Stock positioning bug, affects most of my projects.

[calebktaylor]

So I have my models in a reasonably complicated fixturing setup, and the first time I make my manufacturing setups  the stock positioning is correct, then I switch back to the modeling environment and move the model around a few thousandths, then when I go back to the manufacture environment the simulation stock box won't update, this has been a real nuisance turning moving things around in a complicated setup into a bug-ridden multi-hour ordeal to get the right behavior I want, and I'm pretty sure the WCS origin wasn't updated after editing a model and re-selecting the new wcs origin and posting to the machine in at least one instance.

This bug sucks and effects at least a couple of my projects, I can't figure out a reliable way to get the stock to reposition into the correct location with the same WCS, it's like it's caching/saving the old location and won't change it.

Attaching two pictures, in one the result that I'm stuck with, in another I have switched the X axis WCS direction via the checkbox and then it is fixed.  There is basically no other combination of settings in the stock tab of the setup edit dialog/menu that will produce the correct, updated result.  I can't even make a new setup and copy and paste the old toolpaths into it to get the stock box in the right position, still uses the old position.

There should also be an easier way to move the model around in the manufacture environment so the model + stock box size is in the right place on the fixture and most of these bugs could be avoided.

The WCS bugs after posting were in the Y axis.  For instance, the square milled into the fixture in the pictures at the very corner of the piece is 10 thou below the surface of the face, and after posting with the WCS on the corner/point of the -10 thou edges I needed to modify my zero -10 thou on the machine to actually get the right result.  Quite a real problem.  I modeled the edges (circled in pic) after creating the manufacture setup, then modified the model.

[paul.clauss]

Hi @calebktaylor 

Thanks for posting! Would you mind sharing one of the files in which you have observed this behavior? I've been unable to recreate the problem with the stock not updating to the new location of the component to be machined.

What type of stock are you using (relative sized box, fixed size box, etc)? To make this easier, I would recommend including a modeled stock body (to use with "Stock from Solid") within the component including the bodies to be machined - this will ensure you can move the component to move both the stock and model body. It will also allow you to use points on the stock to create joints to fixtures.

[calebktaylor]

sent a PM with the link

I've been through a couple hundred bucks of aluminum across my fixtures figuring out where to actually set the WCS zero to get the right result, hope you can fix it.

[paul.clauss]

Hi @calebktaylor 

Thanks for sending the file over. I have a few questions here:

1. I don't see the "GRIP LSIDE TOP G55" setup pictured in your screenshot in the file you sent over. Is this issue occurring in another file as well?

2. The surface milled into the corner of the fixture appears to be 15 thou below the face - could this be the cause of some of the errors at the machine? Does this cut exist at the machine? If so, why?

3. In the "Fixture Trigon WCS" CAM setup, the stock is defined to ignore the corner cutout, but the WCS is placed on a point created by the cutout. Are you actually zeroing to that point at the machine? I would suggest just going to the stock corner.

I am a bit confused by the forum post and may need some detail in the form of a screenshot illustrating you triggering the problem behavior to fully understand what is going on here. If you'd like to send me a phone number via direct message I am happy to do a call and remote session.

I found that the stock positions updated as I would expect - please see the video below. It is worth mentioning that the best practice in Fusion 360 is to use Joints to parametrically position the model. I would also recommend modeling the stock bodies within the same component as the bodies to be machined and using "Stock from Solid" - by doing so, you can avoid any stock update issues by moving the component (containing the machined body and the stock) rather than only one of them.

If you can list the steps you use to trigger problems in your file or record a screencast showing the problem on your end, I am happy to continue investigating.

[paul.clauss · Accepted solution]

Hi @calebktaylor 

Thanks again for posting! I was able to trigger the problem.

This lies in what is selected as the body to be machined in the CAM setup - the object definition being moved needs to match the object definition for machining in the CAM setup.

For example, if you select Components as the "Model" in the CAM setup...

• Moving Bodies will trigger all toolpaths for regeneration and the stock will be shown correctly.
• Moving Components will not flag toolpaths for regeneration and the stock will get stuck in the previous location as you described.

If you select Bodies as the "Model" in the CAM setup...

• Moving Bodies will trigger all toolpaths for regeneration and the stock will be shown correctly.
• Moving Components will not flag toolpaths for regeneration and the stock will get stuck in the previous location as you described.

In short, it appears that you need to move Bodies rather than components to have the stock update appropriately.

I agree that any combination should flag the toolpaths for regeneration and move the stock to follow the bodies being machined and have logged internal ticket CAM-13896 on this behavior. While development works out a long term solution, being sure to match the type of object defined in the CAM setup when using the Move command should help you avoid this issue.

Including stock bodies within the same component as the model for machining and moving the component will also work as both bodies will move with the component.

With design history enabled (information on timeline/history/parametric mode vs. direct modeling mode here), the component move will trigger regeneration and move the stock if a "capture position" is completed (the move is confirmed). Move commands in direct modeling should assume that a "capture position" follows and update. So, you could also work around this by working with design history enabled and ensuring that all component move (or drag) commands are followed with a "capture position" command.

Example in video below. Please let me know if you have questions - thank you for bringing this to my attention. This is a confusing issue!

CAPTURE POSITION/HISTORY ENABLED EXAMPLE:

[calebktaylor]

Thank you very much, that helps me tremendously.  But the WCS bug is real, and it is almost worse.

[paul.clauss]

Hi @calebktaylor 

Thanks. I'll still trying to understand the WCS bug - it looks like the code is all posting correctly.

Which setup is this error occurring in? Which toolpath?

In the "fixture trigon wcs", "pierson pallet trigon wcs", and "bottom fixture"  setups, the WCS is set to the corner of the modeled geometry. The face position on the model doesn't add up - the Y-offset between the outer and inner face is 0.015", not 0.01". This seems inaccurate, as you mentioned moving the y-offset at the machine -0.01" at the machine.

I don't think this is the error at the machine if your machine measurements are correct - I think G54 is not being reset between running the "fixture" and secondary "fixture trigon wcs", "pierson pallet trigon wcs", and "bottom fixture"  setups.

Where is the machine actually zeroed to for the G54 offset? The inner or outer faces?

For toolpaths in the "fixture trigon wcs" setup (and the "pierson pallet trigon wcs" and "bottom fixture" setups), it should be set to the inner faces. In the "fixture" setup, the WCS is set to the corner of the stock. This is shown below:

Both of these setups use WCS Offset #0 in Fusion, which means they both refer to G54 at the machine. Are you certain that G54 was reset between running these setups at the machine? If you moved the y-offset (or any offset for that matter), it would make sense that the "fixture" setup was run, G54 was not reset at the machine and then the "fixture trigon wcs" or "pierson pallet trigon wcs" was run. 

This also adds up when considering the incorrect offset in the model. If the corner pocket is actually milled to a depth of 0.01" (instead of 0.015", as in the model), the code would run correctly after offsetting the y-offset -0.01", the x-offset +0.01" and the z-offset -0.01" if G54 was set to the corner of the stuck like in the "fixture" setup. 

If the actual dimension between the inner and outer faces on the physical fixture is 0.01", the code from the "fixture trigon wcs" setup would still be off by 0.005" in the y-direction. 

I hope this helps. Certainly a puzzler.

[calebktaylor]

I can't confirm it actually occurs in the project I sent you because I don't have a manual NC note added to adjust the Y axis zero.  I would prefer not to upload all of my projects though.  If you can try to recreate it, all I did was establish the WCS at the edge of the rectangular fixture, model in the sunken faces at the corner, machine them (with wcs still appropriate), then change the WCS to the inner point.  It shows in the CAM that the WCS is at the right place, but that is not true after posting at the machine.  It's possible my fixture is off, but I don't think so.

Sorry I can't be more helpful.

[paul.clauss · Accepted solution]

Hi @calebktaylor 

Thanks for the response. In my tests, things appear to be working correctly in the software and I am confident this is an issue with offsets being incorrect at the machine. 

If the faces on your fixture are only offset 0.01", the model for the fixture is certainly off.

I hope you are able to get this sorted out!