Scaling Mendel

I am making RC airplanes, planning on making a mendel to self print various parts, some of which will not fit into a standard prusa, specifically wings,
according to my plans some of the wings are as long as 1 meter.
my question is: did anyone try to make such large objects with ABS/PLA and had anyone made an enlarged copy of a mendel?

Edited 1 time(s). Last edit at 01/25/2012 07:18PM by artemen.

No one would try and carve an entire wing out of balsa, and you would probably not want to print an entire wing for the same reasons, but I can certainly see the attraction.

The Mendel design can certainly scale, but it is not the best design for scaling. Look just at the y axis. The build platform is about 200mm deep. In order to address the entire print bed, the machine frame has to be about 300 mm, and the smooth rod diameter about 8mm.

If you now wanted to scale that up to 1000 mm...

Let me just briefly say that the Y-axis is not the ideal axis to scale, and I am aware of that. Now back to the program.

... you will need a 1500 mm frame, and the smooth rod will need to scale as well, because an M8 1500 mm rod is going to bENd a lot. Let's assume that we are going to not go to use an M40 rod, but rather use some kind of fancy rail product that has structural support. Now it's the price that is growing exponentially, and the design starts looking less like a Mendel. And there you have your answer. If you print a 1 meter wing on a Mendel, it's not a Mendel anymore.

My motto? Work with what you have.

Wouldn't you need a 1100 frame?

Nopheads new design would probably work given its use of MDF. Although you would need at least 16mm rods.

Would rotating the object on the bed give you enough room for most of the parts? That darn geometry and triangles.
It is also possible to join shorter printed items. ABS is 'glued' with acetone, which is avail quite readily.

Without a heated chamber, any single part over 200mm be it ABS or PLA will have bad warping issues.

Scaling a machine to 1meter that still makes good quality parts at a reasonable speed (even ignoring any warping issues)is a HUGE challenge.

Just making a flat platform over that distance is tough.

Superglue is a friend of ABS so I would highly suggest just gluing smaller parts together.

---

www.Fablicator.com

I am starting to think that maybe we should sticky some of these commonly asked questions to the top.

We've all thought about the potential for a large scale printer, and we've either posed this question ourselves or read a thread like this one. When your just starting out it just seems logical, just build it bigger, and you start to wonder why nobody is printing go-carts or couches or whatever. So I'm not surprised to see this question over and over. However, can't we just sticky one, so that we don't get burnt out answering the same questions again and again?

With a rail line type rail and using a carriage with bearings on 3 sides similar to a Sells Mendel scaling the X axis is possible and redesigning the Y carriages to something similar would work too, now where to source some small (20mm) rail line

---

__________________________________________________________________________
Experimenting in 3D in New Zealand

One big problem making large things would be the cost of the plastic filament. I think you would need to switch to a granule extruder to make it practical.

---

[www.hydraraptor.blogspot.com]

A long thin machine would be much easier to engineer, say 200 x 1000. You could use linear rails for the long axis with open bearings. This way you could support the axis mid way.

Theres an interesting article on rapid prototyping model aircraft that might be of interest. I have already suggested to them on their forum that the reprap project might be of interest to keep the "Men in Sheds" approach that their first plane had (The PARIS paper plane video is on the right of the page).

LOHAN

They are using ABS powder to rapid prototype their aircraft, but there is a call for a more easily reproducable home method of production. This could be the reprap.

There are a few people here that reduce the temperature of the heated bed after a few layers, has anyone tried cooling the part to room temperature after a few layers, and using clamps to hold the partially finished part to the bed.

Would this reduce warping?

Edited 1 time(s). Last edit at 01/26/2012 06:43AM by martinprice2004.

That would increase warping. Warping happens as layers contract relative to each other after becoming solid, if you allow the part to cool the new layer is solidifying in contact with an already cooled and contracted layer, if you keep the part as hot as possible, the new layers are cooling onto expanded layers and relative contraction is much smaller. That was the whole reason that heated beds and heated chambers were developed in the first place.

Just a few more thoughts.

A wing section would probably mean that it would need a lot of support material if made horizontally as both sides are usually curved. Perhaps a wing would be better made vertically. Less warpage and perhaps it could be supported at intervals. A central carbon fibre tube could be glued into the finished part down through a vertical hole. The fuselage could be made in a similar vertical way, but in two large pieces joined at the centre.

@AndrewSmith
Yes that is a sound theory, so the part contracts uniformly at the end of printing, but a cooled bottom layer resists warping better than a softer warmed base layer, so probably wouldn't lift as easily. Continuing your thought, you would probably get the part splitting half way up (I sometimes get this on taller thin walled parts in ABS). Has there been any actual trials to confirm that this would happen?

Perhaps a part could be produced on a hot bed in separate pyramidal sections, then filled in at the end of printing. It would require more complex g code to avoid clashing with the part, but it might be feasible.

I did a few trials in my early days designing parts with a cross cut base layer which is a similar sort of theory and it did seem to reduce warping.

larger bed sizes will need a chamber to prevent curling of plastic. heated beds can only do so much.

here's another thread with more answers: Large-scale 3d printer


It will take so much effort to build a larger FDM machine that you might as well spend the effort on a design that can potentially scale much better. I think a UV resin printer like lemon curry would scale very easily. Just make a longer z axis and a larger resin pool, and move your projector further back. you'd have a loss of resolution because the pixel size would be larger, but the print time would be the same no matter what x-ysize part you want to print. (print time would obviously increase with larger Z) the biggest problem would be the weight of plastic as you scale. at some point, the weight will be too great to keep it attached to the z platform.

Of course UV resin is very expensive, but you might save money going this route than trying to scale up FDM.

Edited 1 time(s). Last edit at 01/26/2012 11:01AM by Buback.