Calibrate for "widening" objects?

Hi,

so far I think my calibration is done. But I have still some trouble with objects, that get wider (or smaller) with the print. For example spheres, that start small and then later layers are attached to the "side" of the layer beneath.
I hope you know what I mean

These new layers seem to deform very fast and cooling is not helping somehow.
Is there a faq for calibrating these things?

thanks!

Get straight sides is a matter of getting the bed temperature and cooling right. It's easier with PLA than ABS.

---

[www.hydraraptor.blogspot.com]

what has bed temperature to do with objects that are being built into height?
you mean, I should cool more?

From what I understand, objects with a small contact area tend to have later layers printed upon them quickly. This results in the underlying layers not cooling sufficiently before the next layer is added. The layers stack in a semi-molten state and deform as a result.

I have tried various things like enabling the carriage fan at start of print to printing multiple copies of an object to increase the layer print time overall. I do not know whether reducing the bed temperature a few degrees would also help but that runs the risk of reducing adhesion if the bed is not scrupulously clean.

It seems we are both at the stage of wanting to advance our printing abilities and we are discovering the practicalities of the process.

Regards,
Neil Darlow

---

I try to write with consideration for all nationalities. Please let me know if something is unclear.
Printing with Mendel90 from fedora 25 using Cura, FreeCAD, MeshLab, OpenSCAD, Skeinforge and Slic3r tools.

neildarlow Wrote:
-------------------------------------------------------
> From what I understand, objects with a small
> contact area tend to have later layers printed
> upon them quickly. This results in the underlying
> layers not cooling sufficiently before the next
> layer is added. The layers stack in a semi-molten
> state and deform as a result.


I tried to find settings in slic3r to have those low-contact walls print more slowly but did not find anyting.
Hm, wouldn't it be possible to just let the extruder and fan "go round" each layer a second time just to cool things down?

muelli Wrote:
> I tried to find settings in slic3r to have those
> low-contact walls print more slowly but did not
> find anyting.

I have not explored slic3r yet but the only setting I can see is the Minimum print speed on the Filament Settings|Cooling tab.

> Hm, wouldn't it be possible to just let the
> extruder and fan "go round" each layer a second
> time just to cool things down?

Skeinforge has an Orbital cooling method which, I believe, performs something of that nature but I have read comments elsewhere that it does not work well.

If only the lower layers are deforming, it might be a case of lowering the bed temperature a little.

Regards,
Neil Darlow

---

I try to write with consideration for all nationalities. Please let me know if something is unclear.
Printing with Mendel90 from fedora 25 using Cura, FreeCAD, MeshLab, OpenSCAD, Skeinforge and Slic3r tools.

Quote
neildarlow
I have not explored slic3r yet but the only setting I can see is the Minimum print speed on the Filament Settings|Cooling tab.

Right tab, wrong setting

"Slow down if layer print time is below" is the setting for slowing down the layer if it is too small to allow it to cool, but the problem with this implementation is that ALL the moves are scaled down, so if your layer is small anyway then the hot nozzle just hangs over the molten plastic for longer, making the problem worse. The best approach is to just load up the bed with more items so that the nozzle is spending time away from your object and giving it time to cool before coming back for the next layer.

This is a common miscomprehension. The heat carried into the object by the flow of molten plastic far exceeds that radiated from the tip so slowing down does help, not make it worse.

---

[www.hydraraptor.blogspot.com]

Not disputing you Chris, but I'm intrigued to learn more!

If the temperature of the plastic coming out of the nozzle is at a fixed temperature, and the time taken to lay that same layer of plastic is increased, why doesn't that translate to having more heat being applied to the object? I know it's the same volume of plastic, but isn't that plastic being heated for longer?

When I said that the problem gets worse, what I had observed was that the top layer appears to stay "liquid" for a lot longer if my object was tiny, to the point where on some PLA prints (pyramids are a good example!) I'd see what appeared to be a large liquid ball of plastic forming near the point/cap.
Slowing down the print didn't provide noticeable improvement, but if I put more of those small objects onto the bed - just enough so that no layer is less than the 15 second threshold, then the problem doesn't occur at all.

Yes the amount of heat going into the object in the form of hot filament is constant but when you slow down the rate it goes in is less and heat is lost from the object at more or less constant rate, so its temperature drops.

It might not be possible to slow down enough to make a single pyramidal point with PLA but it certainly is with ABS. Also I am not sure how slow Marlin can go. I think there may be a minimum speed.

---

[www.hydraraptor.blogspot.com]

Ok I think I follow. I've only had blobs appearing on PLA, and only when printing solitary pointy objects, regardless of how slow I dial my printer.

In Slic3r there is a setting for minimum print speed (the setting that Neil refers to above in fact!) - the default is 10mm/s but now I think about it I can't remember if I ever noticed that kick in.

Certainly on those pointy pyramids it would make sense (to me anyway!) that if my layer only takes a couple of seconds to print even at the minimum speed, then it would wait out the remaining layer time before starting the next layer. I guess you't then get ooze so it's not a perfect solution either...

The other important factor to remember is that our toroidal fan ducts deliberately don't point at the nozzle. Printing the tip of a pyramid is hard because eventually the travel distance gets so small, the cooling air never actually passes over the apex.
This is the other reason why printing multiple copies helps a lot. Not only does it slow the printing down, but also it forces the cooling fan to pass over the object.
If you don't want multiple copies of the object, then maybe investigate the Prime Pillar option under the Support setting in Kisslicer.