Suggestions for achieving better side-wall finishes on hardened tool steel?

aarongough · May 8, 2020

Hey guys!
I have an operation that I run on my VMC (Fadal VMC15) where I tend to find tiny 'streaks' around the edges of the part. With a fresh tool the surface finish is great, but after a relatively short amount of time (usually 5-6 parts) I start getting 'streaks' where it's obvious the tool edge is breaking down a little. I am cutting 62-63HRC tool steel, currently running a 4 flute AiTIN coated cool (3/8" diameter, 1" LOC) and I use different vertical sections of the edge for roughing and finishing to preserve tool life as much as I can. I previously used a 6 flute tool, but I found the extra expense of the fancier tool didn't make as much sense as simply changing the cheaper 4 flute tool more often.

The 'streaks' are not very large, it basically looks like the tool edge is breaking down and leaving tiny raised stripes around the part.

In the past I have experimented with using electroplated CBN grinding pins to try to get better finishes. Ideally I would love to CNC jig grind these parts (for appearance), and I have tried something akin to a 'chop grinding' toolpath on my VMC but it is hard to get the kind of finish I would expect without dressing the stone in the machine, which I don't want to do because of the dust.

Are there any other tricks I could use to get a more aesthetically pleasing finish in the machine? I have thought about running a rubber grinding point or one of those tiny flaps wheels around the part, but I worry about abrasive dust getting everywhere and creating a taper or rounding the edges over. Dimensions on the part are relatively wide (+-0.0005" or so) but I need them to look as good as they can, and to clean up perfectly during light vibratory tumbling... Currently I hand finish all these edges by wet sanding to make sure there are no streaks left.

Here's a photo of the edge finish. Part on the top was run with a brand new tool, part on the bottom is after about 5-6 parts. I know the difference is really hard to see in a photo, honestly the difference is small even in person but I need to try to get the edge finish as good as I can:

Any suggestions greatly appreciated!

Thanks in advance!
-Aaron

aarongough · May 8, 2020

Also you can see my Fadal leaves small steps when interpolating the very large arc moves that make up the outside shape of these parts... Smaller arcs (like a 3-4 diameter circle) look perfect but large 20" diameter arcs do not. If anyone has suggestions to fix that I'm all ears! I have tried posting those moves as linear moves with no arcs with no difference, same goes for smoothing settings. I am using Fusion 360 for CAM. I have tried the same code on a couple of Fadals with the same results so I don't think it's an issue with this particular machine...

I'm sure something like a Makino F5 would make mince-meat of this finish, but I gotta work with what I've got!

lowCountryCamo · May 8, 2020

The geometry you are programming off could be the problem. Could you reduce the tolerance of the geometry?

couch · May 8, 2020

Are you running your machine in G8 or G9?

Try your finish pass with a spring pass at a different Z height. If you're repeating your finish pass anyways, running it at a different Z height will allow a hopefully better part of the tool to clean up the high spots left from nicks/wear in the cutting edge.

lowCountryCamo · May 8, 2020

When I program hard metals I use 2 identical tools. Rough and finish. When finisher gets ugly it then becomes my next rougher. This has worked very well in the past.

EOLSON · May 8, 2020

First, make sure your flute length and holder are absolutely as short as possible and that you are running sufficient air so that re-cut chips are not chipping the edge of the cutter.

To get the best finish possible I would run 3 of the same tools, rough, semi-finish, and finish. On the semi-finish and finish tools bump your height offset .005 or so before running each part, to use a different part of the edge. Finish tool should only take a few tenths max and a spring pass or two at different height. As far as your steps in interpolation, changing the tolerance on the toolpath to .0001 works for me in Mastercam (makes program massive though). I would look at your code and see how large of straight lines it is outputting, then make a setting change in Fusion and post again and compare. If code shows a difference and part does not, you are probably SOL since the steps can only be so small on a machine.

nateacox · May 8, 2020

Try 6/7/9 flute from Helical or Imco. We've had amazing results hard milling with those. You need more flutes, and very small side step over.

mmurray70 · May 9, 2020

Yes as other said use another tool for finish. I would rough and finish with one tool at different depths like you are doing to save a tool change, then take a spring pass or maybe 0.001 off with a brand new tool. After you get some wear pop in a new finisher and use the old one as your rougher.

Also check tolerance in CAM, even the best machine can look like that without setting very small tolerances.

TeachMePlease · May 9, 2020

So, using your parameters, I used the tool finder on Mikron's website. They gave me this tool: NS Tool-ID-Mikron Tool

Mikron is expensive... But man, they are really up there in terms of performance. It might be worth talking to your tool rep to see if you could get one of these to test (dunno how your rep treats you).

Delw · May 9, 2020

aarongough said:
Hey guys!
I have an operation that I run on my VMC (Fadal VMC15) where I tend to find tiny 'streaks' around the edges of the part. With a fresh tool the surface finish is great, but after a relatively short amount of time (usually 5-6 parts) I start getting 'streaks' where it's obvious the tool edge is breaking down a little. I am cutting 62-63HRC tool steel, currently running a 4 flute AiTIN coated cool (3/8" diameter, 1" LOC) and I use different vertical sections of the edge for roughing and finishing to preserve tool life as much as I can. I previously used a 6 flute tool, but I found the extra expense of the fancier tool didn't make as much sense as simply changing the cheaper 4 flute tool more often.

The 'streaks' are not very large, it basically looks like the tool edge is breaking down and leaving tiny raised stripes around the part.

In the past I have experimented with using electroplated CBN grinding pins to try to get better finishes. Ideally I would love to CNC jig grind these parts (for appearance), and I have tried something akin to a 'chop grinding' toolpath on my VMC but it is hard to get the kind of finish I would expect without dressing the stone in the machine, which I don't want to do because of the dust.

Are there any other tricks I could use to get a more aesthetically pleasing finish in the machine? I have thought about running a rubber grinding point or one of those tiny flaps wheels around the part, but I worry about abrasive dust getting everywhere and creating a taper or rounding the edges over. Dimensions on the part are relatively wide (+-0.0005" or so) but I need them to look as good as they can, and to clean up perfectly during light vibratory tumbling... Currently I hand finish all these edges by wet sanding to make sure there are no streaks left.

Here's a photo of the edge finish. Part on the top was run with a brand new tool, part on the bottom is after about 5-6 parts. I know the difference is really hard to see in a photo, honestly the difference is small even in person but I need to try to get the edge finish as good as I can:

Any suggestions greatly appreciated!

Thanks in advance!
-Aaron

Aaron
fadal thats a Fadal finish when
your drives are not tuned, your gibbs are just a tad loose or ball screw has some slop. if you look at the straight edge it is nice so I dont think your getting any vibration transfer from your pulleys, tensioners or belts.
one way to get it out is to use a seperate tool for finishing.
slower feed rate and less rpms might help also along with dry pass but generally that helps if there is Vibration transfer coming from the belts and pulleys.

add'd: a 5 flute or 7 flute will help.
I am assuming your program isnt polylines for the arc.

BROTHERFRANK · May 9, 2020

Send some test pieces and geometry down if you like. We can cut on a Brother for you. No strings attached. Many knife companies run Brother machines.

Check out this 1/2" high side wall in Titanium:

That was finished with the same tool that roughed that is hanging out 5.5"

TeachMePlease · May 9, 2020

BROTHERFRANK said:
Send some test pieces and geometry down if you like. We can cut on a Brother for you. No strings attached. Many knife companies run Brother machines.

The only string attached is you KNOW he's gonna want one after he sees the finish

Wsurfer · May 9, 2020

TeachMePlease said:
The only string attached is you KNOW he's gonna want one after he sees the finish

Lol, not to mention cycle time reduction.

DanielG · May 9, 2020

How are you cutting: dry, mist, flood, etc.?

For improving the surface finish and limiting your hand operations, you might want to talk to Brush Research or Weiler Abrasives. I don't know if there stuff would work without doing too much corner rounding, but it might be worth a call.

Silent_Man · May 9, 2020

If the large radius you are cutting is aligned with either x or y it might produce a stair step effect where x moves a bunch, y moves a little. You can rotate the part 45 degrees so both x and y are moving closer to the same speed and have an easier time interpolating the curve. Hope that makes sense.
No guarantee that will solve your problem, its hard to see what you mean exactly from the picture.

Sent from my SM-G930P using Tapatalk

Delw · May 9, 2020

aaron

the left side of the attached pic is generally spindle vibration transfer issues, like pulleys tensioners etc etc..
the center is usually gibs thrust bearings drives not being tuned etc etc
the right side is a feed issue combined with the center pic issue but not alwasy.

D Nelson · May 9, 2020

That looks like toolpaths to me make sure you are converting to arcs. And tighten your tolerance up
Don

Sent from my iPhone using Tapatalk Pro

aarongough · May 9, 2020

BROTHERFRANK said:
Send some test pieces and geometry down if you like. We can cut on a Brother for you. No strings attached. Many knife companies run Brother machines.

Thanks mate! I would take you up on the offer, but I don't think I want to torture myself like that... I already want one of your machines but can't have one so it would just make the longing worse

In the future I may look into one more seriously but at the moment I don't think it would be the right move unfortunately!

aarongough · May 9, 2020

Silent_Man said:
If the large radius you are cutting is aligned with either x or y it might produce a stair step effect where x moves a bunch, y moves a little. You can rotate the part 45 degrees so both x and y are moving closer to the same speed and have an easier time interpolating the curve. Hope that makes sense.
No guarantee that will solve your problem, its hard to see what you mean exactly from the picture.

Sent from my SM-G930P using Tapatalk

That is very good feedback, thank-you! I have actually talked with a friend who battled this exact issue on a Fanuc controlled machine, he found the same solution: rotate the part by 45º to make sure both axes are making larger movements and his issues went away...

Hadn't thought about that applying in this case but you're totally right it probably does!

aarongough · May 9, 2020

Delw said:
Aaron
fadal thats a Fadal finish when
your drives are not tuned, your gibbs are just a tad loose or ball screw has some slop. if you look at the straight edge it is nice so I dont think your getting any vibration transfer from your pulleys, tensioners or belts.
one way to get it out is to use a seperate tool for finishing.
slower feed rate and less rpms might help also along with dry pass but generally that helps if there is Vibration transfer coming from the belts and pulleys.

add'd: a 5 flute or 7 flute will help.
I am assuming your program isnt polylines for the arc.

Delw said:
aaron

the left side of the attached pic is generally spindle vibration transfer issues, like pulleys tensioners etc etc..
the center is usually gibs thrust bearings drives not being tuned etc etc
the right side is a feed issue combined with the center pic issue but not alwasy.

View attachment 288047

Hey mate!
My machine is a linear way machine, so no gibs. I did just recently tune the servos following the procedure in the maintenance manual and got the following error pretty much identical (+- 2 counts) on all 3 axes during movement, but I cannot say that I have all the settings on the drive correct as they don't go into a lot of detail unfortunately... The X ballscrew has 1 tenth of backlash, and the Y ballscrew 2 tenths, z axis is 1 tenth. I have correct values set in the compensation table...

I have tried a few different feedrates in the past and found that going slower actually made the 'stepped' appearance much worse, speeding up seems to smooth it out. I found the same thing when 3D contouring. My initial thought was that going slower would make the surface more accurate but what I actually got was waves in the surface from the servos oscillating around their target point (I guess?), increasing the feed fixed that issue when 3D contouring my bevels. I will try bumping up the feed again when doing the profile to see what difference it makes as it's been a while since I tried that, I think the main limiting factor was the surface finish getting worse from too big a IPT. As you said a 6/7 flute would help there so I will have to revisit the idea of a dedicated finishing tool!

I recently saw another member here using a smartphone app to re-balance their spindle, I have been meaning to try that and you comment makes me think I should bump that up my list of priorities!

If you know any good tricks for Fadal servo tuning I would love to hear them! Overall the machine performs well, but finishes on more modern machines definitely seem better...

Thanks for taking the time to give me your feedback mate!
-A

Suggestions for achieving better side-wall finishes on hardened tool steel?

Stainless

Stainless

Stainless

Cast Iron

Stainless

Plastic

Aluminum

Stainless

Diamond

Stainless

Stainless

Diamond

Cast Iron

Stainless

Aluminum

Stainless

Stainless

Stainless

Stainless

Stainless

Similar threads