Dog Bone Filets - For CNC Joinery

[jamie]

Would it be possible to add a function for dog bone filets, either in the Model or CAM workspace?

I used to use another CAM package that allowed you to select different interior corner types when creating a profile cut tool path. I could select "bisect line" then have the tool automatically "push" into an interior corner by a set amount.

This would save hours of design time creating these filets manually.

[ToddHarris7556]

I'd vote to add this on the CAM side - not as a modeling feature. The actual geometry required is dependent on the tooling used. i.e. if a corner is created with 1/4" dogbones, then the tool is changed to 1/8" cutter, the geometry is overkill. 

A simple 'Corner type' add to the CAM feature would allow the toolpath to vary based on tooling. 

[Anonymous]

If I am not mistaken,

You can do this by writing script in the API.

[AdamEllison]

This is a tremendously useful idea to extend the power of the fillet tool.

It's useful for much more than just cutter clearances but for stress relief as well. Any concave sharp corner in a stressed part should have a fillet of some kind and this geometry is perfect when you have a sharp convex mating interface.

[michael]

I second the vote for CAM side, modeling it would be nice for aesthetics as well but less functional overall, especially when different manufacturing processes may be used (distributed product design, or open-source models), e.g if I design for a CNC machine and someone else has a laser cutter, they won't need the dogbones but I will. 

[nmauldin]

I agree with this! The ability to easily create the dog bone fillets would be a very useful function.

[pludikar]

There's an add-in available from github (here) that does just this.  It's not perfect, but it works very well for most situations.  It certainly saves a lot of time.

[thealucardfromhell]

This is a bad idea because it's too specific.

Instead, why not have a parametric template?

Like, you can make parametric stamps/templates/macros.

You design the operation once, then it generates a tool for you.

Sort of like the addins for generating Helical Gears. But you can do it as a seperate file that can be "inserted" and parametrically modified.

[paul]

If this was parametric - you need 10 pets on your back - AWESOME IDEA - even for sheet metal. 

[cj.abraham]

This would very much be a modeling feature, as the CAM operations will only create toolpaths for existing geometry. The nature of having parametric CAM toolpaths is that they are designed around not violating the model, and creating dogbone fillets with a toolpath on a model that has sharp corners would technically be a gouge. 

This is in contrast to the "Tabs" feature in 2D Contour which "adds" material to avoid areas to machine, and therefore avoids gouges.

[ToddHarris7556]

@cj.abraham I completely understand your comment - this would absolutely be a gouge.

I'll just offer that it's a feature that I became used to in other CAM systems, and sorely miss - to the point where I'll still occasionally export parts from F360 to use this feature. It's a very straightforward 2D toolpath algorithm.

Though I respect where you're coming from, I would submit the exact opposite argument re: modeling - I want the model to reflect my design intent, not my manufacturing constraints. If the design intent is to have a square tab fit into a square slot, then that's the way it should be modeled, IMO. I think it's a fair and defensible argument - One could reasonably argue that a part that is larger than design intent is no better than one that is smaller. 

[cj.abraham]

I understand your viewpoint as well. Usually other CAM packages are not backed by a full-featured CAD platform either, so offering toolpaths with pseudo-geometry input is the only way to implement a feature like this.

[subversivespeed]

This is also useful for sheet metal cutting ie plasma when using thicker metal ie .25 thick and thicker. I use alot of tongue/groove weldments and they tend to jam on the corners, this would be great and also relieve stress in the corners.. 

[Anonymous]

Yes please dog bones! I understand both sides of the modeling vs CAM argument, but I'm firmly on the CAM side for this - because of the reliance on the tool size. Whether it is an add-on or not doesn't bother me but this is a pretty key 2D CAM function! Thanks!

[Anonymous]

This would be extremely helpful for the work we do.  We are currently using the plug in, but it is grindingly slow if you need to create a large number, and creates a separate sketch for each dogbone, which is frustrating to scroll through when you need to find a specific sketch.  They also tend to break easily if you make any changes to your parameters.  I also add my vote to add this on the CAM Side.

[thealucardfromhell]

@AnonymousMaking this CAM side would be extremely retarded:

1. Especially since CAM doesn't ADD anything to the model. It's a modeling FEATURE, it should NEVER be on cam.

2. What's worse is you've already accepted that the modeler is excruciatingly slow and act like it's NORMAL and GOOD that it's slow

Instead of adding dog-bone filets, AutoDesk should make the **** software not slower than a slug. I've got a beast of a PC and Fusion360 only utilizes 3-5% of its power, crashes, lags, slow as heck. It's not the hardware, it's the software, and it's F360 being a pain.

F360 needs to first be recoded to not be trash code-wise, and when it has the performance, dog-bone fillets can be an auto generated plugin thing.

Then the next thing AutoDesk will need to do is add a simpler plugin system or a macro system people can record a macro (i.e. create a single dog-filet) and repeat it via a newly added menu item. But I'm 100% certain AutoDesk won't add this, ever, because they get filthy rich from the cut of the sales of badly coded plugins that are in the F360 plugin store.

The only thing autodesk will ever add is things that'll attract new customers, regardless how trash the software is. They want $$ and they want it now.

[paul]

I don't think anybody's ideas are retarded - ever little bit helps.

[ToddHarris7556]

@thealucardfromhell It almost sounds like you don't appreciate the advantages of being able to do this in CAM 🙂

In terms of fab shops that run lots of 2D equipment, this is a pretty common need - It allows rectilinear tabs to fit into 2D contoured slots.

It's really not difficult algorithmically - it is usually a simple option on 2D contours to define how to approach inside corners.

The argument in favor of CAM-side processing is also sound - the dog-bone feature has absolutely nothing to do with design intent. The intent is simply to end up with an opening that a rectilinear part can slide into. I have a handful of different machines that this could be produced on - the kerf on the laser is different than the cutter diameter of a router. My goal as an engineer is to supply someone with a solid part that represents what I want. The supplier should then apply whatever toolpathing they want in order to achieve that. Maybe they have a higher-horsepower VMC and would use an aggressive roughing strategy. Maybe they're using smaller tooling on a lighter weight Bridgeport. In any event, those decisions are best left to the CAM stage, *in my experience*. Most of the time, we fabricate our parts in-house, but 25% of the time, I might send the part down the road to another shop - I'm pretty sure that if they saw dog-bones in the part, they'd scratch their heads and ask why.

Sorry to hear about your experience with the software. I can only say that we use it pretty much non-stop, and we don't run into nearly the issues that it sounds like you do. We push it in some pretty odd directions, and yes, we've managed to crash it more than once. Having said that, we bang out straightforward 2D and 3D solid parts all day long, with very few problems. 

In any event, it sounds as if you've got other software that's clearly meeting your needs far more efficiently than Fusion. Good for you! Fusion's working out pretty well for us, and we appreciate the ongoing development and enhancement from AD.

[pludikar]

I would vote for this on the CAM side.  As mentioned at the very beginning of this discussion, the dog bone is generally dependent on the size of the tool being used, not the other way around.  You normally want to create as small a dog bone as possible, just enough to relieve the corner of an opening, so (as also mention above) a rectilinear part can fit in.  

In one man shop operations, the person who creates the model design also creates the CAM tool path, but ... in larger shops I would suggest that is rare.  In that case, the person who creates the model design usually knows nothing about machining operations, and therefore cannot dictate which tool needs to be used to complete an operation.  The CAM designer may decide to do rough cuts with one tool and change over to a different size finish cut, in which case this is when a dog bone could be created - the model designer might not know that. Once you put a dog bone in the model, you tie the hands of the CAM designer, and potentially stop him/her from creating efficient tool paths.

When you create a hole in a part, the CAM designer has a choice between pocketing, boring and drilling.  This is a choice made in CAM, not model and is dictated by the tools available, the practicalities of operations, as well as the desired finish - dog bones fall into that same category.  

However, at the end of the day, the Model designer won’t be happy if there are huge unsightly gaps around the dog bones, nor if there are no dog bone either - so both CAM and CAD designers need to communicate with each other.

Thats my 2c worth

[peplinski.john]

CAM, CAM, CAM!!  ToddHarris7556 has it right.  A dog-boning feature should be on CAM side of design, not Model.  Changing to a different diameter tool, for instance, would require that all dog-bone features in Model be changed to accommodate new bit.  Over-cutting (dog-boning) is an optional and automated process in most CAM packages, including cheap ones like CamBam.  Implementing this feature on Model (CAD) side also assumes that the draftsman is also the CAM programmer and knows best how to achieve part.