[GuyUlm]

The only issue with the hypothetical scenario you describe is that these two rails are also tied together with three cross braces as well as being tied together with the gantry. Assuming the guide rail cars have negligible play, the force would be distributed across both side rails.

I absolutely explored building something much smaller. My findings were that the cost did not go down as significantly as I had hoped. My initial plan for the current design was to be able work 1/2 sheets of 4x8 foot material. I entertained going with 1/4 that size, 2 feet x 2 feet, and the cost was roughly half, as my cost for structure is effectively zero.

I may still try something closer to 1 meter by 1 meter.

Want to say how much I appreciate your feedback. This particular aspect of manufacturing is new to me and I love a challenge.

Thank you
Guy

[joeavaerage]

Hi,
take a look at the commercial offerings of 8' x4' CNC routers. Why is it you suppose that they are all made of steel and very heavy walled steel at that and then stress relieved?
Do you think the manufacturers are just going overboard to inflate the price? Or are they trying to achieve a goal?

The only issue with the hypothetical scenario you describe

there is nothing hypothetical about this, its actually what happens....I've seen it....my first machine I had thought would be stiff enough....and to be fair it wasn't bad, but it was
nowhere near good enough despite being all cast iron and steel. There is no such thing as too much rigidity and always ALWAYS scope to improve on rigidity.

Craig

[peteeng]

Hi Guy - The cost of a machine is not really proportional to the size as you have found. If you have the space to do a half sheet machine, I highly recommend that's the size you build. I do think you should flick the belt idea... Keep at it. Peter

[routalot]

I've just skimmed through this thread and now that the pictures are visible,I have to wonder if there is any reason why you couldn't attach additional lengths of the gantry extrusions on edge to increase the section and hence the resistance to deflection.A pair of T shaped gantry beams would surely be an improvement.It would give the drag chains somewhere to live as well.

[joeavaerage]

Hi Guy,
I think routalots idea has value.

How much of this material do you have access to? If you have lots of it then riveting/screwing/bolting several sections together could get you into the ballpark.

Craig

[GuyUlm]

I've just skimmed through this thread and now that the pictures are visible,I have to wonder if there is any reason why you couldn't attach additional lengths of the gantry extrusions on edge to increase the section and hence the resistance to deflection.A pair of T shaped gantry beams would surely be an improvement.It would give the drag chains somewhere to live as well.

I currently have 4 pieces of the extrusion that are about 40 inches long and 2 that are about 84 inches long. I also have numerous random very solid pieces of other extrusion. I'll have to see what I've got that might suit my needs.

[GuyUlm]

Hi,
take a look at the commercial offerings of 8' x4' CNC routers. Why is it you suppose that they are all made of steel and very heavy walled steel at that and then stress relieved?
Do you think the manufacturers are just going overboard to inflate the price? Or are they trying to achieve a goal?

My guess would be that commercial units are going to be built as rigid as the buyer can afford to allow for maximum speed/productivity.

there is nothing hypothetical about this, its actually what happens....I've seen it....my first machine I had thought would be stiff enough....and to be fair it wasn't bad, but it was nowhere near good enough despite being all cast iron and steel. There is no such thing as too much rigidity and always ALWAYS scope to improve on rigidity.

I guess I simply have no frame of reference in terms of how much force is exerted by the motors nor how much the structure is exposed to when in operation.

[GuyUlm]

Hi Guy,
I think routalots idea has value.

How much of this material do you have access to? If you have lots of it then riveting/screwing/bolting several sections together could get you into the ballpark.

I have 4 pieces of the extrusion that are about 40 inches long and 2 that are about 84 inches long.

Would my design benefit if I built the gantry instead with the same extrusion I'm using for the main structure?

[joeavaerage]

Hi
'stiffness/rigidity is king'......actually when I think about it......I've heard the missus mumble that too!.

This is the time that you need to start using an FEA program like Fusion or alternately some experimental measurements like you did yesterday.
Sometimes you might think a particular material or structure will be best...but find as a result of analysis (or experiment) find that it is unfavourable after all.

Of the components of a gantry mill, the gantry is usually the weakest, or rather than weakest, the most compliant, and here compliant means the exact opposite of stiff or rigid.

Unlike other parts of the frame where you may have bracing or other intersecting parts to contribute to stiffness the gantry is held at either end. The sole restraint to flexure,
or compliance, if you want to continue the terminology, is the material and section itself.

Get some materials together and start making some more precise measurements and allow those measurements to direct your design.

peteeng is a mechanical engineer and has a career of using FEA programs and can use them to make very astute judgements. I am an electrical engineer and while
I use FEA programs I am by no means as familiar. It takes some time to learn to use FEA......its a sure case of BS in, BS out, and its not clear at all that its BS going in!
Nonetheless if you are prepared for the learning curve use of an FEA program like Fusion will GREATLY inform your design.

Whether by experiment or computer analysis this is where your design is heading......seat of the pants 'she'll be right mate' tends to have very poor outcomes.

Craig

[GuyUlm]

This is the time that you need to start using an FEA program like Fusion or alternately some experimental measurements like you did yesterday.

I've tried Fusion360 a time or two and have stepped away frustrated. I understand that there are plenty of tutorials available, but it's just so damned complicated. I taught myself to use SketchUp years ago when it was still associated with Google. I know it's a FAR more limited CAD app, but as it stands it's what I know relatively well. There's another web-based app that's supposed to be far more capable, but I can't recall what it's called.

Get some materials together and start making some more precise measurements and allow those measurements to direct your design.

One handicap is that I lack detailed information on the aluminum extrusion that I have. I can only say what its dimensions are and that it's almost always anodized on the exposed face(s). I couldn't tell you if it was 6061 or some other alloy. Some of it is relatively new and some of it is decades old.

When it comes to the gantry, the force applied is primarily to the tool at the end of the spindle doing the cutting, right? I can visualize the spinning tool creating quite a bit of rotational torque as it engages the material. This would be amplified if it's a "hard" material or has inconsistent density like knots in wood. Other than that though, are we trying to apply significant horizontal force as the tool is cutting? If I'm running a router by hand, I don't push the tool hard to its limits; if the tool is sharp it's not necessary.

I've been trying to find images that display the typical stresses on a gantry so I can try to get my head around how to design mine. Sadly, we just found out that my wife's losing her job, which obviously cuts well into any disposable income for the household. While I'll still be designing and redesigning, buying the pricey CNC pieces is on indefinite hold.

[joeavaerage]

Hi,

When it comes to the gantry, the force applied is primarily to the tool at the end of the spindle doing the cutting, right? I can visualize the spinning tool creating quite a bit of rotational torque as it engages the material. This would be amplified if it's a "hard" material or has inconsistent density like knots in wood. Other than that though, are we trying to apply significant horizontal force as the tool is cutting? If I'm running a router by hand, I don't push the tool hard to its limits; if the tool is sharp it's not necessary.

You're not the first to make that argument, ie I don't have to push that hard so my machine does not have to be that stiff. Woefully incorrect. What you may not realise that when using a router you
muscles are actually quite bunched and tensed up, not to 'push' per se but to hold the machine in place.

You can all but bet there will be 100N plus, or 10kgf . That will twist your gantry. The higher the gantry is off the material, and its nice to have six or eight inches so you can put something thick in there
but that also means that the Z axis force of 100n will be that same six to eight inches below the gantry and it will twist it. Trying to build a gantry that resists torque is a never ending battle for CNC hobbyists.
It is very much more troublesome (mostly) than the required 'push' to get the tool to cut.

As far as Fusion goes I was thinking of it FEA ability rather than its CAD/CAM, although I use its CAD/CAM daily. I agree that there is a learning curve, but here's the thing.....learning is free! You may not like
it, but it is good for you. You never know your wife might be a dab hand at it, and you know she'll show you up at it, that's what wives are for!

Fusions FEA could inform you greatly about your design, but so also could the type of simple experiments and measurements I've suggested. One thing that you need to investigate is the properties of your
gantry under torque load, so either get good at Fusion (or similar) or get some gear to make decent measurements and work that way.

Craig

[joeavaerage]

Hi,

all aluminum, at least wrought aluminum has a Youngs Modulus of 70GPa. It does not matter whether its a strong grade like 7075 or a weak grade like 3000 series, they all
have the same Youngs Modulus and that's what is important to designing a stiff machine.

Craig

[peteeng]

Hi Guy - A program like FS wizard will give you the cutting loads. But that does not help you much as its the machine stiffness that's the issue not the cutting forces. When you are trying to do a cut at say 0.1mm per tooth or even less 0.015mm the tool can't move 1/10 or 1/100 of this to complete the cut. Otherwise the tooth gets pushed away from the surface and just skids. So a hobby machine stiffness target is around 20N/um at the tool. So start with that. Machine stresses are so tiny that you can forget about them. Find a machine that does what you want to do and start there. A muse is a better place to start then from a clean sheet... Will save you at least a year in the learning curve...Peter