Creating hollow parts and semi-hollow parts

I would love to see a feature that can shell a part (option of shell thickness), automatically generate an internal support structure (option for density and possibly internal structure geometry), then finally allow for semi automatic placement of resin exit holes and air entry holes (automatically oriented and sized for optimal part quality). This would make the printer more affordable for some people who don’t need fully solid parts.


I was going to make a new topic on this matter but since it’s already been brought up the i’ll bump this up.

It would be highly beneficial if we could get this kind of feature within PreForm. Having to work through meshmixer and then to PreForm really slows done the workflow. Being able to optimize the way prints are hollowed per resin type for best printing compatibility within PreForm will allow for more time printing and making and less time fighting with Meshmixer to get the desired results.

Additionally, if it were within PreForm it could better suggest vent holes based on the print orientation for best draining of resin. This would help prevent resin waste and speed up clean up.

I’d love to see something like this.


Shouldn’t have a need to print solid parts except in rare cases such as a prism or lens or something that requires volume. Many software packages do have offset surface function and when not possible Meshmixer does work but I find it cumbersome. It is free and does work so can’t gripe too much.

I don’t see why down the road Preform couldn’t have that function. It would be handy to have and save hop scotching applications to get a job done.

1 Like

I was also about to request this same feature in a new thread. It would be really nice if PreForm could allow the user to set a simple wall thickness that could default somewhere in the 3mm range. Then have it automatically place vent holes that could be edited just like the Supports.

Thanks in advance Formlabs.

This is something we’re considering but we’re also wary of inflating the capabilities of PreForm and raising the learning curve for new users. In many cases, it makes sense to modularize tasks like this; containing mesh modifications in programs like Meshmixer and print setup in PreForm. Our strategy thus far has been to educate users on the various tools that can be used for mesh modification and make PreForm the best it can be for print setup. We have a tutorial on hollowing parts in Meshmixer if you’re interested.

1 Like

Nice tutorial but its been 3 years and meanwhile Prusa come up with this option in their slicer :stuck_out_tongue:

come on folks, this would be so great to have in Preform, let’s go!! chop chop!



The last comment was in 2020? Formlabs, what is going on? Any progress?

Every slicer that supports SLA printing has this function, except PreForm. The old excuse that “we don’t want to bloat our software, or we don’t want to increase the learning curve”, simply doesn’t cut it anymore.

This is not different that the excuses we got years ago when we asked for the ability to mirror a part. It only took 3 years for that to materialize.

I understand being a little behind the times, but then again, maybe their programmers are not as good as we think they should be… Just sayin’… [hint, hint]


What are you folks printing that would benefit from this?

I mostly just print my own designs, so it’s easy to hollow them out appropriately, during the design process, if I need to.

To me this option would appear to primarily appeal to people that are printing designs made by others, where it might not be quite so simple to hollow out the interior, if all you have is an .stl file and not the underlying data used to generate the .stl.

I print both my own designs and STLs obtained from other sources.

I use Solidworks, and hollowing parts out using the “Shell” command sometimes works simply and effectively, but other times, especially on more complicated geometries, with parts that were joined (Booleaned) together can be a never ending cause of frustration.

I also buy models from places like CG trader or downloaded them from other sources like Thingiverse. On those models all you get is the STL or an OBJ file. Most of those models, depending on the size you print them at will require hollowing, and that’s when programs like Meshmixer come in.

But then again, Meshmixer would require the model be imported then hollowed out, holes placed strategically, then exporting the model back out for printing. Another issue is the model’s complexity and size (number of polys). Meshmixer might take a loooong time to import the model, and manipulate it.

I know that on many occasions I got models that were 5M to 10M polygons, and opening those in Meshmixer and working with them is a real pain.

On the other hand, being able to import the model in Preform, and hollowing it out right there without the import, export, fixing, etc would go a long way to help streamline the work flow.

I use ChiTu Box to import my models, arrange and rotate them as needed. I then hollow them out, ad the drain holes, and finally generate the support. All in the same app, and it’s 10+ times faster than anything I’ve been able to do externally.

So, if Chitu software developers were able to do this, why can’t the FormLabs team do the same?

1 Like

Fusion 360 had its Mesh area completely redesigned a few months back and now hollowing an STL / OBJ is not only simple but also way faster than their long abandoned meshmixer. Holes still need to be manually placed (AFAIK) but the hollowing is now easy and fast.

You can then either simply export the hollowed mesh or convert it to a Fusion 360 body and make all necessary changes before exporting back to STL / OBJ.

No idea if they’ve used and improved the same engine they had on meshmixer but it’s definitely highly optimized now, almost instant on a MacPro.

I’m going to agree with other people here that if Chitubox can hollow a model and add drain holes, there’s no excuse or rational reason why Preform can’t. I just spent 10 minutes looking for it in Preform only to find out that it doesn’t exist via this thread.

1 Like

It’s like Apple: “we (think we) know what it’s best for you and we are going to give you just that”

I’m concerned about adding new features.

Preform has problems as it is; complex stl’s don’t rotate freely, and editing and placing additional supports is harder than it should be. Especially since the support lattice dissappears, leaving only the faint dots to guide you.

The slow rotate is sometimes fixed by rebooting the program, but not always.

Since they also advise rebootiing the printers if printer times are inexplicably long, I suspect their software teams need a lesson on “How to avoid memory leaks in your code, 101”.

I’d love to know how big the team is for PF, and who’s in charge. They seem to be adding GUI features, while neglecting problematic core functionality, which usually involves a deeper dive.

They need to be closer to the users and the UX. What use is priming your printer and remote printing, if the printer is going to throw a warning that you’re at 75% of your tank life, which then needs you to go to the printer to cancel that message?

Even more importantly, i find the lattice is way too complex, and needlessly conservative, even at the lowest settings. I think they may be trying to compensate for the deterioration of the tray by having more supports. Regardless, Ii usually set the density, slope, and attachments at the minimum allowed. Which means, to me, there’s a lot of work to be done

I shell and add drainage in Zbrush, mo.stly, which works pretty well, although if I need a completely smooth shell interior for a mold, that can get complicated.

I’d love a complete package, but shelling is less important to me unless it works a lot better than PF does as it exits. I’d rather some other things got fixed first.

To me it seems that they want to me more focused on the needs or professionals, rather than the individual hobbist or DIY. I had asked a similar feature in the part (to be able to move a part below the print level so as to print only a part of the model - like I can do in Cura), but more or less the answer was to split the file on the CAD software and then import the file in preform.
They seem to not want the user to be able to do many things with the software, not make his own adjustments, in order to deliver a good reputation of print consistency.
I think about a dentist for example, apparently that guy wants to press “print” and get the model asap without worrying about settings.
You have to chose between a closed platform with consistent results, or an open system (another 3d printer) which requires user effort for having good prints.

the above being said, I tend to prefer formlabs’ approach on things because I can put my focus and energy on creating 3D designs, rather than on spending my time to make the printer work as supposed to. (I had used a photocentric in the past and I was spending most of my time troubleshooting the printer)

It would really be interesting to take a census and see how many users here have had this “just press print and it works” experience.

Yes, I understand that by their very nature forums tend to be where users with problems find themselves, but if forum feedback is any indication, most users here do not have such a simple “push play” experience.

Some, after some trial and error do manage to get things straightened out, but sooner or later, they have to deal with some issue or another. The ubiquitous “HP printer experience” is simply not there yet.

I assume the press-to-print cow is producing more milk. I’ve seen a similar track on another company (not for 3d printers), first it was all about the average consumer with support forums and open systems, but after 4-5 years they started focusing more on professional customers and they even closed the user forum. You had support only via support tickets, forum was gone.

You’re kidding, right? :sunglasses:

There’s a continuum.

From fairly easy objects on one of the standard resins which approach the press to play level, to complex objects needed accuracy and fine finishes with one of the harder engineering resins.

This can involve almost daily contact with your assiged support engineer, who becomes familiar with your work. And then some of the higher level engineers, using Pro level support.

If you need it, they will go into a deep dive with you, if necessary, Very far from “press play”, and rightly so, since you’re demanding things at the very edge of what’s possible.

Underlying this, there’s clearly a lot of room for improvement with their QC. I think it’s a cultural problem, and thus hard to solve. It’s not a priority, which is a puzzlement.

For example, even after they got a huge infusion of cash, the Form 2’s were assembled in an ordinary environment, instead of a clean room, or under a laminar flow station. As a result, these machines left the factory with dust on the mirrors.

This ignores basic Demming rules, which have been around for what, more than half a century?

My Press Play comment was in response to agiorgitis comment that Formlabs has “dumbed down” the system in order to appeal to those customers that do not want to be bothered with changing or figuring out setting, who simply want to load a model and press Print.

Unfortunately that hasn’t worked all that well either, because more often than not, just clicking print doesn’t produce a functional or satisfactory part.

At the other end of the spectrum, there are those that would like to have control of the print settings in order to control the outcome.

So why can’t Formlabs have a Beginner, Intermediate, Advanced selection, like other programs have? If you’re a beginner, or simply don’t want to be bothered with many settings, select that profile and print away. As you become more familiar with the program and want to have more control you can unlock more features by selecting the Intermediate or Advanced profiles.

I would think that would satisfy most users here.

As for having “almost daily contact with your assiged support engineer”, I hope that was a joke. I can’t imagine having to have my hand held through every print I make, nor can I imagine a support engineer having to field daily calls from the same users who can’t seem to figure out what to do. I’d tell him to put the damn thing back in the box, and send it back.

I posted this little YouTube video I did of the hollow out process in ChituBox in another thread dealing with hollowing, but I thought I’d post it here too. Check it out.