Second machine design process

[yukonho]

My first machine is very cool. I am making a lot of things on it and use it daily for some production where I have actually paid for the machine with what I make! Very cool.
I want to make my second machine now that I have a better understanding of how these things go together. This machine will be more suited to the parts I actually make, not the ones I thought I was going to make!
Requirements are....
-16" X 8" X-Y travel
-3" Z travel
-As rigid as possible with MDF as main material.
-Potentially made from steel, (I can mig and Arc weld) or Aluminum but only if it is truly possible to make a square and true machine with the facilities I have and within budget.
-I am OK to go with rollerblade bearings and gas pipe, but need a better system than I am using now. I use gas pipe and rollerblade bearings mounted on 90 degree aluminum with bearings on all 4 sides. It is not adjustable to get the bearings riding perfectly on the pipes. Need a better way.
If I can get liniear ways for cheap then I will go that way.
-Dremel tool or Dremel Advantage for spindle.
-Enclosure.
-Probably go with a better leadscrew system than I am using which is 1/4-20 allthread and HDPE nuts. But honestly, it is pretty good surprisingly.
I am thinking of having a fixed Gantry for the 16" Y and moving table for the 8" X axis.
I need suggestions for the ball bearing and gass pipe ways, wether or not the fixed gantry and moving table concept is best for my needs, and if I should toy more with the idea of using steel.
Mostly I will be cutting wood, and possibly some G-10, I use xylotex board and nema23 steppers.
Thanks everyone, your input can help me make the perfect machine for my needs.
co

Similar Threads:

• Advise me on the general design process.
• CNC Programming and Process Design
• Process Design Engineer
• New Machine Design book from Alexander Slocum (of Precision Machine Design)
• design and machining process

[Ken_Shea]

Here is a link to a thread on this forum, may be of some help.

http://www.cnczone.com/showthread.php?s=&threadid=2066&highlight=black+pipe

[freezer]

I just got in my rails. I went with some drill rod:

http://www.use-enco.com/CGI/INSRIT?P...1&PMPXNO=94762

I also got:

http://www.use-enco.com/CGI/INSRIT?P...PMPXNO=5873921

but I dont think i am going to get the bearings to work the way i want. so, i am planning on using the rollerblade bearings on the drill rod... It is much smoother than gas pipe anyway but more expensive....

[yukonho]

Thanks guys.
ken_shea, that link doesn't work. Can you try again?
co

[Ken_Shea]

http://www.cnczone.com/showthread.ph...ght=black+pipe

[yukonho]

After following one of the links in one of the threads you guys provided, I have pretty much settled on a design.
http://www.cnczone.com/showthread.php?threadid=1356&s=
I loved this torsion box idea and I am going to use it for my x and y axis. If someone can let me know how to post a picture of my autocad drawing, I will post it here. I like the drill rod idea, but I dont like the extra shipping and duty to get it to Canada. Does anyone know of suppliers north of the border? I am going to use rollerblade bearings still, not the bushings.
co

[Mr.Chips]

I liked that torsion design too, strong and light. But he said he would make holes in it to lighten it even more.

I'm still working on my machine, slow going it was cold here the past couple of weeks, but nice this week. Slowed down with the metalwork of the Z axis. Woodwork has been my hobby, not done much with metal, but am forced to learn.

Having a working CNC will really be an asset iin building the 2nd.

Keep posting your progress.
Hager

[ballendo]

Hello,

Just a note or two on torsion boxes.

They get their strength from the SKIN and the glue line TO THE SKIN. A torsion box without a skin is NOT a torsion box! And also not very strong, relatively speaking.

Next, if you ARE going to use a torsion box construction, do yourselves a favor and lighten it up! Using 1/2" inner parts and 1/4" skins (or even 1/4" for everything) will be about the same strength, but obviously MUCH lighter. FWIW, the only reason to use the thicker inner material is it makes them a bit easier to fasten together. (you DO want thicker material at fastening points; or waht an aircraft designer might call "hard points".)

The design shown uses eggcrate construction, which is fine. But NOT necessary for strength. You can build these things MUCH more easily using butt joints. Use long strips one direction and short squares/rectangles the other. No fancy joints necessary...

Ian Kirby first brought torsion box construction to the woodworking crowd through an article in Fine Woodworking magazine. You can look up his article if you want to hear what I'm saying from a "better" source. (FWIW I have used torsion box construction for years in furniture, and ALSO in my first CNC machine, back in '89...)

Modelers might realise that once they put the skins on their airplane wings, everything stiffens up. Same principle... (and think how thin those parts are!

Another example of torsion box construction we should all be familiar with is hollow core doors... Most of those use cardboard! for the inner parts, with 1/8" skins. Try and twist one! So you see, you don't need the thick, heavy materials to use torsion box design

Ballendo

[ger21]

Ballendo, the pictures he linked to are my machine. It IS a torsion box, I just didn't have the skins attached in the pictures, as I was just test fitting everything together. I have a question for you regarding weight. Is it better to have the lightest possible gantry, as was my original thought, or would a heavier gantry help to dampen some of the vibrations. I know the lighter one will have faster acceleration, but I also thought about filling the torsion box with possibly sand to absorb the vibrations. I'd like to here your thoughts on this, as you seem to have been around the block the most times. Thanks.

Gerry

[ballendo]

Gerry,

That is a GREAT question!

Welcome to the wild, woolly and wonderful world of machine design

You want both...

This is an example of the sort of tradeoffs ALL machines require. The best machines are those which choose the answers "correctly". Correct-ness is determined by the design aims and goals, moderated in the real world by physics and price.

Huh?

Ya gotta choose for yourself which is more important, based on your needs and expected results. Factoring in the price you can pay--both monetary, and what the economists call "opportunity cost"--Which just means if you're doing one thing, you can't be doing something else...

Is there a way out? Well, can you increase damping without increasing weight? How would you do that?

It's best to have the STIFFEST gantry you can, at the lightest weight at which the required stiffness can be maintained.

Nowadays, stiffness has replaced the old standby of mass in commercial items in many arenas. The places where mass is still used relate to cost, or customer acceptance. Down side of stiffness is that you will have a very specific resonant frequency. If this is in the "wrong" place, you'll be cursing Think aluminum softball bat versus wood...

So now that I've "engineered" all over the place, let's talk specifics: Can your drives accomodate the added weight, and still run the gantry at your desired speed? If not, i'd look at the "way out" mentioned above...

Hint: there ARE materials which could provide effective damping that don't have the weight of sand. But they're generally expensive.

Ballendo

P.S. Personally, I'd go for stiffness, and light weight.
(THEN I'd deal with any resonance issue which "spring" up -pun intended) There are other places you can improve stiffness first, IMO. See below.

P.P.S. Another question: Does the sand continue to "absorb" vibrations, once it is fully packed--FROM vibration? Does it ever GET fully packed? If it does, is it still sand? (functionally speaking)

P.P.S. I saw the "corners" to accomodate your intended skins in the photos. Only commented because others may not know or take the time to know what a torsion box really means... It might not surprise you to find that many will look at the picture and decide torsion box means "eggcrate half lap construction"...

P.P.P.S. Wanna make your machine stiffer and add adjustability too? replace the single 1/4 angle with 2, and lose the brass bearing spacer/extenders. By moving the one or the other of the angles, you can dial in the y slide--and it's "carried" z. You "would have" also had shorter distance between the top and bottom of your y/z sliding parts. And so a shorter gantry moment. Now you could just increase the thickness of one or both of the top or bottom. Could even make a torsion box with the added space!

P.P.p.P.S. For others duplicating your machine, I'd replace the aluminum with an upper torsion box; or at least add vertical stiffeners. Is the aluminum epoxied to the BB ply? If not, do so. IMO,that is. It will allow the alum to add the stiffness it can. Which if its not glued is not what it "could be".

What's the overall travels of the machine?

Originally posted by ger21
Ballendo, the pictures he linked to are my machine. It IS a torsion box, I just didn't have the skins attached in the pictures, as I was just test fitting everything together. I have a question for you regarding weight. Is it better to have the lightest possible gantry, as was my original thought, or would a heavier gantry help to dampen some of the vibrations. I know the lighter one will have faster acceleration, but I also thought about filling the torsion box with possibly sand to absorb the vibrations. I'd like to here your thoughts on this, as you seem to have been around the block the most times. Thanks.

Gerry

[buscht]

I have used the kind of insulation foam that comes in an aresol can to fill up the inside of a torsion box to improve the dampening.

You need a pretty airtight box though, otherwise the foam squirts out all over.

[yukonho]

Wow, a guy gets a good nights sleep and wakes up to find all this!
Questions and answers.
First, I am aware that the skin is where the strength comes from (I am an airplane guy) I plan on using 1/4" skin on the Y box and 1/4 on the bottom of the table and 1/2 or 3/4 on the top of the table for fixturing purposes. I had planned to fill the boxes (of the eggcrate) with ground up tire material. I thought about great stuff foam, but I dont think it will absorb the vibes nearly as well, and I have access to the tire stuff.
Regarding one of your p.s' Could you please further elaborate on
--"Wanna make your machine stiffer and add adjustability too? replace the single 1/4 angle with 2, and lose the brass bearing spacer/extenders. By moving the one or the other of the angles, you can dial in the y slide--and it's "carried" z. You "would have" also had shorter distance between the top and bottom of your y/z sliding parts. And so a shorter gantry moment. Now you could just increase the thickness of one or both of the top or bottom. Could even make a torsion box with the added space! "--
I cant picture what you are saying here.
I am quite pleased because I had already settled on many of the things suggested here. Thinner eggcrate, simple rectangular sides and simple cross blocking with a thin skin.
Can anyone let me know how to post a screen picture of my autocad drawing?
co

[ger21]

Ballendo, it's not done yet, because for some strange reason, my wife insists that I finish the whole house remodel I started 3 years ago. Almost there, though.

The X-axis will be about 60" long with a little under 4 ft. of travel. The gantry is about 40" wide with about 32" of travel. Z travel is about 8". I'm still deciding how high to make the ganty above the table. My original drawing has it around 5 1/2". I wanted to be able to reach the table with 1" of cutter out of the collet, and with 3" out of the collet, have the cutter clear the gantry. The gantry will be driven by a Xylotex driving 2-250oz-in PacSci steppers turning 2 start 1/2-8 acme.

Gerry

PS
I read a while back that you were writing some CNC books, what's the status of those? Thanks.

[ger21]

From AutoCAD you can export a bmp. Then you'll need to convert it to a jpg as the file will probably be too big. Under File>Export.

Gerry

[yukonho]

OK
Light blue is the endplate. Red will be several of these along the length of the tube or drill rod or whatever I end up using, yellow is torsion box/table. Dark blue is aluminum angle and bearings. I have some 1/2" X 1/8" aluminum bar shown that will be attached along the edge of the table to align the angle.
comments?
co

[yukonho]

My only question at this point is how do I get some adjustability in there to make sure the bearings ride perfectly on the rods and everything is tight and aligned?
co

[ger21]

How about running some bolts through the red pieces to push the pipe together. I built mine the opposite way because it's easier to hold the pipe in place. I just looked again, and on your endplates you can make the 4 bolt adjutment blocks that you see here all the time. And after everything is adjusted, mount the red pieces to hold everything from moving.

Gerry

[yukonho]

I did something similar to yours for the Y axis. With the X, I wanted the rigidity of the torsion box for the table. I thought about mounting the rails to the box and having the bearings stationary, but decided against that, no advantage.
I am wondering if I should just make it all rigid, no adjustments at all. If I cut it all on my CNC it should (I hope) be accurate enough. Other than that, your idea of aligning it all first and then mounting the support blocks seems best. For some reason I dont like those 4-way chucks. Cant say why really, I would rather have a precise and accurate press fit. I am certain I can get the holes cut in the end plates within a couple thou.
Did you thread your Al angle to bolt them in place? I plan on using 1/8 angle and am worried about threads stripping. If I mill a flat and use a bolt with a flat enough head I can just bolt them in place from the outside.
Keep suggestions coming. Especially about the table concept.
co

[InventIt]

in respect to moving table designs...

Anyone ever think to mount the pipe to the edge of the table and the roller bearing to the frame of the machine? Just the reverse of what your drawing shows. That way the support (bearings) are always under the gantry.

Would that work?

[yukonho]

That is how it is done on moving gantry designs. This is a moving table design. I thought about it, as I mentioned earlier, but decided against it. It just doesn't feel right for some reason. Possibly because a large part of the table would be unsupported as you move away from the centre of travel. It would also be somewhat more difficult to allign the rods. Wonder what others think about this...
co