Knurling

[pfleming]

I would like to suggest both diamond and straight knurling be available as modelled features able to be machined or printed, as the current procedure to create a diamond knurl is very complicated and even somewhat trial and error. The advent of 3D printers has meant things like replacement (or in my case original) knobs are very fast and easy to print and can be highly effective rather than just temporary solutions. However there is presently no way to easily knurl the surface such that they replicated replacement machine knurled or moulded knobs.

[pfleming]

I am crossing over from conventional machining to 3D printing and would like to see more mechanical parts and finishes directly available in the modelling environment from drop down menu to be avaliable to all versions of Fusion irrespective of the underlying platform.

A common surface feature in machining is knurling a surface and I would like to suggest this as an option directly available in much the same way that "threads" currently are, providing both diamond and straight knurls. However knurls are used in conventional machining because they are easy to generate and 3D printing isn't subject to this limitation. Hence perhaps this concept could be developped further as a "surface finish" sub-menu that could incorporate other coarse surface finishes that may not necessarily be limited to just knurling but make available other repeating patterns of surface texture?

[roambotics_scott]

You mentioned gear generation in another idea http://forums.autodesk.com/t5/fusion-360-ideastation-request-a/gear-generator/idi-p/5598516 I'd actually argue that both of these and threading should be handled in the same way - not quite modeled but layered on top of geometry so they don't need to store the fully modeled geometry (knurling in particular would have a lot of complexity) and they can be easily changed or removed down the line.

[pfleming]

I'm not sure of the intricacies of the software itself, that's way out of my league, but a surface texture would need to be cabable of being printed, therefore it would need to be not just cosmetic. From a layman's perspective I would agree, it may add considerable complexity to the model, I have no idea, however at the end of the day knurling is, by definition, only small triangles.

[roambotics_scott]

I don't mean a surface texture but something procedural (think; instructions to build the details, not the details, themselves) that is layered on top of the actual design.

Threading also doesn't seem like it should be a huge deal but the modeled threads gum up the works and they seem to make machining more difficult because they're usually cut with an endmill (~ drill bit). Similarly here you'd use a knurling tool rather than cutting the knurls, so you really want the instructions to do that rather than the details.

All of that said / at the end of the day, the end-user shouldn't need to know the difference or care 🙂

[pfleming]

I think we're really referring to different desired features. What I'm referring to are projects that are to be printed as the manufacturing technique and not conventionally machined, so the features need to be part of the model in order to be exported to the slicer for printing. Instruction to a machinist on the other hand can be included as comments on a drawing. It would seem that what you suggest would be of value in rendering, however I am more concerned with the final product.

[roambotics_scott]

Should be doable for both cases. The procedural stuff would be able to output a fully modeled geometry / mesh that you can export and 3D print or even cut if you wanted to, and it could just export the base geometry and instructions (with the added benefit that you can change your mind about the details very late in the modeling without having to muck about in the timeline).

[keqingsong]

This sounds like a great idea to be made into reality with our custom API and scripting functionality. I'd suggest you post this idea in the API section of our forum and see if anyone is able to help. http://forums.autodesk.com/t5/api-and-script

[roambotics_scott]

Totally

https://forums.autodesk.com/t5/ideastation-request-a-feature-or/knurling-and-surface-finishes/idi-p/...

[pfleming]

Yes sorry, this is my second go at this suggestion. I'm hoping for more support this time.

[roambotics_scott]

No reason for you to apologize. Your previous one is «archived» with 12 votes which under the current policy should make it something that'll be in a future version or at least it should be under more serious consideration.

[pfleming]

My idea of a gear generator as a menu item (instead of a script) got 48 votes, was posted about the same time, and I thought would have been very easy to implement given it had already been demonstrated in Alpha. That hasn't been implemented either.

In my opinion there's a strong demand for both the above from the 3D printing side of things, so I do hope they're implemented soon.

[Anonymous]

+1 for knurling as a modeled feature.

[adamvs]

Yes, +1 on this.

[brianrepp]

Merged your previous submission with this one

[promm]

@pfleming,

Thank you for your idea, I am changing this back to Achieved since it automatically change when the two ideas where combined.  We archived this initially because machining a knurling is machined using a specific tool and modeling the knurl is not necessary.  With that said we are looking into having a visual option for a knurl, but this does not meet all the requirements for your idea.  This is why the idea is staying archived.

Regards,

Mike Prom

[pfleming]

Mike I don't want to appear rude, but your post made absolutely no sense at all. Has it been archived or achieved?

From what I can decipher from it, you're saying that this suggestion has been "achieved" because a tool is used to form a knurl, and therefore no modelling is required. That would have to be the most bizarre logic I have heard for a long time. I quite clearly gave an example of a 3D printed part, for example a knob, that requires a knurled surface. No tool is involved and surface modelling is completely pointless. It would seem that the team there have become infatuated with the idea that Rendering is where the future lies, in which case I wish you all the best of luck for that!!! 

On the other hand if what you meant was in fact "achived", ie it's all too hard and we're going to ignore it, then I would question what is so difficult about a simple faceteted surface model? There aren't even that many facets in a knurl! I quite clearly gave an example of a 3D printed knob, that has no tool in the build process. This would be a very common thing in a project or replacement part, and "one click" feature modelling will go a long way to both speed up a process and/or make it more appealing to a broader audience.

I have to say I'm very disappointed with your response.

Regards

[promm]

@pfleming,

I had a typo with Achieved, it should have been Archived.  While difficulty is one of the factors that is considered when evaluation priority for a feature, in this case it did not influence the decision.  Machining a Knurling surface is done with a specific tool and one could argue that it does not need to be modeled.  For 3D printing, you could manually make a cut and then pattern it around the surface creating a knurling affect.  Here is an example of how I used Coil to crate a knurl, you could also create a loft cut and pattern it the same way for a sharper angle.

http://autode.sk/28SRiRl

Cheers,

Mike Prom

[pfleming]

@promm

Ok thanks Mike, clearly my suggestion is falling on deaf ears here, as it would seem you're repeating back what I've said is a shortcoming, and claiming it's a simple process. I very much appreciate the effort you've gone to in putting up a sample of a "knurl" as a process, but I think you'll agree your example looks absolutely nothing like a generated knurl, as the surface texture is not just cosmetic and forms a functional feature. THAT is essentially the issue in my opinion. In order for it to look like a generated knurl, or some sort of close facsimile to it, requires quite a lot of trial and error in my experience, and the result is essentially not parametric when a solution is finally found.

I don't know where AD is going with F360, and it's clearly none of my business anyway. But I had the impression in the earlier days it was aiming, at least in part, at the 3D printer market. If that is indeed the case, then I'd respectfully suggest modelling features like this as a "one click" process is essential. Knurling is a very common surface texture seen in many manufactured products that require some form of finger manipulation, and it's not like this is a request for something obscure and rarely seen.

In regard the difficulty in generating knurls, in terms of the surface modelling, that's beyond my expertise and I honestly have no idea of the demands it may place upon the modelling kernel. My understanding is this is not however a trivial point. Nevertheless the math behind a knurl itself is primary school mathematics, and I'm surprised the "difficulty" in implementing this is even mentioned. If it's possible to manually generate by trial and error something that looks kinda, sorta, roughly ... not really, something like a knurl, then clearly it's also entirely possible to do a much better job by automatically following established mathematic rules and actually virtually generate an accurate knurl.

Obviously it's AD's train set, and the company's to do with this suggestion as they please. However unfortunately I remain completely unmoved from my original position in regard knurling, and am disappointed this has been brushed aside.

[roambotics_scott]

This feels a lot like gears to me in that the math isn't *that* hard and the complexity needn't be *that* high (especially compared with what can be done in games today), and both may seem a bit opaque and nit-picky to someone who wouldn't need them, but they're both kinda a big deal for the people who care.

I'm still baffled about that involuted gear video from years back. That actually looked like the start of something substantial, then there was a long delay, and, eventually, we just got a script with a text based UI that didn't do much.

I'm sure it doesn't fit into the near term roadmap but maybe it'd be good to build a procedural framework where you can easily build things like these with a nice UI without involving the core dev team.

There are probably only a handful of basic operations and selectors needed to do almost anything, and once you have those you can easily build timeline aware modifier layers to cover knurling, arbitrary threading, almost any type of gearing, etc

[lennarthennig]

@promm

Since this is machined with a custom tool as you said, it would be amazing if we could actually specify the knurl for the machinist in the drawing module. 

Please support my idea: https://forums.autodesk.com/t5/fusion-360-ideastation/add-knurling-as-a-surface-for-2d-drawings/idi-...