Bottom layer lip

[Labern 773]

One of the issues with 3D printing is getting a nice bottom layer that sticks well onto the glass.

To do this you end up squashing the first layer into the glass which creates a sharp lip on the bottom of your prints.

It's then standard practice for a lot of designers like myself to put a small chamfer on the bottom of your part to remove this lip.

But a problem is when you find a nice design on youmagine (or other evil sharing sites ) and the designer did not put a chamfer.

To I ran some tests and found that if you set a negative horizontal expansion value on the first few layers then you can get a nice square edge on the bottom without this lip.

I had to load 2 designs and use per object settings to achieve this but it would be a good feature to implement into CURA.

It really needs 2 different horizontal inputs to get it perfect. One for the bottom layer and one for the second layer as it's also a bit wider than the rest of the print.

This is an upside-down photo (bad photo I know) of a print with the default settings and showing the sharp edge.

This is a photo with -0.3 horizontal expansion on the bottom layers.

[yellowshark 153]

Nice one; I need to give that a try. I rarely print 3rd party designs so if I know I need to really squish the part down, I automatically reduce the dims of the bottom .3mm of the model and then it comes out fine. Of course if I can use a brim then all that is not needed.

[DaHai8 74]

PrintBite - I never print without it.

[Labern 773]

This is not about adhesion methods, it's about removing the sharp lip on the bottom of your prints.

[yellowshark 153]

Actually I think it is. I think what @DaHai8 was saying was that if you use PrintBite you do not have to squish the filament down so hard and so the dimension of the lower layer will not be extended - ofc I may be wrong!!

Edited March 15, 2017 by Guest

[DaHai8 74]

That's correct. You no longer need to press the first layer into the bed if you use something like PrintBite. It naturally sticks without special effort and comes off just as easy after the bed cools.

[Labern 773]

That's true. But if it's not squished then you tend to also get very visible lines of the print path.

I actually have my own methods of initial layer sticking.

But this is mainly biased towards the UM3 using auto bed leveling and default settings, Even UM2 following the bed leveling procedure using the card and not having to use any of these products or methods.

You could solve this by one setting in CURA that users can adjust as needed.

[yellowshark 153]

That's true. But if it's not squished then you tend to also get very visible lines of the print path.

 

That is true too

[neotko 1,417]

 

That's true. But if it's not squished then you tend to also get very visible lines of the print path.

 

That is true too

 

Unless you use a lower extrusion for the first layer and a lower layer height. Squishing less material to get the same effect, adhesion and no deformed parts. I do it every day for my flat objects.

[yellowshark 153]

Interesting, I can understand reducing the flow to get lower extrusion, but not so clear on the lower height initial layer. If I reduce initial layer from .3 to .2 the nozzle to bed distance will be reduced by .1 by Cura and so the "pressure" caused by the nozzle to bed distance will still be there?

[neotko 1,417]

I suppose depends how you calibrate the printer, I do my 3 printers by eye, and adjusting the first layer extrusion on s3d got a nice middle point where I don't get deformation and keep a nice first layer.

The by eye is now, long ago I used a 0.08 feller gauge and set that as 0. Then the rest is done on s3d.

Edited March 15, 2017 by Guest

[yellowshark 153]

Hmn well I suppose one answer is that once I get the distance sweet I then never touch it, i.e. irrespective if which part I print I am always using the same distance. BUT if I do have to change the distance (there is one part that this applies to) that is done via the z-offset command and so that is stored in the gcode file. If I kept changing the physical distance then I would need to document the changes separately and manually otherwise when I reprint a part it will come out wrong.

[neotko 1,417]

Hmn well I suppose one answer is that once I get the distance sweet I then never touch it, i.e. irrespective if which part I print I am always using the same distance. BUT if I do have to change the distance (there is one part that this applies to) that is done via the z-offset command and so that is stored in the gcode file. If I kept changing the physical distance then I would need to document the changes separately and manually otherwise when I reprint a part it will come out wrong.

 

Well, you could just reasign a different first layer extrusion % for different first layer heights, that's what I do. For fast daily prints I go with a 140/140% first layer, this increases the amount of extrusion, and height, so I get a nice effect, using less material. For delicate stuff I do a 80-100% height /80% extrusion. Ofc the trick, to make this work, is not to use a 0.40 nozzle (setting) but using a 0.33-0.35. This forces the nozzle to go over the same area, or close to it, so the flat area of the nozzle flattens the area it just passed by. I do different % for different layers, 0.1 0.15 works better since 0.1 need's to print sooo slow to avoid problems (because I do this with greentec and adhesion is something to keep under control). For 0.15 is easier, since you get more mm3 for the first layer, so is easier to get a natural flattening from the nozzle pressure..

Dunno, when you get use to it, is just natural.

OFC sometimes I go lazy mode and notice too late my error so I need to cancel, clean restart. But you get the hang of it. That's also why is important to do the same calibration always, because this way to control the first layer just by the slicer and you can get repetition. The % all depends on your calibration, then you readjust.

[yellowshark 153]

Ooh some nice stuff there, very clever. Unfortunately us mere mortals without S3D and layer height procedures cannot do that, I think!