calinb
-
Posts
305 -
Joined
-
Last visited
Content Type
Forums
Events
3D Prints
Posts posted by calinb
-
-
It's pretty common for CAD tools to export meshes that cause problems with other tools and it's why netfabb offers repair tools and Cura has "fix horrible." You can try meshlab for repairs too, but I've never mastered it.
I've had pretty good luck exporting from Geomagic (formerly Alibre) and I've exported some very complex models. The more complex the solid model (or the more quick and dirty "hacking" that's been done on a model), the more likely one is to encounter problems in an exported mesh. The solid operations that are most likely to cause problems are the direct editing functions (push/move faces/surfaces, delete faces/surfaces, etc.) or boolean operations. I still use them, but I'll redo or create a bunch of sketches to avoid doing these operations on a solid model, if possible. Sometimes a problem associated with these operations doesn't pop-up until much later--after the model "grows" larger during development.
@calinb: Thanks a lot for your help. Your comments definitely gave me a lot of food for thought as I try to understand why and where MOI would generate a broken STL model when exporting solids. I'll keep on digging and let people know if I find anything interesting to report.
-
Thanks for the model. It greatly helps us to help you!
Here's what I did:
1. Opened with netfabb (you can download the basic Studio version for free still, I think, or you can try the online version).
2. Default repair.
3. Cut out just the problematic bathroom (in the middle of the walls).
4. Default repair + remove degenerate faces.
5. Scaled to fit on Ultimaker bed.
6. Exported .stl.
7. Opened .stl and sliced with Cura.
In examining the gcode with both Cura and Rep-Host, it looked okay!
With this large model, you might want to cut-out the rooms as separate models, print them separately, and bond them together later too (easy to bond ABS, but not as easy with PLA, and impossible with nylon). However, if a strong mechanical bond is not required, you can probably just use silicone sealant to "bond" the parts--even if printing with PLA or nylon.
Next I tried your entire model:
1. Opened with netfabb.
2. Default repair (no degenerate faces this time).
3. Exported .stl + repaired (split) manifold edges (this is the operation that can cause problems with Kisslicer).
4. Opened with Cura and scaled to fit on bed.
5. Sliced with Cura.
A cursory viewing of the gcode toolpath revealed no problems.
Based on my relatively brief experiments, I think I could print this model just fine. Don't give up! Despite what many 3D printer marketers would like you to believe, 3D printing ain't like printing your family vacation photos on an inkjet (and I don't see that situation changing anytime soon). Relatively recently in life, I've learned to fairly competently run a lathe, milling machine, and 3D printer (Ultimaker Original). The 3D printer had (has) the steepest learning curve--at least for me. And I'm a "computer guy" too.
Keep us informed of your progress and good luck!
-Cal
@calinb and @IRobertI: Thanks for your comments. There is one small red area, but it's in a completely different part of the model.
I don't know if it's a Cura issue or not. I've enclosed below a link to the stl in case someone with better skills than I have want to have a look: https://dl.dropboxusercontent.com/u/109695611/2014-08-14%20---%201ST%20FLOOR.stl
The model was scaled down to 0.12 in Cura.
-
It seems like the 297 mm belt (297MXL025G) is probably best for an Ultimaker Original. My long belts are getting long enough that it's becoming difficult to take up the slack with my clothespin spring tensioners so I'm planning to order new long belts from B&B. The minimum order is $35 and they also offer a discount for 10+ quantities of the same belt SKU. If I order three long belt sets (12 belts), the price works out to $19.92 / set total. (Including the B&B shipping and handling charge of $10.07).
If anyone in the U.S. needs new belts, I could package a set or two in a $5.95 USPS small Priority box (which includes up to $50 insurance) and mail them to you. At $25.87 / set, you'll be out of pocket about $20 less vs. placing a single set order with B&B. Please PM me before the end of this week (August 23rd, 2014), if you are interested in a replacement set(s).
-Cal
-
Yes--try fixing any errors, but I've actually had cases (in Kisslicer) where the netfabb "fix" was worse than the problem, because it caused new perturbations in the tool path, though the model passed error-free after the "fix."
I would also try another slicer. You can't have too many of them!
You could also cut this problematic area out of the model to reduce test print time, but the problem might go away from just doing that. (You can cut the model with the free to download netfabb.) To eliminate test print time altogether, load the gcode and preview the tool path in Cura (View mode >> Layers) or Repetier Host (Preview tab).Please report your findings here. It might be a Cura issue or bug.
-
As it's stuttering on the way home at the beginning of a print job, I can briefly toggle the power switch and quickly re-start the print job with the Ulticontroller. Then it continues on its way just fine, regardless of the position of the Z-stage (regardless of whether it has to travel 200 mm in Z or only 5 mm in Z to finish homing). It NEVER exhibits this problem when a print is started immediately after power-up (again, regardless of whether it has to travel 200 mm in Z or only 5 mm in Z to finish homing.)
I've removed the leadscrew many times and the stage moves with no perceptible resistance on the bearings. I can also easily turn the leadscrew coupling with my fingers and screw the stage up and down manually.
My Ulticontroller had a number of assembly defects, which I corrected. There might still be something wrong with it.I guess the next step will be to disconnect the Ulticontroller and run the printer from my laptop. If the problem remains, I'll probably need to replace my Arduino Mega 2560 board, unless someone here reports experience with a specific component that causes this problem.
There have been similar problems recently that were fixed by replacing the vertical rods and the bearings that go on those 2 rods. It's probably too late to get a warranty replacement. But there are some new things coming out of UM to be announced probably Sep 19 at MakerFaire NYC... I'd wait until then.
-
A "master reset" (power cycling--even extremely briefly) clears the problem and it will not reoccur until after the print completes and the next print is started. The stuttering correlates to the termination of one print and beginning of the next print and it is completely reproduceable.
Alternatively, if I power-up the machine from cold and move the stage using the Z-movement control under the "Prepare>>Move Axis" Ulticontroller menu, the stuttering will occur when the machine homes to print the first print. As before, cycling the power will clear the problem.
These symptom don't seem to correlate to a driver overheating to me but, if I can find the adjustment, I'll try to turn the current down.
Nick, did you increase the feed rate of the Z axis?
My guess is that the Z driver is overheating and in Nicks case increasing the feed rate and moving the bed up and down at that increase rate is overheating the driver (barely). It momentarily shuts down and quickly cools just enough to restart.
For both of you, try DECREASING the current.
-
My power supply brick just died (not quite 2 years old) and I replaced it with a higher quality unit. Unfortunately, my hope that the Z stuttering would go away was not realized. Has anyone found the cause of this problem yet? The period of my machine's stutter is much longer than Nicolinux's machine's problem. I also don't have to nearly bottom the stage. It will do it on nearly every print when homing, if I don't power down between prints.
My machine has always been plagued by it. It's actually been plagued by many electrical problems from the day I received it, which have consumed countless hours of my time in the course of correcting them--all except this one, which continues to elude me.
-
A heated bed for PLA is very compelling, because you can use hot swap glass and there's no muss, no fuss, and it is very fast to remove the plate and start a new print on a new plate. As soon as the plate cools, the parts fall right off!
-
Or maybe print the sides in nylon. I mostly print nylon these days. Also, I just changed one of my company's rapid injection molded parts, produced at Protomold, from ABS/PC to nylon. Nylon is inherently self-lubricating and low wear. Though our mold was designed for ABS/PC and I was initially concerned about the higher shrinkage of injection molded nylon vs. ABS/PC, it's not an issue in our part.
Protomold was very fast too. All I had to do was order more parts in nylon, pay the bill, and scrap the old parts.
/ :(It looks like ABS is a mistake for the UM2 feeder. It seems to slowly wear, which causes ABS dust to end up in the hotend and clog the hotend. We're working to get the supplier to change the material, but it's not going as fast as we hoped.
A quick workaround is printing the sides in PLA, and use those instead of the ABS sides.
-
Yes. As you've learned, no single slicer is all things! It's good to have and learn them all.
As my nozzle warmup is completing before a new print, I manually prime my nozzle to pressurize it and clear any jams or crud from it. I do this with the stage about 50 mm or more below the nozzle at ~x=0 / y=0. I turn my UM "Classic" (original) big wheel with my left hand while force feeding plastic up into the feeder with my right hand. A nitrile or latex glove can improve one's grip on the filament. I push hard on the filament--hard enough to nearly lift the frame from the bench or maybe even harder sometimes, while holding the frame down with my left wrist. Once plastic flows nicely and with good quality (no bubbles, steam, or discolored chunks or impurities from prior shutdown "heat creep"), I use my right hand to turn the big wheel more slowly or even intermittently until my UM homes. Before it finishes homing, I clean any blobs from the bed and nozzle and make sure only a short, single, clean thread contacts the bed at the home point. I've become adept at doing all this without getting burned (or burned only superficially
) and I even hold the extruding and cooling filament clear of the bed while feeding filament with my right hand at the big wheel--right up until the hot extruded thread contacts the bed.I look back on all the early jams I had in my hot end when I first started printing (PLA) with my new UM. I looked for problems in the hot end design and knurled/checkered drive stud and pressure "foot" and I did find some dimensional problems in my V1 (pre-Bertho style) extruder (feeder), but I now know that much of my newbie frustration could have been avoided with a few simple tricks, like starting runs per the above. After printing many, many spools of PLA, ABS, and nylon, I never have jams anymore.
Thanks Cal
I'm glad I'm not doing something stupid. I guess for this model I'll just need to use KISSlicer - but KISS has other issues like not priming the extruder by drawing skirts :sad:
Cheers
Roger
-
Well--at least it's so blazingly fast, I can slice, view, and spin gcode of my own models implementing my own support structures and iterate my solid model CAD very quickly!
There is an 14.02-RC1 at http://software.ultimaker.com/Cura_closed_beta/ which has an extra configuration setting for the support angle, adjusting this one generates more or less support. Might help.
Still, Cura's support material sucks. I know it.
-
I'll wager that I've submitted more tickets to netfabb than anyone else. I told them I was giving up on reporting more bugs until they clear the open ones I've already reported. I also told them I would not buy more of their software until they satisfactorily address and close my open tickets. Oh yeah--I submitted a number of bug tickets for their "Pro" software when I was running their 1-month trial of it too. Haha-guess I'll be so much the richer, because they'll never work on these tickets!
Complaint stated (above), I'll also report that I've have excellent support from one of their programmers, directly. He actually implemented one of my requests and sent the executable program to me for testing. Sure--this happens all the time in the open source community (source tree archives are setup for it) but it's rare in commercial software. Also, I love netfabb for what it is--the most sophisticated and professional-looking slicer and mesh repair/viewer software to date (and it's been stagnant for a over a year while the other projects march onward). I'm very sad that it has never realized its full potential with bug fixes, updated features, and excellent documentation.
If you'd like to post the text from your open ticket here or send it to me via PM, I may be familiar with the issue and, possibly, suggest or work-around or provide some guidance (netfabb documentation is also sorely lacking and out of date).
-Cal
I called the support direcly (i come from germany), because there was no reaction of an support ticket since 2 weeks. I got the answer that this is still normal and they ask me what i expect i have bought only the Ultimaker engine licence. For a faster reaction time i have to pay more money. My support Ticket is still open with no reaction since 16.09.2013.
-
Roger,
I wish I could, but I've given up on Cura support and I design it into my models, whenever I wish to use Cura. Sometimes it works, but usually Cura misses important overhangs and, despite my hacking of the .py file, as Daid suggested, I can't get the results I need.
I agree that Kisslicer support is quite versatile and good. My favorite is actually the netfabb UM engine, but I can't say that I'd recommend netfabb, because it is expensive, buggy, difficult to learn, and poorly supported. It does, however, feature far more control over how your print is accomplished than any other slicer. If you ever find yourself looking for layer-by-layer control over a variety of custom and independent extrusion "types" (as I call them) it's the tool for the job. Oh--BTW, netfabb support works great for PLA right out of the box, but it took me quite a deal of effort to devise effective support settings for ABS.
Sorry to be of no help to you, but I wanted you to know you are not alone.
-Cal
Could someone advise me, how I can get Cura to generate support material for this model?
Thanks
Roger
-
My UM can print 210 mm x 210 mm, but I was never able to quite achieve full. symmetrical, and square factory-published UM-spec travel after using any of the measurement and alignment tools.
The large 8 mm "axes" rods are fixed. There's nothing you can reasonably do to adjust their alignment. If they are not nearly square in the frame, you have a problem that I've not experienced and I will not address it here.
Therefore, one can only twist and load the small rods, using the pulleys and belts, to maximize the travel of the head at the machine's mechanical and switch limits. FWIW (your mileage may vary), this is roughly how I adjust the alignment of my small diameter rods. I use an iterative process and, now that I have a feel for what's required, I can adjust the rods in about 10 minutes after replacing all the belts.
First, the goals.
Goal #1: Adjust the pulley positions along the small rods such that the long belts remain parallel to the large 8 mm rods as the print head moves throughout its travel range.
Goal #2: Adjust the limit switches to produce their "click" just as the print head reaches its mechanical limit.
Goal#3: Adjust the pulley positions such that the above switch limits are reached at both ends of the print head's complementary positions. For example when adjusting the x=max limit switch, the switch's audible click should occur just as the head reaches its maximum x-travel, regardless of whether the head is maintained near either the y=min or y=max positions.
Once you achieve the above goals, your UM is optimally adjusted. The process requires trial and error iterations using the following techniques:
1. While moving the print head though its full travel in X or Y and holding the other axis steady, watch the moving long belts. As the head approaches its travel limit, the distance from the block to a pulley will decrease and tend to visibly pull the belt to the side, if the pulley is not correctly aligned along the length of the large 8mm rod. Move the head back and forth between its limits while sighting a moving long belt against its nearby parallel large 8mm rod or the edge of the top "window" in the frame. Adjust the pulley positions until the belts run parallel. Once you achieve Goal #1, you will probably not need to do this step again, unless the position of a pulley along its rod is disturbed in the next step...
2. With one set of long belt pulleys (four pulleys) slightly loose, apply a twisting force to the small rod (push in opposite directions on each end of the rod) and tighten the pulleys. (It helps to be an octopus but you can use clamps or jam blocks, if you only have two hands.) Changing the angular position of the pulleys on the large rods, in this manner, will "bias" the belt tension and change position of the end of the small rod and its mating bearing block. Using this technique, you can eventually realize Goal #2. Alternately, it is sometimes best to just adjust one side at a time (only loosen two pulleys at a time) or one side only.
The process is much easier to demonstrate than describe. I will try to find time to take a video of the procedure after I install my "direct stepper drive" parts and ditch my short belts. I've had the parts for a long time but never seem to find the spare time to install them.
-
Sounds like you have addressed this too, Bertho.
I think avoiding the heat creep is important at shutdown. I'm not sure that retracting the filament after the cool down period helps at all, though I mentioned it as an option. (My code, above, doesn't retract). I think that retracting the filament might actually cause problems, on rare occasions, because small fingers or bridges of filament can be formed at the end of the Bowden tube/Teflon tube, which might fold over and jam the next time the filament is driven into the hot end. I just don't know, but I know that avoiding heat creep on shutdown is a good thing.
calinb lists many good points above.
I too like to avoid the heat creeping up the extruder after shut down. To avoid it I slowly feed some extra filament while the head normally cools off. That way cold filament is pushed in so the heat never creeps up and later creates a possible jam at start-up.
-
[...] If I'm not around, I use shutdown G-code to at least purge the head, reduce the temperature in stages (feed at ~10C steps) and retract the filament before it cools. [...]
I'm interested in the way you set this up - could you please share/explain how you do this??
Sure. From extensive destructive testing of my parts using a torque wrench to measure failure loads, I know my Ultimachine black ABS requires high nozzle temperatures to achieve good layer bonding strength. (I often print it at ~260C). The heat creep at shutdown can create problems for the next print run. Try something like this:
G91 ;Set to relative modeG1 X2 Y2 Z0.2 E-5 F10800 ;Suck back 5 mm filament to minimize blob on the partG1 X0 Y0 Z4.8 F500 ;Keep moving awayG90 ;Set to absolute modeG1 X209.00 Y209.00 F5000.00 ;Move to prime zoneG92 E-5 ;push previous 5mm filament back into nozzleG1 E1 F500 ;Extrude 1 mm net for cool-downM109 S250 ;Cooldown--wait for 250 CG1 E2 F500 ;Extrude 1 mm moreM109 S240 ;Cooldown--wait for 240 CG1 E3 F500 ;Extrude 1 mm moreM109 S230 ;Cooldown--wait for 230 CG1 E4 F500 ;Extrude 1 mm moreM104 S0 ;nozzle cooldownM84 ;disable motors
-
Based on my BKM (Best Known Methods) I don't know that there's any room for improvement. When I need optimum (usually perfect) adhesion, I use the following surfaces.
PLA:
Hot-swappable glass plates on warm (~70C) precision aluminum Mic-6 bed plate. With a little heat, it's very convenient and easy-peasy to get perfect adhesion with PLA. When cool, parts just pop off by themselves
Though not essential for PLA, heat is so compelling I can't imagine not having a heated bed for PLA!ABS:
ABS "juiced" Kapton tape stuck to hot-swappable glass plates over hot (~120C) precision aluminum Mic-6 bed plate. I use a heated build "tent/chamber" and ambient air temp of up to ~65C. The control of temps, fan, and overall process can be very complex and trying, if perfect adhesion of thick and blocky parts is desired. For printing anything other than very small and thin parts in ABS, heat for the bed and build chamber is essential. Just look at the Stratasys patents!
Nylon:
1/16" Garolite LE or CE (LE or sanded CE is best) bonded to aluminum Mic-6 bed plate using water soluble wood glue. The thicker Garolite is often badly warped from the manufacturer--too warped to machine and use without the Mic-6 plate. The Garolite can be renewed, when necessary (not very often) by soaking the plate in a bucket of water. The 1/16" sheets are flexible and bond to provide a reasonably flat surface, which can be further leveled by sanding on a large plate of thick glass (inherently flat). For the best adhesion, a glue stick can be applied to the Garolite and blotted with a terry cloth before it dries to provide a little texture to the glue surface, improving its performance. A wet sponge works well to redistribute the glue between print runs. I have not found heat to be useful with nylon.
-
I don't have a UM-2 but I've printed ABS, PLA, and nylon extensively for nearly 1-1/2 years with my UM. Here is what I've learned.
1. ABS is more prone to jams and partial jams (slowdowns and under-extrusion) than PLA or nylon. You can expect ABS to under-extrude as a matter of course when running medium to high extrusion rates. Illuminarti has studied the problem extensively. The faster you extrude ABS, the more it under-extrudes. To some extent, I often compensate for this property in the slicer. If you learn how to use netfabb well (very time-consuming) it provides more flexibility for doing this than any other slicer. With netfabb's UM Engine, you can define many extrusion "types" and tweak the extrusion fudge factor up when extruding at high rates. It's also unfortunate that netfabb appears to be an unsupported commercial product nowadays and many bugs remain in the ap. Actually, the UM Engine has never been very well supported by netfabb GmbH.
2. I was disappointed when I first saw the very clean and sleek-looking UM-2 filament feeder design. Support for manually feed (forcibly pushing up on the filament with one hand while turning the feeder "big wheel" with the other hand) can greatly reduce the likelihood of jams and also get that most important first layer off to a consistent high quality start. With ABS, I try to be around when the print completes to force feed filament before retracting the filament from the hot end and shutting down. If I'm not around, I use shutdown G-code to at least purge the head, reduce the temperature in stages (feed at ~10C steps) and retract the filament before it cools (sometimes--don't know whether this step helps or hurts). Immediately before every print, I manually force feed the filament to both "clear" and pressurize the hot end to normal operating pressure and temperature.
3. I made a hobbing tool for my lathe and I have made a few conventional hobbed stud feeders in various tooth pitches. The best of them provide significantly more power than the standard UM "knurled" or "checkered" rev. 3 hobbed stud, but they are more prone to clogging than the original stud and they are difficult to clean when they do fill with plastic. Currently, I'm using the original rev. 3 stud, because it cleans up with a few quick swipes of a toothbrush and really never clogs to begin with. I simply slow down when/if it starts to slip too much and under-extrude.
If I were still printing a lot of ABS, I'd probably upgrade the entire feeder mechanism, but I mostly print nylon these days, sacrificing a little print resolution for improved strength and chemical resistance, as compared to PLA or even ABS. At normal nylon extrusion rates, the filament feeder drive capabilities are not a factor. I build manufacturing tools and commercial parts with my printer. It supplements my company's small machine shop. I don't print garden gnomes.
4. I've studied many filament feeder designs. Someday, I'll build a Rollerstruder:
http://www.tridimake.com/2013/04/rollerstruder-filament-feeder-driver.html
including its very large and aggressive hobbed bolt:
http://www.tridimake.com/2013/03/which-hobbed-bolt-for-filament-feeder.html
Counter-rotating, dual-drive feeders also have merit, but they are more complex to design and build.
If I ever get a UM-2, I'll hang a UM-original feeder on the back or perhaps the Rollderstruder.
Just my $0.02!
-
This bug appears to have crept-in sometime between 13.10 and 14.01. I used Rep-Host to confirm the tool paths and the following.
My Ultimaker can print over a full 210 mm x 210 mm range in X/Y and I often use all of its potential. When Cura 14.01 Machine Settings are set to 210 x 210, Cura reports that a 209.09 mm x 209.09 mm model (as reported by both Cura and netfabb) is too large. It also complains when the Machine Settings are set to 211 x 211. Settings of 212 x 212 are required before Cura 14.01 will accept my 209.09 mm x 209.09 mm model and build gocde. Cura 13.10 works properly and centers the model when set to 210 x 210.
However, setting the bed width and depth to 212 x 212 in 14.01 does not result in a successful work-around for Cura's spatial pessimism, because the oversize machine settings improperly "bias" the placement of the model in a positive x and y direction and the print cannot centered manually on the bed. It is impossible to locate the model closer than about 1 or 2mm from the X=0 or Y=0 edges. With respect to supporting maximum-size models, it would be nice if 14.01 could be returned to the capabilities of 13.10.
I heuristically (and painstakingly) found that, for my 209.09 mm x 209.09 mm model, a work-around can be realized in 14.01 by enabling the Machine center 0.0 option, setting the bed size to 421x421, and manually moving the placement of the model to the maximum available X/Y location (the upper right corner). Without the use of Rep-Host to check the resulting gcode, this is not a reliable procedure.
Finally, regardless of Cura's intentions to not let me print outside of its calculated box, I'd prefer to be able to override Cura's limiting behavior when models are right at (or just past) the hairy edge of being over-size, because a slight excursion "out of bounds" for a brim can be safely handled by the printer x/y stop switches. In my case, I usually build my own brim and bed adhesion features (and support structures too) into my models, but Cura 14.01 refuses to build 209.09 mm x 209.09 mm gcode at all when set to 210 x 210.
-
I found the standard support to work with PLA and I mostly used weak fluff and fluff back when I printed PLA.
I had to customize the settings for ABS. Here's an example, if you can copy and paste the following into a .reprapstyle file, you should be able to try it (for ABS). I wish I could figure out a way to simply attach the file here. :(
<root>
<name>09: Ultra quality - FILLED OBJECT</name>
<outercontourwidth>fine</outercontourwidth>
<minregionarea>0.30000001192093</minregionarea>
<maxcontourlayers>2</maxcontourlayers>
<upskinminwidth>3</upskinminwidth>
<downskinminwidth>3</downskinminwidth>
<fillingangle>45</fillingangle>
<upskinangle>45</upskinangle>
<downskinangle>45</downskinangle>
<upskinlayers>25</upskinlayers>
<downskinlayers>30</downskinlayers>
<fillspacing>-0.125</fillspacing>
<upskinfillspacing>-0.34999999403954</upskinfillspacing>
<downskinfillspacing>-0.34999999403954</downskinfillspacing>
<upskinfeaturesize>0</upskinfeaturesize>
<downskinfeaturesize>0</downskinfeaturesize>
<supportcalculationlayerthickness>0</supportcalculationlayerthickness>
<supportdefaultangle>30</supportdefaultangle>
<supportstrongangle>40</supportstrongangle>
<supportweakangle>20</supportweakangle>
<supportdefaultspacing>2</supportdefaultspacing>
<supportstrongspacing>1</supportstrongspacing>
<supportweakspacing>3</supportweakspacing>
<supportminimumdiameter>0.5</supportminimumdiameter>
<supportlayerdistanceondownskin>1</supportlayerdistanceondownskin>
<supportlayerdistanceonupskin>1</supportlayerdistanceonupskin>
<supporttransitionlayercount>15</supporttransitionlayercount>
<supportpartconnectionlayercount>10</supportpartconnectionlayercount>
<minimalhatchlength>0.80000001192093</minimalhatchlength>
<supportfluffoverlap>4</supportfluffoverlap>
<supportpartlockvolumeoffset>1</supportpartlockvolumeoffset>
<regiondefinition1_outercontour>
<filamentdefinition>half_layersize</filamentdefinition>
<speed>snake</speed>
<width>fine</width>
<widthfactor>1</widthfactor>
</regiondefinition1_outercontour>
<regiondefinition1_filling>
<filamentdefinition>default_layersize</filamentdefinition>
<speed>slow</speed>
<width>thin</width>
<widthfactor>0.85000002384186</widthfactor>
</regiondefinition1_filling>
<regiondefinition1_upskin>
<filamentdefinition>default_layersize</filamentdefinition>
<speed>slow</speed>
<width>medium</width>
<widthfactor>0.75</widthfactor>
</regiondefinition1_upskin>
<regiondefinition1_downskin>
<filamentdefinition>default_layersize</filamentdefinition>
<speed>slow</speed>
<width>fine</width>
<widthfactor>0.85000002384186</widthfactor>
</regiondefinition1_downskin>
<regiondefinition1_innercontour1>
<filamentdefinition>default_layersize</filamentdefinition>
<speed>slow</speed>
<width>fine</width>
<widthfactor>1</widthfactor>
</regiondefinition1_innercontour1>
<regiondefinition1_support>
<filamentdefinition>default_layersize</filamentdefinition>
<speed>slow</speed>
<width>fine</width>
<widthfactor>1</widthfactor>
</regiondefinition1_support>
<regiondefinition1_innercontour2>
<filamentdefinition>default_layersize</filamentdefinition>
<speed>slow</speed>
<width>fine</width>
<widthfactor>1</widthfactor>
</regiondefinition1_innercontour2>
<regiondefinition>
<name>Normal</name>
<minarea>-1</minarea>
<maxarea>-1</maxarea>
<zjumpoffset>0</zjumpoffset>
<id>5</id>
<temperature_scale>1</temperature_scale>
<upskintype>0</upskintype>
<downskintype>0</downskintype>
<seamtype>1</seamtype>
<outercontour>regiondefinition1_outercontour</outercontour>
<filling>regiondefinition1_filling</filling>
<upskin>regiondefinition1_upskin</upskin>
<downskin>regiondefinition1_downskin</downskin>
<innercontour1>regiondefinition1_innercontour1</innercontour1>
<support>regiondefinition1_support</support>
<innercontour2>regiondefinition1_innercontour2</innercontour2>
</regiondefinition>
<override>
<name>support1</name>
<speed>1.39999997615814</speed>
<extruder>0.75</extruder>
</override>
<override>
<name>support2</name>
<speed>1</speed>
<extruder>1.5</extruder>
</override>
<override>
<name>support3</name>
<speed>1</speed>
<extruder>1.5</extruder>
</override>
<override>
<name>support4</name>
<speed>1.39999997615814</speed>
<extruder>0.75</extruder>
</override>
<regionselection>
<type>bylayer</type>
<threshold>2</threshold>
<threshold>5</threshold>
<threshold>9</threshold>
<threshold>725</threshold>
<item>
<type>useregion</type>
<regionid>5</regionid>
<temperature>0</temperature>
<heatedbed>100</heatedbed>
<fancontrol>0</fancontrol>
<speedoverride>50</speedoverride>
<extruderoverride>100</extruderoverride>
</item>
<item>
<type>useregion</type>
<regionid>5</regionid>
<temperature>100</temperature>
<heatedbed>100</heatedbed>
<fancontrol>0</fancontrol>
<speedoverride>100</speedoverride>
<extruderoverride>100</extruderoverride>
</item>
<item>
<type>useregion</type>
<regionid>5</regionid>
<temperature>100</temperature>
<heatedbed>75</heatedbed>
<fancontrol>0</fancontrol>
<speedoverride>100</speedoverride>
<extruderoverride>100</extruderoverride>
</item>
<item>
<type>useregion</type>
<regionid>5</regionid>
<temperature>100</temperature>
<heatedbed>75</heatedbed>
<fancontrol>0</fancontrol>
<speedoverride>100</speedoverride>
<extruderoverride>100</extruderoverride>
</item>
<item>
<type>useregion</type>
<regionid>5</regionid>
<temperature>100</temperature>
<heatedbed>75</heatedbed>
<fancontrol>25</fancontrol>
<speedoverride>100</speedoverride>
<extruderoverride>100</extruderoverride>
</item>
</regionselection>
<customization>
<partinformation>\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ </partinformation>
<changetoextruder>\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ </changetoextruder>
<changefromextruder>\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ </changefromextruder>
</customization>
</root> -
I'm all thumbs when using gloves that protect me from heat! :(
Yeah--I always prefer to risk finger burns than risk my hotend brass parts and it comes apart so easily when it's hot!
-
Joergen seems to be very busy these days, but I'll ping him about possibly responding to your query in this thread. Joergen did report the throughput to me but he also drills out his nozzles (or uses larger ones from MG) so I'd rather he answer your question here and provide any details himself, if he has time to make a quick post.
Can anyone from Ultimaker, or anyone with a UM2 already who has analyzed it in detail (Joergen?) give us some sort of idea on the maximum volumetric throughput of the extruder at at least some typical printing temperatures for PLA and ABS?
It would be very helpful when trying to diagnose folks' extrusion-related problems...
-
Yeah--I do scaling with xsteps and ysteps all the time using the Ulticontroller. There is no single "calibration" that works with all filament materials, slicers, models and processes (which I tune for the model), or even all bed orientations of asymmetric models. I rarely find it necessary to tweak zsteps in order to realize my accuracy goals but I have a log of xsteps and ysteps for nearly every model I print or plan to print repeatedly.
I've found that the test models I printed when I first obtained my Ultimaker were a waste of time. Now I only print (or partially print) tests of objects that I actually need. Generally speaking (and given that it squirts molten plastic), the UM is quite accurate right out of the box. Any further improvements in accuracy must be hard-won on a print by print basis, or at least extrapolated between very similar prints and processes.
Sure--when modifying some element of the UM design, as you have done, some calibration tests make sense, but I'd probably still print something that's useful too and take measurements from it.
BTW, one recurring challenge is to find a good accuracy compromise between inside and outside dimension on prints. Typically, settings are a trade-off between the two!
I often must be picky because I'm printing tools and commercial mechanical products and often attempting to hold dimensions within a very few mils. I'm not printing garden gnomes.
Sometimes I use my machine tools to finish projects that I print, where I can often hold dimensions to 0.001 inch.Spot on !
The Ulticontroller proves awesome yet again. Control -> Motion ->X-steps/mm, then back one screen and "Store memory"
That is going to make fine calibration pretty sweet.
-
Yeah--I always prefer to risk finger burns than risk my hotend brass parts and it comes apart so easily when it's hot!
If you disassemble while it is below 180C you may break it. Especially the brass nozzle - do not attempt to unscrew it without it being at least at 180C. It is very delicate. Well -- delicate if you are using a wrench. Very strong if using your fingers.
You could also cut this problematic area out of the model to reduce test print time, but the problem might go away from just doing that. (You can cut the model with the free to download netfabb.) To eliminate test print time altogether, load the gcode and preview the tool path in Cura (View mode >> Layers) or Repetier Host (Preview tab).
Useful Cura Trick (Sparse + Solid Infill)
in UltiMaker Cura
Posted
I print nylon and it's tough or impossible to make solid parts that don't shrink and warp too much, due to the high level of cooling shrinkage. Sparse infill reduces the problem greatly, but sometimes results in a part that's not as strong as I need (often the parts are too easily compressed under load).
Using Cura, I found that I could increase the density of fill all the way to 100% (solid) in areas that require more strength by designing-in thin hollow "cuts" internally in my solid model. In one model, I placed 0.002" parallel cuts 0.050" apart and Cura didn't "low pass" filter them out.
It obediently attempted to print all the tight parallel walls of the cuts and the result was perfect 100% infill, but only where I needed it (and I used 20% infill elsewhere).
I don't know how to do this with any other slicer and I think it's pretty it's a pretty slick and useful capability of Cura!