Flex compensation devices in large gantry routers?

[Splint]

Hi,
I was wondering if there is any method used on large gantry routers (like those used in the marine and aviation industry to make molds) to compensate for flex in the body of the machine itself. I imagined something like a laser system or load cells which could feed data to a plc which could modify the servo encoder signal accordingly. I've seen various machine which make claims of amazing accuracy but I dont beleive these figures can be true "at the tool" considering the length of the z axis columns.

Cheers
Splint

[paul_m_hutton]

most of these use a spoil board(mdf) or some other porous material between the vacuum table and the work. the vacuum can hold the work in place by sucking through this material. If it is nessesary the operator will level this board by running over its entire surface with a large diameter router bit (flycutter). This will at least make the gantry and the bed parallel.

[Splint]

Thanks for the reply Paul, I was more thinking of flex in things like the z column and the gantry itself. A good example of what I'm getting at is in the picture below, the z column is long and skinny, I cant imagine how it would be accurite without some sort of compensation for flex.

Splint

[JerryFlyGuy]

Splint, it all depends on what your cutting w/ it. Using that on an mdf type mold would produce amazing results.. why? Because the grain of the mdf wouldn't show anything under 5-10 thou of variation. Now it if your cutting alum or steel [ neither of which would be rated for this machine pictured] it would be a whole other story. If they were really careful, too really small cuts and really low feeds they might be able to do some alum but.. pretty pointless.

Machines of this porportion, used to cut steel and alum are HUGE, I've seen video of a machine which was hoging alum out of a block which was [guessing ] 10'L x 6'w x 4-5' high.. the block was on a moving table while the machine was fixed.. to mill it. I bet the machine weigh'd in at over 100ton's.

Bottom line.. it all depends what your using it for..

Jerry

[paul_m_hutton]

splint- I can now see what you meant! I'm sure that thing is more rigid that it looks(at least I would hope so!)

[NEATman]

There are ways to compensate for error. We do it everyday where I work. An Agilent laser is used to determine where the error is, and it is "mapped" out. So, in the picture that you show, if the horizontal beam sags, the Z axis would move up and down as it traverses left to right, to compensate for the sag. There is a table with preprogrammed Z moves depending on the position of the left to right carriage. It's very simmilar to mapping out the linear error on a ballscrew in Mach3. There are also other ways to compensate for this. You can have a laser interferometer setup that is the X and Y feedback, instead of encoders. This way, even the deflection in the machine if somewhat corrected for by the servo loop.

NEATman