5-Axis Gouging

[Reko]

Hi,

 

I have a 5-axis groove that works okay except for 4 areas that transition radically at bends in the part.

 

Here are a few pictures that I hope show how radical the fanning is in the 4 areas.

 

Once again, I am finding that the part backplots and verifies good... but the actual part has gouges in these areas... I don't think it is a singularity problem as I have had in the recent past, but the A/B-axis swings are large in these areas.

 

I tried swarf and curve with the same results.

 

Not sure what else to try.

 

Any suggestions would be appreciated.

 

Thanks.

[Jeremy Herron@DBS Solutions™]

What is your max angle increment set to?

[Guest MTB Technical Services]

Try using the fanning option.

[Reko]

What is your max angle increment set to?

 

I don't see a "max angle increment" but my angle increment is set to 3, which is the default.

[Reko]

Try using the fanning option.

 

Okay, I will try it... but can you describe what fanning is... the help file is a bit vague.

[Chris Bloedorn]

Maybe trying using morph between 2 curves. A similar part was posted by someone else and I had to rebuild the groove surfaces to get a decent toolpath. The surfaces overlapped and where generally problematic.

[Reko]

Maybe trying using morph between 2 curves. A similar part was posted by someone else and I had to rebuild the groove surfaces to get a decent toolpath. The surfaces overlapped and where generally problematic.

 

Yes... I tend to agree... the geometry is pretty radical in those 4 areas.

 

Will try.

[Guyinthedesert]

Which toolpaths are you using? On the newer toolpaths, under advanced federate control, you can slow down the feed when going around edges. I use it when doing impeller blades to get a nice smooth path, otherwise when the tool goes around the front edge of the blade A/B swing around wildly.

[Oscar R.]

I usually get great results when setting tighter tolerance on the "Curve following method" (cut tolerance to about .0005" and maximum step to about .010").

Also check your filtering. I set mine pretty tight (.0005" or more.).

Lots of code but it will work great.

[Reko]

Which toolpaths are you using? On the newer toolpaths, under advanced federate control, you can slow down the feed when going around edges. I use it when doing impeller blades to get a nice smooth path, otherwise when the tool goes around the front edge of the blade A/B swing around wildly.

 

I tried swarf, curve and morph between curves (a lot of settings here!) but they all swing pretty fast around A/B axis moves which is when it gouges.

 

Which ones would you try? Settings?

[Reko]

I usually get great results when setting tighter tolerance on the "Curve following method" (cut tolerance to about .0005" and maximum step to about .010").

Also check your filtering. I set mine pretty tight (.0005" or more.).

Lots of code but it will work great.

 

I will try that, thanks... but where is "curve following method"? Or, did you mean wall following? thx.

[Guyinthedesert]

I tried swarf, curve and morph between curves (a lot of settings here!) but they all swing pretty fast around A/B axis moves which is when it gouges.

 

Which ones would you try? Settings?

 

I have good luck with parallel to multiple curves and morph between 2 surfaces. Under utility/ surface radius based feed rate optimization, I set the 4 feeds to 100, 10, 10, 10. I hadn't had any problem with gouging, but when A/B swing around an arc at full speed, I really don't like the way it looks. At 10% it's much smoother.

[Jeremy Herron@DBS Solutions™]

I usually get great results when setting tighter tolerance on the "Curve following method" (cut tolerance to about .0005" and maximum step to about .010").

Also check your filtering. I set mine pretty tight (.0005" or more.).

Lots of code but it will work great.

 

In addition to this, I usually run my angle increment to .1-.5

[crazy^millman]

I keep going back to my favorite toolpath for this type of stuff. 5 axis Curve with Vectors add or subtract as many as you need to dial in the movement you are looking for. Yes work drawing the vectors, but I have done some very tricky stuff over the years doing it that way. If you can send me a file I would be glad to do a narea and show you what I think might work.

[Oscar R.]

I will try that, thanks... but where is "curve following method"? Or, did you mean wall following? thx.

 

This is one of the settings in the multi-axis "Curve" toolpath. Click on the "Cut Pattern" right under "Trim to Stock" and you see those setting right around the middle of the box.

 

 

 

In addition to this, I usually run my angle increment to .1-.5

 

+1

 

I keep going back to my favorite toolpath for this type of stuff. 5 axis Curve with Vectors add or subtract as many as you need to dial in the movement you are looking for. Yes work drawing the vectors, but I have done some very tricky stuff over the years doing it that way. If you can send me a file I would be glad to do a narea and show you what I think might work.

 

+1

[Reko]

Ron, you've got mail. Thanks.

[Oscar R.]

If you can send me a file I would be glad to do a narea and show you what I think might work.

 

I might be wrong but the looks like the same file from another one of Reko's posts on the Singularity issue he was having.

I believe he loaded the file on to the FTP site.

 

I also posted a working file for it last week, I believe...

[Reko]

I might be wrong but the looks like the same file from another one of Reko's posts on the Singularity issue he was having.

I believe he loaded the file on to the FTP site.

 

Hi Oscar,

 

No, this is a different part for a similar assembly. I tipped the last part 5 degrees and it solved that problem.

 

This part, however, no matter how I tip it, has those 4 trouble spots that gouge.

 

Very frustrating.

 

It's crazy how it comes out so perfect in verify, but the huge swings of the A/B axis do weird things.

 

I feel like I need to find a way to smooth the toolpath in verify, then it will work better in the real world.

[Reko]

This is one of the settings in the multi-axis "Curve" toolpath. Click on the "Cut Pattern" right under "Trim to Stock" and you see those setting right around the middle of the box.

 

Okay... I was in swarf, but now I see it there in curve... that really tightens up the cutterpath... I will be anxious to see how that works.

 

We pulled off this part to get some other work in while I was getting some help, so I will likely get back on this tomorrow or Wednesday.

 

Thanks guys.

[Aaron Eberhard]

Reko - what type of machine are you cutting on? If it's a head/head machine, and you're not using G43.4 (tool tip compensation), then what could be happening is that the tool isn't being compensated in between vectors..

 

Most likely, you're just not running enough vectors, though.. Like was suggested above, set your angle step to .1 or so, and you'll probably have better control. Most likely what's happening is that the machine isn't controlling your A & B (or B & C, or whatever) accurately enough during the large multiaxis moves, so feeding it finer vectors will give the accell/decell routines on the control an easier time of controlling that motion.

[Reko]

Reko - what type of machine are you cutting on? If it's a head/head machine, and you're not using G43.4 (tool tip compensation), then what could be happening is that the tool isn't being compensated in between vectors..

 

Most likely, you're just not running enough vectors, though.. Like was suggested above, set your angle step to .1 or so, and you'll probably have better control. Most likely what's happening is that the machine isn't controlling your A & B (or B & C, or whatever) accurately enough during the large multiaxis moves, so feeding it finer vectors will give the accell/decell routines on the control an easier time of controlling that motion.

 

It is a table/table... Haas Trunnion.

 

Cool thanks. I will try all of the above, and I will definitely post the solution when I get a good toolpath.

 

Thanks all

[crazy^millman]

Back at you.

 

The chains were not consistent for the upper and lower chains. I also went in and created points on the chains to use for the sync method. Once that was done the area I did for you it looked very nice. The Geometry is not very nice and a part like this is always tricky sometimes you cannot take the rails at face value and where I did some work make the chains more even and the points just give you some nice control most people are not aware they can use on 5 axis toolpaths.

 

Off to see a customer catch everyone later.

[Reko]

RESOLVED!!!

 

I was a feed rate problem... the A and B axis' were speeding up too fast in those 4 area's.

 

I tried pretty much everything suggested...

 

-re-drawing geometry

-selecting by points instead of entity

-curve instead of swarf

-tightening up settings to output huge amounts of code

 

...nothing worked... until we slowed the feed rate down in those 4 spots.

 

The final solution was to cap the whole part's feed rate at F150.

 

This Haas (5X Trunnion) is only 6 months old and can max out at 999.99 inverse feed rate.

 

I guess my question is... why would it deviate from a programmed path top feed or a lower feed rate???

 

BTW... huge thanks to everyone that helped... this board (and everyone here) is awesome!

[Guyinthedesert]

I believe the max feed for a Haas trunion is 40000. That may be your problem

[Reko]

Not sure what you mean, guy.

 

The Generic Haas Trunnion Post is set to a max federate of 999.99

 

That is exactly how fast it was going in those corners.

 

Once we slowed it to F150. the toolpath smoothed out and cut to the correct dimensions without gouging.

 

At F40000... OMG!!!

 

Not sure what to think there.