Supports. How dense?

[Thunderhawk]

The answer may be in the mega thread but it's hard to find things there, so....

Anyone have some tips on supports for something like a HO scale locomotive cab? I used chitubox and light supports which wasn't a good plan as it was very difficult to get the supports out of the cab interior. Also had some welded to the outside of the cab along with a few failed supports. Made three cabs and three noses in one shot and really only can use one cab.

Will be going to heavier supports but my question is along the lines of how dense do the supports need to be? There was a veritable forest inside the cab which obviously isn't fun to remove. And some of them went through the windows making it even more of a bear.


Here is a pic of the cab. Very pleased with the surface finish.

This is with Monocure grey at .02 and the settings;

[Lemosteam]

One solution might be to design the supports yourself as part of your 3D model.

[Thunderhawk]

One solution might be to design the supports yourself as part of your 3D model.

That's what I am considering. Just curious at what distance to set the supports from one another to prevent sagging without going overboard. Very pleased with how this came out other than the supports. It looks molded.

eta-This is what it does on medium supports. Note it doesn't feel the need to support the windows and they came out fine on the test shot. The roof is rather nuts though.

[rodsup9000]

  I add the supports manually in photon slicer. Depending what it is, I get by sometimes with very few supports.

This is auto-support on 25% density











 And this is the way I printed 30 of them

















  You could do something like this with your cab.  I used heavy around the bottom and medium inside.


 

[Thunderhawk]

Supports are one of the big benefits of the Carbon machine. As it doesn't stick to the FEP they don't need nearly as many supports or as big.

I thought I saw something where someone was trying to make a Photon work similar with an oxygen permeable layer. That would be something.

btw-When making "normal" things, the Carbon printer is nowhere near as fast as they claim. Works pretty fast for mesh like they have in all of their promotional stuff but otherwise it's hardly faster than a Photon.

Back to the Photon, I noticed it wanted to add supports as shown above for little ledges and such. I took those off and had no issues. I know the roof will need some support but what I'm getting seems excessive.

Will have to experiment. Just hate the wait as even at 05 layers the cab and nose is about 4.5 hours.

[IronPenguin]

I'm having similar problems with too many supports in Chitubox.
I'd suggest doing them manually, but do an stl with just the top of the cab roof. Clone it and try different combinations to see how few you can get away with.XGKTET

[narrowminded]

I have just used the standard slicer and since the first print have placed all supports manually. 

For the first few prints, using 30% density, I would look at their version of what support might look like (automatic) to get a sense of what they thought it should be and then clear those and start placing manually.  When using the automatic slicer, the first thing that became apparent was that they didn't get the supports everywhere that they needed to be and also didn't give any consideration to what the part would look like when completed, therefore no consideration to where the supports were attached.  Example, placing a support right in the middle of a bearing surface, or right through a delicate edge, or on the exposed face instead of the directly opposite hidden face.   By placing them manually these considerations can be made and size and exact placement can be adjusted to the need.  Also you can consider the part shape and where it might be rigid and where it might tend to sag and then place the supports accordingly.  One thing I discovered was that often fewer supports were required but also that in key areas, thin corners for example, additional support was needed to prevent sagging or worse. 

Then consider the size of the supports.  When placed frequently enough it seems that the engagement diameter of the "top" adds no additional benefit past .6 diameter and is what I'll use whenever the surface/ space allows.  Down to .5/ .4 is still pretty robust and I've used less but if needed that small I might increase the frequency a little.  It's a judgement call.  One advantage to keeping the engagement at .6 or less is that none of them seem to be so strong that they damage the surface they are attached to, ripping out a chunk of the part on removal.

Then consider the engagement depth dimension.  I leave the default at "0" because I'm generally engaging square on flat faces so additional engagement depth accomplishes nothing.  The engagement is already full diameter.  When near an edge it prevents the support from extending down the edge of the part it's attached to, effecting that face dimension or finish.  Where I might increase that number is engaging a round surface where I think I need the "grip" and it won't deform the surface or be visible on removal.

Then there's the "mid" section.  Again, I have used .6 most often but if the support will be very tall and is one of the key supports for stabilizing the whole part, that could be increased.  A judgement call.  If there are delicate supports that are so close to each other that the mid overlaps, tying in to the adjacent support, then I might reduce the mid to just engage that adjacent support and sharing the stability with it.  Both become more stable.  This might appear on supports to things like small rivet detail.

These are just some of the things that I've arrived at with my own limited experience.   There are still a lot of judgement calls but hopefully these things will help in the early learning stages... which most of us are not far from.  
 

[narrowminded]

This is with Monocure grey at .02 and the settings;

I see a heading and defined number for the "speed" of the Z travel, not just the time.  Does that actually increase or decrease the "Z" travel speed?  If so, I could use that.