Router flywheel to quiet chatter

970 Views 22 Replies 5 Participants Last post by Souza1969, 8 mo ago

doorstory

Discussion Starter · 8 mo ago (Edited by Moderator)

I am considering building a CNC router. I am in the wrap-my-head-around-it stage.
Has anyone seen the use of a flywheel added to a router to improve the cutting? If speed is fairly steady, I think there would be rather little impact on energy. It's possible that some favorable damping could be obtained through flywheel construction, I haven't really thought about that. Are such flywheels built into any routers now, that you know of?

The idea is the added inertia would even-out the load experienced by the motor. Obviously the motor rotor has rotational inertia: the question here is could such an added flywheel allow for a given motor to improve it's cut. My assumption is that the flywheel does not need to be very large or heavy, though I haven't made any attempt at evaluating that.

This is somewhat related to the flywheel action of a large-diameter bit, but in that case the cutters' relative distance from center mean that the router bit body alone can't comprise sufficient rotational inertia. This explains their sometimes-difficult behavior.

The major con is the added inertia with speed changes. If that's not a big problem, a smaller motor might be coaxed to cut like a bigger one.
It could counteract vibration more directly, reducing damping capacity expectations on the rest of the gantry, etc.

While fine-tuning might be necessary (size & weight), there might be wide tolerances for acceptable performance.
The ideal place to mount it might be the other end of a double-shaft motor. A safety shroud could be a good idea in some cases.

Any thoughts? I would be extremely interested if anyone is inclined to try this out, it would be a while before I do.

Thanks!
Keith

Edit: oops, I thought I'd seen that people were using stepper motors as routers on some DIY CNC routers, but I misunderstood. I think the rpm isn't there... right? The ideas all still apply.

See less See more

1 - 20 of 23 Posts

The Hobbyist

· #2 · 8 mo ago

I have wondered why someone hasn't created a balancing wheel for many tools, such as bench grinders, and routers, and essentially ANYTHING that spins quickly and has a tendency to become slightly out of balance for any given reason.

These CENTRAMATIC balancing wheels are used on wheels of vehicles and motorcycles. They are a flat plate with a circular tube. Inside of the tube are tiny ceramic balls that move freely around the interior of the circular ring. They work well to balance loads, since an unbalanced load has speed differentials between the segments that rotate faster (around a larger orbital path) and those segments that follow a smaller orbital path. The beads move around inside and gravitate to the slower areas, thus balancing the wheel.

I use DYNABEADS on my motorcycle tires and trailer tires. They are awesome for maintaining the balance of wheels through their lifetime. As the tire wears, or a rock takes out a chunk of rubber for example, the beads move around inside of the tire to compensate for that change in the centrifugal force of that small segment of the wheel, replacing the weight of the rubber chunk that was removed.

I suppose a balancing wheel COULD be made for a router spindle, and I am guessing that it would improve the performance of the router, if someone would only INVENT the dang thing!

Here is the company that makes the balancing wheels.

Joe

http://www.centramatic.com/balancers.rhtml?customer_id=546-629-9909&gclsrc=3p.ds

See less See more

The Hobbyist

· #3 · 8 mo ago

The Hobbyist

· #4 · 8 mo ago

I am a big believer in DYNABEADS.

Reactions: 1

doorstory

Discussion Starter · #5 · 8 mo ago

I've heard about dynabeads or similar products. Cool demo video. A self-balancing flywheel might be easy to spec, though I'm not sure whether it's necessary in this case. Interesting question. Note that routers are moving a lot faster than tires, I don't know how that would change the math (i.e. bead size or shape, maybe?).

The Hobbyist

· #6 · 8 mo ago

doorstory said:
I've heard about dynabeads or similar products. Cool demo video. A self-balancing flywheel might be easy to spec, though I'm not sure whether it's necessary in this case. Interesting question. Note that routers are moving a lot faster than tires, I don't know how that would change the math (i.e. bead size or shape, maybe?).

I am not an engineer, and I do not play one on TV.

Joe

doorstory

Discussion Starter · #7 · 8 mo ago (Edited)

I want to learn more about this kind of balancing, I wonder if it works well with liquid or if you want the particles to "lock in" to some degree. I imagine particle size, shape, density, and total quantity might all have some contribution. I realize also I was calling this dynamic balancing, but that really means something else.
Regards,
Keith

Stokestack

· #8 · 8 mo ago

Wouldn't anything that anything that adds mass to the spindle be bad, since it must change directions quickly and precisely?

Reactions: 1

Souza1969

· #9 · 8 mo ago

doorstory said:
I am considering building a CNC router. I am in the wrap-my-head-around-it stage.
Has anyone seen the use of a flywheel added to a router to improve the cutting? If speed is fairly steady, I think there would be rather little impact on energy. It's possible that some favorable damping could be obtained through flywheel construction, I haven't really thought about that. Are such flywheels built into any routers now, that you know of?

The idea is the added inertia would even-out the load experienced by the motor. Obviously the motor rotor has rotational inertia: the question here is could such an added flywheel allow for a given motor to improve it's cut. My assumption is that the flywheel does not need to be very large or heavy, though I haven't made any attempt at evaluating that.

This is somewhat related to the flywheel action of a large-diameter bit, but in that case the cutters' relative distance from center mean that the router bit body alone can't comprise sufficient rotational inertia. This explains their sometimes-difficult behavior.

The major con is the added inertia with speed changes. If that's not a big problem, a smaller motor might be coaxed to cut like a bigger one.
It could counteract vibration more directly, reducing damping capacity expectations on the rest of the gantry, etc.

While fine-tuning might be necessary (size & weight), there might be wide tolerances for acceptable performance.
The ideal place to mount it might be the other end of a double-shaft motor. A safety shroud could be a good idea in some cases.

Any thoughts? I would be extremely interested if anyone is inclined to try this out, it would be a while before I do.

Thanks!
Keith

Edit: oops, I thought I'd seen that people were using stepper motors as routers on some DIY CNC routers, but I misunderstood. I think the rpm isn't there... right? The ideas all still apply.

How are you handling your E-Stop..... you'd need to incorporate a mechanical break. Now you have account for that mass at dynamic load on your gantry.... assuming your using a AC motor.....which then gives very little control over speed and torque.

Next idea.....

doorstory

Discussion Starter · #10 · 8 mo ago

We're talking about a small flywheel. Yes, it will slow down acceleration: start up, stopping, and changing speeds. How much does that matter? Given a material and a bit, how often are we changing speed? Why do we care about the E-stop, if it takes an additional second? We're talking about weight equivalent to a slightly larger bit.
In terms of added weight: I'm guessing an ounce or two, not a lot of weight. The trick with flywheels is diameter and speed, total weight isn't exactly a helpful way to think about them generally (but see below).
Such an arrangement won't interfere with translational movement, except in the "total weight" sense, so if total weight is a very small proportion of the current weight of whatever is being moved (either router/z-carriage, entire z-axis, or z+x-axis, I think): in all cases I'm not sure how much difference a few grams would make?

Souza1969

· #11 · 8 mo ago

doorstory said:
We're talking about a small flywheel. Yes, it will slow down acceleration: start up, stopping, and changing speeds. How much does that matter? Given a material and a bit, how often are we changing speed? Why do we care about the E-stop, if it takes an additional second? We're talking about weight equivalent to a slightly larger bit.
In terms of added weight: I'm guessing an ounce or two, not a lot of weight. The trick with flywheels is diameter and speed, total weight isn't exactly a helpful way to think about them generally (but see below).
Such an arrangement won't interfere with translational movement, except in the "total weight" sense, so if total weight is a very small proportion of the current weight of whatever is being moved (either router/z-carriage, entire z-axis, or z+x-axis, I think): in all cases I'm not sure how much difference a few grams would make?

I few grams makes a ton of difference, you're not talking about a static load, it's moving under load as well.

I don't know maybe my 30 years designing machinery is kicking in, but have you looked at the consumer and industrial woodworking machine market lately??? We've got saw that detector skin and within a millisecond the blade is safely out of the users way. If you look at any CE, UL, or OSHA machine safety regulations you have less than a second to safely stop all motion on any machine when the e-stop is pressed.
If you want to build something for your personal use, that's cool. But to even consider this as a commercially available item all this stuff has to be considered.

doorstory

Discussion Starter · #12 · 8 mo ago

Stokestack said:
Wouldn't anything that anything that adds mass to the spindle be bad, since it must change directions quickly and precisely?

As noted: it's very little net weight, though yes that will change responsiveness, roughly equivalent to putting a bigger bit in your spindle.
What's important to a flywheel is energy storage, and that's only indirectly related to net weight. Spindles are often spinning very fast, which is flywheel heaven, lots of energy storage. That will make the entire assembly resistant to rotation (precessional motion blah blah), but it won't effect translation (normal x-y-z movement).

Souza1969

· #13 · 8 mo ago

doorstory said:
As noted: it's very little net weight, though yes that will change responsiveness, roughly equivalent to putting a bigger bit in your spindle.
What's important to a flywheel is energy storage, and that's only indirectly related to net weight. Spindles are often spinning very fast, which is flywheel heaven, lots of energy storage. That will make the entire assembly resistant to rotation (precessional motion blah blah), but it won't effect translation (normal x-y-z movement).

So do it and try it with a router.... effectively an inexpensive spindle. Hell just try it with a shaft collar. A flywheel adds rotating mass and that decreases the response time of the motor as well. And I'm not trying to piss in your Wheaties. I may have experience as a machine designer, but when it comes to drives I'm limited to a few vertical lathes I designed for Zildjian Cymbal. Those were serious DC spindle motor. We were spinning 18" ingot material into drum cymbals.

doorstory

Discussion Starter · #14 · 8 mo ago

Souza1969 said:
I few grams makes a ton of difference, you're not talking about a static load, it's moving under load as well.

I don't know maybe my 30 years designing machinery is kicking in, but have you looked at the consumer and industrial woodworking machine market lately??? We've got saw that detector skin and within a millisecond the blade is safely out of the users way. If you look at any CE, UL, or OSHA machine safety regulations you have less than a second to safely stop all motion on any machine when the e-stop is pressed.
If you want to build something for your personal use, that's cool. But to even consider this as a commercially available item all this stuff has to be considered.

These are interesting points. I haven't noticed people shaving grams off their Z-axis, the design is not seriously weight-optimized in that regard, it seems to me... right? The aluminum plates alone could lose a lot, if done properly, I think. The X-axis gantry could a lighter beam size than the Y, probably, depending on machine dimensions, since it's doing something different. The design isn't really optimized in these regards, trading off ease and cost of construction I think (which are reasonable trade-offs).
Interesting points on E-stop, I hadn't been thinking of it in such fine detail: the flywheel would be intended to have rotational inertia on the order of the energy storage/inertial contribution of a somewhat larger bit (like a flycutter for example): does that mess up E-stop timing? I hadn't thought carefully about this part, thanks for diving into it a little. It will be some time before I can even do a proof of concept, so I'm not really worried about commercial viability, and of course publishing the entire idea online isn't the ideal way to proceed in that regard, haha. OTOH, I'm not interested in creating a hazard, even for my personal use!
It looks like the actual E-stop requirements is more like "dangerous movements and operations of the machine will be stopped as quickly as possible without creating additional hazards." There's no absolute timing requirement, I believe. But I'm not trying to be argumentative: if/when I even get to proof this idea, how it changes E-stop timing is a great question (as well as effects on translational speed, beyond cutting stability, which was really the genesis of the idea).
Thanks for this piece. Interesting!

See less See more

doorstory

Discussion Starter · #15 · 8 mo ago (Edited)

Souza1969 said:
So do it and try it with a router.... effectively an inexpensive spindle. Hell just try it with a shaft collar. A flywheel adds rotating mass and that decreases the response time of the motor as well. And I'm not trying to piss in your Wheaties. I may have experience as a machine designer, but when it comes to drives I'm limited to a few vertical lathes I designed for Zildjian Cymbal. Those were serious DC spindle motor. We were spinning 18" ingot material into drum cymbals.

A shaft collar, at least those I'm familiar with, are too compact/close to the shaft. Little rotational inertia contribution, relative to weight. The other thing is, I'm trying to figure out how to test this, and I think it might be hard in a hand-held router.
The intention is to moderately improve cutting stability, giving better finish and perhaps allowing some extended capacity. That's hard to measure in wood and especially freehand, things just aren't stable enough. And here I am without a CNC yet! But if/when I get to try it, I'll post my results (including if it fails to help!).

Edit: but yes, either a customized shaft collar approach OR a customized collet nut (!) might do it. I mean, what could go wrong with a flywheel collet nut???

doorstory

Discussion Starter · #16 · 8 mo ago (Edited)

There's also a chance, btw, that there's much less leeway in sizing this for it to work properly than I've assumed. That is, it might only help when designed specifically for the material/feeds/speeds of a particular situation. Which might make it less helpful, or not.

I don't really expect this. I think bigger will be better, wrt to cutter stability/cut quality & capacity improvements, as long as it doesn't bog down the spindle. But I could be wrong on this point. There's even another possibility, that building a little damping in to the flywheel through construction & materials, could make a difference.

I feel like all of this ought to be a discussion in a 1930's machining manual, but I can't find any reference to it. But I might not know where to look. I'm not trying to reinvent the wheel... haha. Sorta literally.

malb

· #17 · 8 mo ago

Electrical/electronic engineer here. The armature of the motor (universal motor for router or high frequency induction for a spindle) has a significant mass anyway and effectively functions as a flywheel. Both styles of motor operate reasonably smoothly and with minimal vibration unless the user installs something that is out of balance. Adding something with a significant potential to be out of balance, and canterlevered beyond the bearings, then trying to operate the system at between 10000 and 30000RPM is more likely to create issues than solve them.
Do you have access to a dynamic balancer that can handle that sort of speed range and can be configured to drive the entire system at that speed? If you do, your day job probably involves design, development or maintenance of jet engines, if you don't how are you going to balance the system? That is the only way to be sure that the components you add are balanced to the tolerances required.

doorstory

Discussion Starter · #18 · 8 mo ago

malb said:
Electrical/electronic engineer here. The armature of the motor (universal motor for router or high frequency induction for a spindle) has a significant mass anyway and effectively functions as a flywheel. Both styles of motor operate reasonably smoothly and with minimal vibration unless the user installs something that is out of balance. Adding something with a significant potential to be out of balance, and canterlevered beyond the bearings, then trying to operate the system at between 10000 and 30000RPM is more likely to create issues than solve them.
Do you have access to a dynamic balancer that can handle that sort of speed range and can be configured to drive the entire system at that speed? If you do, your day job probably involves design, development or maintenance of jet engines, if you don't how are you going to balance the system? That is the only way to be sure that the components you add are balanced to the tolerances required.

Remember that all this is about chatter. Which is sort of by definition out of balance dynamic forces. Yes the armature (including shaft, and all moving parts of the motor) are acting like a flywheel. If I bolted on a bigger (higher rotational inertia at least) motor, would you expect smoother cutting? Or not at all? With no other changes to system rigidity? My guess is yes, but am I wrong?

doorstory

Discussion Starter · #19 · 8 mo ago

doorstory said:
Remember that all this is about chatter. Which is sort of by definition out of balance dynamic forces. Yes the armature (including shaft, and all moving parts of the motor) are acting like a flywheel. If I bolted on a bigger (higher rotational inertia at least) motor, would you expect smoother cutting? Or not at all? With no other changes to system rigidity? My guess is yes, but am I wrong?

Let me try one more analogy: why do you bolt a flywheel to any motor? Why do you have one on your car (in your transmission, for example)? It's to smooth out power delivery. The perturbations, rather than coming from a piston motor, are coming from a 2-headed cutter (or 1, or 3...) smacking against the material, scraping through the cut, and then possibly popping out into free air before doing it again. Cut quality is going to reflect how smoothly the blade negotiates those rapidly changing power requirements. This is just that. It's not trying to outweigh your motor or anything: just add a little pulse of umph (possibly nonlinear: acceleration or jerk or what?) at the sequence of moments you most need it. Anyway, that's how it seems likely to work to me. Still, I'm tempted to believe I'm missing the obvious b/c... it seems too obvious, people would have been doing this for decades if it helps. I just can't see why it wouldn't? Do machine tools never include flywheels, really? I mean, lathes have fairly impressive rotational inertia, partly depending on the workpiece.
And yes, balancing. Yep.

See less See more

doorstory

Discussion Starter · #20 · 8 mo ago

doorstory said:
Remember that all this is about chatter. Which is sort of by definition out of balance dynamic forces. Yes the armature (including shaft, and all moving parts of the motor) are acting like a flywheel. If I bolted on a bigger (higher rotational inertia at least) motor, would you expect smoother cutting? Or not at all? With no other changes to system rigidity? My guess is yes, but am I wrong?

Oh, your point about cantilevered beyond the bearings is a good one, I sort of missed that.

1 - 20 of 23 Posts

This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.