Ratio of mass between gantry and base table

Hi, I'm in the aluminum-milling router design process. Desires for rigidity have led me to a fairly large gantry, 200x150x12.7 steel tube, which plus the Z axis will weigh 80-90 kg. I'll be using a granite surface plate as the bed, 600x900x100, which weighs in around 190 kg. There will be a stand support, which might weigh another 50 kg or so, but it got me to thinking: what is an acceptable ratio of mass between table and gantry? I'm only planning on cutting around 2000 mm/min with rapids at 5000 mm/min, but if I go faster with servos, at what point does the gantry mass cause problems relative to the table mass?

I wasn't sure what you were saying at first.

Based on your recent conceptual design, you have a fixed gantry with a moving table underneath.

OK, I see now, fixed gantry and fixed granite plate for the bed, actual moving table is 1" aluminum plate.

So you're worried about it accidentally tipping over?

Sorry, I am looking at changing design to a moving gantry. I want to be able to have a fixed workpiece so I can trim carbon fiber laminates in water immersion. It's also getting a bit bigger as I've already surpassed original budget in the process of using higher quality linear motion components, so may as well add some future-proofness in size while at it.

So the question is about the potential for a heavy moving gantry to shake the table and impact surface finish if the table isn't heavy enough.

I wouldn't worry about if your gantry is too heavy for the table as long as everything is nice and rigid.

You can always bolt the table to the stand and the stand to the floor, or add a bunch of weight to the bottom of the stand.

Good point. Itd be easy to make the lower shelf of the stand a tray and pour a couple inches of concrete in. Thanks

Quote:

---

Originally Posted by NIC 77

I wouldn't worry about if your gantry is too heavy for the table as long as everything is nice and rigid.

You can always bolt the table to the stand and the stand to the floor, or add a bunch of weight to the bottom of the stand.

---

Are you really proposing to machine carbon underwater with an electrically operated machine?I have a vision of water droplets getting into all sorts of places and doing damage.If its just flat sheet cutting for the outlines of parts its much better to use waterjet cutting and home building those is quite an undertaking.

Quote:

---

Originally Posted by routalot

Are you really proposing to machine carbon underwater with an electrically operated machine?I have a vision of water droplets getting into all sorts of places and doing damage.If its just flat sheet cutting for the outlines of parts its much better to use waterjet cutting and home building those is quite an undertaking.

---

I had the same vision. Then I watched some YouTube videos of people doing it.

https://www.youtube.com/watch?v=NV61lAH9NnA

What happens to that black carbon fiber water? Do the fish eat it? Is that even an issue? Would be nice to have some kind of filter to discard the carbon goop and recycle the water? You definitely don't want to be breathing it in.

I'd be tempted to use sandbags instead of concrete for weight at the bottom of the stand. Makes it easier to move the machine in the future.

Hi Cat - To go back to your original question. Its all about accelerations not speed. Say you have a 90kg moving gantry that accelerates at 0.1g. This means its inertial load is 9kg. Doesn't sound like much but 9kgf can shake the table. I have been playing with some dynamic software and when the gantry moves at 0.1g the Z axis right at its end is wobbling at 1g plus so although the average accel that you set at the controller maybe 0.1g others parts of the machine that can wobble (like the Z axis) will be moving at high accelerations. So the takeaway is that everything has to be as stiff as possible. By the way machining CF under water is a bit of an overkill you really need a very good dust extractor and a mister. They don't machine aerospace parts underwater. Cheers Peter

I've done a lot of experimentation milling cf. I was able to get the best edge finish and best tool life by milling submerged. But after dealing with the hassle of a submerged setup, I've decided milling dry is worth swapping tools out more often and just using really good dust collection. After trying many tools, it's not worth spending extra money on fancy diamond coated tools. Cf will still eat it in minutes. Cheap Chinese carbide burr cutters are the best option in my opinion. Can be had for about a buck a piece, swap every 20 minutes of cutting. Cutting dry also gives the option of mdf vacuum table fixturing.
Kinda off the main subject, but wanted to give my 2 cents since I've dealt with that headache. As for your heavy gantry, build the stand underneath with good diagonal cross bracing and bolt it to the floor.

but if I go faster with servos, at what point does the gantry mass cause problems relative to the table mass?

As Peter wrote, it's about aceleration. I'd suggest you can move the gantry quite briskly if you do your research and find a controller (or servo drives if there's no 2nd loop) that has an adjustable S-curve accel/decel function. You'd be surprised at how hard you can ramp the velocity up/down if the motion profile isn't jerky (trapezoidal). Motion profile all depends on how you're going to control the motion - closed loop just to the controller, closed loop just to the drives, or an inside/outside loop.

Not that I'm recommending you get Clearpath servos, but they have a rather interesting Youtube video on jerk & settling times with conventional vs. tuned accel curves.

And you can work on putting the gantry & Z-axis on a diet. Even if the whole thing isn't shaking (sandbags are a great idea), the gantry may be flexing in the middle from inertia. If you lower the Z-axis weight it will help everywhere - assuming the Z isn't flexing from a huge cantilever and tool side loads.

Hi All - On all the machines I've analysed the Z axis cantilever has been the main culprit with vibration. Its rarely made stiff enough so go UBER on the Z. Yes "jerk" which is the rate of change of acceleration is the issue. Some controllers are smoother than others, just like wines I expect.. Peter

Didn't mean to abandon this thread, my email notifications must not be working.

Great point about s-curve. That's a priority for me and is why I'm looking at the Centroid Acorn. Seems to have the best motion planner among diy options, even their manuals extensively cover smoothness settings w/r/t contour milling. Was going to use the Buildbotics controller on Pete's advice, but some digging through their forum revealed that the Raspberry Pi has trouble with large files with lots of short moves, which is exactly what 3d contour milling entails. Too bad, as that solution is drastically cheaper than others. But I want to get it pretty close to right the first time.

As far as gantry mass, I wont make this thread all about that as I had another one with extensive discussion about tube sizes. But after running a lot of spreadsheet calcs and learning a bit of simple FEA with Fusion, in the context of a 600mm span, most any rectangular or square steel tube 200x150 or larger won't be the flexy point as long as the ends are capped; the z-axis is the tricky part.

If it doesn't constitute too much thread drift, I would be interested in more discussion of carbon fiber milling. Ice done a ton or abrasive cutting of carbon, and my experience is the more water the better. Diamond wheels that last many hours in a heavy water flood (like under a tap) last only single minutes when used dry, so dry with dust extraction is not an option in my book. Not to mention the fact that enough water eliminates the health hazard. Flood coolant (water or water/ethanol) would definitely be sufficient, but my moving gantry design with rails on the underside of an upside-down surface plate (a la Datron M8Cube) also solves the problem of ballscrew/rail contamination with flood coolant, so more reason to go that way. (I'll post this design soon) But a tub of water is much simpler and more effective yet. However, what I've learned in the process is that designing a fixed gantry machine is about 10x easier! ;) So if there's some experience here about managing flood coolant re carbon fiber, I'd be interested to hear. Generally sounds like a mess dealing with water filters etc, though there are some good diy design using whole house water filters which you can get at 5 micron or so.

Cheers and thanks for the dialogue