would be greatly appreciated if someone could share a quick video of the lower layers of this object being printed, showing the looping that forms the tendrils. It would give a better indication of how the extruder nozzle should be performing when printing this delicate piece. If you could pan to where your fan/cooling is positioned it would be an added bonus. I have managed to print this object but it’s fairly ugly! Currently resembles, hmmm, two spider crabs playing twister!

This intensive shared focus on a specific object for a short duration, and the sharing of settings seems to be an excellent way of learning and improving printer set-up. Thanks to all involved.
Regards,
NumberSix

Edited 1 time(s). Last edit at 06/15/2012 03:46AM by NumberSix.

---

[numbersixreprap.blogspot.com]

Some raw extra boring vids. I have a third, but its upload failed. Is this what you wanted or further up?
[www.youtube.com]
[www.youtube.com]

0.17 layers. 2mm retract. 10mm wipe. 1 perimeter. 15 second cool. 1h27m actual. KISSlicer released today.


I think I need to tweak my settings for the latest version. Seems a bit light on plastic output now.

Edited 2 time(s). Last edit at 06/14/2012 11:49PM by billyzelsnack.

Thanks Billy, That's great. You're using a gentle print rate, and there's no sign of any curling or pushing of plastic, which I'm running into. You must have some cooling going on of the layers just printed also.

Once I climb out of the base into the tendrils, they remain a bit 'jelly like' and get moved about a bit by the nozzle printing the next layer. I think I need to insulate my hot-end a bit more so I can apply more cooling to the 'just printed' layers. I notice you have your hot-end well wrapped up.

Does the fan on your extruder push air downwards to cool the print also or have you another fan to one side?

Must say, I'm very impressed by your little machine! You are getting very precise Z control with belt drives, and it's print quality is superb.

I've taken a look at KISSlicer but all those settings and slightly different terminology just frightened me off for now. I'll push on with Slic3r for the moment.

Tks for posting the video clips. Much appreciated.
NumberSix

Edited 1 time(s). Last edit at 06/15/2012 04:11AM by NumberSix.

---

[numbersixreprap.blogspot.com]