Diamond ghost patterns under top layer when ironing

I keep getting ghost diamond patterns visible in my top layer when ironing and cant figure out why. I have tried everything but cant seem to shake them. I am trying to print flat bases to support art we make - the bases are 105x105x8 with a smooth finish and these underlying artifacts keep messing them up.



The finish is amazing and super smooth, from a distance they look perfect however up close the diamond pattern becomes very visible. I have 2 x X1C and it happens on both of them - i have just swapped out both hotends and extruders and done full calibrations on both along with full service (clean rods / tension belts etc).

I narrowed things down to the internal bridge support layer that is generated between the infill and the top shell layers inside the model causing issues that are compounded up to the top layer, however at 100% infill no bridge is required so the top layer is clean of defects but the patterns still persist.

I am printing at 0.08 layer height using Arachne with Bambu Matte Black PLA that is dried and am using Bambu Slicer 1.9 final release

Quality:
Ironing on top surfaces at 30mm speed and 40% flow which gives me a great finish

Strength of the base is 3 walls

Strength - Shells
I am using defaults here so monotonic line for top surface, monotonic bottom surface and Rectilinear internal solid infill pattern

Strength - Sparse infill
100% infill with Rectilinear pattern

Strength - Advanced
Infill combination is on and Ensure vertical shell thickness is off

Any suggestions would be amazing !

Here is the link to the file… if anyone wants to take a look

Thanks

My ironing has been a nightmare lately so I might use this as a test print and see how mine fares.

EDIT: what have you added the pause for btw? layer 96 I think it is

sorry about the pause you can remove it - that is the layer after the internal bridge is laid down after the solid infill and before the top shells when you have less than 100% infill. it lets you see if the bridge sticks to the layers underneath when printing - the bridge gave me no end of pain so decided 100% infill was worth the time lost on bad prints from pillowing etc.

there doesn’t seem to be a way to disable the internal bridge when using high infill…

Aside from the triangles, it’s a beautiful looking print.

You’ve correctly diagnosed the problem as related to infill. Rectilinear is a poor choice for this type of print. Gyroid my help or perhaps even these three:

image

Lightning is probably best for creating random-like patterns but either way, you would not be addressing the underlying problem.

The issue that you correctly diagnosed is due to material shrinkage occurring and therefore you’re getting bleed-through of the underlying pattern. Have you tried to increase the bottom or top layer thickness? I would double or triple it. That will provide a sturdier surface to combat the shrinkage that occurring underneath the compromised layer.

What about not using 100% infill and instead pausing before the top layer and filling it with sand, then resuming.

Sorry nothing to do with the ironing issue but it would be a cool experiment :slight_smile:

actually tried this with our other bases that these stick into and its a nightmare as we need to do lots of them… tried sand / clay / plaster and turned out to just be easier and cleaner to print with 55% infill to get the weight we needed.

1 Like

thanks ! ok good to understand what’s happening here with the shrinkage. Will add more top layers and see, currently its 9 + the top ironing with the defaults, will go up to14 and see.

Tradeoff of more top layers is any defects in the top layers then also get compounded upwards… looking for perfection here !

Do i need more bottom layers as well - as that will just ramp up the print times and my infill is already at 100%… arent the bottom layers at the bottom of the print so under the infill - trying to understand how that will help !

This is what the larger bases look like that these plug into - no issues here with the ironing around the edges (we dont iron the inner top). We print the name tags and glue them on with gorilla glue with a mold that the base plugs into, tried different ways of doing this and this was best finish and fastest solution.

1 Like

Actually more top layers will not detract from quality. In fact, one trick I sometimes use is to create a solid but instead of creating solid walls by having 200+ wall layers, I may opt for top and bottom 200+ with walls remaining at the 2-wall support. Try it out with a 10mm primitive if you want to test results.

Here’s what that might look like with the layers exposed during preview. Note I have outer walls view disabled in the preview for more visibility.

image


Solid Top Layers

Solid Walls

Note how there are no gaps in the way the filament is laid down. Now please understand, you may give up outer wall smoothness using this technique depending on the dryness and properties of your filament. Wet filament, even wet PLA is less forgiving to smooth outer walls. To compensate, Use the outer/inner function as opposed to the default inner outer function.

The other thing you can experiment with is infill pattern. Here you can even go a little dangerous and us a less than solid print. The underlying pattern will still be there but your eyes will tend not to perceive it. I kind of like the Hilbert curve for this purpose.

BTW: Try this on a smooth plate with multicolor silk filament and you will get some very cool affects on the side facing the plate.

1 Like

got it - will try a print with shell only no infill and see how that comes out. I tried the hilbert but ironing goes crazy on top of it and we want smooth matte surfaces to match our bases - however i really like what you have it here with the hilbert and multicolor silk it would be a great idea for a different style of base to go with the matte ones we have…

actually got it - hilbert for infill not top layer :slight_smile: will try that as well, also going to try the support cubic and see - at maybe 50% infill for the small bases ?

trying to get away from the bridging. I saw online a option where you could set bridging to 0 and disable it when having lots of shell lines but i cant seem to find it anywhere in the slicer. its called:

Internal bridge support thickness

cant see any way to get this…,.

Well… if you like that, let me offer up a tip so that you don’t waste filament. You may have noticed that there are two kinds of multicolor filament. The first kind is where they vary the color every 10 meters of filament. I found those to be very prone to clogging because the manufacturing process is obviously a ‘welding’ of two filaments and at the weld joints is where the problem occurs. It also doesn’t produce the cool pattern above.
Here’s an example of bad multicolor


It was so bad, I sent back a partial spool and got all of my money back. Later, the seller tried to bribe me to take down my review.
https://www.amazon.com/dp/B09J4N793P/

The good stuff is tri-color where they actually extrude more than one color in the strand itself. It doesn’t have to be silk either, it can be anything just so long as you have multicolor in the filament strand.

https://www.amazon.com/dp/B0CF539T17/

1 Like

awesome !

any idea how to find “Internal bridge support thickness” ? want to see if turning this off will help with my infill…

I do not know the answer to that question and it may be something that you may find within the slicer but is not necessarily supported on a Bambu Printer. At least in Orca Slicer you might find it. I stopped using Bambu Studio a long time ago in favor of Orca.

However, here’s a tip. Did you know that there is a search feature baked into Orca Slicer? It may also be in Bambu.

Here’s were to find it. Go into the global settings and hit the magnifying glass.

Then start typing in the search box. All the terms will come up that are in the slicer. If you click on it, it will bring you to that sub menu and start blinking that choice.

1 Like

What shows up in studio…

awesome been trying to figure out how that works… moved form orca back to bambu with 1.9 as orca creates a lot of stringing in my prints for some reason whereas bambu gives me better cleaner walls and no stringing when printing slow outer walls.

still cant find this function however - its referenced a lot online in posts on reddit.

so now this makes no sense to me really… the infill is gyroid with 15 top layers / 7 bottom. The ironing is in straight lines and the last shell layer is at 45 degrees.

i still see diamonds in the ironing that look like they are at 35 degrees so they dont match up to either the shell pattern or infill pattern or shell infill either which is also at 45 degrees…

the only thing they seem to match is the angle of the bridge layer however they are very uniform and very neat… and that layer is now 15 layers down so how could it be showing through so neatly here

also dont seem to be able to change the angle of internal bridges just external so i cant rotate it and make it straight.

I’m getting a sense that you may not be using the full diagnostic tools at your disposal.

As an example, are you using the preview slider to review how your layers are being laid down? Your screen snip doesn’t show that.

The slider will reveal each layers underneath thus taking the guesswork out of what’s going on under the top layer.

Color scheme checkboxes can also be turned on and off to make it easier to view each kind of line type.

Here’s an example of what I mean when you disable both inner and outer wall. You can of course to the same thing with the top surface which should allow you to see what the pattern is directly beneath that surface which is affecting your print.

Did you also know that you can set the wall layers to 0 which will print the model without walls and allow you to actually see what’s physically happening beneath the surface. It’s the real world equivalent of unchecking the line type box in the preview mode.

2 layer top surface example

10 Layer top surface example

These techniques really take a lot of the guesswork out of diagnosing a filament issue versus a model issue versus a settings issue.

Also, there is something in your most recent image that gives me pause. What I see that was not present in the first photos are extrusion issues. I’m assuming you tuned the flow rate and pressure advance. Am I correct in that assumption? Because the output shown below doesn’t look like it. You shouldn’t have these kind of gaps. What’s more, you’re only having them on one side. Bed leveling might be warranted here a second time. But more importantly, in order to rule out belt or stepper motor issues here, it would behoove you to rotate the model 30degrees and see if the lines change location and orientation. And last but not least, I assume your door is closed and that no draft is blowing over the model.

And on the subject of Orca and Bambu Studio. If you are sharing the same profiles and 3MF files, there really isn’t a logical reason why you would be getting a different outcome. I say that because the underlying default core logic is identical. You may want to start from scratch and move forward with a newly tuned filament profile just to reset what the before and after would look like. It’s just a suggestion.

______________________________________________________

Cutting into sections

Also, I feel bad that I didn’t mention this before but have you used the cut tool to extract a section of your model in order to create a smaller model? The idea behind this is that by cutting up the model into sections, you can experiment with different settings while at the same time, saving both time and a whole lot of filament.

1 Like

yes i use sections when testing at the start but large flat surfaces then give different results when ironing so i fine tune with the master base at full size.

so removing the bridge with 100% solid infill removes the diamonds every time - no idea why they come through even with so many shell layers but go figure.

i had changed the extruder and hotend before doing these tests so doing a flow calibration now - was using 0.01 which looks good on the models.

EDIT: did the flow calibration and its suggesting 0.02 for my K factor so will try a reprint now

1 Like