The Gantry Myth

[ger21]

I make fiberglass parts for vintage racing motorcycles. The effort of making plugs and molds by hand has gotten real old real fast.

So my machine needs to be large enough to accommodate bodywork that could conceivably be 3 ft. wide x 4 ft. long x 3 ft. tall (racing sidecar fairing for example). But if I'm going that big, it would be nice to be able to make bodywork molds for my mid-engined car project (roughly the size of a Miata).

My material of choice is MDF and although my glued up blocks are frequently hollow, they may still weigh several hundred pounds.

Due to weight considerations, I had intended to move the cutting tool in X, Y, and Z, keeping the material in fixed position rather than move the table.

Is this a totally stupid way to go? I'm not getting a clear signal from the group on the advisability of long Z travel from the gantry.

Thanks,

CC

I think you'd be much better off with 8-12" of travel, cutting your plugs in thinner layers then assembling them. You'll still have some handwork from joining the seam though. But I think you'll run into serious clearance issues with a taller gantry, especially on steep vertical areas. Router bits are only so long.

[txcncman]

How about 48" Z stroke?

http://www.dmscncrouters.com/images/...g%20Tables.pdf

[jharvey407]

I have seen some CNC machines that are built for mold making that have very long, 12" or more Z travels.

It all comes down to the design.

Start by defining your work envelope. Then calculate the cutting force required for your work. Then design a gantry and Z axis that is stiff enough to withstand the cutting force.

I make fiberglass parts for vintage racing motorcycles. The effort of making plugs and molds by hand has gotten real old real fast.

So my machine needs to be large enough to accommodate bodywork that could conceivably be 3 ft. wide x 4 ft. long x 3 ft. tall (racing sidecar fairing for example). But if I'm going that big, it would be nice to be able to make bodywork molds for my mid-engined car project (roughly the size of a Miata).

My material of choice is MDF and although my glued up blocks are frequently hollow, they may still weigh several hundred pounds.

Due to weight considerations, I had intended to move the cutting tool in X, Y, and Z, keeping the material in fixed position rather than move the table.

Is this a totally stupid way to go? I'm not getting a clear signal from the group on the advisability of long Z travel from the gantry.

Thanks,

CC

[compositesgal]

Thanks to all for the feedback, pro and con.

Gerry does have a point, though. Router bits are relatively short and the more vertical the surface, the more difficult it becomes.

Sounds like I need more than 3 axes to achieve my overall objective.

CC

[datac]

Thanks to all for the feedback, pro and con.

Gerry does have a point, though. Router bits are relatively short and the more vertical the surface, the more difficult it becomes.

Sounds like I need more than 3 axes to achieve my overall objective.

CC

It takes a whole lot of "machine' to avoid excessive flex when you start pushing beyond 12" of z stroke. If I were attempting to make a deep cutting axis like that FOR WHAT YOU need to do, I would be building a spindle that IS the Z axis. It wont be simple, but I have seen it done.

How do I explain this... Your goal has to be as narrow top view "footprint" as possible so you can do typical 3d deep plunge toolpaths without having anything like a conventional router hanging out in the way. If you just take a typical 6" long bit in a traditional router, the achievable angle of approach for a deep cut is very limited.

So, make the sliding portion of the Z axis out of very thick wall ROUND tubing, and put a machined and balanced shaft with a collet system on the bottom, right thru the middle of it with high end bearings properly mounted on each end. This round tube can slide up and down on properly sized, conventional THK style bearings and rails, with the rails mounted on the tube for maximum travel. The new spindle shaft is driven by toothed belt and pulley from the very top, where whatever motor you pick to drive it is also mounted or clamped to the side of the tube.

Now if you have access to a really good machine shop, you could have the bottom foot or so tapered down, or slimmed down to narrower tube to minimize the footprint near the cutting tool. The linear rails start a bit higher than this narrow portion.

Some calculations have to be made or some good seat of the pants understanding applied when it comes to the size of this tube so it does not flex. If you were cutting foam plugs, this could be lighter duty. Again, you really need a thin profile of the spindle in order to get your longest bits to run along the vertical sides of what you are cutting without a router in the way.

[compositesgal]

Thanks, Chris

I *think* I can visualize what you're describing, but a picture would be most helpful.

I happen to have a small machine shop (lathe, 48" mill, both with DRO) and am reasonably proficient.

More information (awaiting the flames)... I'm planning to use a heavy duty foredom tool with a shortened shaft (minimize shaft flex) in order to keep the tool head as small as possible.

I'm also investigating 4/5 axis solutions with automated path mapping that would keep the tool center line orthogonal to the surface. I refer you to the following doctoral dissertation Tool Path Generation and 3D Tolerance Analysis for Free-Form Surfaces (http://txspace.di.tamu.edu/bitstream...pdf?sequence=1)

One question... (Dammit, Jim. I'm a statistician, not an engineer). Will FEA software (a) generate vectors orthogonal to an irregular surface and (b) provide them in a usable file format?

I agree that this approach is highly ambitious, but it's piqued my curiosity and I feel compelled to follow it; at least as far as my skill allows.

CC

[datac]

Found a photo... weak, but the best I have. Its almost from before the internet era !

The idea was developed by a fellow I met named Dave...... Not something I "invented".

Notice that this was made specifically for FOAM plug cutting, and worked well. The spindle "footprint" is comparably non-existant with a long bit compared to a typical router. The main tube slides in a traditional round linear bearing, and the top is controlled via THK.

I believe this was a 2" diameter tube... may have been 2.5". Don't recall !

The shaft in the middle was NOT a simple home brew shaft... it did have to be properly hardened and balanced.

[ftkalcevic]

Then calculate the cutting force required for your work.

Any links on how to do this? The only information I've ever found was a reference to a very expensive book - Which I can't afford, so I've never tried.

[langwadt]

Here's a prime example of working above the bed.
CNC router - badog cnc

And here's a better solution to go down into the bed,far better design.
CNC Router Plans for Sale

The second needs dual drive, on for each side of the gantry, the first can probably do with one central drive under the table

and as far as I can tell, how the cutting forces will work on the linear bearings
will be quite different

So both have pros and cons