Questions & concerns with Additive toolpaths

[harry.doldersum]

I have two 3D printers & normally use external slicers (mostly Simplify3D, via menu function "Make - 3D print". 

Looking at the ongoing developments in Fusion 360's Manufacture environment, I was getting interested in using this for slicing?

As both my printers are unsupported as yet, I took the generic set to see how it would work out for me....   An object that would take approx 11h in my normal slicer was stated to take approx. 150 hours using the generic processor.  

So, in order to compare settings & maybe be able to tell where an issue might be, I also downloaded a processor for an Ultimaker 2 (as my smallest printer is pretty close to that model & design). I loaded a CAD model that would run for about 3 1/2h as per Simplify3D's estimate - in the Ultimaker 2's settings, it calculated 6h 23min.  Close to twice the actual time...?

Interestingly, there were warnings here:

- Support speed reduced to max extruder speed

- Sparse infill speed ... (same)

- Solid infill speed ... (same)

- Internal perimeter speed ... (same)

So, that might point somewhere...   But when going through the settings for the Ultimaker 2 & its filament settings, I can't find an entry for max extruder speed? Am I overlooking something or it the max extruder speed a derived value? If so, where do I need to look in order to evaluate?

As said above - at the moment, there aren't any processors for my printers, but I noticed an application was released to help people develop their own processors as required: I'm considering to use that solution. But before I'd embark on such a trial & error journey, especially considering I'd might be risking my large delta printer's integrity, I'd really like to understand what I'm doing...  😎

[harry.doldersum]

Never mind, found it with help of a video by Joel Telling (3D Printing Nerd), where he's introducing the manufacturing environment & going through the settings of a Prusa printer model. 

Based on this, I adjusted some of the parameters to align with my S3D settings: the model now approaches the same printing time as S3D would estimate, so that sounds good.

[harry.doldersum]

Right, progress is being made. 😊

I've now gone & made the 2 configurations for my printers, where

• the machine is based on the AutoDesk generic machine.
• the post processor is created with the Additive Post Configuration Utility.

There will be some testing & tweaking before I'll actually use them, but things seem to be on track so far... 😎

At this moment, there is a point of concern, though?  I'd like to understand if there is development work going on, in order to improve the currently available post processor scripts, as (some of) the current ones aren't giving results that are up to par with their standalone alternatives?

I've put out a result from my 1st attempt and imported that .gcode into Simplify3D's preview: it shows there, that there are some unwanted extrusions in the .gcode. These additional drawn strings (see screenshot below) as visible in this preview screen will actually be printed out - as they're part of the .gcode - even though these shouldn't be there.

For clarification, I've also added a link to a 3D Printing Nerd video of last March, where Joel ran into the same & when actually printing the .gcode, these strings were indeed produced. It of course is nothing that a sharp knife can't solve, but Simplify3D, Cura or most other stand-alone slicers wouldn't do this. 

So, are there ongoing developments to improve the scripts, in order to get rid of these random extrusions?

Test result - random material drawn across the part, during a travel move (?).

https://www.youtube.com/watch?v=QWFRh-MdKxo&t=559s

[thomas.stock]

Hi @harry.doldersum ,

First of all, apologies for the slow response, many of us Autodesker's have been away over this festive period, but most of us are now back.

In response to your first post these warning messages pop up when certain print setting values exceed that of the machine. However, in some of the machines, there are some limits which are not accurate and so this leads to these warnings being generated, hence often slowing down the print speeds and leading to a larger overall print time.
Apologies for this, and these improvements will be mae in a future release.

With regards to you 3rd post, I'm glad to see you're progressing!

Just a quick question about this comment you mentioned:

"At this moment, there is a point of concern, though?  I'd like to understand if there is development work going on, in order to improve the currently available post processor scripts, as (some of) the current ones aren't giving results that are up to par with their standalone alternatives?"

Do you mean the simulated gcode aren't up to the standards of externally generated simulated gcode (like from other slicers such as CURA/Simplify3D etc)?

Sorry to see you have travel moves that are being shown as extrusion moves, I'd be interested to hear if this is what you also see when you print?  - Like in the '3D print Nerd's' video?

Perhaps you could share your .f3d filefor this project (if you are comfortable with this) and I can have a look?

If this is still an ongoing issue I apologise and we will work on a fix.

By the way, what version of Fusion are you running?

I attach an image showing how you can find it: 

With regards to ongoing development, I can confirm that there is continued ongoing development across all aspects of Fusion FFF, and we will work to improve any issues our customers find.

Thanks for bringing the above to our attention.

Thanks,

Tom - Fusion Additive Team

[harry.doldersum]

Hi @thomas.stock , 

Thanks for your reply, much appreciated.

My Fusions 360 version is: 2.0.9512

I've attached the F3D archive - if that part of the Fusion 360 model file is included, this would be a version with the standard Ultimaker printer settings. (I copied the model & added the Ultimaker settings to it, in order to look if it also had the same extrusion issues as the versions I generated for my own printers).

Also attached, is a ZIP archive with the mentioned Ultimaker .gcode for this model. When importing this .gcode into S3D, it shows the mentioned extrusion across the model. I've also attached a screenshot, where I have focussed on the actual layer where the issue occurs & you can see how the extrusion pattern travels across to the other side in order to fill a patch in and then travel back to proceed where it had left off. 

As my own generated processors are yet untested & I don't have an Ultimaker printer myself, I haven't printed this out using the Fusion 360 generated .gcode (I haven't gotten around to fully precheck the generated code and I'd really prefer to make sure, that the start- & endcode is safe & I don't end up instructing my printers to self-destruct). 😁 

However, I can assure you, that if the .gcode's preview in S3D is showing an extrusion, it will be produced - I have no doubt there.

Please let me know your findings, when reviewing the attached - I'd appreciate to understand what's happening here & why. The weird thing is, the rogue extrusion is only occurring once, in a single layer. Again, it's nothing that a sharp knife can't solve for 3D printing, but why is it doing this? Are other unexpected travel moves possible? Could this then lead to collisions?  These are questions that I need answered, before I can put this code forward to a 3D printer. 

Thanks for looking into it, I appreciate it. 

[harry.doldersum]

Hi @thomas.stock 

I was curious, if there has been any progress on the above thread during the last year?

Thx! 

[thomas.stock · Accepted solution]

Hi @harry.doldersum ,

Apologies for the delayed response!

I tried your Fusion Project with the latest post and print settings and could no longer see that issue - see atatched image from Simplify below:

Many updates/improvements have been made to FFF since (especially this March release), so it seems the issue has been resolved.

I attach the updated Fusion Project and the gcode file:

[harry.doldersum]

Hi @thomas.stock ,

That's excellent news, much appreciated. The simulation looks great in S3D.

I'll recreate the posts for my machines using the most recent libraries, then - to make sure I've got all the latest stuff onboard - and will try again. 

Thanks!

[thomas.stock]

@harry.doldersum , Great!

Let me know if any issues.