yes, sorry I forgot to post the link to the actual page, it’s the “the Router way” document in this section
EDIT: I meant this pic and text specifically:
2 Likes
Cool -thanks! Here’s some more of the text “Chip Load
Once the correct tool geometry is chosen, the proper chip load is the next consideration. In mechanical-plastics machining, the recommended chip load range is 0.004 to 0.012 ipt, which results in an excellent finish and acceptable productivity rates (Figure 3). This narrow range imparts the finest finish through the continuous generation of properly sized or curled chips. Inadequate chip load can lead to knife marks, which adversely affect the finish. O-flute tools with a high rake and low clearance help eliminate knife marks by slightly rubbing the part during machining." Sounds like friction on the outer cutting edge bevel is used to melt/smooth the workpiece with their specialized endmills. Acrylic cuts real nicely with a laser too!
Could it be that some manufacturers only provide cutting recommendations because they’re designed for that particular speed? The 1/8” endmill recommendations would result in feeds of 50- 75 IPM, cutting forces of ~3.2 - 4.7 lbf, cutting powers of ~ 29 - 44 Watts, and MRRs of ~0.78 - 1.17 cuin/min. Doubling the speed to 25,000 RPM would enable a doubling of the feed rates and MRRs while maintaining the same forces and cutting powers. Halving the chiploads by reducing the feedrates to 50 - 75 IPM at 25,000 RPM would reduce the recommendation’s cutting forces and powers by a factor of 2.
1 Like
[quote=“Julien, post:28, topic:15778”]
what I hear is that friction and deformation are causing heat (nobody disagrees here right?), heat needs to go somewhere, some of it in the chip and some of it in the workpiece, but anyway the surface layer is just a lot of future chips, so in the end what matters is how fast one can get hot material out of the way, be it with lots of very thin chips or fewer thicker ones. @WillAdams told it better than I just did though [/quote]
IMO the best way to cut acrylic is to minimize heat generation by minimizing chiploads (unless something like Aluminum HSMing occurs.) Since the melting temperature of acrylic is much lower than Aluminum, any HSMing would occur at a much lower endmill temperature.
It appears that early on in this discussion BW wisely gave up on his “deformation” argument. All of the heat is caused by friction. It tries to melt the cut edges but virtually none of it is conducted (goes) into the plastic. Any residual heat will be conducted into the endmill. Here’s plastic experts’ recommendations for cutting/routing acrylic. “The spindle speed required to produce a satisfactory edge is 10,000 to 20,000 rpm. A smooth, constant feed rate of 10 to 25 feet per minute is required to prevent localized heat buildup, which will cause smearing or gumming of the cut edge.”
This corresponds to 120 to 300ipm, make that 200ipm max on the Shapeoko. They do not explicitly specify number of flutes, but they advise 2 fluted tools so let’s assume that. This gives a min chipload of 120 / (2 x 20.000) = 0.003", and max of 200/(2x10.000) = 0.01", so relatively thick chips and quite in line with earlier discussions and with Onsrud [0.004" to 0.012"] recommended range ?
I don’t disagree that the theory calls for lower friction to get lower heat and that thin chips are a way to get that, but many, many people have experienced good cuts in plastics with thick chips, as long as one is feeding fast enough. Maybe the chipload does not matter so much as the fact to get that tool moving out of there, fast?
Apparently they were referring to 2 flute endmills, so you’re right. But, what’s “good for the goose [pro machines] is[n’t necessarily] good for the gander [Shapeoko users]”. Higher than necessary chiploads require higher than necessary cutting forces and produce more heat.
“I don’t disagree that the theory calls for lower friction to get lower heat and that thin chips are a way to get that, but many, many people have experienced good cuts in plastics with thick chips, as long as one is feeding fast enough. Maybe the chipload does not matter so much as the fact to get that tool moving out of there, fast?”
Many of those people apparently use only the tips of their endmills, causing them to wear out prematurely. Is that because of BW’s misguided recommendations or is it compensate for the extra force required to achieve the high chiploads?
Lower chiploads = lower heat = better results.
“Because it is a thermoplastic material, Plexiglas® acrylic sheet softens when heated to its forming temperature. The frictional heat generated by machining tends to soften the material in the immediate vicinity of the cut, and causes gumming and sticking of the tool or tearing of the plastic if excessive heat buildup occurs. When proper speed,feed and cutters are used,machined Plexiglas® acrylic sheet surfaces will have an even, semi-matte surface, which can be brought to a high polish by sanding and buffing”
Phew, I leave on a short trip and come back to a lot of heated discussion here. I’ll answer your initial question to me first:
Thinking in chip per tooth provides a more holistic approach to milling materials. It combines both rotation speed, and flute count. This makes translation to other materials and machines much easier since you don’t need to ask for 3 variables as it already accounts for that. When someone tells you they cut maple at 0.05 mm/tooth, it contains more information than that they ran at 500 mm/min and yet takes approximately the same amount of characters to write out. This does still not tell the whole story (ADOC, RDOC, endmill diameter, etc.) but it’s a step in the right direction.
As for all the other conversation, I’ve done tribological (two materials rubbing against each other) calculations at work more for static considerations but similar concepts at at work here. Let’s consider the thermal energy balance:
(My primary source of non-experiential information is an excellent paper by Abukhshim, Mativenga, and Sheikh on “Heat generation and temperature prediction in metal cutting: A review and implications for high speed machining”)
Created sources of energy
Friction from the endmill rubbing the material (kinetic energy into heat energy)
Shearing heat from the endmills shearing through the material (kinetic energy into heat energy)
Since those are the two sources of heat energy in milling, let’s talk about what happens to it. There are 3 possible options here:
- It goes into the endmill
- It goes into the chips
- It goes into the substrate
Let’s talk about each of these.
1 - It goes into the endmill
This is a significant portion, although smaller than what people believe. In metals, roughly 10-30% of heat generated from cutting transfers into the tool. Given the lower thermal conductivity of plastics (by around 2 orders of magnitude [steel = 16 W/mk & PMMA = 0.25 W/mK]), that number is easily less than 10%. This means that most of the heat does not make it into the endmill but instead is transferred elsewhere.
2 - It goes into the chips
A common belief, and although real, is not the largest component. Let’s get into a breakdown of how cutting happens.
Shown below is a diagram of a tool cutting material. There are primary, secondary, and tertiary heat generation zones. In our common vernacular, we think of the primary zone as the shearing/cutting zone, the secondary zone as the friction/rubbing zone, and the tertiary is rarely mentioned in hobby machining at all. Heat generation into a chip happens in the primary and secondary deformation zones. The secondary deformation zone accounts for roughly 20-35% of the heat generation that the primary zone does, and given what we mentioned in section 1, most of that heat goes into the chip. Now, this is a substantial amount of heat that a small chip has to hold and not melt onto the endmill so the heat transferred into the chip must be less than what it will take to get the plastic to its glass transistion temperature (where it becomes very soft). If you are getting your flutes full of melted plastic, you need to speed up to increase the heat capacity of your chips to resist melting.
3 - It goes into the substrate
This is the last place the heat can go and accounts for the majority of heat generation in plastics especially. The heat generation happens in the primary and tertiary deformation zones, and given how plastics have much larger plastic deformation zones than metals, the shear-heating within the material is large since not only is the area by the chip being heated but material a few millimeters in is being heated as well due to thousands of deformations per second moving the polymers around. The heat generation here is at least 50% of the total heat generation and most certainly higher given the low material removal rates we run at and plastic material being cut.
The big importance to all this is that the amount of heat generated and put into the substrate (to contribute to heating and potential melting) is dependent on the material removal rate. The more material is removed and heat pulled away by chip removal rather than substrate heating, the less likely you will encounter a situation where your substrate begins to melt. This makes us hobby machinists need to push our comfort zones and actually take larger chiploads in plastics to avoid problematic situations.
Source: “Experimental study of the temperature field generated during orthogonal machining of an aluminium alloy.”
P.S. I didn’t touch on chip removal, friction affects from coatings, and more here. Please, look up published articles on these things. Nothing we are discussing here is new within the machining community and there is a lot of good science behind it all.
5 Likes
tl;dr
My comments above are in direct contradiction with @gmack’s comments a couple up where he says this:
It appears that early on in this discussion BW wisely gave up on his “deformation” argument. All of the heat is caused by friction. It tries to melt the cut edges but virtually none of it is conducted (goes) into the plastic. Any residual heat will be conducted into the endmill.
Not all the heat is caused by friction and in fact, most of it is not and most of it is caused by plastic deformation of the material and shear-heating and rubbing. Very little is transferred to the endmill. Take larger chiploads to reduce melting of the remaining substrate.
2 Likes
and then @The_real_janderson dropped the mic 
Very good summary, I’ll go and grab popcorn for the upcoming discussion
3 Likes
Do you have anything to support this assertion? OOPs, I didn’t see you earlier post. Let me read it first? Sorry!
Yep, absolutely. Also, I did pull some information from the sources of my primary source, I just didn’t want to cite everything. There are some pretty fascinating articles out there on this subject and you can get lost in what they discover.
2 Likes
Thanks! Can you cite something that addresses end milling rather than orthogonal milling? How about something that address thermal insulators rather than conductors? 
This is all intruiging.
I for one, would like to know if @LAllen8795 got better results with a higher chipload. I have a few acrylic projects coming up
Also interested in what results Lonnie got.
Not to fuel the ongoing debate, but :
happens to also go for chiploads > 0.004" (3 flute, 10.000RPM, 135ipm). He uses a three flute BUT these are not slotting cuts, so chip evac is less of a problem.
Interesting that he used GWizard to get a “warm and fuzzy” feeling about his speeds and feeds for polypropylene, even though it uses a 11.11 cuin/min/HP K Factor for all “Hard Plastics” (including PVC!) But, it’s somewhat reassuring that GWizard uses 20.83 cuin/min/HP for “Soft Plastics” (including Acrylic), which is pretty close to the 20 cuin/min/HP I measured in my testing! Funny, I would have thought that bullet proof polycarbonate would be considered a “Hard Plastic” rather than “Soft” as GWizard classifies it. 
P.S. do banks in France have bullet proof teller windows like the gun crazy US?
OOPs - forgot to mention that his actual chipload was 0.003"!
Nah, they don’t. Your question brings back a memory of the one time I was in L.A., first night there and I had to go and get some cough medicine for my wife, I had to talk to the guy through an intercom and then he passed the meds through a tiny window 
Back to plastics, and since we have hijacked this thread anyway: I have a large piece I need to cut in HDPE soon, I’ll make a note of the power drawn on the VFD while milling, and see what K factor I get to compare it to those values.
2 Likes
Awesome! Did you get an IR Camera yet? GWizard classifies HDPE (and similar UHMW-PE) as a “Soft Plastic”. It looks more like a “Hard Plastic” to me - but I guess you’ll find out soon enough! 
1 Like
Nope, I’ll try with my kitchen IR thermometer first:
I’m going to fill my enclosure with black chips soon, sorry Ikea cutting board.
2 Likes
Interesting source for plastic. I wonder if it’s cheaper than the $2/lb. that I pay for remnants.
Onsrud says that both UHMW and polycarbonate are soft plastics. Regarding their specialized endmills they say “This tooling has been designed to attack soft plastics with a high rake and low clearance geometry that actually carves the material.” And I thought that all endmills did that! They also recommend conventional rather than climb milling and seem to suggest that at least the larger diameters might have balance issues if operated faster than recommended.
2 Likes
I wish I had a good source for large & thick pieces of plastics, but I have not found any local one yet, so in the meantime ebay and amazon and ikea are my sources. This one is 20$ for 5lb, twice as expensive as what you pay, but it’s 17 ½ " x 11 ¼ " x ¾ ", and I can’t seem to find raw HDPE plates of that thickness on (French) Amazon for cheaper than this.
Than Onsrud page is interesting,
- “The increased feedrates associated with the heavier chiploads increases productivity and dissipates heat”. It’s not clear from the sentence whether the heat dissipation effect comes from higher feedrate AND/OR the heavier chiploads
- the recommendation about using a single conventional cut instead of climb roughing + conventional finishing is the most interesting, I wonder if anyone here can concur that specifically for plastics this is true.
1 Like
[Thread hijacking continues, until Lonnie reappears]
Soooo, since I upgraded to a spindle I didn’t have time to rebuild a dust shoe yet, but I still wanted to get cutting, so I figured let’s run this, put HDPE chips everywhere, it will be fun. Turns out I had to stop the job 5 minutes into a 38 min job, because:
So I would have LITERALLY filled the enclosure with black chips, had I continued.
Since this thread was way too serious so far, I just wanted to make you all laugh a little
A little vacuuming later:
which is the beginning of
which is the beginning of my future allmighty floating dust shoe.
Yes, adaptive is a total overkill in HDPE, but what can I say, I love long chips and those spirally movements.
I went for a chipload of 0.07" to try out that Onsrud range discussed above for a change.
O-flute cutter, 15.000 RPM (just because I was unsure this specific endmill was rated for 24.000RPM), 40ipm, depth of cut 0.5", 2mm optimal load: I was also looking at the current on the VFD, but…the amperage drawn was so low it did not even change when the cutting started. It might as well have been a hot knife in butter.
I’ll speed up everything and retry, but only once I have hacked a temporary dust shoe
8 Likes