Thoughts on buy a used Colombo spindle of unknown history?

[wizard]

I've heard this point of view and the opposite one. I'm not saying you are wrong (or right), just that there are various opinions and I have no way of knowing who is right as everyone sounds like they could be when you know as little as me.

The only thing you can cohere is to learn some of the basic equations and get a grasp of the individual spindles power and torque graphs. Once you start to understand that info you can then see how it applies to your machining needs.

I am not confusing finishing and roughing btw. I understand the difference and that high speed machining with smaller bits has been adopted by many manufacturers to improve efficiency and whatnot. As I said, in an ideal world, I would have both (and a super-model wife). The opposing opinions start when you have to choose one or the other for people like me...

Actually this is pretty easy, you need the high speed capability to have any chance at all of doing the molds you are implying you will be doing. That is you need the small tool diameter and the associated high speed to be able to do the features you likely want in these molds. As such you may have to sacrifice bulk material removal rates to make sure you can do high quality detail work.

if if I was milling steel, everyone would agree I needed a CNC mill. If I was only doing wood, everyone would agree it should be a CNC router. Aluminum has the world divided like "fushion cuisine".

"Doing Aluminum" has so many meanings to so many different people that it is hard to make a judgement on what they mean when they say I want to machine aluminum. That might sound like a cop out but the fact is one guy will be perfectly happy with a flimsy router while the next guy really needs a vertical mill to accomplish his goals. What I think we have here is something in-between, that is a router that is more robust than a wood origin machine but maybe not as accurate as a VMC.

Robustness or maybe better stiffness, is very important to a machine the twill be doing mold work because it directly impacts the quality of the final result. I might add the ability to properly follow a path is also a big deal in final quality. I'm confident that a router type machine can be built to do what you want, it just needs to be built better than the average wood working machine.

Anyway, as I've decided to go with the high speed option, it sounds like that is consistent with what you would have recommended so there is nothing for us to disagree on (at least in this group). I have stopped asking questions on practical machinist as all the opposing opinions have started sending me a little insane. People can't even agree what high speed machining means.

Well the issue with high speed machining is that it is often about technique not so much spindle speed.

In any event it sounds like you are on the right track here

[Goemon]

The only thing you can cohere is to learn some of the basic equations and get a grasp of the individual spindles power and torque graphs. Once you start to understand that info you can then see how it applies to your machining needs.

Actually this is pretty easy, you need the high speed capability to have any chance at all of doing the molds you are implying you will be doing. That is you need the small tool diameter and the associated high speed to be able to do the features you likely want in these molds. As such you may have to sacrifice bulk material removal rates to make sure you can do high quality detail work.

"Doing Aluminum" has so many meanings to so many different people that it is hard to make a judgement on what they mean when they say I want to machine aluminum. That might sound like a cop out but the fact is one guy will be perfectly happy with a flimsy router while the next guy really needs a vertical mill to accomplish his goals. What I think we have here is something in-between, that is a router that is more robust than a wood origin machine but maybe not as accurate as a VMC.

Robustness or maybe better stiffness, is very important to a machine the twill be doing mold work because it directly impacts the quality of the final result. I might add the ability to properly follow a path is also a big deal in final quality. I'm confident that a router type machine can be built to do what you want, it just needs to be built better than the average wood working machine.


Well the issue with high speed machining is that it is often about technique not so much spindle speed.

In any event it sounds like you are on the right track here

The point you made on the build quality and stiffness of the machine is the one that has made choosing a spindle so confusing. From day one the advice I have received is to make the machine as strong, stiff, stable and robust as possible for matching aluminum. I have done my best to follow that advice for the frame, gantry and table etc. following that advice for the spindle is not as easy though.

when I looked into what "high speed machining" gear companies were using in larger mold shops, it generally wasn't the router type spindles that we are talking about here. Their 40,000 rpm spindles often look more like this:

Vertical Machining Center Hyundai-Wia ‎HI-MOLD560 - CNC Machine

It's often the guys that use kit like that which are the ones who will tell me that you can't achieve a good finish on aluminum using a router spindle. It's not because they aren't fast enough. They say the bearing and build quality is not robust enough to properly handle the vibration and other extreme forces involved in metal cutting and they include aluminum in that assessment. Again, I'm not saying they are right (or wrong).

I do however get a sense that cutting aluminum using a CNC router spindle is not their core purpose. It's often pitched as more of an afterthought like it can handle a bit of that if you don't push it, or "some light aluminum engraving" after "wood and plastics".

There are also the Datron type high speed spindles which look more like the router spindles we are talking about here but they go on $200,000 machines and they seem to be in a different league.

Anyway, all that is academic because I can't afford a 40hp 40,000 rpm machining center milling head, or a Datron 50,000 rpm milling spindle. In my price range, it's about finding the option with the least amount of compromise while accepting that there will be some extra hand finishing to compensate for my machines shortfalls.

As you mentioned, sacraficing some roughing speed seems like an acceptable compromise for me as I am not looking to start a mold shop.

I have some ideas for adding a second low speed spindle at a later date too, if it turns out that I need the low speed torque. Low (ish) cost BT30 and BT40 mechanical spindles and gear boxes seem to be freely available on eBay...

[louieatienza]

I've heard this point of view and the opposite one. I'm not saying you are wrong (or right), just that there are various opinions and I have no way of knowing who is right as everyone sounds like they could be when you know as little as me.

I've been working with various types of CNCs for the past 20 years, from engraving machines for the awards industry, to lasers, to plotters, plasmas, a few Bridgeport conversions, as well as my own router builds, mill, and 3D printer. So I'd say I know at least a little bit more than you think. I just tried to explain to you why both views are correct, and why VMCs have the rated power they do. They don't use anywhere near their rated power during roughing operations - the HP is a function of the spindle speed.

I am not confusing finishing and roughing btw. I understand the difference and that high speed machining with smaller bits has been adopted by many manufacturers to improve efficiency and whatnot. As I said, in an ideal world, I would have both (and a super-model wife). The opposing opinions start when you have to choose one or the other for people like me...

You HAVE both with a 7.5HP spindle. Just in a smaller scale. I did say this was the better option of the two. Using smaller bits at higher speeds and feeds is more about surface quality and reducing the amount of polishing needed afterwards.

if if I was milling steel, everyone would agree I needed a CNC mill. If I was only doing wood, everyone would agree it should be a CNC router. Aluminum has the world divided like "fushion cuisine".

Back when I started out building my own CNC router, there was no such thing as cheap Chinese spindles and VFDs. I wanted to make guitar bodies, but I also had a need for cutting aluminum. And I eventually figured it out on my router - except a lot of what I found on my own ran contrary to what was told to me then... and there was not a lot of people then saying anything about anything. It was the fad to use hardware store pipe, skate bearings, hose and hose clamp for couplers, u-bolts for router clamps... Most everyone told me you have to run the router at max RPM, take cuts less than .01"deep, yada, yada, yada. I did find that lower than max RPM helped, but the speed control on the router didn't really cut it. Then I discovered that they make single-flute spiral bits specifically for routers on aluminum, and shared my findings here. Now it seems the de-facto standard for smaller bits on aluminum. I even mentioned I got the Amana spiral-"O" bits from ToolsToday, now they advertise here. Then Roman Black came out with the SuperPID, and that opened a new world for me as I could now use 2-and 3-flute endmills at the proper lower spindle RPMs without having the router bog down. Nowadays you can get the spindle and VFD cheaper than a router without the SuperPID. The last piece of the puzzle for me was learning about how CAM plays a role, and why the CAM I was using them (meant primarily for wood) wasn't giving me the results I wanted in aluminum - primarily how the tool buries itself in a corner effectively increasing tool engagement. This would mean running the machine slower than possible to eliminate the tool plowing too fast into a corner. Once I comprehended how certain CAM can actually adjust the tool engagement, feed, and speed, based on situation, I started making better parts. This cost me quite a bit, but I've made it back and then some. Nowadays Fusion360 is free... but it didn't exist when I started.

Anyway, as I've decided to go with the high speed option, it sounds like that is consistent with what you would have recommended so there is nothing for us to disagree on (at least in this group). I have stopped asking questions on practical machinist as all the opposing opinions have started sending me a little insane. People can't even agree what high speed machining means.

I commended you on the decision yet it seems you are the one arguing the point!

As to high speed machining - it's not just about running the machine at the fastest federate with the fastest spindle speed. There's other names for it - waveform machining, adaptive machining... but there are some basic concepts about it.

First is the concept of constant engagement. The stepover (or radial depth of cut or rDoC) is constantly adjusted by the CAM so that the spindle gets a constant load. Since there are less to no spikes in spindle load, theoretically you can run at a higher federate because you don't have to run the machine slower to account for things like inside corners where the tool engagement can double or triple, depending on the angle. As to corners, the CAM will make the tool "nibble" into a corner.

Second, the all hard stops are all but eliminated. If you look at a typical offset toolpath, you'll see there are many "corners" as the tool spirals outwards. Those corners, aside from the aforementioned increased tool engagement, requires the machine to decelerate in and accelerate out. With "high-speed" machining, arcs and sweeps are calculated by the CAM for all motion - even retracts and entering in better CAM. Why? So the machine can run at the highest speed it can. And the side benefit is that it puts less stress on the machine. So as an early adopter in the DIY world (I think I'm the first guy to use high speed machining on a wood-framed router) I found I could push my machine harder than if I used conventional toolpaths, since I didn't have to worry about hard stops and the tool digging in from the intended cut line.

Third, it allows what is called "side-milling" or "peel-milling." Conventionally, the depths of cut were kept relatively low mainly to avoid overloading at the corners. This did allow faster feedrates. But what happens is, it uses the tip of the tool, which in fact is the most inefficient part of the tool. The constant engagement toolpaths allow for a much deeper cut than previous, though at a decreased stepover. However, the side of the tool cuts way more efficient and cleaner than the bottom, and because of the constant spindle loading, the machine can be run faster, which further pulls heat away from the tool and into the chip. Which has the side effect of increasing tool life, which ends up increasing finish quality.

Some CAM even do arc-optimization. Many CAM, for 3D toolpaths, will use point-to-point moves. Most controls have a "look-ahead" feature that will constantly check the angles between these moves and try to keep "constant velocity" through the angle if it is below a certain threshold. Some CAM, however, will arc-optimize the g-code - that is, fit arcs whenever possible where these groups of small moves are. This can hugely decrease the program size, and can make the machine run a lot smoother and faster.

[louieatienza]

It's often the guys that use kit like that which are the ones who will tell me that you can't achieve a good finish on aluminum using a router spindle. It's not because they aren't fast enough. They say the bearing and build quality is not robust enough to properly handle the vibration and other extreme forces involved in metal cutting and they include aluminum in that assessment. Again, I'm not saying they are right (or wrong).

The typical router only has two bearings, and since they're designed for handheld use, the lower bearings only designed to handle light thrust loads. Also many of them have housings made of thermoplastic, or even with some of the aluminum bodied routers, the bearing is in a plastic sleeve.

Maybe the shopping "experience" for some of this stuff leaves little to be desired, but realize this is still a cottage industry. A lot of these Chinese spindles and VFDs weren't readily available at rock-bottom prices like they are today. Most of the industrial linear motion components could be had inexpensively in the day, but you had to know manufacturer's codes to know exactly what you were getting, and sometimes you got lucky on some NOS parts. There was a time when it was still in vogue to buy a unipolar stepper motor drive KIT which you had to solder yourself, and it didn't cost much less than what some digital drives cost today. Heck, even servo drives are relatively inexpensive compared to what some stepper/drive combos were just less than a decade ago. There was a time when guys built CNC hobbled from garage door track, drawer slides, gas pipe, hardware store threaded rod, and harvested steppers and unsupported linear rod from old plotters, even photo-etched their own circuit boards to make their own stepper drives... There are still folks that do that and I think it's pretty cool. But today, more companies are "opening up" to the prosumer/hobby world. You still have to arm yourself with knowledge.

[wizard]

The point you made on the build quality and stiffness of the machine is the one that has made choosing a spindle so confusing. From day one the advice I have received is to make the machine as strong, stiff, stable and robust as possible for matching aluminum. I have done my best to follow that advice for the frame, gantry and table etc. following that advice for the spindle is not as easy though.

That is good, you have some new ideas that I want to see realized in a DIY machine.

when I looked into what "high speed machining" gear companies were using in larger mold shops, it generally wasn't the router type spindles that we are talking about here. Their 40,000 rpm spindles often look more like this:

Vertical Machining Center Hyundai-Wia ‎HI-MOLD560 - CNC Machine

Those spindles aren't cheap either.

It's often the guys that use kit like that which are the ones who will tell me that you can't achieve a good finish on aluminum using a router spindle. It's not because they aren't fast enough. They say the bearing and build quality is not robust enough to properly handle the vibration and other extreme forces involved in metal cutting and they include aluminum in that assessment. Again, I'm not saying they are right (or wrong).

Well they are right to an extent. But here is the thing when they hear "router" they are probably thinking a DeWalt router clamped to an axis with hose clamps. The Chinese spindles are a step above that and you can reasonably go a step above the Chinese spindle if you have the cash. Does that mean a Chinese spindle is perfect - nope just that it is. bit better than the average handheld router.

I do however get a sense that cutting aluminum using a CNC router spindle is not their core purpose. It's often pitched as more of an afterthought like it can handle a bit of that if you don't push it, or "some light aluminum engraving" after "wood and plastics".

Well you don't know what their motivation is. Further we really don't know what your quality requirements are. The only thing you can do here is build to the best of your ability.

There are also the Datron type high speed spindles which look more like the router spindles we are talking about here but they go on $200,000 machines and they seem to be in a different league.

Anyway, all that is academic because I can't afford a 40hp 40,000 rpm machining center milling head, or a Datron 50,000 rpm milling spindle. In my price range, it's about finding the option with the least amount of compromise while accepting that there will be some extra hand finishing to compensate for my machines shortfalls.

The building of any machine is all about compromises! Nothing is perfect even the HLV at work, a work of art that it is, isn't perfect. In the end you have to run the machine with in its capabilities. In that regard it doesn't matter if you buy the machine or make it yourself, there are inherent limits to what a specific machine can accomplish.

As you mentioned, sacraficing some roughing speed seems like an acceptable compromise for me as I am not looking to start a mold shop.

If I understand your requirements and probable machine capabilities correctly, it is the only choice you have. The sacrifice might not be as bad as you think if you employ some modern CAM applications and embrace "high speed" machining. Maybe I can dig up a link or two but I'm about to go to work so that might take awhile.

I have some ideas for adding a second low speed spindle at a later date too, if it turns out that I need the low speed torque. Low (ish) cost BT30 and BT40 mechanical spindles and gear boxes seem to be freely available on eBay...

The more you search the more spindles solutions you will find on the net. That simply due to the wide variety of needs and applications out there. I'm not sure you will need such as mold making doesn't offend benefit, at least not for the size of the machine you are talking about.

[wizard]

Here are a few links that might help with the concept of high speed machining:

1. https://www.canadianmetalworking.com...hining-defined
2. https://www.cnccookbook.com/high-spe...eds-and-feeds/
3. https://www.slideshare.net/nikhilkas...-machining-hsm

Finding info on the net related to HSM is a challenge however a bit fo searching will turn up a few more sites. The important thing is to understand the tit is just one technique for running a CNC mill. How effective it will be for your molds and your machine can't be said at the moment but there is a good chance that it is the path to follow simply due to lighter tool loads.

[Goemon]

Here are a few links that might help with the concept of high speed machining:

1. https://www.canadianmetalworking.com...hining-defined
2. https://www.cnccookbook.com/high-spe...eds-and-feeds/
3. https://www.slideshare.net/nikhilkas...-machining-hsm

Finding info on the net related to HSM is a challenge however a bit fo searching will turn up a few more sites. The important thing is to understand the tit is just one technique for running a CNC mill. How effective it will be for your molds and your machine can't be said at the moment but there is a good chance that it is the path to follow simply due to lighter tool loads.

Thanks. I'll check them out. I know I have a lot more to read up on for this topic.

My quality requirements are way above what people would normally expect from any type of small CNC machine (unfortunately for me). Ideally, molds for carbon fiber parts should have a mirror finish. Even if I wasn't constrained by a low budget, there would be a requirement for a lot of hand finishing and polishing.

The pics on this thread show a good example of what I aim to achieve (after a few weeks of polishing):

https://forum.snipershide.com/forum/...l-empire/page2

I have seen plenty of videos where CNC routers leave a nice finish on aluminum for the edges cut with the side of the end mill but I have yet to see any where they leave a nice finish for the surfaces cut with the bottom of the end mill. As you know (far better than me), they all leave a serrated tool mark pattern. In the videos on YouTube, the tool marks look worst where smaller end mills are used to cut wide flat-bottom channels (like I'll be doing).

I am hoping that by investing a little more for a quality spindle brand and building a stiffer frame that I will be able to achieve slightly better results than your average homemade YouTube machines but I am not super confident in this. I am fairly sure there will be a bunch of manual labor that I won't enjoy at all.

It's easier to achieve a mirror finish on the carbon fiber molds I use now because I can do all the polishing on a plug which is made of softer materials and has a more easily accessible surface to sand before I make the mold. If the plug surface is good then the mold cures ready to use. Hopefully the aluminum molds will save me time and effort when in service to make up for this.

My understanding is that polishing aluminum molds is a whole new topic with an equally long learning curve....

[ger21]

As you know (far better than me), they all leave a serrated tool mark pattern.

That's a function of the toolpaths. Smaller stepovers will remove the scallops, at the expense of longer run times. Using the largest tool possible will give you the smoothest finish.
But a ballnose tool will never give a mirror finish, as the center of the tool is doing more dragging than cutting, as the velocity of the cutting edge at the center is near zero.

[louieatienza]

That's a function of the toolpaths. Smaller stepovers will remove the scallops, at the expense of longer run times. Using the largest tool possible will give you the smoothest finish.
But a ballnose tool will never give a mirror finish, as the center of the tool is doing more dragging than cutting, as the velocity of the cutting edge at the center is near zero.

With a 5-axis machine and CAM, the CAM would tilt the head about 15° to eliminate that zero-SFM condition at thr tool tip... could also be done on a mill with a tilting head provided the mold geometry allows it...

[skrubol]

If you've got large flats, you're best off using a bull-nose end mill (flat with radiused corners,) as the scallop from a ball-nose is going to be hard to polish out. If the flats aren't parallel or perpendicular to the spindle that won't help you though. Five axis machines are nice for this, but quite a different realm..

[Goemon]

With a 5-axis machine and CAM, the CAM would tilt the head about 15° to eliminate that zero-SFM condition at thr tool tip... could also be done on a mill with a tilting head provided the mold geometry allows it...

I looked at that a while ago for this reason but the mold geometry wouldn't allow the flat bottom channels to be cut from any angle except straight down.

This is also the reason why I was thinking of buying an R8 taper milling head - so I could use an end mill of exactly the right width (where possible) and not have to worry about stepovers. The Tormach 10,000 rpm 770 head would have allowed that but at the expense of power, speed and ease of installation.

I have since discovered an adapter that allows the use of using larger diamater end mills on a CNC router spindle. The spindle I am getting has a seperate electric cooling fan and apparently allows full torque and continuous use at speeds of 3,000rpm (or less) while still having a top speed of 24,000rpm. I don't know how well they work but I plan to experiment. I'll be able to use end mills up to 7/8".

[Goemon]

If you've got large flats, you're best off using a bull-nose end mill (flat with radiused corners,) as the scallop from a ball-nose is going to be hard to polish out. If the flats aren't parallel or perpendicular to the spindle that won't help you though. Five axis machines are nice for this, but quite a different realm..

Are bull-nose bits capable of cutting sharp 90 degree wall channels? I am open to trying any type of end mill that will save me having to sand mold cavities (which is my least favorite job ever).

A 5-axis set-up is probably not on the cards any time soon for me. It wouldn't help for any of my current mold patterns and I am already spending more than my budget for my spindle.

Also, I spent a great deal of time and effort in trying to build a super rigid frame to make it as accurate as possible. It kinda seems like that effort would be wasted if I stuck my expensive spindle on a cheap 4th and 5th axis head (which is all I could afford right now). Last time I looked (which was recently) quality 4th or 4th and 5th axis set-ups are all still expensive with no options between the entry level cheap Chinese hobby stuff and the multi thousand dollar pro kits unless I build it myself.

I do have plans for a 4th rotational axis but that is not going to be used for mold-making. I'll be making some nice walnut wood rifle stocks and maybe some aluminum chassis systems but not for a while.

[louieatienza]

A bullnose is basically a square endmill with a small radiused flute end, so the corners they produce would not be exactly sharp but have a fillet; albeit small like .020" or .010" or such... No matter what you do there will be some polishing needed. It should be obvious that smaller tools leave smaller tooling marks, which is why small tools are used for surfacing 3D contours. For flats, a bullnose works well, but also a square endmill if your spindle is perfectly trammed. It's not so obvious that low accelerations cause the tool to "dig" more at the start of the cut, leaving a different "finish" on the part than after the machine reaches the set federate, requiring more polishing.

Of course 5-axis is not practical in this scenario; I mentioned that as how in the industry, that's how they avoid using the tip-center of a ball endmill, since the issue involving using ball endmills was brought up. On a Bridgeport and many smaller column mills, the spindle head can be tilted, which would move the cutting part of the ball endmill away from tip center to a more efficient part, which would leave a cleaner part. But with a spindle that's locked vertical, it means taking smaller stepovers, and using the correct toolpath strategy. There is no rule, however, against using different tools for the finish passes - straight or bullnose endmill for the flats, and ball endmill for contours. Provided the CAM you choose can do that.

It wouldn't be necessary to use a 4th axis for rifle stocks. This can easily be done with 3-axis and the lowliest 3D CAM software with one flip.

I'm sure it can be done, but it may not be best practice to mill a 7/8" slot with a 7/8" mill, since that would involve full tool engagement throughout the entire cut. It may be better to use a 1/2" endmill and spiral down the sides of the slot. This allows the tool to throw chips away from the cut line, where with a full width tool, the chips have nowhere to go but be recut, which is one of the worst scenarios you could have. Which is by the way another benefit of high speed machining and climb milling - the chips are thrown away from the cut line.

As wizard says, technique plays just as important a role. One could screw things up easily on a high 6-figure machine. The right CAM could make all the difference; in fact high speed machining actually benefits more the smaller, less powerful and rigid machines. Most huge powerful VMCs can cut any which way they want because they're so powerful.

[skrubol]

File:End mill types.jpg - Mindworks
The bull nose or flat with radius end mill is basically treated as a standard flat endmill, although you can get them with large enough radiuses to basically be a ball mill with a small flat on the tip. If you need sharp inside corners, they won't do it for you, other than that I try to avoid ever using flat end mills without a radius or chamfer. First reason is tool durability. The sharp corners are generally what fails on an endmill that chips. Round those off and they're much tougher to chip. Second reason (for finishing) is they make stronger parts. Sharp inside corners are weak points in a machined part (if you've ever done FEA on a part with a hard inside corner you'd see the stresses go to infinity in simulation.) Third reason (for floor finishing) is that they leave a little bit better finish than sharp corner flat end mills. If your mill/router is even a tiny bit out of tram, each pass will leave a low spot. With a sharp corner end mill this will show with hard lines you can feel with a fingernail. Radiused corners smooth these out.

For 5 axis mold making, unless your machine is huge, you need to go 5-axis head. I don't think the cheap chinese hobby stuff is in this market, so yeah, probably 10+ times your total build budget for just the head.

[skrubol]

It should be obvious that smaller tools leave smaller tooling marks, which is why small tools are used for surfacing 3D contours.

Not sure that I agree with this.
Small tools are used when they are needed to make a feature. You can't cut a 1/8" radius with a 3/8" ball mill.
For the same absolute stepover, smaller ball mills will leave larger scallops.
Also, using the point of the ball mill requires lower feeds, not smaller stepovers.
I agree 100% about different tools for finish passes. For a many-hour mold cut you might use 3 roughers and 8 finishers. That's assuming you can accurately zero the depth of the finishers though. If you can't zero to tenths (of a thou) probably more like 2-3 finishers, as you don't want many overlapping toolpaths outside of corners that are easier to blend.

[louieatienza]

Not sure that I agree with this.
Small tools are used when they are needed to make a feature. You can't cut a 1/8" radius with a 3/8" ball mill.
For the same absolute stepover, smaller ball mills will leave larger scallops.
Also, using the point of the ball mill requires lower feeds, not smaller stepovers.
I agree 100% about different tools for finish passes. For a many-hour mold cut you might use 3 roughers and 8 finishers. That's assuming you can accurately zero the depth of the finishers though. If you can't zero to tenths (of a thou) probably more like 2-3 finishers, as you don't want many overlapping toolpaths outside of corners that are easier to blend.

While your statement may be true about scallop height, you wouldn't make the same stepover with a small endmill that you would do with a larger one. It would be a percentage of the tool diameter. One could argue that using a larger ball endmill, with the same stepover as would with a smaller endmill, uses more of the tip of the larger tool proportionately relative to that of the smaller tool. I would concede though that for wider, sweeping surfaces it may prove to be time-saving, but only experimentation would yield the best technique for the particular part.

If you can't tilt the spindle head, using the tip is unavoidable where the slope of the surface is relatively shallow - zero to maybe 20-30 degrees depending on the DoC. But what can be controlled is the amount of material removed near the tip versus the rounded section, relative to the tool's size. The CAM should be able to detect flat areas, and you can machine them accordingly with the appropriate straight endmill. For the curved areas, if the surface is a smooth sweep, I may choose a planar (zigzag) strategy, following the flow of the contour. But for more complex stuff I like to use constant-offset, and Z waterline for steeper non-90 degree walls. I work a lot with wood as well, so in that case I do not zigzag as I prefer to always climb mill; there I would use a constant offset path.

I don't think it would be too difficult to zero out the different finishers, since the top of the mold should have already been surfaced flat. As to overlapping toolpaths, that all depends on CAM also. On the roughing side, I use stair-step reduction, so I can machine down the max DoC the machine can handle, then the machine works back up in increments of that DoC. Then if necessary I can go back and rest-machine with a smaller tool, which only machines the areas marked by the "rest-robot" in simulation. Finish passes can be a little trickier, though with a good CAM it's pretty simple to isolate certain regions/features and finish them with the best strategy. I can even set thresholds for example - "steep wall" machining would only machine areas where the minimum and maximum slope is defined. Finally "pencil tracing" around the inside corners saves a bit of polishing in some harder to reach areas...

Another thing to save time would be to buy off-sized cast aluminum tooling plate like Mic-6. Since it's already Blanchard ground, it saves the step of surfacing, and if needed it doesn't take a lot of work to polish.

[Goemon]

File:End mill types.jpg - Mindworks
The bull nose or flat with radius end mill is basically treated as a standard flat endmill, although you can get them with large enough radiuses to basically be a ball mill with a small flat on the tip. If you need sharp inside corners, they won't do it for you, other than that I try to avoid ever using flat end mills without a radius or chamfer. First reason is tool durability. The sharp corners are generally what fails on an endmill that chips. Round those off and they're much tougher to chip. Second reason (for finishing) is they make stronger parts. Sharp inside corners are weak points in a machined part (if you've ever done FEA on a part with a hard inside corner you'd see the stresses go to infinity in simulation.) Third reason (for floor finishing) is that they leave a little bit better finish than sharp corner flat end mills. If your mill/router is even a tiny bit out of tram, each pass will leave a low spot. With a sharp corner end mill this will show with hard lines you can feel with a fingernail. Radiused corners smooth these out.

For 5 axis mold making, unless your machine is huge, you need to go 5-axis head. I don't think the cheap chinese hobby stuff is in this market, so yeah, probably 10+ times your total build budget for just the head.

You reminded of a key point that I was forgetting when I asked about milling sharp corners. These are not just a weakness for milled metal parts, they are also a weakness and a real source of manufacturing headache for carbon fiber parts too. As I'll be starting projects as cad files instead of physical plugs, I'll have the opportunity to remove problematic features like that. I.e. I can round off sharp corners and save all kinds of grief.

[UgraCNC]

Hi Goemon,

Your education is in good hands.

The knowledge shared by Wizard, Scrubol and Louiatienza is highly valuable and expressed in better language then I use

These guys know what they are talking about.

I learned quite a lot by reading this topic.

Regards,

Andy

[UgraCNC]

Upss... duplicate.

[wizard]

I took a look at the thread you linked to above that had a picture of a mold you are trying to achieve. The first thing that came to mind is that there are no flat spots on that mold so worries about flat spots really is not required. Almost every surface on the mold is some form of blended curve or a "flat" area that is at an angle relative to your cutting tool. I don't see where a lot of concern needs to be leveled at the flat areas in the mold.

An interesting statement made in the thread is that they went with a mirror finish to make mold release easier. I would have thought it was for appearances. In a couple of place in the thread something came up missing in translation. In any event it is my understanding that in boat building they apply a layer of epoxy to the molds so that it can be smoothed out, polished if you will for better release and surface finish. This makes me wonder if such approach is possible here and if it would save time.

In any event a lot of hand polishing could be reduced by having the CNC machine do most of the work for you. I know this is possible to some extent but all of those curves do offer up challenges.

I still come back to the question of how smooth do you really need. Some stocks I've seen have checkering and other features molded in so they must have a solution. It is possible the stocks where something other than carbon fiber, its been awhile, but that was the impression i was left with. This brings up another issue, for many of use a super smooth stock is a slippery stock and not desirable at all. This brings up the idea of embracing the small end mills to pattern the stocks surface on purpose. You would need to be creative with tool paths and patterns to make this effective. However if done right it would be as effective as checkering in wood. You would still need some polishing to deal with tool marks but otherwise scallops become a solution instead of a problem. You would still need to polish a good portion of the stock in a regular manner though.

What I'm trying to do here is to think outside the box and not get hung up on how everybody else does it.

Another consideration that might be worth exploring is to make the spindle head manually tilt-able. That might help with texturing operations with a ball end mill but introduces a host of problems on its own. A second spindle fixed an an angle might work just as well.

Another option would be to do the design and ship the design off to a mold maker. Yea a lot of money but he is likely to use a 5 axis machine or at least a high end machine that can handle this without question or speculation on our part. This probably wont fly if you really want to DIY this router but from a business perspective it could make sense. Basically you would be making rifle stocks already if you went in that direction.

Finally one thing we do in the injection molding business is to put high precision surfaces with fancy curvature on inserts that get installed into the larger mold base. This has advantages but in your case leads to far more complex and heavy molds. The point here is that you can increase flexibility with a multi piece mold.

One more final here, have you considered 3D printing? Literally in the last year I've started to see parts come into the plant that are obviously 3D printed. Often these are small limited run parts for fixtures and instrumentation. Often 3D metal parts are cheaper than machined parts if they are of the right complexity. A full length rifle stock might not be possible (not sure what max sizes are) but you may do fine with a multi piece mold. It is hard to say if your level of complexity will make such a move cost effective.