Analog not fitting properly in models

[giedental]

Wondering if anyone else has had similar issues and has any suggestions.

We recently switched from the WhipMix model resin to testing out the Keystone Ultra Plus. We did a print with a Neodent GM 4.5mm and when I dropped the analog inside, it has this extra open gap around the edges.

I don't have a sample of my WhipMix resin, so it's kind of difficult to judge what's going on here. So, I printed out 2 new models using the ASIGA DentaMODEL and cured them in 2 different units, the ASIGA Flash (for 20 mins) and my Wicked CureBOX (20 min @ 60 deg). The results were the same with an open gap where the analog can "play around".

Does anyone know what could be the cause? Am I not curing long enough? Is my printers calibration off? Is it the resin? I could of sworn we didn't have these issues previously.

 

[KingGhidorah]

If it’s 3shape just lower the analog to model spacing setting. It’s normally set to .1, but try setting it to .01 and go from there. As far as the cause I do not know, but as long as it’s consistently printing with a gap like that, then changing that parameter will consistently change it so it prints without a gap.

 

[KingGhidorah]

I’d also scan in your printed models and compare it to the original STL just to make sure there’s nothing crazy going on

 

[giedental]

Ah I totally forgot there was a setting for that... I remember seeing it somewhere... will need to poke around to find it again.

Also a great suggestion to print, scan, and compare! Not using my brain today. Thank you!

 

[tuyere]

The scale factor of the resin might not be dialled in- print a calibration cube of a known size and measure it post-curing and see how far off you are. I'd guess your parts are oversized, if it's printing bang-on with other resins then tweaking the scale factor in the material profile is a good fix. If everything is oversized then you'd wanna tweak the build area measurements, but that's its own process that needs a lot of back-and-forth.

 

[KingGhidorah]

Ah I totally forgot there was a setting for that... I remember seeing it somewhere... will need to poke around to find it again.

Also a great suggestion to print, scan, and compare! Not using my brain today. Thank you!

Settings button is on the right hand side, it’s like the third tab down, all the way on the bottom of that tab

 

[giedental]

The scale factor of the resin might not be dialled in- print a calibration cube of a known size and measure it post-curing and see how far off you are. I'd guess your parts are oversized, if it's printing bang-on with other resins then tweaking the scale factor in the material profile is a good fix. If everything is oversized then you'd wanna tweak the build area measurements, but that's its own process that needs a lot of back-and-forth.

Thanks for this suggestion too! I printed a 20x20x20 cube yesterday and it seems like my print is very accurate.... maybe a tiny tiny margin of error on the Z axis, but still very darn close to perfect. (dentaModel resin, cured in Wicked curebox for 20min @ 60 degrees)

So with that confidence, I'm going to reprint the models and adjust the 'analog to model' distance and hope for a better fit.

 

[tuyere]

Out of curiosity, what did it measure as? For a 20mm cube I don't tolerate more than +/- 0.1mm, but I can usually do significantly better than that.
And yeah, Z heights are usually tricky because the base layers are very heavy, the trick to getting an accurate measure there is to measure from two points that are both off of the build plate. On my own cal-cube design I add a shallow groove all around the cube maybe 2mm off the print bed, and then use calipers to measure from the top edge of the groove to the actual top of the part; this way you can exclude the base layers from your measurement entirely, giving you a much more accurate feedback on 'true' Z-axis accuracy.

 

[Car 54]

Out of curiosity, what did it measure as? For a 20mm cube I don't tolerate more than +/- 0.1mm, but I can usually do significantly better than that.
And yeah, Z heights are usually tricky because the base layers are very heavy, the trick to getting an accurate measure there is to measure from two points that are both off of the build plate. On my own cal-cube design I add a shallow groove all around the cube maybe 2mm off the print bed, and then use calipers to measure from the top edge of the groove to the actual top of the part; this way you can exclude the base layers from your measurement entirely, giving you a much more accurate feedback on 'true' Z-axis accuracy.

If you use supports with thin connectors that come off pretty cleanly, would that work?

 

[tuyere]

Nah, in my experience the underside of parts 'floating' on a bunch of supports is always fairly uneven and varies a lot from nominal, you end up with a bunch of bumps or cones where each support meets the model, because it can't transition instantly from support to big flat surface, it takes a few layers for each column's spreading cone of cured material to connect and produce a contiguous surface. where do you measure from, the tops of the bumps or the spot where the full block of material starts? Unclear. In comparison, a groove in a vertical surface that's a milimetre or two deep will produce a flat upper edge easily and consistently because the material isn't trying to bridge nearly as much open space, and the geometry of the overhang is the same across the entire model so it gives you a very consistent edge to indicate off of.

 

[Car 54]

Thank you @tuyere that make sense.

 

[giedental]

Out of curiosity, what did it measure as? For a 20mm cube I don't tolerate more than +/- 0.1mm, but I can usually do significantly better than that.
And yeah, Z heights are usually tricky because the base layers are very heavy, the trick to getting an accurate measure there is to measure from two points that are both off of the build plate. On my own cal-cube design I add a shallow groove all around the cube maybe 2mm off the print bed, and then use calipers to measure from the top edge of the groove to the actual top of the part; this way you can exclude the base layers from your measurement entirely, giving you a much more accurate feedback on 'true' Z-axis accuracy.

I don't have a digital caliper to get you a precise number, but I would guess it was just about 0.1 mm.

So I adjusted the "analog to model" setting in 3Shape from 0.1 to 0.01mm and I still had the same amount of space around the analog. Being a little annoyed at this point, I decided to change it to -0.015 mm, and uhh yeah. Now it fits like a glove.

 

[tuyere]

Glad you fixed your issues, but... Order a decent Vernier caliper ASAP, it's a basic tool for being able to assess your own work in a quantifiable way. You can get a basic digital one for less than $50, or you can get an analog one for like $15 that's just as accurate but isn't as quick to read. That's nothing for a tool you'll constantly find useful once you have it.

 

[giedental]

Glad you fixed your issues, but... Order a decent Vernier caliper ASAP, it's a basic tool for being able to assess your own work in a quantifiable way. You can get a basic digital one for less than $50, or you can get an analog one for like $15 that's just as accurate but isn't as quick to read. That's nothing for a tool you'll constantly find useful once you have it.

Ah yeah what I meant was I have an analog one, not a digital one.

 

[tuyere]

No problem with that, the Vernier scale is basically as accurate as most digital calipers if you know how to read it properly, I just assume most people won't wanna figure out how to read this:


vs. a digital readout they already understand.

 

[bigj1972]

Why didn't the rep tell us we needed good calipers and settings? He said it was more accurate than impressions. Just push the button
and 10 minutes later, have a 3D cast.

 

[Doris A]

Why didn't the rep tell us we needed good calipers and settings? He said it was more accurate than impressions. Just push the button
and 10 minutes later, have a 3D cast.

Because he wanted to make a sale!