Customizable Z resolution

now that we have different resolutions in supports and model , its possible have different resolutions in the model?
just like slash 3d printer

https://www.kickstarter.com/projects/644653534/slash-the-next-level-of-affordable-professional-3d/description

sorry for my english!! :wink:

This is a feature we recently released in PreForm 2.6 and 2.7. The raft and support structures are printed at a larger layer resolution to speed up prints. Here’s a link to the release notes.

EDIT: I now notice that you mean different resolutions throughout the model rather than on rafts and support structures specifically. This isn’t something we have planned for the near future but it’s an interesting suggestion!

1 Like

This seems limited in nature, you would only see significant speed increases if entire Z layers can have their resolution reduced. If you have a model with some high res features on one edge while the rest of the slice can be 2-4x thick you are only saving the time to cure the resin, you still need exactly as many peel cycles as you did before.

So, in it’s current form you save a lot of time with the low resolution raft, but only very minor time savings on printing supports every other layer because you still need to peel every layer. @Frew Out of curiosity do you have any test data of identical prints before and after the change?

Given my first point I think this feature would encourage inexperienced users to use poor model orientation to try to save time while the veteran users would likely ignore the feature except in obvious use cases.

TL;DR - Interesting, but I’m not sold.

1 Like

You’re correct that most of the times savings will be with the base layers and the time savings with supports are pretty negligible. I don’t think we did any tests with a side by side between the updates because they should be exactly the same but if you’d like, I could give it a go.

I don’t think the update necessarily gives new incentives to use more dubious model orientations to reduce printing time. Getting a bit crafty with orientation (for better or worse) always had the capability to reduce print times and this isn’t exclusive to the updates. The update will allow for holistically lesser print times but the magnitude of the effect of creatively orienting will remain the same.

1 Like

OK, then, how about restoring 0.2mm layers for FL Clear resin, like on the Form1?

1 Like

Definitely, if you find yourself with time on your hands. Quality should be identical yes, I was curious about time savings. Start identical prints on pre and post update printers (doesn’t even have to be done at the same time), and provide times for raft completion and model completion. We should expect the new raft to print in roughly half the time and ??? decrease for the remaining model. You can get a good overall time savings via the PreForm estimation but that doesn’t answer how much of the time savings is from the base and how much you actually save with the supports during the actual model.

Not really all that useful, but would be cool to know.

I agree that the 1/2 res support update doesn’t change behavior at all, I was thinking more of the varied supports in the actual model as mentioned by the OP. Think of a shield with intricate details on the front. The surface details could be designated as high res and the rest of the shield as low res. Now when you look at various orientations of this object you will see a huge decrease in print time for the specific orientations which allow the low res layers to be printed alone but this is arguably not an ideal orientation for the object. My apologies for the lack of diagrams.

Could you explain more what you mean by “you are only saving the time to cure the resin, you still need exactly as many peel cycles as you did before.”

My understanding was that there is a cure and peel with each layer that is printed. However, is there no peel cycle unless there is a certain amount of layer area (such as the base where there is a lot of area vs. supports very little area)? That would definitely explain why the base takes 50% of my print times typically, but for some reason I have never seen this explained anywhere.

For a part that I typically make (different for everyone of course) I have the following on preform estimation with Clear V2, Form 2, print has a 2.8mL volume.

Res - print time
.1mm - 1:18 (315 layers)
.05mm - 1:48 (545 layers)
.025mm - 2:49 (1006 layers)

Since the base and supports are the same for all three with new firmware, that means it is taking an hour longer to make my actual part when I go from .05 to .025 (right?). If i only wanted 100 of the .05 layers to be .025 resolution, I would be adding 100 layers to my print, so 645 total.

If you assume it is a linear relationship between layers and time, 645 layers would be a 2:00 print time roughly, savings of 49 minutes over full .025mm resolution. Savings would be even greater if you do .1 for most and .025 for a portion. Savings depends upon how much you want higher resolution too.

Most stuff I make I have small zones where I want higher resolution but don’t want those print times or the wear on the machine. It would definitely benefit my printing, can’t speak for others.

Hey, that’s a quick and easy example of time savings. Previously doubling resolution doubled layers and therefore print time. I like your logic of time/resolution trade off, and proposal of selecting specific layers, in their entirety, to print at higher resolution.

As I understand it (I welcome better explanations) the base layers take so much time because of two reasons. They are cured more extensively than model/support layers which leads to a slower cure pass and potentially multiple cure passes, and there is an amount of compression involved with the starting layers to ensure good adhesion to the build plate. .1mm is not truly .1mm in the base (not sure exactly how many layers are compressed). At the risk of stating the obvious this is because the base needs to stick to the platform and needs to support the forces of the peel cycle for the entire print, but it doesn’t need to be pretty so it can risk being over-cured and dimensionally inaccurate.

With the 1/2 resolution areas being discussed layer heights are 2x the nominal resolution. In the current implementation you only see this when your slices have supports and model geometry in them at the same time. Those 1/2 res. areas (supports) are only cured every other layer but each layer still requires a full peel cycle due to the higher resolution of the model. So back to my original statement, you are only saving <1/2 the time it takes to cure the supports, no peel time is saved. What you are suggesting (full high res layers) would work fine. My point was that time savings on mixed resolution layers (a section of the model that has high resolution on the right and low resolution on the left for example) would be minimal ie slightly less than <1/2 the cure time of the low res areas with no peel cycle time savings.

I hope that helped, I am utterly useless without napkin sketches or whiteboards. =]

1 Like

Ah yes, that makes sense now. It is only low resolution when it is PURELY supports. If the layer has your model and supports in it, it will be your model resolution. Thanks.

IMPORTANT:

A customizable Z resolution - means different resolutions throughout the model is necessary for professional users like dental technicians.
So if you really would like to sell your printer for dental parts, please read this:

Here are two samples why it’s important to use different layer heights.

First picture is a dental model, height 24mm.
For good quality you need the 50ym resolution in Z for 14mm (grey area)
The green area can be printed at 100ym in Z for same result.
So the layers in different resolutions are:
50ym: 469 layers
50/100ym: 369 layers!!!

Second picture is a more extreme sample - the modeldie, height is 21,2mm, but quality area is only 5mm! For best result at the sharp edges, 25ym in z is recommendable.
25ym: 848 layers
25/100ym: 362 layers!!!
only 50ym: 424 layers (not the best choice)

You see, with different layer hights it’s possible to print objects in highest quality much more faster - in dependence of the model highest resolution could be faster than middle-quality and in some cases you can print in less than half time with same results.

3 Likes

Great example! Makes perfect sense with that application.

1 Like

Hi there!
Thank you for your valuable insights.
Be sure, that your feedback finds our technical development.
We will definitely have a look at all feedback and evaluate further steps.
Best,
Florian

2 Likes

Hi DentalUser, great advice for higher speed production, only slight issue IMO would be the need for two print runs ?
One for the model and one for the preps ?
Would it not be easier/quicker to raise the main model to a level where the plane between high and medium/low resolution is the same ?
We are having issues with the removable preps being loose in the sockets on the models, I’m thinking its a settings adjustment in our 3Shape program, do you have any experience with this ?
cheers
M

Hi Ambo,
you’re right, if you print more parts at same time, the plane between different resolutions is at the same hight.
Note: If Formlabs will implement this feature, the different resolutions are always for the whole built-area, not single objects.
But it is possible to optimize this in z with supports.
My pictures are only samples.
The parameters for a perfekt Geller-model are tricky, you have to optimize them for your printer. I’m using exocad, so i don’t know how to do this in 3shape, sorry