Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you: No adverts like this in the forums anymore. Times and dates in your local timezone. Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance. Email notifications for threads you want to watch closely. Being a part of the oldest wargaming community on the net. If you are already a member then feel free to login now.

So I got a failure that's rather interesting to me in that it seems to almost have failed and yet I've got a full rafter of supports just, no model!




Phrozen Sonic Mini 4K
Phrozen Aqua Grey 4K Resin

Settings:
Layer Height 0.02mm
Bottom Layer Count 8
Transition Layer Count 0
Transition Type Linear
Exposure Time 1.6s
Bottom Exposure Time 30s
Light off Delay 9s
Bottome Light off Delay 9s
Bottom Lift Distance 6mm
Lifting Distance 5mm
Bottom Lift Speed 60mm/min
Lifting Speed 50mm/min
Retract Speed 150mm/min

Anti-aliasing on
Grey Level 4
Image Blur off


So I will note I did think the number of supports in a few areas were too few (like that curved tendril in the foreground) and I did a few more. However not a great many and adding them was tricky since the pre-supported model file shows as a full model including the supports, so the new supports had to come from the "outside" a bit, which might well mean that they didn't help as much as I might have thought. I think the overall angle of the model is good, its more the supports that I think have let me down.

It appears to have failed in two parts, once on the foreground tendril that never formed its lower part, and then once again when it seems that a line from base to body formed and was, I suspect, too great a body all in one go without sufficient support, to hold onto it and it pulled away. Taking one formed tendril with it and then blocking the printer from casting right, resulting in the blob of resin that pooled and never let any more of the beast form. I would assume that going back to the unsupported model; changing the angle to the same as in this and then adding my own supports and focusing on the back and lower body to increase support there, would help; along with supporting the foreground tendril more.


I'd be interested in any here have any thoughts, if the presupport versoin show above shows you any clear errors/omissions or if you might suggest another reason for failure that I've not considered.

Those are afwully modest supports.

Also are you sure you're exposing enough? They look cured but your resin was designed to work at around 4 seconds...?

I did some testing yesterday (was going to post it up here but imgur was spending the evening being broken). The Ameralabs test image was showing generally good at 1.8s and some possible underexposure in some of the vertical segments. So for this I cut down to 1.6s. I did cast a deer head on the same plate as this, which I'd added more supports too after 2 failures from its pre-supported version before and that cast perfectly fine. So I think the 1.6s is doing well and seems to be similar to times that others with the 4K resin have.

Yeah, I'd put that down to too few supports and/or the supports being too thin. Once it hits the bulk of the model, the suction from the plate lifting pulls the model straight off the supports.

Yeah, those supports look definitely inadequate to me. Sometimes, pre-supported models aren't really done well enough, depending on who did the supports and if they even tested them (and even then, which printer, resin, settings, etc. were used is often unknown to us).

The angle on the body of that cat definitely is creating more suction force than I think those supports can handle. Adding some more would probably help. Is the pre-supported model only an STL, or do you have a .chitubox version available as well? If you can, use the later.

In addition to the exposure settings, the actual supports (diameter, etc, and especially contact depth) might need tweaking. Again, if that's just an STL, you won't really be able to change those, but I highly recommend going through some of Greg's videos, 3dPrintingPro on YouTube, for a lot of great info on supports.

The thing about supports on a resin printer, versus a FDM, is that they don't need to "hold up the weight of the model." They need to be able to hold against the pull-force of popping off the FEP, and also be able to hold the weight of the model against gravity trying to pull it away from the build plate. Tiny supports that might work for an FDM style print, where it doesn't have to keep the weight of the model from "floating up off the build platform" will often fail for resin printers, where it has to hold the whole thing as the build plate elevates.

Edited to add: I'm almost at the point now where I'd rather get unsupported models and do them myself than get poorly done supports, especially if that pre-support is only available as an STL so I can't inspect/edit the supports themselves.

I had this problem a lot when starting out, it was down to two things for me.

1- exposure times were too low. I'd done the exposure calibration prints and detail tests and such, ad was getting good prints from manually supported files but no luck with presupports from most places. On a whim I tried whacking up the exposure from 6.5s to 11s and suddenly everything was primting perfectly. I've since come down to 9.5s for .05 layers as the Elegoo resin works well with that, the NOVA one I started with seemed to need longer times. Which is a long-winded way of saying that you might want to try a bigger jump up, go too high and you'll lose some detail but get it right and you'll find a point where the supports hold up and the loft is good.

2- Lift Speed. I did change it in Chitubox but my Photon wasn't (and still isn't) reading it right and defaulting to a crazy high 3m/s, or 180mm/min. These days I just go into the setting on the printer and manually set it to 0.8mm/s as the print is starting, and again, it made a huge difference I was essentially printing on so-called 'vroom' settings and getting terrible results until I figured out the Photon wasn't reading the lift speed properly, so you might want to check that.


Presupports are only as good as the person who did them, so while you're dialling in settings, maybe try something like this. Artisan Guild's presupports are among the best in the business, so this freebie would be a good test where you know that if it fails, it's not an issue with the file or supports, but the settings.
https://www.thingiverse.com/thing:4756383

I think the obvious problem is that you are printing a displacer beast. The supports should all be positioned about 3” off to the side of where you think you are printing it.

Thanks guys, I'll give the freebie model linked above a try and also start to look at the greg videos mentioned about supports.
So far all the models I've picked up have come both with and without supports, so I've always got a "blank" model to work from without any supports to add my own too. Indeed its pretty standard it seems to supply non-supported as well as supported models (often as not the patreons tend to deliver the former before the latter anyway).

 Nevelon wrote:
I think the obvious problem is that you are printing a displacer beast. The supports should all be positioned about 3” off to the side of where you think you are printing it.

I did consider this as a possibility, however I'd hope that any displacers sold are at least half trained to not vanish before they are printed

When using the 'light' supprts, that is fine, but go ahead and manually add a few 'medium' supports at critical spots to help keep all in place and not break away - such as the lowest areas, some of the floating arms and the bulky part of the body.

Also, if printing at .02 you may find AA is not really needed. On this type of beast the details are fairly smooth, so no harm, but on a super detailed mini the AA can actually blur out some of the fine details you want to keep and is why you are printing at a low layer height to begin with.

I really dislike pre-supported files and always prefer to set my own supports.

Print peeled off at the tips of the supports. Looks like suction force (@FEP) > tensile strength (@support tips).

Increase exposure, use thicker supports, and/or slow lift speed.

You also need a lot more supports attaching to the underside of the print especially if you're printing .02mm. As you see on the picture of the mush, each layer is paper thin. It needs lot more 'substance' to latch onto in order to be stiff enough to receive the subsequent layers.

Displacer Beast TAMED!



Or at least captured!

Adding way more supports and it cast without any issues, though I've now learned a few more important things

1) Medium and Large supports are great, BUT unlike light supports, they tend to snap and leave a small intent into the model surface. Indeed my Displacer Beast looks a bit pock-marked to say the least! Now part of this I can address with perhaps using fewer and using medium and large to support long towers of supports which then break off into lots of smaller ones that attach to the model itself. Another way is to not be lazy and to use clippers, though considering how complex the model becomes with its own shape and with supports all over the place I'm thinking that might require some clippers even more fine than the normal hobby ones (at least to make life easier).
I can probably cover over many of the dimples with careful scraping and greenstuff application, but in the end I'd rather not so I've got to improve my use and removal from medium and larger supports.

2) Exit holes might look big on the computer, but can end up way smaller in reality which makes them into a nightmare of a tiny leaking point that seeps out wet resin and IPA which makes it take a lot longer to clean them up. Granted this is a smaller model than the deer I cast before which was hollowed out, so everything gets a bit smaller.

3) Microfibre clothes are not helpful when cleaning up models - far too liable to get fibres all over the place. I really need to get a couple of long bristled fairly large to smaller brushes to wipe them clean and to get into the tight corners.


Also this makes me all kinds of happy



This is just what I wanted from the 3D printer - fine detail even with insane enlargements (that's a full 1:1 macro shot). And yet I know it can do sharper and even finer edges if the file supports it so that pleases me greatly. Esp when considering that quite a few of these models are a wee bit larger than the normal GW or otherbrand scale models. Indeed I suspect I could scale it down by half and still have a good sized model and preserve the details.
Seeing this makes me all kinds of happy with the machine.

That looks pretty good overall, congrats Overread!

You might tweak the "contact depth" settings on the medium/heavy supports. If they're too deep, they're more likely to cause those pockmarks when breaking off. The other option is of course to not snap them off, but go more meticulously with clippers (sprue cutters work just fine).

I watched a video from 3dPrintingPro where he shared his support settings, and here's what he uses (probably crap formatting, but you should be able to figure it out)


Support Type Contact diameter Contact depth Upper diameter Lower diameter Connection length
Light 0.20 0.02 0.20 1.20 3.00
Medium 0.20 0.02 0.40 1.10 3.00
Heavy 0.60 0.30 0.60 1.20 3.00

I've done some prints with this, and they work great.

Edit: https://www.youtube.com/watch?v=AIFRpG5V5vQ&ab_channel=3DPrintingPro has an update I think that I haven't tried yet.

Glad you got it working! A fine print, and yes, the details you can get from these is mind-blowing when you start getting decent prints.

A few tips for supports (which are basically cribbed from the 3dPrintingPro videos, so you might have come across them already)

- Orientation is key, but if you get it right, most of the time you can print with any heavy ssh supports (the ones doing the actual work of keeping the model attached, versus light which generally just give islands something to attach to) on the very bottom areas, like soles of feed or,the underside of an arm. In other words, not visible from the usual viewing angles, and not damaging very visible areas like that rear thigh where you've got some heavier damage (though nothing putty or a dab of superglue won't mend!)

- As mentioned above, the 3dPP support settings work fantastically in Chitubox, I use them with almost no tweaking and very rarely get fails, even though they're a fair bit smaller than the Chitubox defaults. On a related note, you'd be amazed at what a few heavies can do. I was printing a werewolf from Titan Forge a while back and adding in maybe four heavy supports at the very bottom of the presupported file was the difference between a complete fail and a perfect print.

- A few seconds in warm water after washing and before curing makes support removal a much neater process, leaving less damage and if you've got the, set up right, just pulling away with no need for clippers. Just be sure the model is washed thoroughly first so you're not handling it with liquid resin still clinging to it, and you're golden.

- It's a matter of preference. but hollowing is sometimes more hassle than it's worth. Anything smaller than your average ogre or troll I don't tend to bother with hollowing out, especially if it's got spindly limbs like that Displacer Beast, as I like to keep drain holes to the undersides of the feet if I can. Sure, it might use a bit more resin, but on a lot of things it's only saving pennies worth, and that's offset by the extra hassle in prepping the files and post-processing for me.

Excellent!! You might find you do not need any heavy supports at all and only minimal medium, as the light supports do the bulk of the work.

Thanks guys!
Valander thanks for the link, I've seen a few videos by him and he's always clear and informative. One time I'll really have to go through his channel.

Interestingly in looking at his suggested changes it seems he goes for a thicker stem than chitubox but then mixes that with much thinner actual contact points. I can see the logic in that, the thicker stem is where you need the strength whilst the contact point is where you want the lighter hold and support and you make up for that with more connections if needed; trusting the thicker stem.

I'll certainly muck around and give the settings he suggests in that video a try; I really like the idea of thinner contact points but thicker stems and, with my pock-marked displacer, I can well see value in messing with the thickens and depth of contact points.

Yup, I've learned a lot watching his vids. Even the reviews are fairly informative, if not necessarily directly relevant if you didn't buy that machine.

And as much as he prints (he's got multiple printers running pretty much non-stop), I kinda trust his opinion there.

 Overread wrote:
Thanks guys!
Valander thanks for the link, I've seen a few videos by him and he's always clear and informative. One time I'll really have to go through his channel.

Interestingly in looking at his suggested changes it seems he goes for a thicker stem than chitubox but then mixes that with much thinner actual contact points. I can see the logic in that, the thicker stem is where you need the strength whilst the contact point is where you want the lighter hold and support and you make up for that with more connections if needed; trusting the thicker stem.

I'll certainly muck around and give the settings he suggests in that video a try; I really like the idea of thinner contact points but thicker stems and, with my pock-marked displacer, I can well see value in messing with the thickens and depth of contact points.

The only thing I disagree with him on after experience is his advice to not rotate the model 45 degrees is crap - at least in my experience on my printer.

NOTHING warps my models like trying to print any kind of flat section all at once, no matter how well I support it, and the light supports are so completely unobtrusive as to be nearly totally invisible. I've been rotating every single model such that the maximum number of surfaces are 45 degrees and they've been coming out 100% less warped than when I was taking 3dpp's advice and just printing the models standing straight up and supporting from the bottoms of the feet.

There just tends to be a lot of objects that end up large and flat on your typical model when you print them standing straight up - pouches, backpacks, gun stocks, sometimes arms/elbows, and all of those warp super badly if you try to print like that no matter how meticulously you try to structurally support them.

The prime position for me has been model tilted back and rotated, structurally supported typically off of one heel and one other feature and then detail supported with hyper-small light supports on every other feature.

I agree an angle to a flat area always works perfectly for me and also putting things on an angle will typically make the height less, so shorter print time.

Also, you need to look at your support depth as well as your connection point. Sometimes you can have the end connect point taper to a small point to minimize the pock mark, but your support depth is set so the support goes in too deep, so the end of the support touching the model is actually buried into the model and does not do what you want, which is a light touch on the surface.

I really do like the latest version of Chitubox, as when hitting the basic auto-support function will also add some very tiny supports in needed spots, which are much smaller than even the light supports - really great for super tiny overhang details where a normal support would be overkill. You can always add or subract whatever supports you need, so I have the auto setting % set fairly low, otherwise it may add so many supports it is ridiculous.

I think the "rotate 45 degrees" is a decent rule of thumb, but you really need to look at each model to determine its best orientation. Sometimes, yeah, that's going to be "straight up". You want to try to reduce peel force as much as you can by minimizing cross sections (and some models, you'll make this worse by just blindly rotating 45 degrees), and optimize your support structure, both for actual support and for trying to hide the contact points to less visible areas on the mini (not always possible, especially if there's crazy overhanging details).

But yeah, a lot of people in various groups I follow seem to immediately do the "you need to rotate it 45 degrees" as a first response. "Conventional wisdom" and all that.

45 degrees from horizon is a "general purpose" angle for most prints. As others have noted, the actual angle required for best prints will depend on the 3d model.

Various test prints would show that striations begin appearing at about 30~45 degrees from horizon. Shallower angles may face drainage issues, and more upright (therefore taller) prints face longer print times.

I wish Chitubox would let you rotate whilst in the add attachments menu option. I've been reading around and a few use the 45degree angle ,but also look at the red zones highlighted as needing support and try to move the model into a position where there is the least red on show. Thing is you can't rotate and view the reds at the same time (because the software doesn't allow you to rotate/change scale and preserve the supports at the same time)

I warmly invite everyone to try Lychee Slicer, it's actually built with a sensible workflow in mind.