3D cement printer

Looks great!

You could print a Dutch hot tub!


Looking forward to following your progress.

How are you pumping the cement?
Are you using calcium aluminate cement to speed curing?, good for cold weather.
Coat your garage floor with some acrylic floor sealer!
Can you provide a link to electrically assisted accelerated curing/hardening?

Have you explored using a Moineau pump?

An idea on Moineau pumps
[www.thingiverse.com]

Moineau stepper extruder
[forums.reprap.org]

Calcium aluminate cements
[en.wikipedia.org]

calcium aluminate cement
[claystore.alfred.edu]

Calcium Aluminate Cement Concrete, Tip Sheet
[ftp.dot.state.tx.us]

Requirements for using calcium aluminate cement
[www.fomento.es]

Edited 1 time(s). Last edit at 12/20/2013 06:15PM by A2.

AndreyR that is awesome. I would love to play with a cement printer. Will you be releasing your design?

---

WWW.ZATOPA.COM - Your Place for high quality 3D Printing Filament and accessories

This is great stuff your doing but I'm curious how you handle the two conflicting requirements of:

1) The cement needs to be fast setting so you can build up layers taller without collapsing the earlier layers because they have not set yet.

2) The cement should have enough working time so that it can move through the extruder without altering the flow
rate or blocking up the extruder pasages.

Awesome work though, and something I'm definitely going to watch closely.

Quote
AndreyR
After several months of experimentation, I finally made a custom extruder which works in the way it needs to for the cement printer (stable on any speed). This was not an easy task, and I made 5 different versions before the current working one. In the photo, you can see the very first print. Unfortunately, my garage workplace is close to freezing right now, and it would be hard to print several layers in one trial. But it will be even more interesting to experiment with cold conditions. I hope to be able to make these prints within the week, and when I do, I will post pictures or videos of my final results.
Thus, the first version of the prototype is done, with the prototype working just fine. My idea to make a cement printer based on the Rep Rap project and using attainable parts has finally become a reality.
Now on to version two.

yes definetely post pics/videos

this is very cool.

Quote
JamesNewton
hackaday has an interesting article about Paste extruders including concrete:
http://hackaday.com/2013/10/16/3d-printering-pastestruders/

Quote
A2
Have you explored using a Moineau pump?

For the extruder, I'm indeed using a progressive cavity/Moineau pump. Unfortunately, I can't use any plastic parts; I have to use metal because of the very high amounts of torque in the pump. I tried using various motors, but with each new pump design, I had to increase the power. I finally went up to a 120V/7.5A motor, and this is still not enough. I have come to the conclusion that for the extrusion of a cement/sand mix, a huge amount of power is needed.

Quote
jb92563
This is great stuff your doing but I'm curious how you handle the two conflicting requirements of:

1) The cement needs to be fast setting so you can build up layers taller without collapsing the earlier layers because they have not set yet.

2) The cement should have enough working time so that it can move through the extruder without altering the flow
rate or blocking up the extruder pasages.

Compared to different materials like plastic or clay, the cement mix is the toughest material to be using for extrusion.
The cement/sand mix has been studied for a long time, but for the use of extrusion, it needs to be treated like a completely new material and studied all over.
I don't know of any cement mixes with an extremely short setting time. But for the cement mix, if it is not touched or vibrated, each layer does have enough time to become hard enough to layer on top of, especially when it is warm out and sun radiation and wind speeds up the drying/curing time. My biggest problem at the moment is pushing the cement through the pump with its high viscosity. All existing pumps can use medium-viscosity at a high speed, but for the cement printer, a high viscosity at low speed is needed, and this is a new field of research. I have attained good results; however, I will keep advancing my extruder/pump design.

Quote
AndreyR
I don't know of any cement mixes with an extremely short setting time.

My biggest problem at the moment is pushing the cement through the pump with its high viscosity.

Calcium aluminate accelerates the hardening of Portland cement, which allows you to mix a watery slurry.
Look at the links I posted above.

Maybe you could pump it by adding calcium aluminate to help it harden/cure faster?
To prevent the wall from slumping, extrude a multiple layered wall with a small diameter extrudate.

Video, go to 1 min, he mixes a soupy/watery batch of concrete with calcium aluminate, and it sets in ~3 min.
[www.youtube.com]

Quote
A2

Quote
AndreyR

Calcium aluminate accelerates the hardening of Portland cement, which allows you to mix a watery slurry.
Look at the links I posted above.

Maybe you could pump it by adding calcium aluminate to help it harden/cure faster?
To prevent the wall from slumping, extrude a multiple layered wall with a small diameter extrudate.

Video, go to 1 min, he mixes a soupy/watery batch of concrete with calcium aluminate, and it sets in ~3 min.
[www.youtube.com]

definitely i have to try this! Now I am using calcium chloride and some other accelerators but it helps to set only 2 time faster.

looks like not easy to find that cement, at least in my minnesota aria... I dont see any supplier here.

Edited 1 time(s). Last edit at 12/22/2013 10:10PM by AndreyR.

@AndreyR ,

I'm not able to find a USA mfg..
You might have to make a few phone calls.
A few companies selling it premixed, see below.

There is also some thing to increase the fluidity of cement,
which when added to a watery slurry containing aluminate could make pumping much easier.

Another video of calcium sulfoaluminate
[www.youtube.com]

To speed up the hardening:
calcium sulfoaluminate (hydraulic tetracalcium trialuminate sulfate (C4A3þ))
Rapid Set, Concrete Pharmacy® FAST

To increase the fluidity:
Rapid Set 2.12 oz. Concrete Pharmacy Flow Control

QUIKRETE® FastSet™ DOT Mix (No. 1244-56; with Corrosion Inhibitor No. 1244-53)

Sakrete 50-lbs Fast-Setting Concrete Mix
Sakrete Cement Mix Accelerator

ctscement dot com
homedepot dot com

Alternative names are:
monocalcium aluminate (CaAl2O4, CaO · Al2O3, or CA in the cement chemist notation)
Calcium aluminate cements
Calcium sulfoaluminate
Hydraulic tetracalcium trialuminate sulfate (C4A3þ)
Aluminous cement
High-alumina cement
Ciment fondu

Calcium aluminate cements
[en.wikipedia.org]

Edited 1 time(s). Last edit at 12/23/2013 12:26AM by A2.

The following idea is published under the LRC licence, it is free for all personal use and no longer patentable (unless it was patented before).

Wet Sock

This idea would make it possible to use a lot simpler mixes as they are constrained while they set slowly and with maximum strength. It will also allow the use of low cement mixes in some applications.

---

Kalle
--
Lahti, Finland
The only stable form of government is Open Source Government. - Kalle Pihlajasaari 2013

The following idea is published under the LRC licence, it is free for all personal use

For commercial use, who owns the license, the concept outlined has been around for at
least 15 years to my knowledge.

---

Random Precision

Quote
johnrpm
The following idea is published under the LRC licence, it is free for all personal use

For commercial use, who owns the license, the concept outlined has been around for at
least 15 years to my knowledge.

JohnRPM, Thanks for the heads-up, any ideas of where there is published information. Obviously prior art belongs to others and their licence model would apply. I would love to read up more and see if any of my ideas are innovative and attribute the prior art correctly, I do not wish to claim any rights that already belong to others.

---

Kalle
--
Lahti, Finland
The only stable form of government is Open Source Government. - Kalle Pihlajasaari 2013

Update?

30cm in 8 hours @32F/0 C

WOW!
Can I send you a CAD file of the Dutch hot tub!

That is just neat! Congratulations!

What are you going to do for a version 2? How big? What do you think you will need to drive it? Let me know if I can help.

---

[www.linistepper.com] Open source stepper motor drivers.

Oh man, oh man, oh man, I cannot stress enough the importance of your work

It's beautiful to see it come this far, and your last pictures are only a promise you are not even halfway, but you will be able to make this work and be a shooting star! At least, I hope so. It's not about becoming famous, but imagine the possibilities! Did you think about scale yet? That bathtub looks like a very good scale to me!

Actually, A2, I will have to build my own bathtub like that one for next summer...

Quote
A2
WOW!
Can I send you a CAD file of the Dutch hot tub!

Not a bad idea, actually. I'm seriously considering printing something like this for an experiment. But I cannot use CAD files for the cement printer yet because Slic3r has made things more complicated for simple design like a box, tub or even wall, and I have to program g-code manually, like 50 years ago.

And thank you for the enthusiasm, Ohmarinus. Printing with cement seemed like a simple idea in the beginning, but to turn it into a reality has been much more complicated than I expected. There will have to be many, many hours/days/years put into working on this idea to turn it into a flawlessly working device. I have a lot of work ahead of me.

Edited 1 time(s). Last edit at 01/02/2014 05:26PM by AndreyR.

Quote
JamesNewton
That is just neat! Congratulations!

What are you going to do for a version 2? How big? What do you think you will need to drive it? Let me know if I can help.

Thank you, James, for your advice to increase the voltage. I increase to 36 Volts and promptly received the results I was looking for. The movements are stable/even and the motors work smoothly. In the next generation of my printer, I’m planning on continuing to use your drivers. I think they have enough power to move the heavy parts if I increase the voltage to 50. I made a simple PSU- just transformer, bridge, and capacitor. Your driver proved to work very well. Under extreme circumstances, the chip may overheat, but can handle a heavier load than is stated in the manual. The one thing I can recommend for those that are planning on using them is don’t waste time on rare plugins, just solder two wires to drivers ”step and direction” and one “ ground” between all drivers and ramps.direct and step
[plus.google.com]



I have not yet been able to use a step motor to rotate the progressive cavity pump in the extruder, and I am temporarily using a DC motor. In the next generation, I will have to somehow use a more powerful step motor. I need at least 500-700W, or even better, 1000-1500W at 200-500 RPM. I found a motor similar to this, but it costs $750. I will try to use SanjayM the given advice about Gecko-Drives and DC motors, basically transforms a DC motor into a stepper motor, but currently don’t have any idea which motor to buy, and I don’t know how to convert horsepower to Watts in this case.

Edited 1 time(s). Last edit at 01/02/2014 07:29PM by AndreyR.

You can probably get even more out of them if you switch to a CPU cooler (with a fan) instead of just a standard heatsink. They really are amazing little drivers.

1 horsepower = 745.69 watts

But it looks like the DC motor works ok for you... why change?

---

[www.linistepper.com] Open source stepper motor drivers.

I'll design you a Dutch hot tub


What is the wall thickness, and the actual layer height?
That will help me design the Dutch hot tub so you don't have to compromise your slicing.
How many person, or would you like the design to be near the size of the original?

About 10 years ago I designed a Dutch tub after the original, but I modified the interior walls.
The walls were curved to put near zero stress on the lower bones of the back.
The numbers came from an engineering ergonomics book, and from my studies of airplane seat designs for cross country flights.
This modification will increase the diameter.

I'll work with you on the design. I'll need to know what some of the limitations of the bot are.
For example, the legs supporting it, might it be easier to add them after the main tub has been formed.

When you're ready let me know, I understand that you are still in the experimental stage.

I'm sure someone reading this thread will know of a slicer workaround so you don't have to hard code thousands of way points.

Edited 1 time(s). Last edit at 01/02/2014 11:56PM by A2.

@AndreyR,

Quote
AndreyR
But I cannot use CAD files for the cement printer yet because Slic3r has made things more complicated
for simple design like a box, tub or even wall, and I have to program g-code manually, like 50 years ago.

Is this what you're looking for?

slic3r auto-calculated extrusion width problem

To place your extruder passes closer together.
Under Filament Settings>Filament>Extusion mulliplier
[forums.reprap.org]

Amazing project!
Simply awesome!

On the curing, why not try heat? Lots of heat.
You could put a heat gun or two blowing at the tip of the nozzle from 2 sides, or 4 at each side, that way you get some heat applied to cement coming up ahead and on the wake of just extruded.

As i understand, cement cures in function of water "leaving" it, ie. that means heat + air circulation is the quickest means to lower the relative humidity of the surrounding air and thus more water is getting out of it faster, ie. faster curing. Is this correct?
Also as heating would be easily controllable
You could also have on each side some IR heating lamps increasing temperature all around your print for faster curing.

This project gave me some ideas for my extra large plastic printer.

Quote
PulsedMedia
Amazing project!
Simply awesome!

On the curing, why not try heat? Lots of heat.
You could put a heat gun or two blowing at the tip of the nozzle from 2 sides, or 4 at each side, that way you get some heat applied to cement coming up ahead and on the wake of just extruded.

As i understand, cement cures in function of water "leaving" it, ie. that means heat + air circulation is the quickest means to lower the relative humidity of the surrounding air and thus more water is getting out of it faster, ie. faster curing. Is this correct?
Also as heating would be easily controllable
You could also have on each side some IR heating lamps increasing temperature all around your print for faster curing.

This project gave me some ideas for my extra large plastic printer.

Cement does not cure from water leaving it, cement can cure underwater. Heat can speed up the rate the cure process happens which might help. A calcium chloride spray on the printed layer may also be an option.

---

WWW.ZATOPA.COM - Your Place for high quality 3D Printing Filament and accessories

Quote
A2
I'll design you a Dutch hot tub

What is the wall thickness, and the actual layer height?
That will help me design the Dutch hot tub so you don't have to compromise your slicing.
How many person, or would you like the design to be near the size of the original?

About 10 years ago I designed a Dutch tub after the original, but I modified the interior walls.
The walls were curved to put near zero stress on the lower bones of the back.
The numbers came from an engineering ergonomics book, and from my studies of airplane seat designs for cross country flights.
This modification will increase the diameter.

I'll work with you on the design. I'll need to know what some of the limitations of the bot are.
For example, the legs supporting it, might it be easier to add them after the main tub has been formed.

When you're ready let me know, I understand that you are still in the experimental stage.

I'm sure someone reading this thread will know of a slicer workaround so you don't have to hard code thousands of way points.

Thank you, A2. I would be happy to print the hot tub and ill post my thoughts later. The major limitation at the moment is the weather; unfortunately, its too cold to print.

Quote
JamesNewton

But it looks like the DC motor works ok for you... why change?

As you can hear in the video, the rotation speed/RPM varies because the cement mix is not the same consistency throughout the entire extrusion and that's the major source for the discrepancy in the layers' width. And as you can see, the extrusion speed need to be reduced on the corners. Thus, a step motor has to be involved. I'm thinking I need about a 1000w step motor, but one that is as compact and light as possible.


here are two videos of the my cement printing experiments

[www.youtube.com]

[www.youtube.com]

Edited 1 time(s). Last edit at 01/06/2014 12:08PM by AndreyR.

Awesome work

you might improve the finish of your walls by taking the slack out of the chains in your axes... depends how much stress your framework can take though

Quote
AndreyR

Quote
JamesNewton
But it looks like the DC motor works ok for you... why change?

As you can hear in the video, the rotation speed/RPM varies because the cement mix is not the same consistency throughout the entire extrusion and that's the major source for the discrepancy in the layers' width. And as you can see, the extrusion speed need to be reduced on the corners. Thus, a step motor has to be involved. I'm thinking I need about a 1000w step motor, but one that is as compact and light as possible.

The regulation of your existing extruder motor should be quite possible with something like my friend Roman's "SuperPID". It's more or less the same as controlling the speed of a cutter on a CNC machine and that is exactly what a PID does. You can change the speed programmatically via PWM or other signals from the controller. That is fully supported by G-Code in general, but I'm not certain it is supported by your current breakout hardware. It certainly isn't supported by Pronterface, so you would have to switch to Mach 3, or EMC2. They should still be able to accept G-Code from Slic3r or whatever, but you may need to pre-process it to change the encoder step codes into spindle speed codes.

---

[www.linistepper.com] Open source stepper motor drivers.

Quote
JamesNewton

Quote
AndreyR

Quote
JamesNewton
But it looks like the DC motor works ok for you... why change?

As you can hear in the video, the rotation speed/RPM varies because the cement mix is not the same consistency throughout the entire extrusion and that's the major source for the discrepancy in the layers' width. And as you can see, the extrusion speed need to be reduced on the corners. Thus, a step motor has to be involved. I'm thinking I need about a 1000w step motor, but one that is as compact and light as possible.

The regulation of your existing extruder motor should be quite possible with something like my friend Roman's "SuperPID". It's more or less the same as controlling the speed of a cutter on a CNC machine and that is exactly what a PID does. You can change the speed programmatically via PWM or other signals from the controller. That is fully supported by G-Code in general, but I'm not certain it is supported by your current breakout hardware. It certainly isn't supported by Pronterface, so you would have to switch to Mach 3, or EMC2. They should still be able to accept G-Code from Slic3r or whatever, but you may need to pre-process it to change the encoder step codes into spindle speed codes.

Thank you, James, as always, you point me in the right direction. Looks like SuperPID is what I need. Now I have a complicated situation. I have too previously thought it would have been a good idea to build the printer with CNC hardware. It's not too difficult to switch hardware, but I have spent around 2 months studying software/firmware for 3D printers, and thus I know nothing about CNC software/hardware/firmware. Looks like linuxCNC is what people generally use? I read about it, but am still confused about how to install it.
Can anyone point me in the right direction with how to install it and which hardware to buy? I'm not a programmer so also any links with a step-by-step toolchain installation process would be helpful.

On a different note, I've had an idea for a different cement printer I can make in the future, and it will need a different/modified firmware. I will have to switch from Cartesian to polar coordinates. Is it possible to modify firmware on LinuxCNC/EMC 2?

Well, the good news is that CNC firm/software and 3D printer firm/software are more alike than different. They look different, but are doing almost exactly the same thing under the hood. There is a firmware package called grbl that will make a RAMP's system into a CNC controller, and EMC2 or Mach3 isn't really that different from Prontrface once you get down to it, they take g-code and send it to the firmware.

I'm not sure EMC2 is even required when working with grbl, you should just be able to use any terminal program, (e.g. putty) to send the g-code to the controller. Done. Since you are already a g-code expert, that should be easy. ,o) There are g-codes to set spindle speed, and a quick google for "grbl pwm" seems to indicate that support for that is in works or complete. Start your research there.

EMC2 / Mach3 is typically run on an old PC with a parallel port and the port pins are directly interfaced to the drivers. The software directly generates the step, direction and PWM commands. no RAMP's / firmware needed at all. You can just hack up a printer cable and ohm out the wires, or use a breakout board (like my 4 Axis board ...
[www.ecomorder.com]
shamless plug.... ,o) If you want one, say the word and I'll send out a kit. Anything to support your excellent work.

Then you apparently just configure EMC2 to send the PWM output to a spare pin on the parallel port. In this example they are going to pin 9, but you should be able to use any output pin that doesn't already have step or direction on it.
[www.linuxcnc.org]


I've worked with Mach3 and TurboCNC and don't have much experience with EMC2, but there are a lot of nice people out there who can help with it.

---

[www.linistepper.com] Open source stepper motor drivers.