Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Nova3d Elfin 2 resin 3d printer Ask Me Anything (AMA) + updated info

9101 views
61 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    November 2006
  • From: Northeast
  • 746 posts
Posted by GraniteRailroader on Tuesday, January 19, 2021 12:58 PM

BNSF UP and others modeler

No, I'm pretty sure the layer height is legitamate. It takes me almost 2 hours to print 340 layers give or take, which is what my well car crew walkway treads have. Thats pretty ridiculous if you look at them via the google drive link I posted. They are very short!



Your printers max resolution for layer height is .025mm with a precision of .01mm accuracy on lift height. You could potentially be losing layers setting it at .01mm in Chitu. Just something to take under advisement. You won't really notice any real difference between .01 and .025, other than time. The 15 micron difference is neglible.

I'm actually surprised at how high your exposure times are. My Creality 002H's (which at stock are overpowered for uv exposure, but are modded now) are significantly less with minimal fails. I'm thinking your issues with prints pulling off are your lift/retract speed from the FEP. My lift speeds generally never exceed 50mm/min, with a 5mm lift distance. This allowed me to lower my exposure times without having to overcure to the previous layer. Are you hearing the "drum" sound when your printer lifts?

The more you print, the sooner you'll want to find one of the flexible magnetic build plates for your printer. They're amazing. If you have the ability, lapping the build plate to fix the imperfections on the surface isn't a bad idea. It'll help with prints failing due to coming off.

Also, you may have luck with cleaning your resin vat, and then wiping it gentily with soap and water. Just a damp paper towel (one of the no-lint/no-fuzz ones!!!) with a small drop of dish soap. This will help you relieve some of the adhesion to the FEP. Generally what I do is after cleaning the vat (which isn't often. I leave mine full 24/7, and just put a cover of seran wrap over them when swapping resins and put the next one on) is clean it with IPA, let dry (you'll see the filmy residue) wipe it down with the damp papertowl and dish soap, let dry, and then "polish" it to get rid of the film with a cloth. The dish soap helps resurface the FEP from the resin and UV exposure.

The warping is only an issue once I remove the small parts from the build plate. I've already fixd any warping that happens prior. Everything prints perfectly straight...



You can cure on the build plate, if you're using supports. On thin flimsy parts (think lithophanes, or scale mesh etc) this is an easy way to help prevent warping prior to curing.

I don't use the wash and cure station (I built my own). I'm also a cheapskate and filter my IPA and cure the leftovers to remove the resin from it.

You sound infinitely more knowledgeable on resin 3d printing than I. I am very interested in microstepping. Would you be able to tell me how to do it? I'm always looking for maximum performance in things I buy, and getting that kind of detail from my printer would be great

I'm just an educated idiot who picked this up fall of 2019. With the pandemic it's turned into a decent small business for myself. Most of these printers are just using images displayed from the slicer, with G-Code control for the lift servo controls/speed/timing. If you're familiar with firmware editting, and want to learn all the G-Code control, you can attempt it. If you're not familiar with that, I wouldn't try it.

You mentioned you can also glue your parts like any normal styrene model? What model glues have you found to work? My tamiya liquid cement won't...



I have a small set of files that I use to rough the edges. With some of the resins the plasticizers don't always react with CA. I've had decent luck with the Gorilla Glue superglue, their exoxy stick (when applied with a toothpick to keep it from getting everywhere), and generic two part epoxies. I don't have a huge selection of glues (just random generic CA) that I've tried on ABS-Like and PP-like resins.

A good way to check your overall performance, and to make a whimsical print, are some of the Benchy tugboats. I've got a fleet of them for the kids.

This space reserved for SpaceMouse's future presidential candidacy advertisements

  • Member since
    January 2014
  • 1,498 posts
Posted by ROBERT PETRICK on Tuesday, January 19, 2021 2:13 PM

There have been a lot of responses on this thread since I last posted. I have read (most of) them, and I will add a little more later.

I am working out the parameters for my Photon S printer using Elegoo water-washable resin. I will soon try the Anycubic-supplied resin just to see how it goes and how much it smells.

Here are CAD renderings of a pile cap for one of my bridges. I posted these photos and a coupla others in WPF last weekend. They are exported from AutoCAD 3D as .PDF files and converted to .JPG files for online viewing. The .STL file is a lot clearer and is user-usable in a kind of animated sort of way. But, no way to upload .STL to this forum . . .

 

And a photo of actual results . . .

Notice I printed directly on the build plate, as I will do for other designs. And notice the slight 'elephant's foot' around the bottom layers. I can compensate for that effect in the CAD file by adding a rabbet around the bottom perimeter. I am still piddling with the settings for this idea, but I think I can get it pretty close. Nothing fell off and nothing got stuck to the FEP.

I am using my actual project designs as test prints for right now, but I will meander over to the webpages provided earlier.

Robert

LINK to SNSR Blog


  • Member since
    November 2006
  • From: Northeast
  • 746 posts
Posted by GraniteRailroader on Tuesday, January 19, 2021 2:43 PM

ROBERT PETRICK

Notice I printed directly on the build plate, as I will do for other designs. And notice the slight 'elephant's foot' around the bottom layers. I can compensate for that effect in the CAD file by adding a rabbet around the bottom perimeter. I am still piddling with the settings for this idea, but I think I can get it pretty close. Nothing fell off and nothing got stuck to the FEP.

I am using my actual project designs as test prints for right now, but I will meander over to the webpages provided earlier.

Robert



The little bit of "elephants foot" around the base of your print is a result of over exposure and bleed through of light.

Your holes on the bottom - how did they measure in comparison to the design?

This space reserved for SpaceMouse's future presidential candidacy advertisements

  • Member since
    March 2018
  • 688 posts
Posted by BNSF UP and others modeler on Tuesday, January 19, 2021 3:14 PM

ROBERT PETRICK


 

And a photo of actual results . . .

Notice I printed directly on the build plate, as I will do for other designs. And notice the slight 'elephant's foot' around the bottom layers. I can compensate for that effect in the CAD file by adding a rabbet around the bottom perimeter. I am still piddling with the settings for this idea, but I think I can get it pretty close. Nothing fell off and nothing got stuck to the FEP.

I am using my actual project designs as test prints for right now, but I will meander over to the webpages provided earlier.

Robert

 

Your solution to the elephant's foot is a good one, I'm guessing you might have found it in a youtube video like I did?

The reason I don't use it is because that much flat surface on the build plate requires a lot of force to seperate, and it requires sanding for me, but if it works for you, more power to you.

 

I also noticed that for an early print, your results look spectacular! What settings did you use? Are you happy with the resin and the print?

I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.

  • Member since
    March 2018
  • 688 posts
Posted by BNSF UP and others modeler on Tuesday, January 19, 2021 3:35 PM

GraniteRailroader



Your printers max resolution for layer height is .025mm with a precision of .01mm accuracy on lift height. You could potentially be losing layers setting it at .01mm in Chitu. Just something to take under advisement. You won't really notice any real difference between .01 and .025, other than time. The 15 micron difference is neglible.

I'm actually surprised at how high your exposure times are. My Creality 002H's (which at stock are overpowered for uv exposure, but are modded now) are significantly less with minimal fails. I'm thinking your issues with prints pulling off are your lift/retract speed from the FEP. My lift speeds generally never exceed 50mm/min, with a 5mm lift distance. This allowed me to lower my exposure times without having to overcure to the previous layer. Are you hearing the "drum" sound when your printer lifts?

The more you print, the sooner you'll want to find one of the flexible magnetic build plates for your printer. They're amazing. If you have the ability, lapping the build plate to fix the imperfections on the surface isn't a bad idea. It'll help with prints failing due to coming off.

 

Also, you may have luck with cleaning your resin vat, and then wiping it gentily with soap and water. Just a damp paper towel (one of the no-lint/no-fuzz ones!!!) with a small drop of dish soap. This will help you relieve some of the adhesion to the FEP. Generally what I do is after cleaning the vat (which isn't often. I leave mine full 24/7, and just put a cover of seran wrap over them when swapping resins and put the next one on) is clean it with IPA, let dry (you'll see the filmy residue) wipe it down with the damp papertowl and dish soap, let dry, and then "polish" it to get rid of the film with a cloth. The dish soap helps resurface the FEP from the resin and UV exposure.


You can cure on the build plate, if you're using supports. On thin flimsy parts (think lithophanes, or scale mesh etc) this is an easy way to help prevent warping prior to curing.

 

 

 

 

 

I don't recall mentioning anything about my prints pulling off on this thread...

Although yes, it was an issue until very recently. I switched to 1.1mm thick supports at 60% density, so now everything stays put on the build plate, but it leaves me with a ton of pock marks under the print after snipping off the supports. I have considered tweaking my lift speed and distance just to tinker, but I might have to try it for this FEP film sticking, which I have had a problem with. It hasn't noticeably affected prints, but I don't want it anyway, I don't hear a "drum" sound exactly, just sounds exactly like the plate pulling up the FEP a bit before it snaps back down.

I have thought about adding a plate to my build plate, as printing directly on the plate without supports would be nice, but often I leave substantial scratches using sharp metal blades to seperate the print. And the provided plastic spudger is no good. I'm confused, however, as to what you mean by lapping the build plate? Also, where can I get these flexible magnetic build plates? And how do you compensate for the sudden additional build plate thickness? Recalibrate the printer's "home"?

Your dish soap idea for FEP cleaing is interesting, but I don't know if I need it at present, as I too do not clean my vat often and just leave the resin in for the next print. My only concern here is that since my resin is water washable, some of the residue from this cleaning process could screw with my resin and subsequent prints...

 

Curing my builds on the build plate has crossed my mind as well, but I have (at least at present) decided against it for 2 reasons. One, I remove all my supports prior to curing (less chipping and potential part damage, everything is more brittle after curing) which means the parts would come off the plate anyway. Two, I'm wary of curing prints on the plate becuase I'm pretty sure that it will make the leftover supports even harder to remove after they have been cured. And honestly the warping doesn't bother me that much. I just tape those walkway parts to a flat clear ruler and cure them like that.

 

Sadly, I have no skill at present in G-code, however, I would still like to see if I could figure out how to "jailbreak" my 3d printer somehow to get that insane resolution, as I intend to sell many of the parts I make and eventually print locomotive shells. If you ever find a tutorial online you think would help novices do this, by all means PM me or post it here.

 

 

I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.

  • Member since
    January 2014
  • 1,498 posts
Posted by ROBERT PETRICK on Tuesday, January 19, 2021 3:57 PM

GraniteRailroader

The little bit of "elephants foot" around the base of your print is a result of over exposure and bleed through of light.

Your holes on the bottom - how did they measure in comparison to the design?

Pretty much as expected. I added a rabbet recess around the interior openings the same as the outer perimeter. If I didn't, the bottom layers would drift into the openings the same way certain alphabet letters would tend to fill in if overexposed.

Yes, elephant's foot is caused by overexposure and bleed through. I am fidgeting with the settings as all part of the 'dialing in' process. Lowering the exposure time for the first five or six layers will help, but the trick is to not lower too much too soon or face the dreaded drop off syndrome. I'll get there. I have printed exactly two 3D resin prints in my entire life, so I'm pretty much at the point on the learning curve where I expected to be at this point.

Robert 

LINK to SNSR Blog


  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Tuesday, January 19, 2021 8:19 PM

BNSF UP and others modeler

 

 
GraniteRailroader


I can use a syringe with some of my photo sensitive resin and a UV pen-light and glue up parts. It takes seconds. The bond is the same material (resin) that it's printed from. Or I can glue them like any normal styrene model.

 

 

 

 

Interesting. You also came up with the idea of just using the resin to glue the parts. That UV pen idea is quite clever, I've just been using my curing lamp. I'll have to try that.

You mentioned you can also glue your parts like any normal styrene model? What model glues have you found to work? My tamiya liquid cement won't...

 

 There's actually a 'glue' sold just like that, complete with a UV pen light. It's really UV curable resin. Here's one:

https://www.amazon.com/Curing-Plastic-Adhesive-Seconds-Anything/dp/B07ZQVH9TR/ref=sr_1_2_sspa?dchild=1&keywords=UV+Activated+Glue&qid=1611109110&sr=8-2-spons&psc=1&spLa=ZW5jcnlwdGVkUXVhbGlmaWVyPUExRENUNFg2TUdKUzNCJmVuY3J5cHRlZElkPUEwMDY5ODk3MVdUNzlGQUI0NE5HViZlbmNyeXB0ZWRBZElkPUEwODM2NzU1M0dGVDU4UDBZQjBDUiZ3aWRnZXROYW1lPXNwX2F0ZiZhY3Rpb249Y2xpY2tSZWRpcmVjdCZkb05vdExvZ0NsaWNrPXRydWU=

                            --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    March 2018
  • 688 posts
Posted by BNSF UP and others modeler on Tuesday, January 19, 2021 9:25 PM

rrinker

 

 There's actually a 'glue' sold just like that, complete with a UV pen light. It's really UV curable resin. Here's one:

https://www.amazon.com/Curing-Plastic-Adhesive-Seconds-Anything/dp/B07ZQVH9TR/ref=sr_1_2_sspa?dchild=1&keywords=UV+Activated+Glue&qid=1611109110&sr=8-2-spons&psc=1&spLa=ZW5jcnlwdGVkUXVhbGlmaWVyPUExRENUNFg2TUdKUzNCJmVuY3J5cHRlZElkPUEwMDY5ODk3MVdUNzlGQUI0NE5HViZlbmNyeXB0ZWRBZElkPUEwODM2NzU1M0dGVDU4UDBZQjBDUiZ3aWRnZXROYW1lPXNwX2F0ZiZhY3Rpb249Y2xpY2tSZWRpcmVjdCZkb05vdExvZ0NsaWNrPXRydWU=

                            --Randy

 

 

 

Oh I see now. I know about those UV glues. I thought he meant he just used the UV light part of the pen, such as those off of "invisible ink" pens.

I would probably opt for just using little bits of resin for now, but the UV glue idea is interesting too...

I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Wednesday, January 20, 2021 10:07 AM

 Yes, I'm pretty sure that's what he meant - just using the printer resin as the 'glue' and curing it with the UV pen. Same concept - I have seen the glue and pen demonstrated somewhere in a video, and that has been around for a couple of years now. As long as the pen produces UV of the correct wavelength to cure the printer resin, it should work exactly the same. Or you could always get a UV LED of the proper wavelength and make your own UV pen.

                                       --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    November 2006
  • From: Northeast
  • 746 posts
Posted by GraniteRailroader on Wednesday, January 20, 2021 11:05 AM

BNSF UP and others modeler

I don't recall mentioning anything about my prints pulling off on this thread...



I misread one of your earlier statements about adjusting times and "pull up" as "pulling off".. Mea culpa.

BNSF UP and others modeler

Although yes, it was an issue until very recently. I switched to 1.1mm thick supports at 60% density, so now everything stays put on the build plate, but it leaves me with a ton of pock marks under the print after snipping off the supports. I have considered tweaking my lift speed and distance just to tinker, but I might have to try it for this FEP film sticking, which I have had a problem with. It hasn't noticeably affected prints, but I don't want it anyway, I don't hear a "drum" sound exactly, just sounds exactly like the plate pulling up the FEP a bit before it snaps back down.



The "drum" sound isn't as sharp on some printers. On some of my prints it's sharp (when the tension is stretched by large flat areas), others you can hear it peal away.

BNSF UP and others modeler

I'm confused, however, as to what you mean by lapping the build plate? Also, where can I get these flexible magnetic build plates? And how do you compensate for the sudden additional build plate thickness? Recalibrate the printer's "home"?



Lapping the build plate is simply sanding and polishing it's surface. They're thick enough, generally speaking, that you can resurface them and keep them in decent condition.

With the magnetic build plates you simply recalibrate your z-axis to zero it out with the additional thickness.

BNSF UP and others modeler

Your dish soap idea for FEP cleaing is interesting, but I don't know if I need it at present, as I too do not clean my vat often and just leave the resin in for the next print. My only concern here is that since my resin is water washable, some of the residue from this cleaning process could screw with my resin and subsequent prints...



The soap once it's dried and wiped off shouldn't give you any issues. It's my go-to for resurfacing the FEP. It will also help with FEP sticking... the dish soap reconditions the surface the same way teflon protects your cooking pans.

BNSF UP and others modeler

Sadly, I have no skill at present in G-code, however, I would still like to see if I could figure out how to "jailbreak" my 3d printer somehow to get that insane resolution, as I intend to sell many of the parts I make and eventually print locomotive shells. If you ever find a tutorial online you think would help novices do this, by all means PM me or post it here.

I was looking for micro-stepping for your printer. I haven't found anything specific but will keep my eyes open.

Aside from all that stuff, you guys should look at the availability of nFEP sheets on Amazon/McMaster Carr/wherever you can find it. PTFE lubricant is also a good option to have for keeping your FEPs healthy.

This space reserved for SpaceMouse's future presidential candidacy advertisements

  • Member since
    March 2018
  • 688 posts
Posted by BNSF UP and others modeler on Wednesday, January 20, 2021 12:33 PM

GraniteRailroader

 
With the magnetic build plates you simply recalibrate your z-axis to zero it out with the additional thickness.

 

 

Ok so 2 questions then. First, since mine comes autocalibrated, all I would have to do is tell it to find the bottom of the printer again? I also noticed that before every print, it does that anyway before going back down to actually start the print... So maybe I wouldn't have to do anything?

Second, are you saying that my resin prints will still stick perfectly fine to metals other than the build plate surface, e.g. a magnet or maybe even a thin sheet? If so, I would be very interested in trying out sheet metal. I know they have flexible build plates for filament printers, so if we could bring that into resin printing that would be great!

I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.

  • Member since
    January 2014
  • 1,498 posts
Posted by ROBERT PETRICK on Wednesday, January 20, 2021 12:53 PM

Here's a photo of my secret design for a curing chamber. Three-gallon stainless steel bucket (seamless, not galvanized). Plywood top (with rectangular hole). 405 nm UV lamp. Bucket with tapered sides and shiny insides disperse and reflect the UV rays all over the place; top keeps the rays inside the bucket to protect human eyes and skin.

Robert

LINK to SNSR Blog


  • Member since
    November 2006
  • From: Northeast
  • 746 posts
Posted by GraniteRailroader on Wednesday, January 20, 2021 1:23 PM

BNSF UP and others modeler

Ok so 2 questions then. First, since mine comes autocalibrated, all I would have to do is tell it to find the bottom of the printer again? I also noticed that before every print, it does that anyway before going back down to actually start the print... So maybe I wouldn't have to do anything?



It looks like it's testing the axis limits, not calibrating, on the video I watched of the startup. Here's your calibration instructions if you were to purchase and make use of a flexible build plate: https://www.youtube.com/watch?v=N_zvCgVjKGg

BNSF UP and others modeler

Second, are you saying that my resin prints will still stick perfectly fine to metals other than the build plate surface, e.g. a magnet or maybe even a thin sheet? If so, I would be very interested in trying out sheet metal. I know they have flexible build plates for filament printers, so if we could bring that into resin printing that would be great!

 


Yes, and no? The magnetic build plates use spring steel. I know from using it myself, that it works out okay. It makes pulling prints off a dream.. Other metals? I'm not totally sure.

There are flexible plates for resin. You could likely adapt one of the other printer's flex plate to work with yours without much issue.

This space reserved for SpaceMouse's future presidential candidacy advertisements

  • Member since
    March 2018
  • 688 posts
Posted by BNSF UP and others modeler on Wednesday, January 20, 2021 1:58 PM

ROBERT PETRICK

Here's a photo of my secret design for a curing chamber. Three-gallon stainless steel bucket (seamless, not galvanized). Plywood top (with rectangular hole). 405 nm UV lamp. Bucket with tapered sides and shiny insides disperse and reflect the UV rays all over the place; top keeps the rays inside the bucket to protect human eyes and skin.

Robert

 

Yep, looks like you got the exact same curing lamp. Nice setup!

You mentioned "protecting humans eyes and skin." Are there any risks with these UV lamps? I wasn't aware other than a couple faint references I've read in the past...

I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.

  • Member since
    March 2018
  • 688 posts
Posted by BNSF UP and others modeler on Wednesday, January 20, 2021 2:02 PM

GraniteRailroader,

Thanks for the info and link! I'll have to check that out. Are the magnetic build plates you are using flexible? What about it makes pulling prints off a dream?

Also, do you have a link to where I can get these magnetic plates?

 

Thanks!

 

Edit: I watched the video, it shows the automatic homing feature like I mentioned earlier. Additionally, my printer is the elfin 2, and the build quality is much better than the one in the video (all metal build plate). It looks as if I could just put an additional layer on my buildplate and autorecalibrate it and it should work? I might have to just find out for myself. -\(._.)/-

I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Wednesday, January 20, 2021 7:12 PM

 The UV exposure to your skin cna cause sunburn, and looking into it can most definitely damage your eyes. Given that I have enough eye issues so as it is, I will be extremely careful around the printer and whatever curing system I come up with, or buy.

 The ayes are strange, when it comes to exposure to damaging light, like strong UV, or a laser, by the time you even notice any pain, it's far too late to prevent damage. There's no warning of impending danger like say poking a sharp object into your arm. I'm considering UV protetive goggles. The lack of any real early warning system int he eyes and my general dislike of visiting doctors of any sort, especially when there is nothing wrong, is why I now have lost most of the vision in my right eye. The was a wakeup call. So much of what I enjoy doing involves being able to see, so now I am a bit paranoid when it comes to vision protection

                                        --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    March 2018
  • 688 posts
Posted by BNSF UP and others modeler on Wednesday, January 20, 2021 10:49 PM

Thanks! I was unaware of this. I will build protection against light bleed into my curing station to prevent further damage to my vision.

I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.

  • Member since
    January 2014
  • 1,498 posts
Posted by ROBERT PETRICK on Thursday, January 21, 2021 2:44 PM

I've printed a couple more pieces. I suppose now is a good time to discuss deposition dimensions and over-exposure, pre-cure and post-cure shrinkage, and long-term changes to finished resin parts occasionally exposed to sunlight coming in through windows.

In general the few parts I cast so far have come out of the printer a little bit larger than designed. Not much, about 0.010" or so, but 0.010" is still more than zero. I've read online and have seen YouTube videos that indicate shrinkage of 3% to 6% should be expected. And everyone seems to think that is not a big deal. But then again, the vast majority of the online 3D Resin printer people are casting baby Yodas or orcs or robust Amazon Warrior Princesses with large . . . umm . . . shields. So yeah, they don't seem to care much about inaccurate dimensions.

My plan for this technology is to design and produce architectural and structural fittings and details, and a tolerance of 0.010" (or even more) is not suitable. Even the idea of 3% to 6% is intolerable. Plus, is it 3% or is it 6%? I can try to compensate, but even after considerable calculation followed by extensive trial-and-error castings, the resulting pieces and parts still might not fit; and what would be the point?

Here's a photo:

Shown is the second pier cap along with another part that is designed to fit into the recess of the caps. I know what the dimensions were designed to be, but even after allowing a 'little bit' more clearance, I still don't know as of this moment (after about 2 hours of print, wash, and cure time) whether the smaller section will slide into the cap. And I won't know that until I remove the pieces from the build plate and try it out.
 
Any comments or suggestions as to how we can compensate for unknown over-exposure-induced swelling and pre- and post-cure shrinking? Any anecdotal or empirical figures regarding dimensions any of you have experienced?
 
Thanks,
 
Robert

 

LINK to SNSR Blog


  • Member since
    March 2018
  • 688 posts
Posted by BNSF UP and others modeler on Friday, January 22, 2021 12:27 AM

Honestly I'm a bit confused by your problem. I haven't had any issues with shrinkage or expansion once I had my exposure settings ironed out. Perhaps GraniteRailroader can chip in, but honestly, I didn't even know this was a thing.

Perhaps my eye is untrained, but all of my models have come out to the correct scale dimensions AFAIK...

Maybe check to see in chitubox if its doing anything funny?

I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Friday, January 22, 2021 8:29 AM

 These low cost printers aren't meant to make ulta-precision parts. If you need tolerances in the thousandth of an inch or in tenths, you need a lot more expensive printer. .01 in HO is less than a scale inch, I defy anyone to say they can see the difference without measuring. 

 The pier and cap, both being on the same print, should be fine. It when you print a bunch of components across multiple prints and then try to fit them together where you may have tolerance issues. I don't think these low cost printers are the answer if you want to regear a loco, for example. They should work, but with the rather loose tolerances, they will probably be very noisy gears. Detail parts, accessories, things like that - these printers should be fine. Again, you're talking less than a scale inch for HO.

 30 years ago, I worked in a machine shop where we made jet engine parts. Tolerances there were a tenth or two - affordable consumer 3D printers can't come close to that - yet. Neither can affordable consumer CNC machines or laser cutters.

                                       --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    March 2018
  • 688 posts
Posted by BNSF UP and others modeler on Friday, January 22, 2021 12:30 PM

That may be true...

But I would also venture to say YMMV between lower cost printer brands as well, which is why I might not be experiencing the problems other might with tolerances in their prints...

I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.

  • Member since
    January 2014
  • 1,498 posts
Posted by ROBERT PETRICK on Saturday, January 23, 2021 5:39 PM

Here're Part B and Part C to go with the photos from earlier. Shown are the pile cap, the pile shaft, and the pile base. Together they make up one of the bridge piers.

They are sitting on a couple of tooling parallels, which are used as references and offsets in machining operations, and are certified to be flat, straight, and square to a tolerance of 0.0002". The 3D resin prints are not nearly to that close a tolerance, but they are good enough for their intended purpose. I have been checking them repeatedly with a digital caliper and they are within 0.010". They started out a little 'fat', but they have shrunk a bit. The shrinkage factor is nowhere near the 3% to 6% figure that I found reported on several websites by 'internet experts'. No idea how they came up with such a number. My guess is that they do not have a firm grasp of mathematics or mechanics, and they pretty much eyeballed it. Then tossed out a wild guess, and went with it. My granny would say that these 'experts' are full of prunes.

Here's the assembled pier:

And one showing the pier in its intended place on the layout replacing the 'place holder' that has been there for a couple of years:

The pieces feel a lot smoother than they look. The plan is to put on two pretty heavy coats of sandable primer and see where that gets us.

Robert

LINK to SNSR Blog


  • Member since
    March 2018
  • 688 posts
Posted by BNSF UP and others modeler on Saturday, January 23, 2021 7:58 PM

Those look great!

Now the million dollar question...

How many hours did it take to print that thing?Laugh That looks like a loonnnggg print time...

I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.

  • Member since
    January 2014
  • 1,498 posts
Posted by ROBERT PETRICK on Sunday, January 24, 2021 12:22 AM

BNSF UP and others modeler

Those look great!

Now the million dollar question...

How many hours did it take to print that thing? That looks like a loonnnggg print time...

The cap and base took about 90 minutes apiece, and the shaft took just over 8 hours. It is 5-1/16" (5.0625") tall, and at 0.050 mm (50 micron) layer height that works out to 2670 layers. Looking back, I discovered that I could have printed two caps or two bases or two shafts in the exact same time it took to print single individual pieces. Lessons learned along the way . . .

Here are the other parameters. Five bottom layers at 50 seconds each for the base and cap. Six bottom layers at 60 seconds each for the shaft (I wanted to make double extra certain it would stick to the plate, knowing that it would be a very long build). For the rest of the layers, 6.5 seconds exposure with 1 second delay. Anti-alias set to 2.

Typical hollow structure for each piece has 1/16" skin and 1/16" stiffening ribs with 1/16" 'top' surface. The shaft has two 1/2" by 1/4" bosses with small center holes that will be tapped and threaded for 10-32 screws to attach the pier to the plywood deck of the benchwork. For the rest of its length (4-1/2"), the piece has the exact same cross section for well over 2000 layers. About half way through, I started thinking that that might not be such a good idea. It meant that the exact same pixels of the LCD screen would fire repeatedly with only a few seconds of recovery time. Which meant they could overheat, and in turn could cause the FEP to overheat, and then the idea that the piece sticking to the FEP could become a real possibility. You can think about a lot of stuff in eight hours.

Anyhow, things turned out reasonably well. I now have a full five 3D prints under my belt. Onward and upward!

Robert 

LINK to SNSR Blog


  • Member since
    November 2006
  • From: Northeast
  • 746 posts
Posted by GraniteRailroader on Monday, January 25, 2021 9:16 AM

I haven't forgotten about you guys. I plan on replying to everything once I get a moment. Just not enough time to get my Honey-Do list done before surgery AND try to write out replies on here.

I'll have plenty of time once I'm laid up for a day or two with the laptop in bed. Confused

This space reserved for SpaceMouse's future presidential candidacy advertisements

  • Member since
    January 2014
  • 1,498 posts
Posted by ROBERT PETRICK on Tuesday, January 26, 2021 11:04 AM

3D Resin Print Project No 2 . . .

Corner turret with witch hat roof for the Maritime Museum. N scale. CAD renderings. Exploded view and assembled view.

 

Robert

LINK to SNSR Blog


  • Member since
    November 2006
  • From: Northeast
  • 746 posts
Posted by GraniteRailroader on Wednesday, January 27, 2021 8:32 AM

Your piers look great. I'm interested in seeing how the corner tower comes out.

rrinker
These low cost printers aren't meant to make ulta-precision parts. If you need tolerances in the thousandth of an inch or in tenths, you need a lot more expensive printer. .01 in HO is less than a scale inch, I defy anyone to say they can see the difference without measuring.
Currently I have no problem reaching .001” tolerances with my LD-002H’s that have calibrated LED drivers. The hobby and pro-sumer market has demanded quality while prices have dropped.

rrinker
I don't think these low cost printers are the answer if you want to regear a loco, for example. They should work, but with the rather loose tolerances, they will probably be very noisy gears. Detail parts, accessories, things like that - these printers should be fine. Again, you're talking less than a scale inch for HO.


What ‘loose tolerances’? There are printers are on the market that offer a .002” tolerance out of the box. Calibrate the LED drivers (or replace them with a quality aftermarket driver), replace a single stepper motor, and make sure you have quality mounting hardware, and you can easily bring your XY precision into the tenths without a problem. For less than $500 / printer including the printers cost.

When I first got the LD-002H printers I had nothing but problems with them. Z-rails that wobbled causing build plates to misalign in the middle of prints, binding on the drive screw for the z-axis lift, LED output that varied with temperature, amongst a ton of other issues. I had some of the earlier models that really had not gone through testing and QA yet.

The solution? Replace the coupler on the drive screw, drive screw, and the thrust nut/washer assembly on the lift. Retighten hardware with a small bit of Loctite applied to counteract the tendency for it to loosen. Add additional cooling (via 120mm PC fans) to the internals, an external current driver for the UV LED’s, and find a stepper motor the would tolerate microstepping so I could have a finer resolution on the z-axis lift. I did a few other things like sending my resin VAT’s out to a buddy’s shop and having him mill the surfaces so that I knew they matched up to the glass of the LCD. A lot of these aluminum resin vats I have found are slightly bowed. No big deal if you are friends with a machinist or can do it yourself.

I’ve currently got two of my machines setup with dental grade resin doing proof-of-concept for our local dentist. We’re rapid prototyping dental parts for them in a manner that would allow same day or next day installation.

My parts will never see that use, but we’re showing them the technology for it. I don’t think anyone is expecting to compete with Lockheed or Grumman with a $250 3D Printer. Same with the folks that have the extruded aluminum CNC mills with belt drives. That’s not the application. But with a little effort you can bring these things down to a tenth’s accuracy with not many issues.

This space reserved for SpaceMouse's future presidential candidacy advertisements

  • Member since
    November 2006
  • From: Northeast
  • 746 posts
Posted by GraniteRailroader on Wednesday, January 27, 2021 10:00 AM

I switched one of my vats over to Elegoo Water Washable White. This is 10 layers at 6 seconds per layer for the bottom, and then 1.2 seconds exposure per layer after that. Layer height of .025.

I highlighted a couple bits in Sharpie so it hopefully would stand out. On the left side "pin hole" test, you can see that only a few of them have not filled in completely. This is a two part problem - lift height not exitting the resin to allow it to drain out (I meant to put 55mm, I only put 5mm) and slight over exposure.

You can see that for the most part the raised cylinders came out as they were suppose to.

In the center, the negative relief versus positive center test came out pretty well.

On the bottom I got a little overzealous with the sharpie (these pain meds will do that to ya!) and filled in some of the detail. You'll notice the angled portions on the left have some broken bits from me being a little rough while washing the part. 
This was on a standard aluminum build plate (not one of my flexible plates).

The only "treatments" to the build plate and FEP are what I have outlined in previous responses.

 

 

 

This space reserved for SpaceMouse's future presidential candidacy advertisements

  • Member since
    March 2018
  • 688 posts
Posted by BNSF UP and others modeler on Wednesday, January 27, 2021 3:22 PM

GraniteRailroader

I switched one of my vats over to Elegoo Water Washable White. This is 10 layers at 6 seconds per layer for the bottom, and then 1.2 seconds exposure per layer after that. Layer height of .025.

I highlighted a couple bits in Sharpie so it hopefully would stand out. On the left side "pin hole" test, you can see that only a few of them have not filled in completely. This is a two part problem - lift height not exitting the resin to allow it to drain out (I meant to put 55mm, I only put 5mm) and slight over exposure.

You can see that for the most part the raised cylinders came out as they were suppose to.

In the center, the negative relief versus positive center test came out pretty well.

On the bottom I got a little overzealous with the sharpie (these pain meds will do that to ya!) and filled in some of the detail. You'll notice the angled portions on the left have some broken bits from me being a little rough while washing the part. 
This was on a standard aluminum build plate (not one of my flexible plates).

The only "treatments" to the build plate and FEP are what I have outlined in previous responses.

 

 

 

 

1.2 second cure time at 0.025mm?!?!

 

I used the exact same benchmark print on my water washable grey resin, and I got my most ideal results at 6 second exposure with a layer height of 0.010mm.

I tried dabbling in the one and two second exposure range later when I printed small parts, but it wasn't working at all on several fronts. Everything from the supports to the part itself was way underexposed. I believe my exposure is currently at 4 seconds.

 

However, you just accidentally answered one of my long standing questions, and that was the purpose of being able to tweak lift height. To drain the resin properly! I never knew that and it was a big issue with my parts. Now I finally know. I definitely need to tweak that so it lifts above the pool of resin each time. Do you have any recommendations to what I should change the lift speed to?

I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.

  • Member since
    November 2006
  • From: Northeast
  • 746 posts
Posted by GraniteRailroader on Wednesday, January 27, 2021 4:35 PM

BNSF UP and others modeler

1.2 second cure time at 0.025mm?!?!

I used the exact same benchmark print on my water washable grey resin, and I got my most ideal results at 6 second exposure with a layer height of 0.010mm.

I tried dabbling in the one and two second exposure range later when I printed small parts, but it wasn't working at all on several fronts. Everything from the supports to the part itself was way underexposed. I believe my exposure is currently at 4 seconds.

However, you just accidentally answered one of my long standing questions, and that was the purpose of being able to tweak lift height. To drain the resin properly! I never knew that and it was a big issue with my parts. Now I finally know. I definitely need to tweak that so it lifts above the pool of resin each time. Do you have any recommendations to what I should change the lift speed to?



Yes, 1.2 0.025mm for the non-base layers.

Are you sure you're at .010 mm, and not .1 mm? 6 seconds for WW Grey seems way too long - even knowing that the Creality's tend to have more UV output than most of the other offerings on the market.

Everything I've read so far on your printer says that it does not support .01 layer height. Try slicing at 0.025mm height with the same exposure you're using now if you don't mind obliging my curiousity for one of the test prints.. and posting the result?

Lift speed and lift height both have influences and outcomes on the final product. I try to keep my resin vat's topped up without overflowing during printing. Lift speed can effect print adhesion to the build plate and FEP. Too fast and you risk tearing the FEP or breaking adhesion to the build plate, to slow and you face the same thing just over a long period of time. There's a sweet spot in the middle that isn't always the same.

Height matters to make sure that your resin has a chance to get away from the printed layers and clear off the build plate. This can help with overexposure a little bit, but not tremendously. On hollow parts like the bridge piers above, it helps get the excess resin out of the nooks and crannys of the print to keep consistent layer and wall thicknesses.

The hold time and delay also matters when using thicker viscosity resins, giving them more time to flow in (or out) between layers so that you get a consistent result.

This space reserved for SpaceMouse's future presidential candidacy advertisements

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

There are no community member online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!