Feeds and speeds - the topic that just won't die

CNC: Altmill
Mill: 1/4" Tapered Ball Nose with 1/16" radius
Material: Basswood (hardwood similar to pine)
CAM: Vectric vCarve

What feeds and speeds do people use for a finishing pass in softwood?

If I follow the IDC chip load calculator for softwoods, I should aim for a chip load of 011"-.013".

But if I follow their bit specs for a similar bit, I should go with a chip load of 0.0016". But this doesn’t take into account the material.

A chip load of 0.01" translates to :
a speed of 10000 RPM and feed of 200 in/min, or
a speed of 20000 RPM and a feed of 400 in/min.

400 in/min is within the Altmill’s specs so should I go with that to reduce machining time ? Or should I go with 200 in/min? Or even slower at 60 in/min as per the bit spec?

Judging from prior posts, it sure seems like feeds and speeds is more of an art than a science…and I ain’t no artist :frowning:

True statement. “Judging from prior posts, it sure seems like feeds and speeds is more of an art than a science…and I ain’t no artist :frowning:

Chipload is only a starting point. At some point, real life experience kicks in and users develop a feel for what works. I know that sounds hazy at best, but it’s true.

Remember that chipload is simply a measurement of the thickness of each chip removed by each cutting edge on each revolution. What it does not cover is every bit as important.

Depth of cut. Even at correct chiploads, it’s easy to see that cutting too deep per pass can overload the bit’s ability to evacuate the chips, resulting in excess heat.

Inconsistency of the material being cut. Sienci’s latest newsletter mentioned that they’re working on an Auto Feeds and Speeds module that will detect the actual load being put on the spindle and adjust the feeds and speeds to compensate for the inconsistently. Let me say here, if this works it will be revolutionary for the market they occupy. Very high end machines have had this, but it will go a long way toward mitigating one of the biggest learning curves for new CNC users.

Tool geometry. A ¼" 2 flute bit is a ¼" 2 flute bit. No. No. No. Two tools that look identical can have very different cut characteristics. Even the same bit over time cuts differently as it wears. The tapered ball nose is a good tool to consider. Is it a ¼" bit or a ⅛" bit? But wait, it doesn’t have a ⅛" cutting edge so it should be more rigid and capable of tolerating higher feedrates without developing chatter. Factor in the depth of cut being typically deeper than a ⅛" tool, more rigid body, etc. and the TBN can cut much faster than the ⅛" end mill. This is exactly why it was developed.

Depending on project, depth per pass settings, material, etc. I would set it to the middle of the settings you mentioned, 200 ipm and then adjust it on the fly by listening to the cut. With a small diameter tool it can be very hard to hear (I turn off the dust collector for this) if it’s sounds like it’s chattering (slow down) or screeching (speed up feedrate).

With some tools, the TBN being one of them, the rigidity of the machine becomes the limiting factor. That’s where you take in the quality of cut.

So, chiploads are a starting point that ALL new CNC users need to be familiar with. They are not the ONLY consideration. We once did a trade show (IWF in Atlanta) where we cut nested ¾" cabinet grade melamine at 4500 ipm in one pass using a 4 flute ½" compression bit (2X chipload). People stood back.

1 Like

I’m a bit with mick on feeds 'n speeds. I see no need for exact figures for my handheld router. I just push it along and see/feel/hear/smell where things are about to go sideways. Since my longmill is the same router out of my hands, I’m left with see/hear/smell. Starting with a speed I know to not break the bit. Usually based on the sienci chart, dialing in the snf rate by shs.

I use a 16th tbn to trace outlines of silhouette pockets I am going to laser etch. I run it as fast as I know the mill not to lose steps, but I never exceed 2mm depth with it, because I need my silhouettes to be as shallow as possible, but deep enough to run a sanding machine over the silhouedge without risking damaging the etch inside.

Tbns can run fast because they kinda are a roughing and finishing cutter in one tool. The deeper you go, the wider the roughing part becomes. Starting at a wee slower speed at the start and speeding up when the first step over is initiated may be a way to run such a tool without actually needing to rough out a complete path with a different tool.

It always feels a bit weird to run such a tiny tipped bit at high speeds, but I have pushed tapers far over their recommended sienci speeds, reducing times by more than half. I feel they can handle even more, but when you see your machine cutting at speeds you normally jogg around at.. you start to feel you’re on the boundary of pussing a tbn into your luck.

If you don’t watch your step and the longmill doesn’t either, things become unique pretty fast.

1 Like

@gwilki Well…it is a deciduous tree. Whether or not that makes it hardwood or not can probably be debated ad nauseam, much like feeds and speeds.

@Chucky_ott I’ve deleted my comment. As with feeds and speeds debates, I should have known better. My mistake.

1 Like

I’d be really interested to know how it goes with the basswood since I also have quite a bit of basswood. This hardness chart seems pretty thorough: https://forestsource.io/wood-hardness-chart

and shows basswood just slightly harder then white pine.

1 Like

@DavidB As I was driving to the cottage this morning, I thought to myself that I should compare the results of various speeds and feeds on a scrap piece of basswood. I have plenty and it would save me from experimenting on my 30"x48" slab. I’ll do that tomorrow and post my findings. I’ll play with the stopover too.

The carve I’m doing is a mountain peak. It’s just a relief carving without much fine detail. I can sand by hand if required to get rid of any lines should I go with a lower stepover.

I just did my first series of tests with the following parameters. It’s on a 4" diameter dome carved in basswood. 1/4" TBN with a 1/16" tip radius. Raster cut at 0 degrees (along the grain). In the picture, starting at the top, from left to right:
5% stepover, 18k RPM, 300 IPM
10% stepover, 18k RPM, 300 IPM
20% stepover, 18k RPM, 300 IPM
30% stepover, 18k RPM, 300 IPM
5% stepover, 10k RPM, 200 IPM
10% stepover, 10k RPM, 200 IPM
20% stepover, 10k RPM, 200 IPM
30% stepover, 10k RPM, 200 IPM

And some close-ups for the 300 IPM feed




Very small difference in quality between 5% and 10% stepover. Perhaps a bit more fuzzies at 10% but the ridges are very small. No noticeable difference between feeds and speeds. Forget 20% and 30% stepover. Ridges are very noticeable and a lot of fuzzies on one side of the dome.

Currently running another test at 5%, 7.5%, 10% stepover but with 400 IPM and 500 IPM feeds. Will post those when done.

2 Likes

I dropped my Z by 1.5mm and went over the 6 domes on the right with different stepovers and feeds and speeds.
5% stepover, 18k RPM, 300 IPM (from prior test)
5% stepover, 18k RPM, 400 IPM
7.5% stepover, 18k RPM, 400 IPM
10% stepover, 18k RPM, 400 IPM
5% stepover, 10k RPM, 200 IPM (from prior test)
5% stepover, 20k RPM, 500 IPM
7.5% stepover, 20k RPM, 500 IPM
10% stepover, 20k RPM, 500 IPM

7.5% has negligeable ridges but there are more fuzzies. That could be just because of the stepover or it could be because of the location of the carve on the piece of wood and the resulting grain pattern there.

If I had to choose based on the single test, I’d probably go with 5% stepover, 18k RPM, and 400 IPM. But I might repeat the test for a 5% stepover on a different piece of wood to confirm.

3 Likes

That’s what I expected. The other factors that will contribute a lot to the cut quality is the species of wood and angle of cut. With the AltMill and TBN I think you’ll be safe to cut really fast on all but very hard woods.

1 Like

Final two tests. One at 5% stepover with different F&S; the other at 400 IPM but with different raster angles.

5% stepover, 10k RPM, 200 IPM
5% stepover, 15k RPM, 300 IPM
5% stepover, 18k RPM, 400 IPM
5% stepover, 20k RPM, 500 IPM





All 4 gave similar results but with progressively more fuzzies in one quadrant of the dome as the speeds go up.

Now with the various raster angle. Same sample as before but with Z dropped by 1mm.
5% stepover, 18k RPM, 400 IPM, 0°
5% stepover, 18k RPM, 400 IPM, 45°
5% stepover, 18k RPM, 400 IPM, 90°
5% stepover, 18k RPM, 400 IPM, 180°





A raster angle of 0° and 180° (parallel to the grain) gives the best results. 45° is the worst. The 0° and 180° cuts were at opposite ends of the board and gave similar results. This would indicate that the grain pattern that I suspected before had limited impact. The 400 IPM 180° cut was at the same location as the 500 IPM cut but gave a better finish. So 500 might be a tad too fast.

That settles it. The carve of the mountain I’ll be doing will have a finish pass at 400 IPM, 18000 RPM, and 0° (or 180°) raster angle. Estimate is 9 hours for a 30" x 48" carve.

3 Likes