The indents and cutouts in the injection moulded part aren't only originally there to save on material cost but more importantly to preserve shape when casting, since the demoulding shrinkage causes thick parts to dimple and warp. Just something engineers get drilled into them pretty hard to always design with near-constant wall thickness. Absolutely love your redesign.
Excellent video, as an industrial designer & professional model maker in the past, that was one of my first takeaways when I started designing for my first 3D printer 10 tears ago. I love making replacement parts and fixing stuff using my design skillset, and working with the limitations of FDM printing to achieve results. You're doing a great job of educating the masses, with those subtle bits of trickery.
Instead of printing moulds you could use salt annealing. You pack the part in salt and then put it in the oven. That way you don't need a mould for each different part. CNC Kitchen tried this with good results. The biggest problem was the coarseness of the salt patterning the surface of the part. Maybe there is a finer material that could replace the salt. Or as you say just design the parts better for 3D printing. Great video. Keep up the good work.
just fyi, injection molded parts are required to not have large areas of plastic, as they will take too long to cool and cause sink marks on the surface, thats why there are "lightening cuts" its not because of material cost.
I love the mad scientist experimentation. So glad to see you redesigned it, because the first thought I had was it was a bad design with weak points. Chamfers and fillets are our friends. Great work.
as a 3d printing project, this is pretty great. as a wood working fix, just use a hard wooden dowel that fits in the hole. most normal dogs are round dowels anyway, and the workmate only uses that weird shape so that you can get a very low edge on one side.
I made a button like that to replace my broken microwave door button. Mine had the same problem with weak tangs but all I needed to do was to change the print orientation so that the layer lines are in the long dimension so if turned out very strong. The side effect was that the face of the button was now lined instead of the prefect flat face it previously was. But that was not a big issue and the button has been in my microwave for months now. It actually did break once but I adjusted something (I forgot what, some dimension) to make it more reliable and now it's good. One of the things I added was to fillet the base of the tang to give it a little bit more material and that made a big difference.
Thats cool! Actually did some "remelting" experiments some months ago as well. I tried embedding printed parts into one-time-use "temporary" molds, being a) compressed salt powder and b) heat-resistant silicone. interestingly, my primary issue as well was plastic "bubbling" out of the form. So I am pretty sure this was not related to the release agent you used. I was able to minimize the issue by tempering the part at 90°C , moisture seems to be the primary issue. But also any air caught inside will expand! And the plastic itself too, to some degree. PC and PETG was affected heavily by this. PLA worked much better. But I never got rid of it entirely. I stopped further experiments because I think I've found an alternative solution for my use case. But I really hope to see more insight and progress on this topic in the future. Would be so powerful to reach isotropic/injection molded quality!
As you mentioned salt you got me curious! I've seen recently some experiments on annealing 3d printed parts and I've been thinking about trying that but with the part suspended in fine salt and heating the salt around it as evenly as possible with temperature sensors close to the part, have you tried something quite like this or was the setup different? Does it anneal the part in a similar manner as they do with injection molded parts?
I’m impressed with your matte PLA version. Matte filament typically has worse layer adhesion than standard PLA, but with your design improvements, those kicked ass. I’d probably used something with glass or carbon fiber though, particularly for a workshop part like that.
@needitmakeit Take high Temp silicone. Print the part. Make a mold from silicone with the printed part. Take a 3d printer Toolhead. Drill a hole on the upside of the oven. Place the mold into the oven. Heat up the oven. Place the Toolhead on the drilled hole. Extrude the filament directly into the silicone mold.
I used the salt method to remelt and fuse printed PETG parts with success. You should give it a try if you haven't already. Small features became rather tough and stretched quiet a bit instead of snapping off.
Why would you even use the pins in the first place ? They are there to make an injection molding process possible but since you 3D print the part, just change the geometry instead. Your final design though is pretty good, well done.
That's a cool idea and I'm glad you tried it even though it failed. It seems making a silicone mold of the part you want to reproduce and casing it in polyurethane would be a better way to go than a 3d printed annealing mold. Your final solution to redesign the part for optimal strength when printed is great.
I personally would love a way to injection mold stuff not because I need something that can only be injection molded but because I have dozens of kilograms of support material and failed prints that I could recycle that way
I use Plaster of Paris to mould my parts. Print the part, completely encase it in plaster, dry the plaster thoroughly, then bake at the melting point of the printed plastic. Chip away the plaster bulk, then soak the part in baking soda to dissolve the remaining plaster. Works great, but requires about a week total and is a total loss for the moulds. You don't get the part distortion that is common with salt or sand baths, and you don't get mould fouling. Still, it's a huge time investment, so you really have to want the parts re-flowed to eliminate the weak layers.
Interesting. Would have been nice to see the same parts printed in regular PLA and tested as well. Carbon fiber supposedly makes everything more brittle.
Very interesting process. I took the simple road to replacing mine by making and/or replacing dogs using a hardwood block for the "Head" and hardwood dowel. I also made custom shaped "heads" for various holding methods. Made a box full in a hour. If they get chewed up, they go in the garbage.
I am really impressed by this idea and the execution! Obviously plenty of room for improvement, but this is so cool. Redesigning the part before using the "mold" would also make the "molding" process easier I think. Don't really need the injection molding voids anymore for example. The results of the melted PETG were surprising to me and also the look was very unsettling. I would like to see some ABS tests with some tweaks to the execution. I was wondering if the part partially melting and getting into all the voids and rough surface of the mold would be an issue. Dialing in the temperature just right may be important. Good stuff!
I haven't tried this, but for separating those molds, you might try designing in some jack screws. Put some more threaded inserts into the face that touches the other half of the mold, but put the clearance hole for the screw through the same half instead of the opposite half. On the opposite half, you might want to put a recess to glue in a bit of metal for wear resistance (a washer for a smaller-diameter screw perhaps). Then, when demolding, thread in a screw and use it to bear against the opposite half.
One trick for stronger studs that have to be printed vertical is to design a small hole in it, far below the surface. Normal stud with fail by ripping the top layers it is printed on, by embedding a small hole it has an anchor point and you get the full strength of the layer adhesion.
As riba2233, what if we used this high temp plastic just as mould. With maybe a cone on the top to serve as a buffer for material, we could just put some plastic granulats and leave it melt and fill the mould in the oven with gravity.
I love matte PLA but I've read multiple times that it (silk too) is weaker than standard. Also, how are you liking the Biqu CryoGrip plate? I've been using a SliceWorx flex bed for months and loving it. I saw the Biqo on a BlackFriday sale, it's being delivered today for a second printer.
I'd torque an M8 bolt or similar through the part with large washers each side. PLA under compression over a large area is absurdly tough, preloads it from going into tension as much, just like they do with pre-stressed concrete for structural things.
You can add some groves to your part and then just add nylon (or any other strong rope) under tension to it and then add some epoxy or CA glue to fix it in place. Should be faster than your approach.
I'd be curious to see what happens if you encased the parts in plaster of Paris rather than a printed form. It may well have been done, but I haven't seen it.
Very interesting.. but why not creating a rib between the two endpoints? They need to flex just a bit, but they do not need to be exact copy of the original molded part. However, love the experiment…thanks you for sharing?
Those are the only plastic parts of my Workmate that aren't broken 😀. It's all the other parts I need! I may try printing some anyway, as I need bench dogs for an MFT bench, and if I can get good prints on my Qidi Q1 Pro I could use the ideas to design what I need.
Honestly, that's seems like an absolutely pointless exercise, since you're okay to print a mold - just cast the part with polyurethane in it and it will be significantly more robust either than pla or petg or whatever kind of thermoplastic which in your opinion needs this kind of treatment
But that's not the point of this experiment. Everyone knows you can print a mold and fill it with a material. Leaving comments that call a video pointless is actually pointless though.
The indents and cutouts in the injection moulded part aren't only originally there to save on material cost but more importantly to preserve shape when casting, since the demoulding shrinkage causes thick parts to dimple and warp. Just something engineers get drilled into them pretty hard to always design with near-constant wall thickness. Absolutely love your redesign.
The last ~5 minutes are the most important part. Design for the process!
Excellent video, as an industrial designer & professional model maker in the past, that was one of my first takeaways when I started designing for my first 3D printer 10 tears ago. I love making replacement parts and fixing stuff using my design skillset, and working with the limitations of FDM printing to achieve results. You're doing a great job of educating the masses, with those subtle bits of trickery.
While using PPS-GF as a mold was cool, highlight of the video was seeing your approach to redesigning the part for 3D printing :)
Instead of printing moulds you could use salt annealing. You pack the part in salt and then put it in the oven. That way you don't need a mould for each different part. CNC Kitchen tried this with good results. The biggest problem was the coarseness of the salt patterning the surface of the part. Maybe there is a finer material that could replace the salt. Or as you say just design the parts better for 3D printing.
Great video. Keep up the good work.
just fyi, injection molded parts are required to not have large areas of plastic, as they will take too long to cool and cause sink marks on the surface, thats why there are "lightening cuts" its not because of material cost.
I think this could work great with the part cast in plaster instead of a printed mold.
Not working well, cnckitchen tried this already and it was a f*n mess 🫨😵💫
I love the mad scientist experimentation.
So glad to see you redesigned it, because the first thought I had was it was a bad design with weak points. Chamfers and fillets are our friends. Great work.
as a 3d printing project, this is pretty great. as a wood working fix, just use a hard wooden dowel that fits in the hole. most normal dogs are round dowels anyway, and the workmate only uses that weird shape so that you can get a very low edge on one side.
I made a button like that to replace my broken microwave door button. Mine had the same problem with weak tangs but all I needed to do was to change the print orientation so that the layer lines are in the long dimension so if turned out very strong. The side effect was that the face of the button was now lined instead of the prefect flat face it previously was. But that was not a big issue and the button has been in my microwave for months now. It actually did break once but I adjusted something (I forgot what, some dimension) to make it more reliable and now it's good. One of the things I added was to fillet the base of the tang to give it a little bit more material and that made a big difference.
Thats cool! Actually did some "remelting" experiments some months ago as well. I tried embedding printed parts into one-time-use "temporary" molds, being a) compressed salt powder and b) heat-resistant silicone. interestingly, my primary issue as well was plastic "bubbling" out of the form. So I am pretty sure this was not related to the release agent you used.
I was able to minimize the issue by tempering the part at 90°C , moisture seems to be the primary issue. But also any air caught inside will expand! And the plastic itself too, to some degree.
PC and PETG was affected heavily by this. PLA worked much better.
But I never got rid of it entirely. I stopped further experiments because I think I've found an alternative solution for my use case.
But I really hope to see more insight and progress on this topic in the future.
Would be so powerful to reach isotropic/injection molded quality!
As you mentioned salt you got me curious! I've seen recently some experiments on annealing 3d printed parts and I've been thinking about trying that but with the part suspended in fine salt and heating the salt around it as evenly as possible with temperature sensors close to the part, have you tried something quite like this or was the setup different? Does it anneal the part in a similar manner as they do with injection molded parts?
I’m impressed with your matte PLA version. Matte filament typically has worse layer adhesion than standard PLA, but with your design improvements, those kicked ass. I’d probably used something with glass or carbon fiber though, particularly for a workshop part like that.
I think what you are thinking of as matt is fiber filled.
@needitmakeit
Take high Temp silicone. Print the part. Make a mold from silicone with the printed part. Take a 3d printer Toolhead. Drill a hole on the upside of the oven. Place the mold into the oven. Heat up the oven. Place the Toolhead on the drilled hole. Extrude the filament directly into the silicone mold.
I used the salt method to remelt and fuse printed PETG parts with success. You should give it a try if you haven't already. Small features became rather tough and stretched quiet a bit instead of snapping off.
Why would you even use the pins in the first place ? They are there to make an injection molding process possible but since you 3D print the part, just change the geometry instead.
Your final design though is pretty good, well done.
Save material too maybe.
Tho the same could be achieved with modifiers instead. Make the thing that flexes 100% solid and everything else hollow.
@@Splarkszter Exactly, for 3D printing, ribs and such are irrelevant.
Last redesign montage put a smile on my face. [ Chef's kiss ]
That's a cool idea and I'm glad you tried it even though it failed. It seems making a silicone mold of the part you want to reproduce and casing it in polyurethane would be a better way to go than a 3d printed annealing mold. Your final solution to redesign the part for optimal strength when printed is great.
I personally would love a way to injection mold stuff not because I need something that can only be injection molded but because I have dozens of kilograms of support material and failed prints that I could recycle that way
Recycling companies are slowly coming up with ways to recycle failed prints and unused filaments as a service but its so slow...
Instead of whacking it open, use compressed air.
I use Plaster of Paris to mould my parts. Print the part, completely encase it in plaster, dry the plaster thoroughly, then bake at the melting point of the printed plastic. Chip away the plaster bulk, then soak the part in baking soda to dissolve the remaining plaster. Works great, but requires about a week total and is a total loss for the moulds. You don't get the part distortion that is common with salt or sand baths, and you don't get mould fouling. Still, it's a huge time investment, so you really have to want the parts re-flowed to eliminate the weak layers.
Fascinating. TY for putting in the time and effort in the experiment, as well as producing a video to share it with us.
Interesting. Would have been nice to see the same parts printed in regular PLA and tested as well. Carbon fiber supposedly makes everything more brittle.
Oh boy you were so close to my idea, using pps-cf as a mold for sunlu pla meta which prints as low as 170c, very cool vide0
Thank you I will print some for my workbench!
Very interesting process. I took the simple road to replacing mine by making and/or replacing dogs using a hardwood block for the "Head" and hardwood dowel. I also made custom shaped "heads" for various holding methods. Made a box full in a hour. If they get chewed up, they go in the garbage.
I am really impressed by this idea and the execution! Obviously plenty of room for improvement, but this is so cool. Redesigning the part before using the "mold" would also make the "molding" process easier I think. Don't really need the injection molding voids anymore for example. The results of the melted PETG were surprising to me and also the look was very unsettling. I would like to see some ABS tests with some tweaks to the execution. I was wondering if the part partially melting and getting into all the voids and rough surface of the mold would be an issue. Dialing in the temperature just right may be important. Good stuff!
I haven't tried this, but for separating those molds, you might try designing in some jack screws. Put some more threaded inserts into the face that touches the other half of the mold, but put the clearance hole for the screw through the same half instead of the opposite half. On the opposite half, you might want to put a recess to glue in a bit of metal for wear resistance (a washer for a smaller-diameter screw perhaps). Then, when demolding, thread in a screw and use it to bear against the opposite half.
why not just change the orientation so that potential layer breaks would be less likely to occur?
Because that wasn't the purpose of this experiment.
In conclusion, redesigning for FDM is better than annealing. Easier too, since if one breaks you can just print another one. Great work.
Those tabs break because they unconstrained, add an additional tab between the two flextures to limit how far they can bend.
One trick for stronger studs that have to be printed vertical is to design a small hole in it, far below the surface. Normal stud with fail by ripping the top layers it is printed on, by embedding a small hole it has an anchor point and you get the full strength of the layer adhesion.
Can you elaborate more on that idea?
Sounds interesting
I've seen lower quality filaments with CF having many voids, that may have been your PETG-CF expansion issue.
I am satisfied with the strength of the parts printed without any filling at all. I fill them with epoxy resin with filler and reinforcement.
What kind of filler? Thanks.
@@michaels3003 Portland cement
What about making a silicone mold for the annealing?
What if you just cast plaster around it before baking? Having a material you can soak off would avoid the oil, however you are exposing to water.
As riba2233, what if we used this high temp plastic just as mould. With maybe a cone on the top to serve as a buffer for material, we could just put some plastic granulats and leave it melt and fill the mould in the oven with gravity.
Or just use a powerful extruder to pump it in :)
Instead of printing the final piece flat, I'd try printing it at 45 degrees with tree support so that the layers aren't the weak point.
Thanks for the video and the files!
You might want to look into powdered salt annealing, instead of printed molds
That wasn't the point of this video.
Do you have a resin printer? I’d be interested in a test using HT resin for the mold.
I love matte PLA but I've read multiple times that it (silk too) is weaker than standard. Also, how are you liking the Biqu CryoGrip plate? I've been using a SliceWorx flex bed for months and loving it. I saw the Biqo on a BlackFriday sale, it's being delivered today for a second printer.
Can you do the same thing with a complete captured plaster of Paris cast?
You should consider using a screw as rebar
I like your way of teaching great video o will be following you from now on
There are releasing agents for the molds you can buy, spray ones.
It’s such a good channel, I’m so glad it’s not another cooling nozzle video.
Put the part in a cup of fine sand or salt, tight the powder, and heat all that on the oven. then just break the sand.
How about making cylindrical holes inside each "finger", and insert a couple of nails or some metal rod?
That is probably what I would do.
I'd torque an M8 bolt or similar through the part with large washers each side. PLA under compression over a large area is absurdly tough, preloads it from going into tension as much, just like they do with pre-stressed concrete for structural things.
Devil Scooby Doo 😂.
Interesting.
Thanks for sharing.
PPS parts really need to be annealed to allow all layer lines to fuse together.
Great video Mike. Always look forward to your next video. And...Thanks for the improved work clamp/dog. How is that QiDi holding up?
You can add some groves to your part and then just add nylon (or any other strong rope) under tension to it and then add some epoxy or CA glue to fix it in place.
Should be faster than your approach.
I'd be curious to see what happens if you encased the parts in plaster of Paris rather than a printed form. It may well have been done, but I haven't seen it.
The plaster will dissolve in water after heating removing one of the bigger issues you had.
How about PU injection in 3d print mold?
Very interesting.. but why not creating a rib between the two endpoints? They need to flex just a bit, but they do not need to be exact copy of the original molded part. However, love the experiment…thanks you for sharing?
Haha… I should have waited until the end of the video…;-)
Those are the only plastic parts of my Workmate that aren't broken 😀. It's all the other parts I need! I may try printing some anyway, as I need bench dogs for an MFT bench, and if I can get good prints on my Qidi Q1 Pro I could use the ideas to design what I need.
Really interesting ideas
An interesting experiment, I didn't think about it. Do the new glacier plates work well?
I literally was about to print one of those work table clamps. spooky coincidence
Injection mold dies are highly polished and that does make a huge difference when it comes to part release.
Is this the guy from that chipmunks movie?
Awesome video
Smart re-design ! 3D printing requires more brain than printing in itself.
Thanks.
You can use a glass tupperware full salt when you don't have enough pps cf 😁
Very good
"ePLA" is said to be tougher then PETG
What about annealing?
The melty PETG CF looks an awful lot like a top down view of a beetle. 😅
Geometry always triumph over material.
try using salt or plaster
Silicone molds would probably work better than rigid molds
Honestly, that's seems like an absolutely pointless exercise, since you're okay to print a mold - just cast the part with polyurethane in it and it will be significantly more robust either than pla or petg or whatever kind of thermoplastic which in your opinion needs this kind of treatment
But that's not the point of this experiment. Everyone knows you can print a mold and fill it with a material. Leaving comments that call a video pointless is actually pointless though.
Matte PLA is an inherently much weaker material, btw
DFM FTW