3D printing

[Iceman 29]

I know next to nothing about 3D printing, except the excellent results from pieces I occasionally buy from Model Monkey, and Reedoak. Of the current generation of 3D printers, what quality results could I expect from a $150 - $300 machine? Could I reasonably expect the results of the above mentioned companies?

3D PRINTER SHOWDOWN: $350 CONSUMER VS $73,000 PRO MACHINE:

https://hackaday.com/2022/01/29/3d-prin ... o-machine/

[drasticplastic]

So...for a printer with good results I shouldn't look for less than $300 USD, or around $500 CAN.

[Fliger747]

As with most consumer electronics, something a year out of date will perform most satisfactorily at half the cost. Until recently I still used the I-Phone Zero. Printers only have to do one thing, print. Perhaps having a learner printer and upgrading at a time that you have defined your needs and abilities would be a reasonable path.

Good Luck! Tom

[wefalck]

One thing I have been wondering about, whether it would be possible to print onto a prefabricated shape, rather than onto the 'building plate'. Zeroing-in might be a problem. I was just thinking that one could use this technique to add (surface) detail to a machined part.

[NavyShooter]

I'm realizing the difference between filament and resin printers - filament (which is what I have) has much lower resolution. 0.4mm wide is the smallest I can do, unless I get a 0.2mm nozzle, and that will drastically slow my print speed, but still not attain the resolution that a resin printer can do - apparently it's 0.05mm.

That's kind of stunning to me - a resin printer has 8x more detail capability than my filament printer.

I've received a detailed model file from someone here that was originally printed on his resin printer, and I've looked at the rather spectacular results he attained, and when I look at the model after slicing, and the detail simply disappeared - anything less than 0.4mm wide was just gone.

It's left me pondering the idea of entering the Resin Printer game now...

[wefalck]

Resin printers for the time being seem to be the way to go for the kind of things we need. In addition, the thermoplastic materials that have to be used in filament printers are not very amenable to mechanical 'post-processing', while the acrylic etc. resins used can be drilled and milled, if needed.

[Fliger747]

I have no personal experience with filament printers, however they have advantages with regards to the size of items than can be printed. Additionally the use of the liquid chemicals has a certain mess and disposal factor. For the small components resin is a good procedure.

Tom

[Iceman 29]

I think the two types are complementary for the moment.

Although the Anycubic Photon Mono X can make large parts, almost as large as a wire printer. and with a much higher resolution.

viewtopic.php?f=13&t=348018

[Iceman 29]

One thing I have been wondering about, whether it would be possible to print onto a prefabricated shape, rather than onto the 'building plate'. Zeroing-in might be a problem. I was just thinking that one could use this technique to add (surface) detail to a machined part.

It seems to me complicated and risky to make the resin cling to a metal part that will necessarily be flat. There will be the 3 to 6 layers of cling that you can't get rid of.

You might as well print on the plate and glue later on the part.

[DrPR]

I have met my nemesis! I have been working on the TALOS missile launcher from the USS Oklahoma City CLG5. The base and trunnion support (body) print just fine.

I am using the same lock and key mechanism I used with the arms for the body to fit into the base ring. This should allow the launcher to rotate on the base without coming off, unless the launcher is rotated 180 degrees to where the key will lift out of the base.

But I have tried repeatedly to get the launcher arms to print without much success. Of a total of about 16 prints only two are useable, and these are of the original design which I later modified to hold the missiles and I am now trying to print.

I mentioned some of the problems earlier when I tried printing the arms horizontally. The "key hole" in the cylindrical part was collapsing. I tried the suggestion of printing the arms with the cylindrical part vertical. No joy.

The vertical print part is on the left, with a horizontal print part on the right. The entire side (closest to the printing platform) of the vertically printed part is a mess. However, even though the part closest to the platform failed, the remainder did print. This had me stumped for a while! Usually if the supports fail there is a gob of resin on the film over the LCD. But not in this case! Nothing remained on the film!

Here is an image of the printed part and supports. As you can see there is a forest of supports connected to the side of the arm that didn't print properly. But the surfaces that these shorter supports connected to appear to have not printed. So how did the rest of the thing continue to print?

By the way, the keyholes and keys printed fine in this vertical position. One problem solved, another encountered!

I think the answer to this problem is that there were not enough supports on the curved surfaces closest to the print platform, or the support attachments were not large/strong enough. So when the print progressed these surfaces started printing correctly. But as the process continued, the cross section area of these conical/cylindrical parts increased with each step, and the load on the support connections increased proportionally as the print stage lifted after each slice. Eventually the supports failed at one end, but remained attached at the other end. This allowed the unattached printed part to flex as it was pressed down close to the film over the LCD. It picked up exposed resin each time, so no resin stuck to the film, but the vertical step tolerance was lost. Eventually enough resin was cured onto the loose part that it no longer flexed, and the remainder of the part printed more or less OK.

So that's my theory. Any other ideas?

I had another print failure with the launcher base ring. Initially Chitubox put about a dozen supports to the flat bottom side of the cylinder. But these failed, leaving the bottom part of the ring with the 3mm thick outer circular part and the interior "X" stuck to the film over the LCD. Then I manually added three or four times more supports and it printed OK. Again, the forces of pushing/pulling the large flat surfaces through the resin were too great for the default number of supports.

The moral of this story is that you can't have too many supports. And the default Chitubox supports are not enough.

Phil

[Iceman 29]

From what I can see the contact area of the tips of the supports is too small, as your part is not hollow, I think, the supports under the part broke when printing the layers with the most surface area, the sussion is too high when peeling the FEP film, too much stress on these supports.

If the part is hollow, the layer surface decreases, which reduces the stress on the supports and improves the surface condition of the object. It is also necessary in this case to make a small hole at the base of the piece to let air in to avoid the phenomenon of sussion.

[BB62vet]

An observation (of sorts):
I've had similar problems but have basically eliminated that one area of concern simply by using MEDIUM supports when letting CHITU do the Platform act and then go in and add manually supports where they are missing. Especially on ends of parts. I use the LIGHT supports on veritical surfaces/edges where the weight of the part is not part of the consideration. Only that the part needs vertical support thru the printing process.

Hope this helps,

Hank

[Fliger747]

Generally the bottom surface on a SLA print will be crap. I printed quite a number of LCVP's and the V bottom needed some sanding. I tried printing vertically but the bottom shows less than any side surface. The outside of the launcher is the most visible surface and perhaps not a good candidate for being closest to the platen. Sometimes printing at some angle is optimum. I have very good luck with printing masts booms and yards as quite thin tubes.

[DrPR]

Thanks for the suggestions.

I think Pascal is right - as the print proceeds the surface area in contact with the film increases and the force necessary to separate it from the film increases. However, in most cases the part does separate from the film without breaking all of the supports (otherwise cured resin would remain on the film). But it is clear that some of the supports are breaking. After printing I can see that some of them have no connections to the printed part.

His suggestion of making the part hollow is a good idea. On the revised part I have two holes in the bottom of the launcher guide rails (the flat bottom of the launcher arm) to accept pins on the TALOS missiles. These holes could open into a hollow interior, and the hollow cavity could also extend through the cylindrical arm. In this way the cross section area would be reduced.

I have been using the "medium" supports in Chitubox. The contact area and penetration depth are also editable for the upper supports. If the contact points are too large they can break off the printed part leaving a pit. More experimentation.

I think this is what they call a "learning curve."

Phil

[Fliger747]

Not having worked with surface modeling, I am not familiar with how interior spaces are dealt with. Using the primitive 3D program I use, the thickness and shape of all surfaces is part of the intentional design. For voids it is necessary to consider fluid drainage during the printing process, trapped resin can remain gooey as it is not available for surface UV curing. On occasion this can eventually seep to the surface along partings between layers. For my LCVP's the bilge compartments required addition of seep holes to the allow drainage. Easily filled! For the king posts on the ARL they were printed vertically as a hollow cylinder. The interior remained goopy and required swabbing out. The masts and boom were then left open at both ends which helped. Blowing air through them after rinsing and cleaning out with a piano wire kept the passage open since the desire was to insert a reinforcing wire. The more fluid resins tended to print the hollow better but in rinsing tended to form bubbles inside which when cured could block the tube. Hence the blowing and swabbing.

Tom

[bwross11]

Coming a little late to the conversation, and honestly, I haven't read the entire chain so if I overstate something my apologies.

Tom, if you're talking about actual surface modeling as opposed to solid modeling, then the part is created of the surfaces stitched together, after which they need to be thickened. 3d printers do not print surfaces that are just surfaces, they have to have the parametric parameters which the slicer can then calculate to.

Phil, just a thought, and again I haven't read the entire string. Have you tried adjusting the settings of the printer? I printed a Mk37 director, with a 2 second exposure time. All the details really popped if they were oriented toward the resin bath. On the other hand, features that were facing the build plate looked like they were melting, and circular features looked like they had been flattened with an iron, and there were layer separations. Increased the exposure time to 5 seconds and now the melting was gone, the circular features were circular. Downside was that some of the smaller features now weren't as crisp. So, I actually adjust the setting almost each print depending on what I'm printing and the orientation due to feature details.

Just my 2cents worth for possible consideration in the conversation.

[Fliger747]

Using my software rather than specifying a thickness, the interior is part of the design parameters which can provide useful structural bulkheads and whatnot. On the pilot house level of the ARL I included bulkheads, tabled, Nav's work area and various vent trunking and whatnot which will never be seen. Foe tidiness I try to remove any surfaces before saving and slicing. This ensures that everything I see is going to be realized in the slicer.

Interesting the changing results with the varying exposure times!

Good info!

[DrPR]

Bruce,

The exposure time may have something to do with the failures I showed. When I first started using this printer I used the default exposure time of 5 seconds. Everything printed OK - unless I screwed it up some way. But Anycubic has a data sheet for the resin I am using (Basic Resin) that says 2 seconds is the right exposure for the printer (Photon Mono).

I used 5 second exposures for the horizontal launcher arm prints and 3 seconds for the vertical prints. These were solid pieces without hollow interiors, and had fairly large cross sections in contact with the film. Most of the horizontal parts finished printing, but all failed in some way and I think that was because there weren't enough supports. All of the vertical prints failed, again I suspect there weren't enough supports for the cross section area.

Long exposures create thicker layers with the Basic Resin. The light penetrates into the resin, and the longer the exposure the deeper the "cure" is. I suspect these thicker cured layers are stronger and perhaps less susceptible to distortion from the forces that pull the resin from the film.

I am hollowing out the interior of the parts so there will be a much smaller surface area in contact with the film, and therefore less force as the resin separates from the film. I'll see if that solves the problem. But I might also use longer exposure times.

[Iceman 29]

If you use longer times you'll lose details necessarily.

Hollowing out the parts has two advantages, reducing the volume of resin and avoiding tearing by putting a reasonable number of supports.

[DrPR]

Here is some interesting "learning curve" information.

I tried hollowing the arms and used the same 3 second exposure time as last time. I started to do this by hand, and then remembered Chitubox had a hollowing tool. And it could also put supports inside and punch drain holes. Here is the result:

It does a pretty good job, and is a lot faster than trying to build the interior surfaces and supports in the CAD program!

So how did it work?

This is the bottom of the port launcher arm. You can see the two drain holes - which are actually mounting holes for stubs on the missiles - and the launcher rails. The "key hole" in the end of the cylindrical part was extended into the interior for another drain hole. This is the best print yet! The port arm was almost perfect.

The starboard arm had some problems. It had the same two drain holes between the launcher rails as the port side, but I added another drain hole on the top of the cylindrical part close to the "key". This part of the arm is inside the launcher trunnion support and will not be visible.

Notice the front of the outboard launcher rail in these two pictures:

The port arm rail end came out square, and this is good.

The starboard arm rail was deformed upward. Bad!

So what's the problem? Look at these two pictures of the Chitubox supports:

Chitubox doesn't add enough supports by default, so I added a lot more, including supports at the very front end of the rail.

On the starboard arm I added a lot of extra supports, but I didn't put one right at the front of the rail. So it flexed as the print was being pulled from the film. Lesson learned!

There was another failure on the starboard arm.

This is the port arm Emergency Igniter (arrow). If the TALOS missile booster failed to ignite when we pulled the trigger we had a problem. It might ignite at any time so we certainly didn't want to take it back into the missile house! When the 4000 pounds of class B explosives ignited it accelerated the 7500 pound missile/booster to Mach 2.2 at a range of 8 miles in 5 seconds. Makes fuel dragsters look like kids toys! The booster exhaust flame was about 75-80 feet long, a real blowtorch! It wasn't something you wanted inside a magazine containing 200,000 pounds of explosives!

The Emergency Igniter arm carried a black powder charge in the can on the end (arrow) and could be swung down behind the booster. The arm rammed the charge into the booster nozzle and then it was ignited. No more problem (unless the Emergency Igniter failed).

The starboard Igniter didn't print correctly.

This is the default Chitubox support position and it worked OK.

This is the starboard arm supports. It was initially like the port side, but I moved the contact point a bit because it leaves a bit of support attached to the thin end ring on the Igniter can. My fault!

So the moral of this story is two-fold.

1. Try to hollow out any parts that have increasing cross sections as they print, like horizontal cylinders. The forces needed to separate the print from the plastic window over the LCD will increase with each step until they are great enough to break any supports. Hollowing the part reduces the cross section against the plastic and therefore the separation forces.

2. Always support the ends of parts. Parts that overhang their supports may flex when they are separated from the film, and will not print correctly.

So now I have 18 bad parts and two port arms that are good enough to use. And a bit better knowledge of how to get a successful print.

Phil