Heated Bed, How Flat?

Hello, Sorry if this question has been asked before and if I'm asking in the wrong place. I spent some time trying to find an answer, but I must not be using the correct vocabulary in my search strings.

I'm building my first printer and I am wondering what people mean when they talk about the print bed being flat. I have been a precision machinist now for almost 40 years, and my understanding of the word flat needs more information I don't seem to be able to find.

I would like this machine to be accurate. But at the same time, I don't want to waste any time fighting to get a level of flatness that's just not needed for good print results. I also understand the need to compensate for changing temps.

So what does the 3D Printing Universe consider to be a flat print bed? +/- 0.0100" per foot? +/- 0.0010" per foot? +/- 0.0001" per foot?

Thanks in advance for any help.

Take care,

David

For best results, IMO you want the bed to be flat to within 0.05mm over its entire surface. Perhaps the easiest way is to use a plate of float glass.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thank you dc42 for the fast reply!

Maybe I should restate my question:

What percentage of layer thickness would be a good target for flatness? I'm guessing/imagining once the first half dozen or so layers get put down, things get better from there...

I've read on this forum and elsewhere that a 0.3mm layer thickness is considered the gold standard for a reprap home built machine. I've also read the marketing material created by commercial manufactures of 3D printers claiming 0.1mm for a layer thickness.

The prints I would like to try and attempt can stand a fair amount of surface roughness and they are very large (24" by 40" by 32" is the envelope of my planned machine). I'm hoping with a thicker layer thickness, I can get things made in faster time.

Sorry, my mind is running on...

David

(24" by 40" by 32")
That is very large

Layer size is usually 25% to 75% of nozzle size

Common extruders use 0.3 - 0.5 mm nozzle (opening at end of extruder where ABS/PLA squirts out)

Probably want to use 3mm filament to feed it

Take a look at "The volcano" with a big nozzle

[e3d-online.com]

As was said many use a piece of window glass or a piece of borosilicate glass

Make your "flatness" measurements off of that?

An etched surface on glass may grab your work piece better

Thank you cozmicray!

I was really relieved when you pointed me to the "Volcano" nozzle site. I had already been wondering if using nozzles much larger than what I normally see others using was something doable. Instead of filament, I'm looking at the extruder heads some here are working with. I think in my case, extruding pellets would be easier, and cheaper for the large parts I want to make. I could also potentially lay down a lot more material faster.

Why am I doing any of this? I have some parts I would like to start casting in aluminum. The dream is to extrude the part on the big printer in PLA. Encase the printed part in investment. Burn out the PLA from the investment. Pour molten aluminum into the void left in the investment. I admit up front there will be a little sanding and finish work on the finished print to smooth it out a bit. I'm hoping to experiment with sand blasting the finished print to deliberately give it a rough sand-cast sort of finish. With some luck, the finished casting will look more like a sand casting than an investment casting.

Why not just sand cast these parts? Because there just isn't a big enough market to justify the VERY expensive tooling needed. If the parts ever become popular, then I can look at real tooling for a more traditional casting process. Until then, I'm hoping I can leverage this world of 3D printing for small runs of some big parts.

My machine will be similar to the designs you see in what I think are called the CoreXY systems (?), but driven a little differently. I recently came into possession of some really cool linear steppers for my X and Y axis. Something similar (but not exactly) to this:



From what I understand, these types of linear steppers can be both precision in motion (with micro drivers) and fast. They don't have much linear force, but from what I'm understanding about these printers on this forum, a printer doesn't need tons of force to move the print head around. Also, the linear steppers eliminate belts and rail system because the motor and rails are all one unit. The Z axis will be a more common direct drive, stepper driven, lead screw on each corner design using linear shafting for alignment. I feel safer about using a stepper on each corner than belts. The frame and major structure will be TIG welded square tubing.

Anywho, I will definitely create a build thread and share everything I do. I have zero plans on building large machines or becoming a business that sells technology to the casting industry. I'm just a one man shop. I'm also a huge fan of the opensource ideology. Things always go better with help from others.

Thanks again and take care,

David

If the bed is not flat then first layer adhesion is often a problem. The 0.05mm I suggested was based on a first layer height of 0.2 or 0.25mm, which is what I generally use. Some people use a thicker first layer than the rest so that bed flatness is less important. So with a combination of a 1.2mm nozzle and thick first layer, you should be OK with a surface flat to within about 0.4mm.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Why not just squirt out wax? lost wax process?

"Solidscape® 3D printers are primarily used to produce "wax-like" patterns for lost-wax casting/investment casting and mold making applications."
[www.solid-scape.com]

Metal Casting with Your 3D Printer
[makezine.com]

You could squirt PVA
Polyvinyl alcohol (PVOH, PVA, or PVAl) is a water-soluble synthetic polymer.
and just wash out your cast.

Yes, PVA is water soluble and could be washed out. Unfortunately it costs a lot more than PLA.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thanks everyone. You have given me lots to think about. The PVA would be perfect, but as dc42 points out, it's super expensive in filament form. Because the pellets are used in water treatment, it's possible they can be purchased in bulk for less. I'll do some inquiring about that and see if anything comes up.

I did stumble across this interesting video of a company that makes a machine that prints out the actual sand molds. Pretty cool stuff:

Sand Printing Casting Mold

Printing wax looks like something worth experimenting with. If an recipe could be created out of easy to find materials... that would really be something. Especially for others.

I am so appreciative for your thoughts!

David

I'm imagining the machine I link to above in the video is somehow printing sodium silicate onto a thin layer of sand, and then immediately hardening it with C02.

Hi,

I am very interested in what you are doing; you may want to look at what we are doing, see

[3dprint.com]
[www.fouche3dprinting.com]

Hans Fouche & I are partners, together with Kobus van der Walt.

I have made pellet extruders for Hans with a 3mm nozzle, putting down 0,5kg/hr and am experimenting with one on a Velleman K8200 (a re-badged 3Drag). Pellet extruders are the way to go, especially for larger nozzle sizes, the only disadvantage being loss of resolution.

See also Richrap’s blog: [richrap.blogspot.com]

Regards and best of luck,

Wally

---

To get the right answer, you must ask the right question!

You do not need 3mm filament to extrude with a 1.2mm or even a 1.5mm nozzle. A 1.5mm nozzle for me extruded a 1.8-1.85mm thick bead using 1.75mm filament. There are a whole slew of problems to overcome when extruding through a larger nozzle. One of them is the heat zone and thermal mass. My 40w cartridge could barely keep up with my modified j-head running 1.5mm nozzle. The heater block simply was not big enough. The temperature went from 240c to 195 if I went much over 30mm/s. My printer is capable of 100mm/s+. Also the 2/3rds rule for layer heigh vs nozzle orifice goes out the window after .8mm - 1.0mm, you can only do about 50% nozzle width in height if your expecting good adhesion of the layers.

Here is my 1.5mm nozzle doing its job. [www.youtube.com]

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

E3D recommend 1.175 filament for their large diameter nozzles...

[e3d-online.com]

Thanks for the info gentleman!

WallyL: I like how simple Richard's design is.

Jaguarking11: Love the video and the great information!

FatFreddie: I find it interesting that the smaller filament sizes work with the large nozzles.

Gives me lots of brain food to digest...

Been spending some time on the design of the machine. I just got some stepper drivers in the mail so I'm going to fire up the linear steppers I have here and measure the kinds of forces and speed they can actually produce. I'll share all in a build thread.

Question: Where should I start my build thread? Under the "Reprappers" section?

Thanks everyone for your thoughts!!!

David