Aluminum surface 6061 T6 moon craters? 1

[DWilliamA]

Hi, first off - I am not a machinest. I am quality guy trying to chase down an answer. from what my machinest tells me, we are doing this right... here are the details.

Extruded 6061 T6 Aluminum 3/4"x2-1/2" stock basically machined to a block shape, lots of flat shiny surfaces. Approximately 1/8" to 1/4" of material from the sides and top taken off. 4 flute cutter, 3800 rmp nice chip load the surface looks really nice - but when you look real close, really really close when the light is just right you can see what looks like a moon crater surface or orange peel surface... actually it looks like a spackle paint maybe. At a glance you see the nice subtle rainbow sheen and the miniature fly cut lines - looks nice. but upon close examination there is this bumpy spackle characteristic

Any ideas on what causes this and how to eliminate it, if possible with machining technique?

I believe that when milling - the cutter creates a shear zone and just under this shear zone is a compression zone. Its almost as if the aluminum is inconsistent in hardness and certain areas compress more than others. This situation didn't just start btw.. - so I don't think it is just bad lot of material. Its just that I am the first person to come along and try to solve the problem.

that said... this condition is causing problems, or is possibly related to problems we have later when the part is Type III hard black anodized. The same type of mottling shows up in the hard black anodize. Unfortunatley, we could hit these parts hard with the ano and etch heavy and anodize thick, but this dulls the surface. We want a nice shiney surface which takes a light anodize so to speak, which leaves the spackling intact.

different cutter - thicker extrusion and cut through more exterior material, heat treat, lower Iron levels, plate stock (if possible) ... ?????

Thank you,

David A.

 

[btrueblood]

Cutting too fast, and/or dull tool, and/or no lube/coolant. Instead of shearing the material with the cutting tool edge, you are "snagging" chips and yanking them off the surface, resulting in local spots of ductile rupture. All of the above is a guess, lacking any visual observation of what you are talking about.

 

[DWilliamA]

Ductile Rupture - I think I like the sound of that :0

It sure makes sense - we have plenty of coolant, new sharp tooling and we do cut fast... fast... to get the nice shiney surface.

Thanks for the tip

David A.

 

[saberblue]

You may be getting a little chip reweld -- check to see if you are getting any built up edge on the tools --

 

[CastMetal]

What is fast? fast for 6061 is in the neighborhood of 800sfm so assuming carbide inserts and 3800rpm you are using a 1.25in fly cutter? .006 ipt would then mean you feedrate is around 91.2 ipm.

 

[DWilliamA]

How fast 'is' fast... good question - I will get these numbers tomorrow Wednesday.

 

[ProgressiveRacing]

I don't believe there is a maximum rpm for 6061T6 to be honest. Generally, in my aerospace machining experience with high-power gantry mills, we worked back from maximum RPM to find stable cutting conditions at maximum metal removal rates. That was before tap-testing and harmonics software came along. 3/4" carbide end mills at 16-18k RPM are pretty common fare these days...

Anyhow, if a good face mill is out of the question, use a fly cutter. A PCD insert with a 1 mm radius is the ticket. Either way, use a cutter with a nice .03-.06 radii on it.

Regards,
Chuck

The Manufacturing Reliquary

http://cmailco.wordpress.com/

 

[btrueblood]

I know a lot of folks who think high-speed, high-rate metal removal from Alum. alloys is the "done thing". It sure looks cool, with all those chips flying around. Makes the boss think you are really accomplishing something. Best finish, though, at a micro level, will always be at speeds/feeds/depth of cuts that result in the most stable cutting conditions. And you can't have a sharp enough edge.

I.e., go ahead and hog the stuff like crazy, but save the last .005" or so for a finish pass, slow the F down, and pump lots of coolant, both to keep the cutting temp low and to clear chips - saber is on the money with the chip reweld idea too, though the finish there looks more like smeared, lumpy cheese than lunar craters.

Or, you can always plan on letting a vibratory tumbler and polish media cover up the mistakes.

 

[DWilliamA]

@Btrueblood

you mention "slow the F down" do you mean slow the "feed" down or the "F%&^" ... as in a swear word, which is kind of the same thing?

The moon crater spackle look doesn't look like smeared cheese either - the pattern doesnt' follow the mill cut direction - fact is, if it had discernable pattern, I would say it runs parallel with the extrusion. Also the pattern is small when freshly machined, the anodized pattern is larger - as if the rupturing you see in fresh metal is the tops of the iceberg. Anyway - this is all guessing on my part. All I know is there seems to be a connection between the alleged rupture surface and the anodized mottled surface, they occur in the same places.

post processing (vibe / polish etc) isn't an option at this point - besides, if the ductile rupture idea is correct than the rupture defect would only be *concealed* by the post processing, and the anodizing process would uncover it. I think this is true because the anodize process already removes 90% of the cutter lines and the mottling shows, and when the cut lines are completely removed (i.e., part is obviousely etched and anodized heavy and thick) the aluminum becomes grainy and dull matte in finish, as if to suggest that the entire shear zone (with ductile rupturing or smearing in it,) has been removed.

thanks again - great stuff

 

[DWilliamA]

....
also

it appears that the moon cratering occurs more often where the cutting tool is cutting perpindicular to the grain of the extrusion. I just noticed this - we have one surface where the 4 bit cutter which is about 2" in diameter is cutting a 1" wide step down, so basically the cutter only covers have of the extrusion as it makes this cut. The cratering happens mostly where the bits are cutting perpindicular to the grain but diminish as the cutter starts to run parallel with the extrusion. There is a nother surface which is larger that gets full cutter contact and the cratering doesn't occure here.

It has always been a mystery why the anodizing mottling occurs along this edge. And now I can see the 'ductile rupture' or smearing right in this area.

getting somewhere I hope

 

[btrueblood]

Slow the feed, yes. Or the $%%, either way.

Trying to envision what is happening with the partial cut. Is the bottom and side of the "step" being machined in this one pass? Can you try slowing the feed across the end grain, or at least doing a finsh pass for the lateral distance across the step (as opposed to the depth)?

The analogy here should be obvious - cutting on materials with a "woody" grain structure does require that you change the cutting technique to avoid tearing/rupturing those individual grains. Think just like cross-cutting a board vs. ripping down the length, you can really fly in a rip cut and get good finish, but you need more and sharp teeth and a slower feed with good rigidity to keep from having a torn finish on crosscuts...

Anodizing is in some ways analogous to corrosion. Pits will grow, sharp corners will get rounded...it's all about how the electric field changes with the local shape of the surface. As you've noted, the smoother the starting finish, the smoother the end finish. Works in plating too.

 

[DWilliamA]

Yes the bottom and side are being machined in one pass.

The part is 2-1/2” wide and 10” long, the top surface is roughly split in half running the long way with an approximate .100” step down. So roughly 1-1/4” wide by 10” long, 2 sides, 1 side about .100” lower. The first cut (top surface) uses all of the 2” diameter cutter and takes two passes. The second level uses 1 pass but only half of the 2” cutter. The cutter travels the long distance. The final fly marks look almost exactly the same on both sides because the second smaller pass on the top level is removed. We see some moon cratering on the top surface, across the entire surface. We see moon cratering on the lower surface but only the outer half of it. I believe our final pass depth for both surfaces is fast and around .015” thick. We routinely see anodizing mottling on the lower surface and only along the out edge.

I am thinking about the cutter travel path – it spins at 3800 with a certain feed rate, fast I am told – does the travel direction matter? Right now we travel the long distance for both levels and it seems to me that the moon cratering that shows up on the lower level, but only on the outside half near the edge about ¾” wide area where the cut is perpendicular to the grain, should show up on the upper level too? I need to find out if the cuts are indeed the same between both levels.

The wood grain structure analogy is good. I follow this and it makes sense.

We are going to try two things - we are going to try machining deeper into the extrusion. instead of starting out with 3/4" stock and machining off only .075" - we will start with 1" stock and take off more material. There seems to be a difference between cuts at different depths indicating the material is not as homogenous as we would think. And we are going to play around with speeds and feed rates too as discussed above.

I have a good camera and macro lens btw…. I need to take some photos of this.

 

[btrueblood]

Um, sounds like it is making sense to you, which is the only important thing. But you mention cutter travel direction, and it brings up another possible problem - does the cutter change milling direction, i.e. change from "climb" milling to "push milling"? Another thing that might be happening to explain top vs. bottom difference, is which cut is made first. If cratering shows on the last cut made, the problem may be part stiffness, i.e. the part is more free to vibrate and cause tool chatter?

 

[DWilliamA]

Both passes are climb cuts.

A test piece was machined, the material was taken down in .050" steps, 5 passes total each with a .045" rough cut and a .005" final pass. rough cut 50 ipm and the 005" final pass 30 ipm. both passes 4200 rpm. we are using a 2" carbide 4 bit inserts.

this supposedly is the same method we use on the production parts - per our machinest, but frankly it looks like the feed rate is slower, the cut swirls are a little finer. Am I dealing with some dishonesty here? None of the test sample had the moon crater spackle - but there was a consistent grain in the aluminum similar to the moon crater spackle, but the grain was long and straight and looked like the exterior of the extrusion. to describe the moon crater spackle, imagine the long grains as a bunch of tiny rubber bands being stretched out just a little, and you go along and slide them up into little teeny weeny chunks and they retract back into little blobs, maybe like little beads of water, or chunks of cheese, but not so perfect shaped. the long grain lines and spackled moon crater marks look almost the same, related anyway.

I don't know if we she should see the grain at all? The production parts had places where they were spackled grain alongside no grain, the test piece was 100% long grain.

lastly - my machinest insists all this is nonsense. he understands shear zone and ductile rupture enough to talk about it, but he thinks the grain is normal, and whatever good luck with that you are wasting your time. I am pretty confident that no one in their right mind would care about this grain stuff if it weren't for the fact that it causes problems when you try to get a shiny black type III anodize. Never had to deal with this before, so in other words, why would he/we know or even care about? Might be the material doing this, and not the machining method, but we can't get better material - cost a fortune to get SAPA and Kaiser to run a special heat for us, and still, we have no clue what that heat chemistry and or heat treat lot variable is. Dealing with the material we get and working out a solution in machining seems like the best approach to start.

In any case thats my lot to deal with. If I have to send these out to a reputable machinest i will do that.

 

[btrueblood]

Yeah, grain is normal, and occurs in all materials except dead-soft castings. What's important is what you do about it, you can't treat different directions in grainy material the same way, and your test seems to confirm.

"Dealing with the material we get and working out a solution in machining seems like the best approach to start."

Bingo. You know it can be done, the trick will be finding processes that work more often than not, and how to adust the process if problems show up. Yeah, it's arcane knowledge useful only to you in this particular instance, but it's still worth something. Good luck, I think you are on the right path. Amazing how slowing down a bit changed things, eh? Maybe he put a fresh, sharp tool in the machine as well? Is there a procedure to ensure that all the bits are properly seated (all cutting edges have the same radius) when bits are changed, and how often are bits changed? Maybe the final pass needs to be made with a finish cutter, and the roughing passes with the cutter that has already made a few passes, i.e. rotate thru a set of two or three cutters - one on the bench getting fresh bits, one used as roughing, one (freshest edges) as the finish cutter.

Too many machinists, in my opinion, are taught about speed and production and throughput. Too few are taught about craftsmanship, and surface quality, and maintenance of tools and edges and machines...

...but that's just one guy's opinion.

 

[4JawChuck]

Try shimming your carbide cutters out 0.001" and one of them down 0.001", this places one cutting edge out of the horizontal chip load zone but down into the face cut load zone to do the finishing. I'm assuming your using a square insert carbide face mill for this. If your using a triangle insert you can accomplish the same thing by shimming two sides of the triangle on all the inserts but only one side on one insert.

Doing this will break up any harmonic vibrations and put the majority of the chip load for the surface finish on one cutting tooth imitating a flycutting operation. You can experiment with shim stock thickness to get the desired result.

I'm betting your "mooncrater" surface is caused by harmonic vibration, reducing your feed and speeds would be required to get optimal results with the modified cutter setup since your essentially doing two operations at once.

 

[dragracer9]

maybe its in the coolant your using causing an oxidation pattern when drying on the material.

you could also try changing supplier for the material.

 

[DWilliamA]

The moon crater thing is very interesting - very hard for some people to even see it. It turns out that cratering doesn't impact the final appearance of the hard black anodize .... IF .... huge if, if the coolant is cleaned off the parts immediately after machining.

We had 100's of parts that were left wet and scummy and these parts literally corroded and etched themselves. THey turned colors and finger prints were permanently embeded in the surface.... I mean permanent.

I took parts down for Ano and personally watched them soak in sulphuric Acid... for 5 minutes and nothing. The finger prints remained and showed through until the end of the process, right through the Nickel Acetate seal.

So.. if we keep the parts clean - the craters don't show. IF we don't keep them clean, the craters show up in the coolant corrosion which shows up on the anodize.

Also, I have contacted the customers engineers (had to bypass sales and QA) and convinced them that it is not practical to expect Aluminum to react 100% perfect and consistent with the anodize process. If you want perfect looking parts you have to post process them after anodize - like scotch bright, or steel wool and sometimes with acetone. So now we accept some naturally occuring irregularities in the black anodize, and we don't spend 5 minutes per part polishing them.

Now if we can just get the Anodizer to get their act together.... They now make great parts, but every once in awhile they spit out 30% of the parts with brown/gold edges where the black dye doesn't take, or a permenant purple hue where the nickel acetate was soiled, or the smut from anodizing didn't deox before the dye.... who knows. Our supplier who has a relative stranglehold on this process for other reasons I can't go into - sucks. Don't get me started. Good folks working for the company, but the company and its commitment to quality is dismally poor and very very frustrating.

Hope someone learns something from this thread.

 

[saberblue]

I have followed this one with some interest and I have an off the wall thought. Check to see if there is any copper (Cu) in the coolant -- you can do a metals specific test or add a drop or two of sodium parathion (sodium Omadine a fungicide that cleats multivalent cation ions in solution) and see if you get a blue or green precipitate. Your Coolant supplier should be able to help you with either or both.

This would indicate that you have Cu in solution which may be coming from the 6061 -- some times during the forming heat treating process Cu which hardens at a higher temperature than does the Al winds up in pockets near the surface and can then cause galvanic corrosion between the Cu and the Al. These are particularly and issue if the yellow metal corrosion inhibitor in your fluid has been depleted. This problem normally shows up as white rust but your cleaning process could be removing the corrosion products.

Having said this are you sure that you are clearing the chips out of the cutting zone so that you are not getting chip re-welding and or smearing. If this is the case when you look at the surface you will be see what looks like built up edge on the freshly cut surface.

If you have about $250 you might what to look at one of the Dino Lite digital microscopes -- they plug into you PC with their own lighting and zoom to more than 200X the soft ware allows you to measure with reasonable accuracy.. Have use one for several years with good luck.

 

[DWilliamA]

@Saberblue...

I have thought about chip re-welding and smearing. the pattern roughly follows the mottle pattern that exists in the raw aluminum stock itself - I figure the moon crater is associated with this somehow. Also, after about 1/4" you machine through this 'crater' zone altogether. The comment you make about copper, heat treatment and surface galvanic corrosion seems to jive with my observation. Interesting.

back on the topic of smearing and re-welding chips.. it just doesn't look like a build up. Its relatively perfect and consistent across all the parts. i would think we would see some variation. I find the concept of ductile rupture to be elegant. Maybe this copper heat treatment galvanic corrosion scenario is enabling a ductile rupture? I am not sure though, what I can say is that moon craters are definitly there but you have get them in the light just right to see them, right at the junction of a bright shine reflection and a shadow. Real tough for some people to see. If smearing and re-weld is easy to see - then that isn't what we have.

I know we have copper in our fluid - we have machined copper with our fluid. On that note - if we clean the parts immediately after machining we appear to eliminate this problem. Except for one thing.... what if this copper related corrosion still occurs to some degree, maybe the finger prints wipe off, but maybe the edges react negatively still? we have been getting ano parts with gold/brown corners and edges where the black dye doesn't take. I am going to look into the copper/coolant thing again. thanks for the tip.