3. What is SFM & Why it's overrated!

[VarkaTheDragon]

All I could think about, while John was going through how to calculate feed per tooth... was what the can of Silly String in the background was used for.

[ChiralSymmetry]

I think I heard the following, but not sure, so take this with salt:

Recommended SFM and IPT are designed to optimize two or three things simultaneously:

(1) Surface Finish

(2) Productivity (machine as many parts as possible per hour)

(3) Cost of production (machine as many parts as possible per tool sharpening or tool replacement). For example, as mentioned in another comment, too fast or too slow will generate too much heat and/or wear out the tool prematurely.

Some of the super high RPMs are recommended because, if they were possible, would help maximize productivity (while maintaining surface finish). Much of the time, these super high RPMs are simply not possible. That just means productivity is limited by the machine, not the cutting properties of the material.

For a small shop, or a home shop, high-throughput productivity isn't important, let alone possible. Instead, surface finish probably matters the most.

If you drop the requirement that productivity should be maximized, then you can use much slower RPMs and still get good surface finish. But I haven't seen any charts optimized this way, for small/home shops. So IPT may be a better way to think about it.

[LogicIndustries]

Hey John, if you really want to understand chip theory and what is going on in the cut, I highly suggest finding and reading a copy of Sandvik Cormant's excellent book on the subject titled "Modern Metal Cutting, A Practical Handbook".

It traces the development of cutting tools from the very beginning up through the cemented carbides of today. Very informative, and very in-depth vis-a-vis how a chip is made.

Used copies come up on eBay and Amazon from time to time for $50-$75, or you can pop for a new copy off Amazon for ~$200.

It's worth the $50, I assure you. If you don't like technical reading, it'll be a little dry, but I found it a fascinating read.

[ChrisDePrisco]

Awesome John, that's a big help! Going by CPT makes a lot more sense; I've always wondered how we should reach the SFM for some of these smaller tools.

[kchigley]

SFM is an under ideal conditions recommendation. Unless you get the numbers from the manufacturer of the specific tool you're using, you're not getting a good picture of what you can do.

For instance, I just received a promotional insert from Seco for cutting low carbon steels. My general goto for turning steel with carbide is 400-800 SFM depending on the alloy. But on this insert, the recommendation was 1600 SFM...which is my baseline for aluminum. Since it was a free insert, I decided to go with their suggestion and it cut like a dream in 1018 hot roll with a beautiful surface finish.

If I had only gone with my baseline (much like your spreadsheet), I would have been leaving a whole boatload of cutting power on the table and not using everything I paid for if I had bought the tool. That's what I use SFM for: the maximum capability of the tool in the material...NOT the absolute target number. Tools will work just fine, for the most part, if they're run at a lower SFM. There are some exceptions to this, but the people who use Tormach machines are not likely to ever see those scenarios.

[ParallaxFPV]

Try U-drilling/turning 17-4, Saf 2205 or titanium. You'll find the manufacturers recommendations are essential for optimum tool life/productivity.

[christophersmith]

Recommended SFM is not derived from a chemistry lab but from cumulative use data that includes a multitude of variables including tool material, part material, lubricant, and so on. While it is frustrating to try to find proper speeds and feeds for smaller tools while cutting softer materials I would never tell anybody that is not important. Aluminum has a very broad range of recommended speeds but that narrows considerably when the materials increase in hardness. Perhaps it would be better to say that chip load weighs more than sfm when machining soft, ductile materials and leave it at that. SFM is greatly appreciated on harder, more unforgiving materials.

[SuicideKang]

My professor just told us to never go over the recommended surface footage on our first year. The thing is materials that need low sfm react different at crazy sfm being cut with hard tool material like ceramic. You can even get away with higher than recommended sfm with chip thinning. I think chip load is where people are really coming short
Example: I have studied suggested drill feeds and speed. I found out that every single drill company suggests the same proportional chip load for each tooth but just change the sfm
How I calculate chip load for a 2 flute drill for ANY common material up to 440c is I multiply the tool diameter by .015 for conservative or hard steels (no jobbers go half) to .02 for fast cutting
Just change the rpms

[mikeatkurzzobelinc.]

I just wanted to kinda bring you back alittle bit. I understand what your saying about about the high end of aluminum tooling which helical does recommend 1600 to 2400 sfm. Now what I want to make clear is yes in aluminum you can basically go with your peak Machine RPM and then do you tooth load calculation. But where it becomes extremely important to know you SFM is when you work with materials including steel. Your machine may do 5100 rpm but if you try and run a half inch 4 flute end mill through a part to make a slot you will ruin it. If you say on the low end its 505 sfm for a slotting cut your speed should be (505*3.82)/.5 which is 3858. And that is just in say a piece of 1018. You start working with stainless 17-4 ph or titanium and you don't use sfm you will throw money right out the window. So I agree with you in the one aspect that yes in aluminum it does not matter but in harder metals it is an important part or programming.

[LogicIndustries]

Two things:

1) SFM does matter and it varies by material. Why? Three letters (words) BUE. What is BUE? BUE stands for Built Up Edge, and is a process whereby the material being cut literally welds itself to the tool cutting edge as it is passing through the material, gradually building up a longer and longer ridge of material extending off the edge of the cutting tool until the length gets too long and the built up material breaks off. When it breaks off, it takes a bit of the cutting edge with it (remember, the BUE is WELDED to the cutting edge), and the process starts over (only with a slightly blunter cutting edge, which makes the next BUE build up faster). This condition is continuous and degenerative in certain bands of surface speed in certain materials. Mild steel and aluminum both have cutting speed bands where BUE is a big problem. So yes, you can ignore the MFG recommended cutting speed (SFM), but you always have to be aware of the SFM band on certain materials where BUE will be a problem. Slower is not always better. Coolant type and application style also influence where the BUE bands land on the speed chart. For instance, BUE is a BIG problem on 6061 alum at 600-700 SFM dry, but if you apply a flood coolant with even a little lubricity, the BUE problem drops off entirely. This is one area where misters and MQL setups are insufficient to the problem. Light mist and MQL air blasts are usually insufficient to ameliorate a BUE problem.

2) Travel speed and therefore IPT are dependent on may things beside just the tool geometry and the material hardness/tensile strength. Feed rate needs to vary according to your tool engagement. Side milling VS slotting, or one diameter DOC VS three diameters DOC, or both at the same time (side milling at one dia DOC VS slotting at three dia DOC) will drastically effect the chip load you want to use when programming that cut. The relationship is NOT linear. Feed rate also needs to vary according to chip evacuation and coolant supply.

As an example, I routinely cut 6061 T651 alum bar stock with a 1/2" dia carbide two flute endmill (std 30° helix, nothing special, no coating) at 650 SFM (that's the top speed of my spindle, 5k rpm), 0.25" DOC (one half dia DOC), full 1/2" WOC (slotting), at a chip load of 0.004 IPT (40 IPM). At this speed, I can trust the machine to run all day with low pressure flood coolant.

If I were to increase the DOC to one full dia, I would have to slow the travel speed down to about 0.0015 IPT (15 IMP) and either increase my coolant pressure or add an air blast nozzle adjacent to the coolant nozzles to prevent tool breakage from chip packing in the tool gullets and chip recutting in the slot.

Twice the metal removal results in a travel speed that is less than one half the original, decreasing overall MRR.

What I'm trying to say is that there are a LOT of variables to consider in this stuff, and a lot of them are interconnected such that ignoring one while varying another can lead to tears (or more probably curse words).

[alonsorobots]

Could you make a video about advice for smaller less rigid and less HP machines like the Avis Benchtop Pro? I’m not sure if your advice on chipload and speeds and feeds applies to it

[tomdchi]

In looking at the specs of systems like the Tormach PCNC systems, I've wondered about that RPM range with smaller end mills/parts in materials like aluminum. Good to know that going with a slower SFM than the standard charts is very doable.

In my case, I'm currently using a Shapeoko 3 with a DeWalt router with an RPM range of 16k to 27k - the opposite problem. It's hard to get the SFM *down* to recommended ranges, even with small diameter (1/8" or 3/32") end mills. Maintaining a reasonable chip thickness without blasting ahead at 100ipm (and deflecting the not-so-rigid gantry) is the problem that goes with that, so I just got a single O flute end mill to try out in aluminum. My take away is "go ahead and give it a try - SFM recommendations aren't absolutes." Thanks!

[earthsurgery]

Where did you get that adjustable chart?

[flikflak]

The about center of the drill is actually not turning at all ( since it's a point and points can rotate)

[lineage]

im running 9, 000 rpm on my diy cnc router for milling hot rolled steel! Going 40 inches per minute, .039 inch width of cut & .39 inch depth of cut. I found that canola oil mist worked best (dont use water), I felt my tool & part afterwards, its cool to the touch & still super sharp!

[Tom-cszf]

Your whole video pertained to machining soft metal. The surface feet per minute is very important for harder metals and metals that produce a lot of heat when machining. You don't need a sliding chart to calculate the RPM, all you need to know is the suggested SFPM and the cutter circumference. And a little algebra. Some metals cant exceed certain SFPM because the speed of the chip removed would cause excessive heat for the cutting surfaces.

[bbarker]

I used a chart that was given to me years ago that got me in the ballpark for manual or cnc machines. We had 6 manual lathes and vertical mills that always used different speeds and feeds for each one even if cutting the same metals. An old machinist told me that no matter what machine you have or how old it is you have to learn what it does or doesn't like and it'll treat you how you treat it.

[mattruth]

3:30 lol I've never seen a dog stretch like that before, but my cat does it all the time

[contraktr]

This is a great series for me to watch, as I've been thrown in the deep end with our new CNC. But please, give that poor dog a pat! ;)

[gsimplecregk]

Dog: "there is a human, in a room, alone, sitting on a chair, talking, articulating... and theres nobody there... hmm... ok, whatever, I'm out a here".