Variable Layer Height prints awful

I’ve been working to improve the top surface quality of my prints by applying variable layer height, but the results have been disappointing. Please take a look at the attached pictures for reference. Additionally, I’m occasionally seeing a burnt layer during the print. I’ve tried adjusting the top layer speed, but the outcome remains the same. I’ve also included the 3MF file at the end of the post for further insight. What can I twitch to have a better result?
I’m printing using a P1S and Bambu Labs PLA.



head2.3mf (451.0 KB)

Welcome to the forum.

The problem is that you don’t have any infill.

Suggest switching to an infill like Gyroid with 20 percent infill.

1 Like

Thanks, is there no way to make it work with Lightning? I need these parts printed quickly, and using Gyroid triples the print time
What about the “burnt” layer, it doesn’t happens all the time but how can I avoid it?

Lightning support isn’t structural, it is designed to be used mostly with more standard-shaped models where it can quickly come out of the side to hold up a smaller area top.

You have a lot of weight being held up with very little support, nothing structural.

You could always play with Height range modifiers where you can choose different support or values of support for different height ranges.

If your model was taller, it might have a better chance of working with lightening support as it would have a greater volume of branches underneath.

Right now, you are simply asking too much of something that is not fit for your specific use case.

1 Like

Welcome to the community!

First, thanks for posting clear pictures and uploading the 3MF file. This removes a lot of the guesswork from helping you.

Here are a few things I noticed between your photos and the 3MF file:

As uploaded, you may have noticed the message that you can’t use organic supports with variable layer height. However, your model doesn’t require supports in its current state, so you should be fine to disable that.

Here’s some suggestions, and I’ve re-uploaded the modified 3MF file:
head2 – modified.3mf (456.9 KB)

  1. Disable supports altogether, and instead, increase your wall count to 3. For angles this shallow, that should be enough.

  2. Experiment with setting the sparse infill density to 1% using rectilinear. Combined with the increased wall thickness, this should be enough, but you may need to try increasing the density to 2 or 3%. This should provide sufficient supports internal to the structure while saving internal nozzle movements and filament thus speed up the print while not sacrificing quality.

  3. You’re using the layer height tool incorrectly. Here’s what I mean: you set the entire model at a 0.20 layer height. However, the way you used the variable layer height tool essentially tries to smooth the whole model. This resulted in:

At the top, a layer height of 0.08 and at the bottom, a layer height of 0.14. This completely overrides all the other layer height settings so even if you set the layer height to 0.28 for the rest of the model, you aren’t getting any benefit. Please rmember, layer height is supposed to only be applied where you have refinements in curves or details, not so much on vertical walls.


Here’s what to do instead:

If you want to pick up speed, set the model to a 0.28 layer height and only use the variable layer height for the top of the curve. You may need to use the smoothing function to make the transition gradual.


_______________________________________________________

While I wasn’t able to match the exact time your original file showed, I was able to get it to 1h:18m using these methods.

_______________________________________________________

Other Possible Suggestions

After making the changes above, here are some other things to consider.

I’m noticing what appears to be a clear demarcation in layer changes. This might simply be due to the angle, but it could also be caused by incorrect temperatures or improper filament tuning.

Two factors can influence this: one is speed. Let’s not dive into using modifiers to correct speed here, as that’s a more advanced topic and will require trial and error. It’s typically only worth the effort if you plan on continually printing a model and want a fully dialed-in print.


On a more practical note, try running an overhang test from Printables. A temp tower using Orca Slicer would be better, but I don’t want to overwhelm this post with too many details. An overhang test is easier to interpret if you’re new.

Select a simple model from the list and print test models at different temperatures. At first, use wide variations to figure out if you should go higher or lower, then dial in from there. For generic PLA, start at 220°C and test at 190°C and 240°C to find the extremes.

Post processing

Sometimes you can only go so far. Then it comes to sanding and painting. Search for posts here on the optically clear topic last year for some tips if you’re really interested.

These suggestions above BTW are from a similar experience I had and and a lot of wasted filament. If any of these suggestions save you you some time, I will have done my good deed for the day. :fleur_de_lis:

I had a project last year that I did that was the skull from hell.

Final print was 22 hours. 90+ hours over 2 months and 3 spools of filament figuring out how. :wink:This was a lot of trial end error.

This was a gift to my daughter-in-law upon her entrance into med school, her reaction was priceless. In the end, the results were worth it!!!

Here’s another one that is more similar to yours. In order to achieve the final smooth surface, I had to resort to post processing(sanding and painting) to achieve the smooth “resin-printer-like” finish. I’m afraid that sometimes, there are simply no shortcuts. :slightly_frowning_face:

6 Likes

wow thank you so much for such a level of details in your responses guys, especially @Olias I learnt a lot with the insight you guys just gave me.

@JonRaymond , your solution works, Gyroid with 15% percent infill was enough, probably could go a little lower but I didn’t make another test. The only problem is that increases the printing time x3.
@MalcTheOracle Thanks for the explanation, now I know how that part works.
@Olias thanks for the detailed answer and all the screenshots and links, it was very helpful. I tried 1% rectilinear, wall count 3 but it wasn’t enough. What it worked is going up to 7-8%. I printed multiple ones and it worked by trial and error. I also tried changing the speed but then the color of the print would change in that section.

1 Like