I recently saw this video by Teaching Tech https://m.youtube.com/watch?v=cbhWni9f980&t=325s&pp=ygUJbm8gcGxhbmFy and I noticed that with the part fan removed the a1 has a lot of clearance from the nozzle tip to the print head
I’m familiar with that video. If you follow the link he posted, it’s cutting-edge math and science. The researcher who published the paper is way ahead of the current state of the art.
Let’s be candid: current 3D printing is really 2.5D printing, functioning like a hot glue gun laying down material layers. This method is straightforward and leverages X-Y motion technology that’s been around since the '60s. GCode itself was developed at MIT in the late '50s specifically for CNC machining.
It took 30 years for that technology to be incorporated into FDM printing. It’s conceivable we won’t see Non-Planar printing for another 10-20 years unless a simplified GCode-like language is created or a strong business use-case drives demand. Until then, it’s just an interesting video, and some Swiss grad student got a great PhD out of it. 
I may seem jaded, but I’ve seen too many “great ideas” hit the market and fizzle. That’s where I got the phrase “A technology looking for a problem to solve.” A smart person identifies an obscure problem and tries to create a market for their idea, hoping others will figure out how to use it. While Non-Planar printing is cool, it doesn’t address more pressing issues with FDM, like drying filament on the fly or detecting layer separation. To those scientists, I’d say, “Fix those issues before chasing your pet project.”
Since my memory goes back to the '70s, I asked ChatGPT to list technologies that made a splash and fizzled within a decade. Here’s what it came up with. I remember all of these and where I was when I first saw them. Google Glass, in particular, gave me a chuckle. When Apple released Vision Pro last year, I noted they were ignoring Google’s lessons. Well… here we are a year later, and that technology is still looking for a problem to solve.
| Technology | Principal Sponsor | Years Active |
|---|---|---|
| Zip Drives | Iomega | 1994–2000s |
| MiniDisc | Sony | 1992–early 2000s |
| Netbooks | Various Manufacturers | 2007–2012 |
| HD DVD | Toshiba | 2006–2008 |
| Windows Vista | Microsoft | 2007–2009 |
| Microsoft Zune | Microsoft | 2006–2011 |
| Google Glass | 2013–2015 | |
| Google Wave | 2009–2010 | |
| Floptical Drives | Insite Peripherals | 1991–mid-1990s |
I do hope it comes to market in the near future. It has massive potential. Especially where aerodynamics are key. But also for durability and looks.
It would considerably level up FDM printing for industry prototyping and hobbyists.
About the ChatGPT examples I couldn’t disagree more as they are only technologies/products that have been superseded by more advanced tech or an alternative approach. It’s not like they were completely useless.
In this case it’s a cutting edge development of an existing technology with no real alternative to my knowledge. Might be superseded after. But in any other case is better than the 2.5D FDM printing we have now.
Orca Slicer has a bit of interest on the GitHub but no idea if anybody is actively doing anything yet [Feature Request] Non-planar slicing · Issue #5053 · SoftFever/OrcaSlicer
I know this is a bit of a necro topic, and the bambu print head isn’t very well suited, but I wonder about non-planar ironing. It would require some fancy dynamic ironing flow rates, but I do think it is achievable as a partial implementation provided the top layer is within the clearance of the nozzle to bottom of print head (lidar etc).
Can anyone confirm that their motion controller supports G01/G03 in the X-Z or Y-Z axes? I know in CnC machining, some controllers support simultaneous 3 axis moves. Aside from segmenting the code into many steps, it would be nice if the feedrate could also be mapped in a curve, but I’ve never seen that on a traditional subtractive machine.
I am almost sure that all recent FDM controllers support simultaneous XYZ motion in G1. Klipper supports it for sure. I am almost 100% sure that the feed rate is measured along the path.
Big limitation is the feed rate in Z for the older machines.
G2/G3 motions in G18/G19 planes is easily tested.
However, if you want non planar printing you better keep the motions in G1, unless you consider splines in the tool path.
But that isn’t even common in the most recent CNC controllers for professional machine tools.
Additionally, modern controllers can calculate smooth spline paths through a segmented G1 tool path.
Handling feed rates is also easier then.
To my opinion non-planar printing is very limited when the tool head stays upright. A 5-axis FDM printer would be perfect for this.
In 5-axis however, the feed rate is more complicated due to the rotary motions.
Also keeping track of the nozzle tip location requires a lot of math if the controller needs to solve that.
All in all I think it will come in time towards the hobby platform.
How much time? The future will tell.
