Chatter when milling aluminum

So here’s what I am doing right now:

My configuration:
X-carve 500mm
24V spindle (not so quiet)
NEMA17 motors running at 2A (doing fine)

Machine has been operational and cutting since yesterday, so total x-carve n00bie here. I did do a couple of successful cuts in both MDF and thinner aluminum already, using a 2mm mill bit.

The aluminum is a 5mm thick cold-rolled aluminum sheet double-sided taped to a 1/2" MDF build plate. It’s going in with 0.1mm cuts at 300mm/min. Disclaimer: I know I’m running it much faster than the default settings. I’d like to use this machine for lots of aluminum cutting in the future, so I’m trying to find out how best to optimize the machine for that purpose.

Toolpaths are standard Easel generation, original drawing in Catia, exported through Inkscape.

I’m obviously getting chatter here. It’s not particularly bad in the video, but when I attempt 500mm/s it’s mill-dullingly bad. What is holding me back here? In the video you can clearly see the chips on the board vibrating, so part of the blame is no doubt the stiffness of the hold-down plate. However, I’m also seeing significant deflection on the mill bit itself, much more than the TIR of the collet. So the frame is clearly flexible as well.

What’s my best course of action to improve this situation? Should I try to stiffen the frame? Should I install a better hold-down plate (I’m eyeing a vacuum plate from if that’s the case)?

I read that you are generating the tool paths with Easel. I am guessing that you chose the Aluminum material selection and went with the default feed rate and depth of cut. If not, can you give your feeds and speeds and type of bit and rpm?

Looks a little fast to my eye for the 24v, but it would be better to look at the numbers.

Yep, it’s the standard Easel settings but with the default 125mm/min changed to 300mm/min. I’m deviating from the standard for testing and frankly because this part would take all day at default speeds.

Edit: Would it make sense to grind down the shaft of the mill bit so that it can be inserted into the collet all the way up to the neck? That would reduce the flexibility of the bit itself. Or is this just detrimental to tool life?

I would certainly look at getting to tool much further into the collet.

I would also look at reducing the feed rate down.

It sounds like, and I’m going purely on the sound, that the workpiece could do with some more clamping. Easy test, start cutting then being very careful, lightly hold down the material with your hand close to the cut. If it gets quieter /less chatter then look at more clamping down.

Alright, well. The piece in this video is still being milled. I do need a second one, so once it’s done I’m going to grind down the mill bit, reinforce the MDF bench with some steel or something (I still have a 500x400x6mm steel plate doing nothing) and try again. I’ll probably check back in like 1.5hrs

Couple tips - search forum for “width of x-axis makerslides” and you will see very simple rigidity modifications. It is a long thread but many folks are performing this mod: Width of gap X axis makerslides

Also - if you can drill a few holes through your stock and screw them to the wasteboard you will be much better off.

Yes - minimizing tool stickout has an amazing effect on your cutting.

I had a the same problem. I upgrade to the DW611 router and spindle as well as the x-axis stiffening and I cut Al like butter in comparison.

I am slammed right now but i will check back on this shortly after I knock out something this AM. But there area a number of things to cover that will help you out. First off please let me know the type of end mill you are using. How many flutes? I understand your desire to not take forever to mill your aluminum, but with a stock XC you are best off working your way up from a slower speed instead of down from a faster one. I am also guessing you are milling with a HSS end mill vs carbide?

I have to step out for an appointment but I will be back shortly and able to give you more advice on aluminum milling shortly. I do a great deal of aluminum milling and I should be able to give you a hand later today to help you out. To give you an idea, here is the type of work I do with my Shapeoko and X Carve.

I have a conference call right now, will answer you after that’s all done. Thanks for the great and fast help by the way - this forum is awesome.

While I’m waiting for lync to connect: Yes, I’m using crappy chinese $2 HSS 2-flute endmills. On purpose, I just assumed that the first thing I’d do on this machine is mess up a setting and ram 5 mills into oblivion within a minute.

Ok, what rpm are you using?

Minimizing the the tool stick out will help, so definitely do that.
Better clamping is good too.
It might sound silly, but you might try lowering rpm and lower the feed rate to 270mm/min.
How many flutes does your bit have?
Is it a spiral up cut?

Getting a beefier spindle will help and doing the stiffening mod will help the beefier spindle, but that doesn’t help you with your next job. If your 24v spindle is cutting other stuff fine without any tell tell signs of being a dud, then this is a problem solvable in the mean time with your current setup.

Here’s another thing you can try before the next job:

  • After inserting the bit into the colette, make double sure that it is dead centered in the colette
  • With a sharpie marker, marke a line on the coupler that your top wrench torques on
  • Now with hands out of the way, flick the switch to manual on
  • Listen and look for runout. If there is any:
  • It will sound like there is more vibration than normal
  • The solid line that the edge of the bit makes as it goes down away from the colette will get visibly more blurry down that line
  • If you hear or see runout then try this:
  • turn off the spindle
  • bring the line that you marked around to the front
  • tug on the bit/colette forward and to the right a couple of times (not too hard, not too soft)
  • hands out and turn on the spindle again
  • if it sounds/looks better then you are heading in the right direction, if not try another direction to tug or push in always with the mark facing forward
  • rinse and repeat until you find the sweet spot where the spindle is running at it’s quietest

Hope that helps.

Truth be told, the 24v 300w spindle cannot handle fast aluminum cutting. Your are going to be stuck with going extremely slow. Also chip clearing is extremely important, recutting aluminum chips will cause vibrations, which leads to chatter. Like @Earwigger suggested, the X-axis stiffening mod can help a lot with the rigidity of the machine.

If you want to get serious about aluminum cutting, you’ll need to upgrade to NEMA23s and the spindle to at least a dewalt (but a VFD would be better). And if you’re upgrade those things, you may as well do the stiffening mod while the machine is taken apart (it costs ~$15-20 in material from home depot, and an hour of your time).

Eric is 100% right that you will need a few changes to your mill to do aluminum. We both do 90% or better aluminum only milling and the stock mill and spindle will not work well enough.

Your end mills will make a huge difference. I HIGHLY suggest milling with a GOOD 3 flute end mill designed for aluminum work. Do NOT use a AlTiN coated end mill for aluminum milling. It will bind with your stock and break for sure. TiAlN Titanium Aluminium Nitride end mills will work, but the best is a MoS2 Molybdenum Disulfide coated end mill for aluminum work with 3 flutes. If you are serious about wanting to mill aluminum on the XC then you need to take every advantage you can in end mills, spindles and mill upgrades.

You will need to follow Eric’s advice and upgrade your steppers to NEMA 23 sizes. You will also need to upgrade to a Dewalt 611 or a VFD spindle. I highly advise the VFD spindle over the Dewalt 611 personally. Though many folks have the Dewalt 611 and it works OK. But if your main focus is going to be the milling of aluminum on the XC then I would go with a VFD .8kw water or air cooled spindle.

You will HAVE to work out some sort of chip clearance system for your setup just like Eric spoke about. Most folks use an air compressor and use it to manually blow the chips out of the way. For larger jobs this is not practical due to cycle times. If you have access to a 3D printer then you can make a mount to hold your system like this:

At the very least you will need to attach the air outlet to the spindle in such a way that it can keep the toolpath clear of chips. A vacuum shoe is good, but the forced air often does a far better job of clearing the toolpath and the chips can be removed by vacuum manually with ease. This will also cut down on the vacuum noise for longer cycle times. Even a simple cage blower fan can be adopted to use forced air to blow the chips out of the toolpaths. I use this sort of system over a compressor since I mill in a room inside my house VS a workshop.

You will need to fasten down your working stock much better than with just double sided tape. I double side tape all my aluminum stock, but my first cycle is to drill fastening holes and then screw down the stock into a sacrificial waste board. This will keep your stock steady while milling and reduce the chatter a bit.

Your stock spindle is not going to mill very fast at all in aluminum. I started milling aluminum on the Shapeoko 2 with a 48V 400W DC spindle at .12MM DOC and about 135MM/M and it took forever, but it worked and allowed me to sell enough products to pay for the costs of the Shapeoko and to buy upgrades for the X Carve. (in full disclosure , Inventables sent me a X-Carve with NEMA 23 motors for testing of aluminum milling). The stock spindles are just not powerful enough to cut at the rate you are desiring to cut at and they WILL fail much sooner than later.

Get back with us with more info about your end mill and such and we can try to help you out but you will need to make some upgrades to your mill if you wish to mill aluminum daily.

1 Like

I can’t help but comment again: @Travelphotog this set up is so freaking cool it should have its own movie.

Thanks! It is just what is needed for me to get my system up and running to mill the way I want to. I need to reprint the forced air clearance holder and give it a bit more support around the spindle. Having the 3D printer really helps to keep thing cleaner on the setup and to try out new ideas. I also need to print some holders for the air hoses to make it a cleaner look by holding them together better.

I am debating doing on post on the controller and a few other mods I did since I keep being asked about them and it is a good deal to type up.

You fixture post was great also. I think it all has to do with milling AL on the XC and the mods you need to make it work. I enjoy the challenge of getting a mill made for wood to work in aluminum on a large scale.

I can see some Chris Bathgate style anodized aluminum clearance and hose holders. Then give your machine its own Facebook page.

If I can ever get caught up on production I can have some fun like that! I sort of miss playing around and anodizing for fun and to goof off. I need to take a few days and relax a bit and get back to the fun stuff on the mill for a change. I still want to mill my 65mm spindle mounts and anodize them! I like the mount I have and it works great, but a nice anodized one will of course look and work better!

Dude, look at this mess. I have never been one for 'aesthetics."

1 Like

For a moment, that looked like strips of duct tape around the bottom. Not a bad idea, might be able to double it up enough to make it soft, yet resist the vacuum sucking it into the blade… :smile:

How much was the tach, how much does it weigh, and how much do you really use it?

That plastic is “carpet runner” and cost $.60 and I have a lifetime supply. It works perfectly, even with a couple legs missing.

Everything is homemade. I started using the tach to find sweet spots to increase speed for really long cuts with various bits in various materials. The Hitachi goes from 7k to 24k so had great variability. It took a WHILE to get the code right so it could measure 24k, but here is a step by step. It only ways a couple of ounces, but I used a full sized Arduino. You can use a microview arduino and it would be the size of a quarter.

1 Like

Alright, the second part has been set up and is milling right now. First one came out pretty well, aside from one snafu at the start (steppers running at 1/4 power and missing steps). But of course the edges were pretty rough because of the chatter. Surprisingly, the mill bit was fine. No real damage, a very small ding on the outside edge of one flute but that’s it. I’ll take pictures of all of this after dinner.

I’ve screwed down the new plate with 4 M4 screws at the corners which at least subjectively feels a lot better. Actually, the tape I used on the previous part definitely came loose in a corner, because there were little chips on the tape when I took the part off. So note to self: tape = definite no-no. I’ve also reduced the shaft length on a fresh 4mm mill bit by about an inch, it is now only just sticking out the collet. The result?

Much better certainly, I’m seeing almost no big chatter anymore. The milling operation is still not what I’m used to with big mills (i.e. ‘smooth sound’, like it’s actually cutting) but it’s certainly not disconcerting anymore. I’m at least confident that this part will come out well tonight.

Still running at 300mm/min (against your advice, sketch42).

I’m aware that the stock x-carve isn’t ideal for aluminum, but honestly I’m pretty impressed it can churn out this pretty large part in 3 hours. The objective with this machine is mostly to have a starting point. As I get enough work to justify the investment, I can upgrade the parts. And I’m very happy for the wealth of advice in exactly what to get.

I’ve got an MQL and chip clearance setup lined up but not installed yet. As I said, the machine has been milling for close to 12 hours and it’s not even 36 hours old yet :stuck_out_tongue: