Do mechanical engineers routinely analyze the heat in new milling operations?

Factual Questions
1

Seeing CNC machines go through steel like butter: do the designers of the tool heads (which are sometimes replaceable I presume) perform tribological analysis of the normal functioning regime of the tool, such as finite-element heat maps? Or is it more of a test and observe thing based on preceding products?

Do some end-user manufacturers do the same (are expected to) before even setting up the machine station, and compare results, and they order the milling heads and associated design constraints (e.g. HP, torque), in big enough projects, of course, paying for new designs or complete systems?

On a smaller scale, I’ve seen cooling fluid (usually water) in some vids. How fine is the analysis before such an “add-on” is deemed necessary?

Or am I looking at this all ass-backwards?

2

I’m not a tool designer. Rather than analyze each cutting operation, I would guess it would be easier (and just as useful in most situations) to define the operating parameters for each type of tool. You would the look at your work, determine the material hardness, depth of cut, material feed speed, etc. and then select the appropriate tool. Only if that didn’t work for your operation, or if there were significant savings or other advantages would you perform a specific analysis.

You might also do it for critical operations. For instance friction stir welding being used for aircraft parts is probably studied in depth before production.

3

First of all, cutting fluid is almost always used when milling. It acts as a coolant and lubricant, to keep the chips from clogging the mill flutes. Heat buildup is important, but that’s generally easily handled by the fluid (or chilled air in some cases). The proper tool is selected by a combination of material to be machined, finished surface roughness desired, and speed of machining. The operation will have a reasonable idea if the tool is going to get too hot, requiring backing off on feed speed or depth.

4

I thought using a cooling fluid was routine … just a pump and a nozzle is a pretty cheap add-on … I use gobs of cutting oil just for your basic tap or die work … dipping my drill bit in a cup of water between each hole drilled … keeps the cutting edges sharp …

5

You don’t need an engineer, just an experienced machinist. Cutting speeds and feeds, depth of cut and chip size are all part of today’s machine shop environment. The tools use inserts that are indexable to several cutting edges and easily replaceable.

I belong to several machining forums and the knowledge among the “real” machinists (unlike myself) is considerable.
Dennis

6

Unless a machine is mass producing the same part it wouldn’t be worth much investment in heat analysis. Tool heads are replaceable, they wear out over time, although modern materials can give them remarkable longevity making machining more practical than ever in some circumstances. The interaction of machine tools with particular materials is well established so case by case analysis would be rare. OTOH analysis of general cases, for instance the ideal tool speed, cutting depth, and lubrication requirements of tool heads with particular materials is quite extensive I assume.

7

I think you’re overthinking things a bit. The average machinist doesn’t do FEA thermal analysis. The selection of the parameters is largely based on pre-computed tables (“feeds and speeds”) with empiricism and some tribal knowledge.

The systems are designed as much for cost and ease of use than anything. The replaceable inserts you mentioned come in a variety of standardized shapes, and are made from various materials (though usually carbide) and coatings. The machinist reads tables (and advertisement material) about which combinations work best with which materials.

Heat is generally thought of something to be avoided rather than being part of the analysis. Use the fluid or air blaster. Check your chips–for steel, they’ll turn blue if they’re getting too hot. Back off if this is the case. The environment isn’t set up to handle hot materials.

The machines are expensive and wouldn’t be custom-ordered except for the very biggest jobs. The machine selection is going to be based on what the shop has available and the needs of the job. If it’s just a small, one-off part, then maybe one of the older, slower machines is used. If there are lots of parts that need some deep cuts at the horsepower limit, then they use the shiny new machine. The shop is trying to do as much as possible for minimum cost and that means not using more machine than necessary.

All that said, I suspect that for really big jobs, the analysis you suggest is performed. Take the body of an iPhone–they pump out so many of these that optimizing the process by 1% has an effect on the bottom line. So they probably run at the absolute limit of everything, and it wouldn’t surprise me if a thermal analysis was a part of it. Though even here it may simply be empirical.

8

BTW, the cutting fluid is not straight water. It’s generally an oil of some kind, though frequently emulsified in water. The kind I use has a milky look to it, and is maybe 10% oil and the rest water. You can tell when the shop is running low since the fluid will get a little less opaque–they’ve been “cutting” it with more water.

As said, air blast cooling/chip clearing is used, though that’s more of an advanced feature (obviously, the machine needs full containment here since you don’t want the chips flying everywhere). For some materials you still need fluid; aluminum is a bit gummy and will adhere to the cutting tool without it.

9

The people who designed the machine would have done a heat analysis, to be sure that they could get all the heat out in normal use.

The people who wrote the original cutting-rate tables would have measured the cutting rate, tool wear, and surface smoothness. People who re-published cutting rate tables might have done temperature checks as part of a re-analysys.

People who design tool bits would have measured tool temperature, and might have calculated it. Manufacturability is important. After that, the end points are tool wear, removal rate and surface smoothness, temperature is only of indirect interest.

10

This.

To be honest, I never really consulted feeds and speeds tables in my machinist days. I would pay attention to the sound of the cutter and the quality of the cut, spraying cutting oil on a regular basis. I look at feeds and speeds in the same way I look at using a light meter for serious off-camera flash photography (another hobby of mine): some folks will do it by the numbers, while most just look at the back of their camera and dial it in until it looks good. A pro studio guy probably will use a light meter though, and a pro machinist working on the clock will run the chips blue.

One day when I was helping a woman at church wash the dishes after potluck and she said “men always use too much soap and all of the hot water when doing dishes”–that sums up the finesse of my technique in the feeds and speeds department.

Efficiency wasn’t really an issue since I wasn’t in a production shop, my machine work was in an area where we made custom parts for R&D machinery, so there was really no clock ticking.

Machinists in a job shop will definitely consult their Machinery’s Handbook and get the correct feet/min for the material they are working with (e.g. high speed steel in aluminum is something like 600-800 feet per minute). They then figure out how fast to get the lathe or milling cutter turning in order to get there.

Here’s a cool video from Abom79 showing some serious blue chips starting at 13:30. If you look closely, you can see as the chips fall away they start out silver, turn straw colored, then purple, blue, then a light blue as they progress through the steel oxidation stages. This is a good thing since they are taking the heat away from the cutting process.

11

I’m a mechanical engineer who specializes in finite element analysis. In short, yes, you are looking at this backwards. (It’s a good question, though).

The first question one must ask when preparing to do an analysis is “what am I trying to learn from this model?” Your post doesn’t ask a clear question for FEA to answer. Heat maps? Well, the pointy bits get the hottest and the center of the tool is the coolest. Are you trying to design a tool that magically wicks heat away? If not, what do you anticipate learning from a thermal FEA model?

What you’re really proposing is a (highly nonlinear and computationally very expensive) structural model combined with an integrated thermal model. The solver would model the cutting action for maybe a half-degree of tool rotation and then figure out the thermal portion for that small time step, cycling back to the structural side when the temperature model substep has converged. The structural side is the heavy lifting; the thermal part is trivial by comparison.

And what will we learn? That the pointy bits are the hottest. Even in Dr. Strangelove’s iPhone example, I’d be surprised if anyone bothered doing that analysis for the reasons outlined above. Yes, 1% faster machining matters in that situation, but it’s faster and easier to get the optimal speed by running a few dozen of the hundreds of CNC machines hogging out iPhone cases at different speeds and feeds to find the best parameters than it is to pay someone like me to spend maybe 100-150 hours preprocessing and postprocesesing that model. Solving the model would take again as long if you had ~12-24 CPUs involved.

This information is absolutely held by experienced machinists, not engineers. I can run a mill, but I’m a total klutz compared to even an apprentice machinist. Machinists and especially tool-and-die makers are true artisans.

Parenthetically, many machinists are stone-cold jerks to the mechanical engineers they work with. There are all sorts of socioeconomic and class issues at work. There’s also the fact that many engineers are patronizing and egregiously undervalue the experience and knowledge that machinists bring to the equation. Engineers tend to be smart, but in my experience a plurality of machinists are easily as smart as your typical engineer.

Personally, I try to show up with the attitude that the machinists know a ton more than I do about machining and I am there to learn. Machinists often have good ideas that I hadn’t thought of. So I try to embrace a collaborative spririt.

About a third of machinists respond and we have a great working relationship. Another third are indifferent and we have a fine, normal working relationship. But one third just hate engineers on principle and are just looking to show up the whiz kid any way he* can. (I’m 43 but look young. I was in my 30s when customers stopped asking if I was doing a co-op, jargon for “college student internship.”) I used to earnestly try to engage with everyone, but after a certain point, that bottom third is worth neither my time nor theirs. Then again, I imagine many machinists could say similar things about engineers.

  • I have met many female engineers but never, ever a professional female machinist. I’ve met women who were good with machine tools, but mostly they were engineers whose fathers were machinists. I’m sure there must be some female machinists out there.