A wobble in tall, thin-walled prints

#1
My art involves designing lots of objects for homes. I play with light and textures, and I print at architectural scales. Though I may prototype at “standard” scale, the gMax is my tool of choice (and only option) for printing at full scale with fit, finish and polish.

Kyle and I worked through an issue related to a wobble that showed itself in a decanter-like test object that I used during my initial shakeout of the gMax. In particular, I discovered that the gantry belt hadn’t been sufficiently tightened during assembly. But the wobble disappeared after I tightened it. But in the course of running other tall, wide and thin-walled geometries the wobble returned. I want to share my thoughts on what I think is causing it.

I ran several tests when the printer would otherwise be powered off. A photo of one of these tests is below. As can be seen, the wobble doesn’t exhibit at the very bottom of a print. This test object tapers outwardly and then back in at the top. I made sure the gantry and bed belts were both tight before starting that job. I think the wobble on this print is particularly bad because the job was laid out using S3D’s Vase Mode. So the walls are VERY thin.

So if this isn't caused by slop in the belts, what's causing it? Does it show itself because the circumference is greater or because printing is happening far above the bed?

My take is that the root cause is distortion of the print. This distortion is caused by the pull on the object that results from the viscosity of the filament being extruded. The tall, (very) thin walled object is being pulled out of true. So the newest layer settles and cools on the layer below it slightly off center from where it should be. As successive layers are also laid down off-center, the harmonic begins to swing in the other direction because the lower layers aren’t where they should be. So the system presents itself as a series of wobbles.

Some more evidence worth noting is that a single shell vase cylinder having a 10mm radius and 400mm height that I printed did not develop a wobble. This tells me that it’s not being caused by a loose belt or a frame that has too much flex. The straight wall and pure, circular profile provided enough object rigidity that each layer was laid down directly on top of the one below.

I wish my tall, thin prints weren’t coming out with a wobble. And I’ll continue to experiment looking for a solution. (Higher extruder temps?) But I don’t believe it’s the printer itself. In fact, the surface quality of that 400mm cylinder is very good all the way up. I’m impressed.

- Bob
 
#2
Bob.. are you using the standard filament spool holder? If so, try another method... I found the stock holder was causing what looked like Z wobble in my prints... Once I did something else (hatchbox spool holder) those issues went away... Others have seen that happen as well...
 
#3
I'll try Hatchbox's feeder mechanism, thanks, that said, why would the feed mechanism affect the tapered shape but not the more narrow, straight walled and equally tall cylinder shape?

I'll report back with results.

- Bob
 
#5
its just the spool holder... its smoother coming off the spool than the stock one....
https://www.amazon.com/HATCHBOX-Spool-P ... hbox+spool

Bob.., i dont know why things are affected sometimes and not others, but im guessing it has to do with the fact that sometimes when it moves LEFT all the way, it pulls a certain amount of filament out, then when it works back center it goes slack a bit and the wobble is lessened, then again to the outside... of course if its pulling off lets say 10mm of filament per rotation, that wobble might only show up every 5 layers because it takes that long to get through the 10mm of buffer filament it pulled out.
Ive noticed this issue in a LOT of printers. The key is finding a spool holder or a feed mechanism that doesnt take an overwhelmingly high amount of force to pull the filament off, which is going to cause the wobble lines.

ill try to print a vase later. shoot me the STL or a link to that one, and ill try that print exactly
 
#6
The wobble never shows up in the lower layers of my prints. So I can appreciate your theory that other printers would perhaps exhibit the same wobble if they were able to print taller objects. Again, that still doesn't explain why the 20mm cylinder doesn't have a wobble. (I've enclosed a shot of the cylinder below.)

I appreciate your trying the vase mode test object on your rig. Note though that I'm able to print the shape with 2 shells without a wobble. (Photo of it is also below.) What fails due to wobbling is the same object printed in S3D Vase Mode.

Warning -- printing even in Vase Mode a big job. But I guess all gMax jobs are big. The vase mode print will take 17 hours.

I've enclosed a S3D factory file below. I'm eager to see the results from your setup.

- Bob

Krause gMax Test Vase.factory
 
#9
Mt head is down this week doing initial design on a new project. So my gMax would otherwise be idle. I've been using this opportunity to print test objects that clarify the cause(s) of the wobble I'm seeing in tall, very thinly-walled objects that I'm printing.

I've created a derivation of my "decanter" shape that has a wave in the profile that I'm printing using S3D's Vase Mode. So it's very thin walled. Though it's still a 24 hour job, I'm far enough along to see that a wobble's not developing.

I'm using the same spool farm arrangement that I've used in the other jobs -- both those that have developed a wobble and those that haven't.

All the evidence that I've seen so far supports my sense that it's the degree to which the already printed layers of an object are susceptible to flex that determines how much wobble I see in tall, thin-walled objects that I'm printing.

The factory file for this wave test print is here.

- Bob
 
#11
The exercise today gave further confirmation of my sense of why tall, thin-walled objects have a wobble. Viscosity. The challenge is working around it. Sometimes I can print more shells. Sometimes I can modify the design like I did in adding the waves so that the object is more rigid. I may even be able to raise the extruder temperature so that the material is less viscous. But sometimes I can't do any of this. It'll be an ongoing challenge, I'm sure.

- Bob
 
#14
For vase mode they wont matter, its all pretty stock temps, fans, layer height etc... pretty close to what I run.
no biggie.
thanks --- its off and running
 
#15
OK... so it happened to me too... could be it be just the perfect angle of vase that the model nearly needs support, but can print without it, or something..
Ive never seen that happen ever :)

 
#16
Yup. Viscosity.

When I first saw it last month I was concerned that it was caused by a "sail effect" -- distortion due to the the object being pushed by the air as it moves forward and back on the bed. But it's not that. Well, at least that's not the primary cause. What we're seeing is the same physics as we would find on every other printer -- if these other printers enabled us to print such tall and wide shapes with very thin walls.

Thanks for running the job on your rig to confirm that it's not simply some condition in my lab.

- Bob