Improving the RepRap Printer

Hi,

I have been reading the following link
[printo3d.com]
and watched the video here:
[printo3d.com]

I think we can improve the RepRap to print more finer objects.We have to do the following things:
1.)One is use heavier materials to build the printer so that it will be more stable
(this would increase material cost )
2.)Make the nozzle hole diameter smaller
(this would make the material not to flow freely out,to over come this we may have to use some sort of a pressure system which would push the plastic material out at a constant velocity)

Edited 1 time(s). Last edit at 10/15/2009 12:36AM by bootstrap.

Yes, you could.

But do you want to ? Lots of people in engineering still use CNC plasma cutting over CNC laser cutting. Plasma cutting is rough, wide cuts with inferiour edge finish and taper, but it is FAST and CHEAP and for larger structures in particular it is the best option.

It is the same with RepRap. What do you want to do with it ?

Lampbus Wrote:
-------------------------------------------------------
> Yes, you could.
>
> But do you want to ? Lots of people in engineering
> still use CNC plasma cutting over CNC laser
> cutting. Plasma cutting is rough, wide cuts with
> inferiour edge finish and taper, but it is FAST
> and CHEAP and for larger structures in particular
> it is the best option.
>
> It is the same with RepRap. What do you want to do
> with it ?


That isn't really a valid analogy, because not only are Stratasys machines higher-resolution than RepRap, they also have higher volume output. They can do this because their printhead moves much more rapidly than the RepRap one. I think in the next several generations we'll see RepRap getting much faster and higher-resolution than it currently is as well - we're not really up against any hard technological stops, it's mainly that no one has had a chance to work on that particular hurdle yet.

Kyle Corbitt Wrote:
-------------------------------------------------------
>
> That isn't really a valid analogy, because not
> only are Stratasys machines higher-resolution than
> RepRap, they also have higher volume output. They
> can do this because their printhead moves much
> more rapidly than the RepRap one. I think in the
> next several generations we'll see RepRap getting
> much faster and higher-resolution than it
> currently is as well - we're not really up against
> any hard technological stops, it's mainly that no
> one has had a chance to work on that particular
> hurdle yet.
>
I think that Nophead is well on his way to matching both Stratasys' speed and resolution.

For my own part, I feel that there is a good case for matching resolution, matching speed is merely going to run the cost of building a reprap up. If you use the analogy of an ordinary printer, would you expect the printer sitting beside your PC to have a production level of a printer hooked to a business's LAN that is used to print multiple copies of large documents?

Forrest Higgs Wrote:
-------------------------------------------------------
> Kyle Corbitt Wrote:
> --------------------------------------------------
> -----
> >
> > That isn't really a valid analogy, because not
> > only are Stratasys machines higher-resolution
> than
> > RepRap, they also have higher volume output.
> They
> > can do this because their printhead moves much
> > more rapidly than the RepRap one. I think in
> the
> > next several generations we'll see RepRap
> getting
> > much faster and higher-resolution than it
> > currently is as well - we're not really up
> against
> > any hard technological stops, it's mainly that
> no
> > one has had a chance to work on that particular
> > hurdle yet.
> >
> I think that Nophead is well on his way to
> matching both Stratasys' speed and resolution.
>
> For my own part, I feel that there is a good case
> for matching resolution, matching speed is merely
> going to run the cost of building a reprap up. If
> you use the analogy of an ordinary printer, would
> you expect the printer sitting beside your PC to
> have a production level of a printer hooked to a
> business's LAN that is used to print multiple
> copies of large documents?


It should be possible to increase speed significantly on the current hardware with a bit of clever software and a slightly nicer extruder motor. I think a few months ago I calculated out what percentage of their potential we're running the current steppers at, and it was something very small, like under 10%. The current system probably can't get to the level of a Stratasys, but certainly a substantial improvement on the current status quo can be made.

IIRC, Nophead said that the Stratasys hits something like 30 mm/sec with a 0.3 mm extrusion thread. Right now Darwin is doing something like 6 mm/sec with a 0.8 extrusion thread from what I've read.

I think that Nop is getting something like 15-20 or more out of his rig and I'm able to get a single axis speed of around 12-15 mm/sec out of a Tommelise with my linear steppers which means something like 17-21 mm/sec for a 45 degree diagonal infil road, which is most of a print job.

The problem I see with really high speeds is that you have to do accelerate and decelerate steps to keep from shaking the cartesian positioning system to pieces. When you are slinging around that kind of mass you have to build a heavier, more rigid framework. All of that is expensive and, for me at least, not really indicated when you are trying to print parts for clones of your system.

Another issue is do you stretch the filament when it is laid down.

I wonder what the Stratasys does?

Right now my machine using ABS is set to not stretch or compress the filament.

With that setting it is the extruder that is slowing me down not the robot. Now that might change with a smaller filament size.


-Bruce

I do 0.5mm filament through a 0.5mm hole at 16mm/s. I think Stratasys do a similar volume with a 0.25mm filament so that would require 64mm/s head movement.

The motors on Darwin will easily do that speed but as Forrest says it might shake it to pieces and the belt might jump. I have yet to try it.

Bruce,
Have you tried stretching the filament? Left to its own devices ABS through a 0.5m hole swells to around 0.8mm so if I moved my head at 6mm I would get 0.8mm filament. Simply by moving faster I get finer filament. I can also move slower, at 4mm/s I get 1mm filament for the base layer of my rafts.

I don't think stretching that amount has any downsides. ABS is very compliant and does not try to spring back while it is molten. When you think about it, it has already stretched that much to get though the nozzle so you are only preventing it from swelling.

ABS is much easier to push through the nozzle than PCL and HDPE. Most of the extruder's torque is used by the friction of cutting a thread into the filament. Given a more efficient drive, and perhaps a bigger motor, I don't think it would be hard to get it to come out at 64mm/s though a 0.25mm hole.

---

[www.hydraraptor.blogspot.com]

brucew Wrote:
-------------------------------------------------------
> Another issue is do you stretch the filament when
> it is laid down.
>
I really wonder about this practice. If the filament is really adhering to the previous layer it would seem to me that stretching either wouldn't be possible or would have the possibility of breaking loose the filament being printed from the previous layer.

Its not really stretching in the sense of changing the volume of the plastic, its just persuading it to lie in a long thin filament instead of a short fat one. Molten ABS behaves like a paste with very little elasticity.

---

[www.hydraraptor.blogspot.com]

Thanks for the clarification, Nop!

And it does stick when laid down in a smaller diameter then its extruded size (see I did not say stretch.

I do not know if it is constantly the same size from start to end if lied down this way? So I guess it is off to speed up my machine a little then...

-Bruce

I have just noticed that parts made on a Stratasys have infill that is wider than the outline. That means the head does not need to move as fast as I thought. I revise my estimate to 32mm/s to get the speed and resolution in the same ballpark.

HydraRaptor can actually move that fast, so with a 0.25mm nozzle I reckon I could match their build speed and resolution. I don't see why Darwin can't do the same as well, it has less mass and more powerful motors.

---

[www.hydraraptor.blogspot.com]

nophead Wrote:
-------------------------------------------------------
> I revise my estimate to 32mm/s to get the
> speed and resolution in the same ballpark.
>
Are you presuming that the infil and outlines are being extruded at the same dimensions. It would see to me that you could increase your print productivity considerably with a single 0.25 mm extruder orifice if for infil you slowed down your print speed and upped your extrusion rate. That way you could do outlines at a sedate pace and then fill things at a much higher rate but nowhere near 30 mm/sec. The gearing down that the latest Mk II design has seems to me to suggest that that might be a good idea.

My current setup is outline and infill 0.4mm high by 0.6mm wide from 0.5mm filament at 16mm/s, pi mm^3/s.

I am proposing that outlining layers are about 0.2mm high by 0.3mm wide and infill is 0.2mm by 0.6mm, both done at 32mm/s from 0.25mm filament. During infill the extrusion rate is the same as I have now: pi m^3/s. During outlining it would be half that rate.

Since most of the time is infill that would give only slightly less build rate than I have at the moment. The outline and z resolution would be twice what I have now and the infill would remain the same width.

I could slow down the outlining back to 16mm/s if 32mm/s caused a lack of detail but that would reduce build rate a little more.

---

[www.hydraraptor.blogspot.com]

I think it is my mistake to start the thread,I did not have much knowledge about the difficulty associated in developing the reprap.All I saw was a finished product and not the research involved in it,when I posted the thread.The inventors are the best to help things out regarding this.I sincerely request people to forgive me,if I have mislead them in saying so.Reprap is really a great solution for low cost 3D prototyping.

I request Dr.Adrian Bowyer sir and the inventors to forgive me,I dint have much knowledge to improve it but just dumbly enthusiastic.

Edited 2 time(s). Last edit at 10/13/2008 09:54AM by bootstrap.

bootstrap Wrote:
-------------------------------------------------------
>
> I request Dr.Adrian Bowyer sir and the inventors
> to forgive me,I dint have much knowledge to
> improve it but just dumbly enthusiastic.
>
LOL! Get a grip! There was nothing out of place or inappropriate about the question you posed. It started a great discussion and even being on the core team, I learned quite a bit about what some of my fellow team members know and are doing that I wasn't previously aware of.

Thanks for starting the discussion.