Requesting Opinions on Mechanics on my Next 3D Printer Build

This will be my second 3D printer. My first printer was a repstrapish machine that prints poorly when it prints and has many underlying design failures. It was made with a combination of a hand drill, a Dremel, and a lot of time.

For my second machine, I am looking for all of the following qualities:
- No / Limited Smell
- Heated Build Chamber
- Quiet
- Working Heated Bed
- No Inductive Sensor Required
- WAY Overbuilt frame
- Linear Rail Motion
- Motors separate from the chamber
- Enclosed Electronics
- Part Cooling Fan(s)
- Easier to transport
- 12" x 12" x 12" (300mm x 300mm x 300mm)
- Better Extruder and Hot end
- Decent Electronics
- Built in PC (raspberry pi or similar)
- Removable Bed
- Connected Lead Screws for Z axis
- Nice Interface
- More integrated Filament Holder
- Looks sorta cool
- Filament runout sensor

And so with my short list of criteria, I started designing my printer in CAD. I use Onshape.

Frame:
For the frame, I have some 15 series black anodized t slot and I plan to arrange it like so:

[gyazo.com]
In most of the corners will be aluminum angle brackets along with tapped end screws in all the connections to ensure strength. I plan to cut the extrusions to length using a chop saw and then carefully grinding the ends with a grinding disc attached similar to a saw blade. The general concept with the frame is to make it sturdy and well overbuilt so that any upgrades down the road are not limited by the frame.

XY Carriage:
For the movement system, I want to do an H-Bot system with linear rails for both axes. I understand the issue with H-bot and the twisting that can occur but I think that with the linear rails and the aluminum support bar the twisting force would be entirely negated quite easily.

[gyazo.com]
For Homing, I plan to have the Y axis home first to the modeled end stop in the picture above, then have the X axis home to an end stop on the same corner except that it would be under the extrusion with the linear rail on it.
Tensioning would be taken care of by the pulleys across from the motors with a simple bolt and nyloc nut.
For the betting, I plan to use something name brand in the homes of preventing backlash and ringing caused by poor quality belts.


Extruder:
The Extruder system will be direct drive using the Titan Aero that is not modeled beyond the motor mounting position in yellow. The Aero will have a simple blower fan attached to one side with a nozzle that spreads the air around the hot end. I am not entirely sure how I will do the wiring yet for the extruder carriage and so any help is appriciated. Should I do a drag chain, long flexible ribbon cable, is there a better solution? I would love some help with this.

Z-Axis:
For the Z axis, I plan to use two 400mm lead screws that are both driven by a single motor. The belt path for the leadscrews would be a V with each line of the V being a separate belt loop that is both connected to the same motor rotor at different heights. The bed will be supported with 400mm linear rails. For homing, I plan to use a bracket on the frame that can be moved up and down and a screw placed in one of the supporting aluminum extrusions for the bed. For heating the bed I will use the same AC heater as my last build but will add a higher amperage SSR so that it is safe to use or make a heatsink for it. As for making the prints removable, I plan to use a printnZ plate that is clamped to the aluminum bed with large C clamps rather than binder clips.

[gyazo.com]

Foam Insulation and Side Pannels:
1.5-inch thick foam insulation as detailed in the CAD model in pink with a reflective coating designed to keep as much of the heat from the bed in as possible.


[gyazo.com]
[gyazo.com]
The exterior shell will be made out of 1/4" Plywood. The back area is for housing the electronics and has a simple door that will have some kind of latch.

Final Pictures:


[gyazo.com]
[gyazo.com]
And... It ended up looking like a cardboard box for books.

What are your thoughts on the mechanics of my design? Do all the mechanical design choices seem solid?
Thanks for your opinion!

Edit: added links under the pictures and the .step file of the CAD model > [cad.onshape.com]

Edited 1 time(s). Last edit at 06/16/2017 11:04AM by Trexation.

Reserved for Updates

Edited 1 time(s). Last edit at 06/15/2017 11:13PM by Trexation.

To bad we can't see the images....
I'd say the 15mm extrusions are too weak for a printer that size. Maybe you have some special frame design in mind with lots off cross braces?
I also prefer MDF instead of plywood, which tends to warp over time.

Edited 2 time(s). Last edit at 06/16/2017 01:30AM by o_lampe.

I think 1.5 inch thick insulation will be far too much. The chamber will be in danger of overheating because too much heat from the bed will be trapped. I haven't built an enclosed printer yet, but when I was printing ABS I found that 2 plastic bags over the printer was sufficient for the bed heater to keep the chamber warm enough to prevent warping. So I am considering enclosing one of my printers using double or possibly triple skin clear plastic sheet.

---

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].

The Extrusions will be 1.5 inches thick (38.4mm) and the plywood is just meant to keep everything enclosed, not to provide structure to the printer.

Edited 1 time(s). Last edit at 06/16/2017 11:06AM by Trexation.

Wouldn't the chamber just reach the temperature of the bed and then stop heating? What would be the issue with that? I think that I could set my bed to 110C for the first few layers and then just drop it down to like 90C and print in an 85C chamber...

Quote
Trexation
Wouldn't the chamber just reach the temperature of the bed and then stop heating?

Yes. But you probably want the chamber temperature to be lower than the bed temperature.

---

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].

Quote
dc42

Quote
Trexation
Wouldn't the chamber just reach the temperature of the bed and then stop heating?

Yes. But you probably want the chamber temperature to be lower than the bed temperature.

Well, I don't think that is a problem since all the motors except the extruder will be outside of the chamber, what would be the other issues? Have you looked at the pictures yet?

Why do you use two belts for the leadscrews? One long belt routed as triangle is easier to tighten.

Heating up the frame and linear rails too much can cause binding from different temperature expansion of aluminum and steel ( think bimetall )
You'd need some clever mounting for the rails, so one end is fixed and the other end can expand, but is fixed in 2 degrees of freedom.

The reasoning behind using two belts is because it is hard to find loops of belts that are over 1 meter. In my design, this is solved by having two belts. Also, the two belt system gives the pulleys on both the motor and the lead screw excellent grip because the will have 60% + of the teeth engaged as compared to 30%.

With the thermal expansion, the steel will expand .324mm over its 450mm length, while the aluminum will expand .623mm, a .3mm difference that will result in a stretching of the linear rails so I am not very worried. Should this make me worried? The temperature change that I used was from 30C to 90C with an LEC of .000012 of steel and .000023 for aluminum.

Expanding aluminum won't "stretch" steel rails. If the rails are linear guides screwed to the aluminum there may be some shifting of the two against each other if the screws allow it. If the two materials were locked tightly together the whole thing would curl with the aluminum taking the outer radius. If the rails are end supported, they will probably slide in their mounts. The real question is will it all remain in alignment after multiple heating and cooling cycles?

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

How would I be able to calculate the curling that would take place? I tried to use the formula on the Wikipedia page for bimetal curvature but could not figure out what the K value for curvature meant. I got a value of -.01447 for K.

Thanks

Very high temperature heated chambers haven't been well explored by hobbyists. You can look to stratasys for guidance, but they tend to go for the highest quality/most expensive solution everywhere so its hard to figure out exactly what is necessary.

I'd be very hesitatnt to place the extruder in the heated chamber. Stratasys either seperates the extruder from the heated chamber with insulating bellows, or puts the extruder in a mostly airtight box and pumps in cold air to keep the motor/heatsink at ambient. I don't think you will be able to print with the heatbreak at 80c.

Stratasys machines also have air circulation in the heated chambers to prevent a temperature gradient from forming. I'm not sure how important this is. I'd personally include a fan cutout to introduce cold air into the chamber so if you find out that the chamber is getting too hot you have options.

I'd recommend studying the Stratasys designs and patents because they use some pretty neat tricks to keep almost everything except the bed and print cold.



Properly mounted rails should not be able to slide and will therefore bend when heated. I did a set of coarse deflection calculations and it is likely in the range of 0.05-0.3mm which is managable.

If your fasteners are tight the deflection will be very repeatable so if you wait for the printer temperature to stabilize you can run grid leveling, or you can assemble the printer hot so that it is flat when running and "bent" when cold. The problem is that the printer could take an hour or more to reach equilibrium temperature. You could also play games such as only fixing the middle section of the rail and leaving the ends loose.

If I were building this printer I would assemble a rail/extrusion unit cold with all the screws and measure deflection when its hot by laying a straightedge on the rail and poking a feeler gauge in the center. If deflection is really bad I'd assemble at 40c and omit one screw at each end to shorten the bimetalic strip but I don't expect that to be necessary.