Speeds, Feeds, and Strategies for Bench Top Mills

Seeing as this area is specifically for bench top mills, generally meaning small to mid range bench top mills in the posts I see, I figured I'd probably be best off asking for S&F's around here. Mainly in that we are never going to hog like the big boys do, and so those of us who get to work on these machines should be better suited for this conversation. Additionally, I converted to TTS in the fall and while that has been pretty good so far, last night I managed to experience the famous TTS tool pull out twice in an hour. Meaning the TTS tool holder slipped out of the collet. I know many people have converted to TTS, even if you doing manual tool changes.

I don't claim to be a S&F wizard, but over the years I have gotten a feel for it. Since I am running a G0704 that is at this point fairly souped I am aware that I can exceed the limitations of my machine fairly easy, but what I am really trying to establish are general rules of thumb for step down and step over, assuming my F&S are giving me the correct chip load.

Now I try to pay very close attention to chip load, for example, I've had great luck in 6061 with a 2 flute 3/8" carbide endmill running at 36ipm and 6000rpm, providing .003" feed per tooth. I run flood coolant. This is what Harvey and all the major manufacturers recommend for chip load. Last night however I ran this same setting at .15" step over and .3" depth of cut it yanked the TTS tool holder right out of the collet. I caught it, the end mill was pretty much squeezed into the aluminum. I wouldn't say that it was chip welded, but the aluminum definitely started to stir a bit. Luckily I was right there. It should be noted that I don't currently have my PDB on as I was actually machining some new parts for the PDB itself, so I used the standard drawbar and TTS. So lack of PDB force shouldn't be an issue, I crank the drawbar down by hand about as tight as I can with the 4" long wrench.

I noted that there was a good bit of chatter, the part may have been too far out of the vise... It was a 1.25"X5"x"5 plate on end in a 4" vise, I was cutting the 1.25" face. Even running at 30% feed the chatter remained, but the tool pullout issues went away. So the thinking here is that I just had too much stepover last night on the 3/8" end mill, that resulted in chatter, and the chatter may have created enough resonation to help the TTS wiggle loose? Will less stepover ease chatter? If I run less stepover can I get away with more DOC? If I add DOC but remove stepover wont I ultimately see similar tool loads?

I hope that this can be a discussion were we share information. I plan to machine a few more parts tonight and I'll report back with results.

So I was hoping to be able to come back to this post much sooner than I was able to, but over the past few weeks there's been a number of issues I've had to work out that have meant I couldn't really do that.

The first one being that I realized my TTS R8 collet, the 3/4" one, was discovered to be protruding too far below the nose of the spindle. It was a very small amount, but it meant the TTS collets were never getting pulled to the face of spindle as needed. This also resulted in increased runout, which would have made the chatter and pullout worse yet. So i tossed a lathe tool in my vise and cut the R8 collet face down by placing a piece of 3/4" aluminum in the collet and tightening it up. That worked awesome, blue chips were flying everywhere. Keep in mind, I'm still not happy with my offbrand TTS holders as they have substantial runout, but that's something I'll address in the future.

Suffice it to say, now that I've got this squared away I am running my feeds and speeds at normal rates. Additionally, I got my PDB working pretty well, I estimate around 3000lbs of drawbar force. The PDB will flatten the bellevilles, meaning I'm generating 4000lbs of force on the bellevile stack at 90psi. So I may be to run even more preload, but I may have to add a few more washers to preserve tool release and leave enough of an opening for the ATC to reliably insert the tool.

Unfortunately I need to work on my spindle a bit more, but once I get that taken care of, I should be fully operational. I'm hoping my chatter issues weren't feed and speeds dependent at all, and instead were a result of these mechanical issues.

So i've progressed a bit further. I got the ATC running, the PDB seems well sorted. I made a bunch of parts, including a large run on a cylinder head that was quite a neat project.

Here's the ATC going: https://www.youtube.com/watch?v=-lLb5XDjep8

in regard to the runout . have you dialed in the taper or are you just dialing off the tool itself . If the collet nuts aren't the bearing type then they are prone to a certain amount of twist on the collet and both can get offset some . what resolves this is to tighten the nut and give it a few taps with a soft hammer on different spots . That should help to set everything into position , then give a finish tightening .
I've got piles of the tormach "style" holders and a few were off slightly but most of them are good , but , when they run out then I loosen them some , give the collet a twist then do what I mentioned above .

Btw nice setup .

I checked the tool holders on their taper. Once I got the collet cut down and the tool holders started pulling up tight to the spindle the runout got better. It's still not great, but it's workable. I'll buy some better tool holders eventually, either the verified ones from Novakon or Tormach ones. I played around with trueing tools like you describe, but am not quite there yet. Once I get a few more small things worked out that will be on my list of standard practices.

Thanks for the compliment, it's still a work in progress... lol

I may be wrong but I recall you saying in another thread that you've beefed up with a strong spindle motor . Did you do an off the shelf mod or was it something you pieced together ? I sort of retired my 704 mill but I might redo it and beef it up

I bought a 2.2kw (continuous) BLDC with VFD from Taiwan. I machined a mount for it with the old motor, and adapted the pulley from that conversion. It uses a small V belt, maybe too small, but I've not had any issues with it. The entire setup is quite quiet and the motor is very powerful. I haven't stalled it or overloaded it yet, so I guess it makes enough power. The conversion was fairly straight forward, but I kind had a jump on it because I got to use parts from my previous belt drive conversion. Cost for the motor was around $800 after import, but it runs 6k and is fairly light. There might be cheaper options, but this one has worked really well for me.

https://www.adlee.com/en/product-552...-AC-Drive.html

I got the umbrella door hooked up again. It damps the ATC carriage nicely. Just makes everything a little less violent on 90psi shop air, and when fully assembled it protects the tools from chips and coolant, which can be quite a problem. Once I get it all assembled again I will take a better video, I had to pull the carousel down to work on the door assembly, and while I'm in there I will adjust some other bits too.

https://www.youtube.com/watch?v=Lrs1AeLWRQM

I've been cranking away on this project. Mostly just trying to get the ATC dead reliable, but I have also machined some parts. I had to machine some flats for a wrench on a shock perch (actually a short run of perches), I modeled them at .7355" and they all came out damn near spot on. Even with the micrometer it was only about +/- .0002". I was really impressed because when checking things out I've realized that my X axis jibs are a spot loose. So I can't complain.

All of that said, the ATC is really getting there. I went through several hundred tool changes today. It's gettin very consistent, but the little hiccups are still filtering in. This is a really good thing. In endurance racing you realize you have to work things until the unexpected happens. I have found a few of those items in these test runs. As they get worked out, the tests runs will get longer, the errors fewer, and eventually the test runs will turn into manufacturing runs. I'm very close.

https://www.youtube.com/watch?v=O_5vVTIhDt0

Just a casual observer, but would it be possible to slow/throttle the speed down as the ATC approaches the spindle for the tool change. It really shouldn't be any faster than the retract speed..IMHO.

That setup is really well done..good job.

Stuart

I've never had any luck trying to get any kind of 'decent' DOC on my mill even with using a .003" chip per tooth. (I use mm, convert to inch, then convert back). I'm limited to 3000rpm max and trying to go 1/2 tool diameter DOC does not work. The amount of bits I've broke is stupid.
I don't go for carbide due to the machines rpm limit, they're too brittle to even lightly hog so now using cobalt instead. I have flood.

If I use a 6mm bit for instance, go 3mm DOC, 0.003" load, it goes bang. If I then go, 1.1mm DOC I can easily go 0.004" load and likely even more. The time difference between the 2 according to simulation is minimal. This is under slotting with a 3 flute.
It's a new build so still playing around atm.

When doing a pocket, if I stick to the 1.1mm depth per 360 degree cycle by adjusting the ramping degrees. I can ramp smooth profile straight down the center then expand the pocket as needed at 0.4mm stepover all at the same 0.004" load with a 0.2mm slower finish.

It's all trial and error to me, every machine is different. I'm still cluless :confused::rolleyes:
With an 8mm bit I've only tried 1.2mm DOC so far at the same 0.004" load.

Quote:

---

Originally Posted by atomarc

Just a casual observer, but would it be possible to slow/throttle the speed down as the ATC approaches the spindle for the tool change. It really shouldn't be any faster than the retract speed..IMHO.

That setup is really well done..good job.

Stuart

---

I think the issue is that on the retract side the linear rails run a bit tight. So when insert action is initiated, there's a bit of a slingshot effect in that at first the carousel doesn't move, but as pressure builds it overcomes that stiction, and it shoots off. If you watch closely, the carousel also is fairly slow to actually overcome the cool grippers on the insert side at times. It's all down to that stiction on the retract side. Likely today I'll pull the thing off the mill and see if I can't get those rails running smoother. Once that's ironed out, I should be able to adjust the need valve on the insert action and slow it down a bit. I will also probably add a bit of time to dwell period on the spindle lifting after the retract.

Quote:

---

Originally Posted by dazp1976

I've never had any luck trying to get any kind of 'decent' DOC on my mill even with using a .003" chip per tooth. (I use mm, convert to inch, then convert back). I'm limited to 3000rpm max and trying to go 1/2 tool diameter DOC does not work. The amount of bits I've broke is stupid.
I don't go for carbide due to the machines rpm limit, they're too brittle to even lightly hog so now using cobalt instead. I have flood.

If I use a 6mm bit for instance, go 3mm DOC, 0.003" load, it goes bang. If I then go, 1.1mm DOC I can easily go 0.004" load and likely even more. The time difference between the 2 according to simulation is minimal. This is under slotting with a 3 flute.
It's a new build so still playing around atm.

When doing a pocket, if I stick to the 1.1mm depth per 360 degree cycle by adjusting the ramping degrees. I can ramp smooth profile straight down the center then expand the pocket as needed at 0.4mm stepover all at the same 0.004" load with a 0.2mm slower finish.

It's all trial and error to me, every machine is different. I'm still cluless :confused::rolleyes:
With an 8mm bit I've only tried 1.2mm DOC so far at the same 0.004" load.

---

If I could readily get metric tooling here I would have gone metric on the setup of my machine. However, I am constantly converting back and forth between metric and imperial so I'm not sure what going for metric would have gained me, because I'd still have to work with imperial stuff all the time.

I think .003" chip load on a 6mm end mill is too high, I'd go .002", or even .0015". I use a lot of .25" (6.25mm) 2 and 3 flute end mills and have good luck with them with that chip load, very long tool life and great finish. Even with a 2 flute end mill, there just isn't enough space in the flute to evacuate a .003" chip IMO. On the other hand. .003" chip load with a .375" 2 or 3 flute end mill seems to run very well. I can run at 100% tool diameter depth very comfortably. If I'm feeling lucky I've gone as much 120%, though I know many claim that 150% should be achievable. For slotting at full radius engagement, I cut all those number to like 10%. I try to stay away from slotting, instead I like to program toroidal tool paths and maintain constant engagement. It's probably just as slow, maybe even slower in some cases, but tool life seems better as I get to use more of the tool.

All of that said, switching to the power drawbar and TTS is still in progress for me. I'm not super comfortable to just the thing fly like I did with R8 and set screw holders. In theory I should be able to get the same performance, but since I built the drawbar and frankly have no Idea what I'm doing, I have decided to sneak up on it rather than break stuff. Of course I dropped a brand 3/8" new end mill out of the spindle when I accidentally released the tool and it smashed the corners off all three flutes, so there's always my fat fingers to break things when my brain isn't making it happen.

Using undersize cutters for internal radiuses is optimal. Using metric end mills for imperial parts and vice versa is very useful.
A 6mm end mill is handy for 1/8" internal radius pocket, instead of a 1/4" end mill. 12mm instead of 1/2" endmill.
3/8" for 5mm radius (I.e. rather than a 10mm end mill).

I have a hard time finding decent price metric end mills in the U.S., or rather, good quality metric end mills at a decent price. That said, that's a great tip and not something I hadn't ever thought about. Now that I use more ER20 it's a possibility, in the past I was well setup with imperial set screw holder so going up or down in size wasn't as easy.

Are you using coolant or mist? I use a home brewed fog buster with a 1/16" outlet (1.5mm) at 90PSI to blow the chips out of the slot.

On a slot like your example I'd use a 1/4" HSS end mill with .060" DOC, .0025" IPT. I'm sure that others can push harder, but I tend to do small lot production and breaking tools eats up a lot more time than taking it easy on the machine does.

On your machine I'd go 1.06 M/M, 1.5mm DOC (1/4 D), and use ramping if you can.

Keep in mind that this isn't an industrial machine. It weighs 1/10th of what even a small industrial mill does. Counterintuitively, It just isn't stiff enough to feed small tools at their limits. The amount of flex in the castings means that a small tool could easily see 200% or more of what the programed feed per tooth is.

I have no problem flogging a 3/8" 3 flute at 6750 RPM, .75" DOC and .004" IPT, but I keep the WOC to .025" to keep the chatter in check.

Quote:

---

Originally Posted by dazp1976

I've never had any luck trying to get any kind of 'decent' DOC on my mill even with using a .003" chip per tooth. (I use mm, convert to inch, then convert back). I'm limited to 3000rpm max and trying to go 1/2 tool diameter DOC does not work...

---

Quote:

---

Originally Posted by ChrisAttebery

Are you using coolant or mist? I use a home brewed fog buster with a 1/16" outlet (1.5mm) at 90PSI to blow the chips out of the slot.

On a slot like your example I'd use a 1/4" HSS end mill with .060" DOC, .0025" IPT. I'm sure that others can push harder, but I tend to do small lot production and breaking tools eats up a lot more time than taking it easy on the machine does.

On your machine I'd go 1.06 M/M, 1.5mm DOC (1/4 D), and use ramping if you can.

Keep in mind that this isn't an industrial machine. It weighs 1/10th of what even a small industrial mill does. Counterintuitively, It just isn't stiff enough to feed small tools at their limits. The amount of flex in the castings means that a small tool could easily see 200% or more of what the programed feed per tooth is.

I have no problem flogging a 3/8" 3 flute at 6750 RPM, .75" DOC and .004" IPT, but I keep the WOC to .025" to keep the chatter in check.

---

I'm using flood coolant with a 350w pump. Gives a very good blast through 4 nozzles. It ended up that powerful I had to make an adjustable bypass line.
I'll have to play around a bit more with it and try those suggested settings. I'd like to have a faster spindle already tbf.

I do use ramping where I can. Been trying some pockets on smooth profile settings circular ramping down the middle. With downward degrees set so that it cuts 1.1mm depth per 360 degrees. Then at the bottom expanding 0.4mm woc to the right circumference with a 0.2mm finish. Gave good results so far and no chips to be seen near the pocket.

Quicker than using multiple depth stepdowns of 1.1mm.
I'm still learning this stuff!

I run flood as well. Currently with 5 nozzles. I'm not sure I'd call it high pressure, but it is high flow. I use a 1400 gallon per hour water feature pump. It's not really made for high pressure. That said, it really puts coolant out. When I went to coolant my results improved across the board. My current enclosure deals with it ever well too. I need to add a bit of a ceiling on the thing as over long runs I do loose something like a gallon per five hours of run time just to mist and evaporation.

https://www.youtube.com/watch?v=LLqZpaXqlsk

I've got the TTS working pretty nice these days. Drawbar pressure must be fairly high as the bellevilles are nearly flat when releasing the tool, so when I do that math that puts drawbar pressure somewhere around 2800lbs (my best guess). But performance is good thus far. I have managed to take a .5" deep cut at 54ipm, .003" per tooth, with a .375" 3 flute carbide end mill. No chatter, no carrying on, just big beautiful chips.

Yesterday I was machining a new shift lever for my RD350 and all went well, but when the coolant turned off it did so permanently. The pump blew up. So to continue getting work done i did some dry cutting but hosed the parts down with the wash down pump. I run two pumps in my tank, one for coolant during milling, the other I run via a foot pedal and it has a hose so I can wash down the enclosure. Turns out I accidentally added a spare pump to my setup. Running without coolant gave me the opportunity to record some video without the coolant obscuring the action. This video is kind of lame, the last 20 seconds are the best part, but sound when cutting is just right for 6061. If you catch the clattering in the video, that's the tools rattling in the tool changer. Not sure what I am going to do about that? Maybe just get used to it.

https://www.youtube.com/watch?v=hQiY0WyoRwA

Also, the new Steep and Shallow tool paths in Fusion are awesome.

https://i.imgur.com/s6bCDj0.jpg

https://i.imgur.com/44kbhzt.jpg