Machining path deviation

[RoboTomo]

Hello,

I have a problem creating straight machining paths in Fusion. I use the Steep and Shallow method for 5-axis lead and lean machining. The machining direction in the Y-axis is fixed, so it would be desirable that the software could create straight lines in this direction, but this is not the case.

1. We can already see visually that in some cases, in the curved area, the line in the middle of the workpiece is going left and right, but ok, this is not a big deal yet (Picture 1).

2. The problem arises when I export the machining path/points that Fusion generated and plot the results. Here, we can clearly see the path deviation since it creates a Zig-zag pattern in the middle of the workpiece (Picture 2). 

I calculate the optimal direction and minimal speed in this Y-direction. I now have to generate a machining path in this direction and verify the minimal speed, but this speed is always incorrect because it is caught in that Zig-zag Fusion pattern.

Why is Fusion doing that and how to eliminate it?

[seth.madore]

Would you be able to share your Fusion file here?
File > Export > Save to local folder, return to thread and attach the .f3d/.f3z file in your reply.

[RoboTomo]

Thank you for your answer. I am attaching the file.

So the goal is to obtain straight machining lines for different stepovers (2, 1, 0,5, 0.1 mm). In this example, we have a stepover of 2 mm and for this case, I tried changing tool diameter, tolerance, and smoothing but was unsuccessful.

[RoboTomo]

I am still struggling with this problem and have additional question.

How exactly are toolpaths generated in this Steep and shallow strategy and if this is different for other strategies? Are they generated directly on the 3D surface with offset curves, or are they generated on the XY-plane and then projected onto the 3D surface?

Thanks.

[a.laasW8M6T]

Hi

I have made a video that hopefully explains what is going on with the toolpath there

2024-08-02 18-14-55.mp4

Video Player is loading.

Play Video

Play

Mute

Current Time 0:00

/

Duration 10:33

Loaded: 1.56%

00:00

Stream Type LIVE

Seek to live, currently behind liveLIVE

Remaining Time -10:33

1x

Playback Rate

Chapters

• Chapters

Descriptions

• descriptions off, selected

Captions

• captions settings, opens captions settings dialog
• captions off, selected

Audio Track

• en (Main), selected

Picture-in-PictureFullscreen

This is a modal window.

Beginning of dialog window. Escape will cancel and close the window.

Text

ColorWhiteBlackRedGreenBlueYellowMagentaCyanTransparencyOpaqueSemi-Transparent

Background

ColorBlackWhiteRedGreenBlueYellowMagentaCyanTransparencyOpaqueSemi-TransparentTransparent

Window

ColorBlackWhiteRedGreenBlueYellowMagentaCyanTransparencyTransparentSemi-TransparentOpaque

Font Size

50%75%100%125%150%175%200%300%400%

Text Edge Style

NoneRaisedDepressedUniformDropshadow

Font Family

Proportional Sans-SerifMonospace Sans-SerifProportional SerifMonospace SerifCasualScriptSmall Caps

Reset restore all settings to the default valuesDone

Close Modal Dialog

End of dialog window.

Close Modal Dialog

This is a modal window. This modal can be closed by pressing the Escape key or activating the close button.

[RoboTomo]

Thank you very much for your answer; you provided a very good explanation! I have been trying all sorts of parameters and did additional tests after your video, and the results are better. However, in some cases, there are still very sharp declines, as in my original post (Figure 2) when we zoom in. 
I noticed that if I use the tool diameter the same as stepover distance, the results also improve.
As you mentioned, the part itself can have an impact on the machining path. Here, originally, I had the .stl part, but I converted it into a solid body in Fusion; maybe it will be better to do machining directly on the original .stl part with a more dense triangulation?

[a.laasW8M6T]

Hi

whether the original .stl will be better I am not sure, it will really depend on the quality of the original.

The toolpath will re-mesh it when it calculates, but it can't Improve the quality better than the original.

[RoboTomo]

Hi, it's me again. After more testing, unfortunately, I have to un-accept your answer @a.laasW8M6T.
At the time, it seemed like it was working, but more detailed analysis showed there were still deviations. Please see the attached file, where the machining path from Fusion is compared to other software, where surface normals can deviate up to 8 degrees alongside the uneven toolpath in my original post.

I am attaching another workpiece; please take a look if somebody can. I can't find the strategy or parameters to create straight machining toolpaths in Fusion 360 CAM software, and I think it's probably impossible.

Conditions for machining path:

- 5-axis machining with tool normal to the surface (Lead and lean option)

- the strategy must offer arbitrary machining direction selection

Currently, I am only able to create those paths with Steep and Shallow strategy but if there is another option that can create straight machining toolpaths under those conditions, I will gladly use it.

I want to get confirmation on why that happens. Mainly, I want to eliminate if I am doing something in the design that results in bad toolpaths or if this is because of the underlying Fusion path planning algorithm.

[a.laasW8M6T]

Hi

As far as I'm aware Steep and shallow is the only toolpath that meets your criteria of Lead/lean and arbitrary direction.

Looking at the PDF there is a significant surface normal error in Fusion compared to the other software.

unfortunately I'm just a user of Fusion so I don't know what's going on "under the Hood" so to speak.

@seth.madore this looks like something Craig might be interested in, may be worth running up the chain.

[RoboTomo]

Ok thank you again for all your time and help. I hope somebody from Autodesk can take a look.

[seth.madore]

I've raised this issue with a couple of people who would be able to speak with more authority on the matter, I'll let you know if/when I hear back from them!

[craig.chester]

Here's an explanation as to what is happening with Steep and Shallow and some methods which might help you to get a superior result

[craig.chester]

It might also be worth stating that both BLEND and FLOW both support Lead/Lean toolaxis control. However, they will not maintain a constant Y value or Parallel passes when viewing down Z.

[RoboTomo]

Thank you very much for your time and help. I don't know how I didn't see this option before, but changing the smooth angle to 0 deg improved the machining for Workpiece 1 ("Fusion_example.f3d") along the 0.005 tolerance. The normal deviation has now improved (Fig. wp1_tolerance), but the downside is that there is a significant increase in points, but I can handle this. 

Applying the same tactics to Workpiece 2 (Fusion_example_WP2 v4.f3d), which is also attached in previous posts, did not yield such good results. When the tolerance is low (0.005), the path drifts out from the surface, and additional turns at the end appear (Fig. wp2_low_tolerance_dev); those turns are still visible when changing the tolerance to 0.1 (Fig. wp2_tolerance_increase).

Would you happen to have any suggestions for this workpiece? Based on the video, it seems that tool diameter can impact the machining since you can not pick a tool diameter smaller than the stepover distance in Fusion. Theoretically, if we set the tool diameter smaller, the contact point would be closer to the tooltip, and the path deviation may improve?

[craig.chester · Accepted solution]

I am unsure exactly why you want what you ask for....

What may help you......

You can cheat the system and use a CHAMFER TOOL with a very small tip diameter (I used a tip diameter = 0.1mm) and a much bigger diameter (I used 12mm for the diameter which enables me to use a stepover up to 12mm). This creates a Parallel Toolpath with Lead/Lean=0 very close to parallel. Fusion file attached. Let me know if this is useful to you.

 

 

[craig.chester]

Switching on CONTACT POINT BOUNDARY produces a toolpath with much cleaner end points.... Use this Document to check the result.

Is this of use to you?

[craig.chester]

 Toolpath is very close to parallel

[RoboTomo]

Dear Craig,

that's exactly what I was looking for; the paths are ideal now! Perfect!

Thank you for taking the time to investigate that problem. You have saved me a lot of time because I was already looking at some other software, but I like the simplicity and versatility of Fusion and I am very glad I can continue using it.

Yes, my demands for this machining are odd, and because all the strategies are based on milling, it was hard for me to achieve the right results. To simplify what we are doing: We are working on a robot hammer peening application, which is basically surface finishing, and there is no material removal like in the milling. Here, the hard metal ball is hammering the surface of the workpiece to improve hardness, smoothness, etc. (https://www.youtube.com/watch?v=E4zfizO65xo). We are calculating the optimal direction and robot machining speed to minimize the machining time, so the catch is that the calculated (designed) optimal directions are straight, and the robot should follow them very closely. Then we pick those optimal machining directions in Fusion, but if there are deviations from the designed directions, as initially, the machining task is not successful. That's why I need nice, smooth, and straight paths without deviations. If interested, I can send our paper, which better explains it.

Thank you again. I would also like to thank @seth.madore to forward this problem to you and of course @a.laasW8M6T who is always willing to help and was also a big part in answering my other questions.

Best regards.

[craig.chester]

So happy that this looks like a workable solution for you 🙂  

Is your robot hammer peening application aimed at hardening the outer layer of molds or press tools by any chance? If so, I know of a large customer who will probably be interested in your work.

[RoboTomo]

Yes, this is meant for the toolmaking industry, but we are still in the research phase. We can exchange some more data in private messages.

Thanks.