Preview layers shifted up or missing

[mechaMecha]

I'm trying to figure out a way to combine thick layers with thinner top layers to get smoother top surfaces, so I modeled a test object consisting of three sections:



The object is 15mm tall; each of the three sections are 5mm high. I used 0.4mm layers for the entire print except on the top surface of each section where I used a single 0.2mm layer. Each 5mm section should be made up of twelve 0.4mm layers (totaling 4.8mm) and one 0.2mm layer, and that's the way I created my processes. However, as you can see from the screen shot in "preview mode," the first and third sections look as expected, but the second section does not. The most obvious problem: there's a gap where the first and second sections are supposed to meet. If you look closely, the 0.4mm layers appear to be shifted up and smashed into the top 0.2mm surface of that section. Is this a rendering glitch? A bug in the g-code?



https://dl.dropboxusercontent.com/u/600679/Test.factory

[dkightley]

To save me repeating something written in another thread......take a look here: viewtopic.php?f=8&t=3802

Essentially, you need to change profiles at a distance equal to an exact number of layers.....and only full layers will be printed - the last layer will not be thinner!

[mechaMecha]

To save me repeating something written in another thread......take a look here: viewtopic.php?f=8&t=3802

Essentially, you need to change profiles at a distance equal to an exact number of layers.....and only full layers will be printed - the last layer will not be thinner!

Thanks for the reply, Doug. I must not be understanding something, though, because I believe I'm doing exactly what you said.

If you look at my factory file, you can see that I have six processes in total, with 2 processes making up each 5mm section in the 15mm tall print. Each 5mm section is made up of twelve 0.4mm layers (totaling 4.8mm) and one 0.2mm top layer (4.8 + 0.2 = 5mm). Here's what they look like:

0.4mm 1: Start box unchecked, stops at 4.8mm (first layer at 100%). Should produce exactly twelve 0.4mm layers.
0.2mm 1: Starts at 4.8mm, stops at 5.0mm. Should produce one 0.2mm layer, completing the first 5mm section.

0.4mm 2: Starts at 5.0mm, stops at 9.8mm. Should produce exactly twelve 0.4mm layers.
0.2mm 2: Starts at 9.8mm, stops at 10.0mm. Should produce one 0.2mm layer, completing the second 5mm section.

0.4mm 3: Starts at 10.0mm, stops at 14.8mm. Should produce exactly twelve 0.4mm layers.
0.2mm 3: Starts at 14.8mm, stop box unchecked. Should produce one 0.2mm layer, completing the third 5mm section.

All of the processes seem to be nice and tidy, starting and stopping exactly where they should and corresponding precisely to the layer heights used. The top and bottom sections look fine, but the middle section is always messed up no matter what I do.

[CompoundCarl]

Use 2 of the thin 0.2mm layers and then it will work fine.

If you use only 1, then the sections that are using the 0.4mm thick layers won't line up cleanly. To illustrate, imagine that you have three 0.4mm thick layers, except you remove the middle one and want to print a 0.2mm thick layer there. If you only print one 0.2mm layer, then there will be a 0.2mm gap right above it until you get to the third 0.4mm thick layer. If you really wanted to work around this, you could just use a few strategic gcode z-axis offsets to move the layers around until they line up. But it's much easier to just use 2 of the thin layers.

[mechaMecha]

Thanks, Carl. I'll give that a shot.

So, if I were using 0.1mm layers at the top of each section, I'd have to include 4 of them?

[CompoundCarl]

Yea, keeping everything in even multiples is the easiest way. You can always manually adjust the z-offsets for the different processes if you don't want to do that.

[mechaMecha]

Thanks for your help!