CoreXY or Moving Bridge with fixed bed which way to go for large parts

Hi,

I am preaparing to build a 3D printer for large industrial parts where the strenght of the part and ability to produce a block faster (not weeks but days or hours) more important than the surface finish. So I choose to design for 1mm nozzle diameter and 0.8mm layer thickness. I aim to print at a velocity of 300mm/s upto 600mm/s. So I should be able to construct a powerfull extruder (possibly not a bowden type but a direct one). Therefore the direct extruder should be capable of producing upto 500 mm3/s of plastifying capacity.

According to my calculations. I must use a Nema42 stepper which weighs approx 4kg. It has 870Ncm. Possible gear ratio will be 1:1 or 2:1

Build dimensions: width =500mm length= 1500mm height=400. I need to produce long but narrow parts. (parts like machine chasis)

1- I wish a fixed bed . But Delta type is not suitable because of unsymmetrical part dimensions.

2- Since I need heavy direct extrusion , I think I can not use belts. I want to use direct ball screws and linear bearings , so that I can go faster with the heavy load.

I am considering a corexy design but the moving build table is a problem. Because I want to be able to scale it in future . Ex: 1mx1mx2m

The main question that I feel hard to answer is: What if I use a fixed bed and use a moving gantry just like this one : [3dplatform.com]


I know that the best construction would be something like : [bigrep.com] . But I believe the bigrep topology moves the whole 3 axis x,y,z over a fixed bed. It is maybe the most robust one but maybe hard to build for larger volumes . In my case I think this design is not scalable (I believe I can not realize more than 1m). For my parts where I have a long length (1.2m) and relative narrow width 0.4m. the bigrep is not practical to build.

I believe the downside of 3dplatform is that I may need powerfull servos to move the gantry when the speed goes over 200mm/s. (I have several nema42 steppers in stock, SI I want to use them.). I think this topology (3dplatform) is the most basic one to build.

What do you think? Which way should I go?. If I suceed I may also need 5mm nozzle and 3mm layer with a longer x axis configuration in the future. I tend to 3dplatform with some modifications But please give your comments/recommendations.

Best wishes.

Note: Important detail for my parts. My parts have a square shape, most of them are long and does not have very small direction changes in short distances. therefore I may need a lower accelaration but higher speeds. this may help to compensate a little bigger heavier moving column.

Edited 2 time(s). Last edit at 01/31/2016 02:04PM by rxpu99.

I think the 3dplatform one cannot or hard to afford for high speed print like 300mm/s. You can imaging that you are moving a gantry by hand in high speed where compare with moving a bar. Moving a bar will definitely be faster than the gantry one. Therefore, my vote will go on BigRep.

By the way, I am not sure whether BigRep is using CoreXY system or not. CoreXY system need to use timing belts to pull the carriage in XY axis.

Edited 1 time(s). Last edit at 02/02/2016 05:37AM by bunbun.

Printing at 300-600mm/s, up to 500 mm3/s... The only hotend that maybe could achieve this is the E3D vulcano, what hotend you will use?

To achieve a printing speed above 300 mm/s . I also noticed that the moving bridge idea is bad. I produces so much inertia. Even with powerfull servo i dont want to run agaist wind.

I changed the system to bigrep.com style. The user fASE-2 in this forum had a good idea. He used a linnmon table as the base. I know it is risky but , I will try to build a prototype with such a table. My aim is to reduce the need for a very strong external chasis. I try to implement 4 plastic guides that can stick and unstick to the legs with the help of a pnematic actuator.

See the attached picture of my prototype: there are 4 long and tight bars and guides along these bars. In the middle you see a platform to build the corexy with a extruder which i plan to use nema23 motor and 2 direct gears to increase accuracy (both sides of the filament will contact seperate gears, and they grip very tight to achieve low backlash) (see picture).

The 4 guide is PE (polyethylene) very slippery material. PE is very tight so tight only 4 leadscrews can move it. In addition to that I plan pneomatics add on to make the guides much tighter when the corexy is busy on the layer. When corexy finishes one layer, it sends a confirmation to move the Z. after this confirmation the 4 guides will be less tight for a couıple of ms to move upwards. Less tight means: They are not locked but unlocked to a less tight state where the leadscrews can move them.

through this structre I plan to acheieve higher speeds. What are your opinions?

The main question is: Is it better that I build the corexy on top (fixed) and move the bed as described above or move all corexy+Z togethar with lock mechanic. which one is better?

What are your opinios abaout this lock mechanic?

Delta printers do not make asymmetric parts, unless they are calibrated badly. If well-done, a delta printer can move its effector to any point, just like a CoreXY or gantry setup.

Should probably system engineer this project.

Start with requirements (needs, wants, nice to have)
rate the requirements to see what are drivers
Include cost, schedule and technology requirements
(is the technology available to do what I want?)
(can you feed filament fast enough, or will you need a pellet hopper and screw extruder?)
mechanical, thermal, electrical, electronic, software, firmware

Development
Build yourself -- contract out
Do you have skills and tooling to build
Schedule --- lead time to get parts.

Then that may lead you to type of printer that would meet requirements

Starting out with design --saying CoreXY with spinning raminfratzer on diagonal
-- stage for frustration and failure

Digital_Dentist a real sage in this area!

Quote
bunbun
I think the 3dplatform one cannot or hard to afford for high speed print like 300mm/s.

A Flex3Drive from Mutley3D is run on a routine basis at 350-400mm/sec and produces consistently high-quality prints. you can see the videos on the website.

there's a guy who made a concrete 3d printer (yes, really) - it extruded concrete and was therefore a moving x,y,z system. if you choose to go with filament (following cozmicray's advice first and doing the systems-level analysis in advance) i would suggest looking at a Flex3Drive in combination with a Volcano E3Dv6 hot-end, those come with 0.6, 0.8, 1.0 and 1.2mm nozzles. the Flex3Drive will give you a 40:1 gear ratio using a worm-drive, and the advantages of a direct-mounted extruder, so you get both speed and control.