Prototyping a slip cast handle mold 

1) I use a heat gun to soften the PLA mold for challenging designs, it lifts right out but is one-time use only. 2) I use Murphy's oil soap in a spray bottle (50/50 with water). Let dry about 10 minutes. Greatly helps with removal in corners and small text. Thanks for your continued experimentation Kent!

I think it's because the plaster is surrounded on all four sides inside the handle and when the plastic shrinks, it shrinks around the breaking part. You can make the ear itself separately, applied in the same way as the separate wall, or maybe a semi-circular ear will be easier. I have the impression that all squares are problematic in ceramics. I'm waiting for the next videos. Now your video blog is better for me than any action series.

May I once again mention boil molds? Release works great when boiling the mold - it only allows you one cast per mold, so there's that. But the finished plaster mold will of course work over and over again. The mold can be made very thin with very little inner supports which cuts down on plastic and printing speed.

Love the video! I agree with the comments on the layer lines - on your larger moulds you sand and prime - basically filling in the layer lines. I assume you do not want a mould release (does it affect the plaster absorbency?) and so think the 'optimum' way would be to sand/prime the inner face of the handle and the base f the mould - which is what the plaster is gripping onto. Even at 'high' settings, each layer has a rounded edge sat on top of another rounded edge and I have found when using 3dprinting to make moulds (for casting concrete - via silicone) I need to do the sanding/priming that you do for your bowls (as I also do not want to use a mould release). However - I do then resort to some silicone (just the handle) and build a regular mould box around that.

Even a high detail 3d print is still a relatively grippy surface for a mold. I think if you just printed out the handle section itself, then put that on a silicone mat, that would allow you to separate it into two pieces

Just 3D print the handle and pack it halfway into clay, make a flat slab by pressing clay into a container and flipping it out. I print aligned registers for mine. Then it can be demolded separately as the back and the handle. You can use silicon from there if you want it to be replicable, I’ve always damaged the 3D print getting it out so far.

I think you NEED to add a mold release. I don’t know why you haven’t yet. 😊

I haven't done any plater molding so take my suggestion with a grain of salt. I think using mold soap would keep the plaster from sticking so much that it will not release without chipping.

Consider switching materials. Make a silicone copy of the handle model. Silicone works best for me with pottery plaster. Urethane is good too, and sometimes cheaper. I also tried and made my own reusable mold material out of gelatin, glycerine and water. Works too, but lower resolution.

One thing to consider is that a handle shrinks far less than a pot or mug or whatever. Say PLA shrinks 1% with the alcohol slush (completely made up number, recalculate with a real one) it'll pull 1mm away from the walls of a 10cm wide pot which gives you a great chance of it releasing. But if your handle is only 5mm wide, it'll shrink 0.05mm, maybe not enough to "unlock" the PLA layer lines from the mold. This is definitely a job for flexible TPU filament, that can deform and let you peel the mold out of the plaster. TPU works on most printers so I wouldn't immediately rule it out - direct drive printers do a better job of it, Bowden printers will tend to have uglier start/stops because of the filament squishing in the tube but still print walls fine. If you have a "ugliness here doesnt matter" feature on your mold and tell your slicer to start/stop there, you should be OK - I use this trick to print TPU on an Ultimaker. Also it makes your molds disposable which isn't ideal, but you don't need to heat PLA to a high temperature to make it flexible - you could bake the mold in the oven to soften the PLA, then carefully peel it out of the plaster. I was doing this with my mug molds before I stole your alcohol slush trick.

I don't think compressive and tensile strength are related much other than they are more when the plaster fully cures. I'm having similar issues, but seems to me that having a perfectly sharp edge is the reason, especially when the plaster isn't fully cured. A sharp corner is inherently a weak point. maybe adding a fillet around the edge would help negate this? Seems like doing this with plaster at a smaller scale like this is part of my issue. In other materials this would be less of a problem. I just now cast a part using a 50/50 of potters no.1 and ultracal.30. The ultracal would probably make a perfect mold as it's very much stronger, harder, and more durable. But I was concerned with using this sort of plaster because it has lower porosity, as its more for casting foundry stuff. I talked to another potter who has run a production slip casting shop for decades and he said doing a mix is what he normally does, as it makes the molds last much longer. He said they also maintain good porosity to absorb water during slip casting. I'm hoping this helps. My molds are similar to yours and don't have a massive undercut or anything weird. Middle part of the handle is probably just locking because of the plaster's expansion. Potter's no.1 has .21% and ultracal 30 has .08%. I'm betting that mixing the two will create a fairly linear change in properties. If so it'll drop the expansion to .145% Seems like a tiny number but, with a mold a tiny number is the difference between locking and popping it out. 😂 On simpler shapes or more rounded shapes like a pot (or a wing skin), it's pretty straight forward demolding. Thanks for the info, i hope you find some more insight into the issue.

You think the layer lines are the problem? Grabbing the plaster? Maybe prime it like the large molds

Love the scientific experimentation here. I'm interested to see you try TPU. I'm fairly sure my printer can handle it but I haven't invested the time or money to try. And I'm a bit worried about the filament messing with my PLA hot ends. I might just get a hot end just for TPU though.

Best video yet! I am always in a similar position as you on my building journeys. Im not a fan of 4 part molds or overenginerd complexity...simple is best. I agree on Tpu not being the #1 choice not because of ease of printablity (its getting much easier with 95a high speed overture) but the end result aesthetics would not be pleasing to me. Tpu has to be melt smoothed and im not interested in burning tires if i dont have to lol. I think the "C" shaped design ending in the pour sprues causes contraction to the inside of the "C" this could makes it pinch as the ends come together especially with the large end sprues. With that being said I would.... Have a pour sprue on the outside/backside of the C. (Uncomplicate the breaking zones) and make a draft angle for the inside of the Cup handle. Tangentially widen the opening distance to create a better overall shape draft instead of so much focus on radial chamfers (those seem to make a new pinching problems also). Since this is not true slipcasting yet. Maybe spray a mold release agent. I have seen affordable ones at $10 a can for cement. The shape combined the detail of part needed renders the Freeze removal process unsatisfactory. If in a bind I recommend heat to soften the PLA and remove. Use a heat lamp for example. Keep doing a great job 👍 Also i have no idea 2hat im doing half the time either lol 😂 All i know is forward lol must go forward. Make them pots 😅

Why not just make the entire mold in the 3d printer? Does it have to be plaster for drying?

Your shrinkage method usually avoids the need for any type of mold release, but in this case it's less effective, as you note. Perhaps a thin coat of soap or slip would do the trick.