[walter]

My point was some posters seemed to be thinking the EG was going to have a big effect on stiffness and I can't see that. Whatever the size, each cross section is going to deflect, and if the engineering goal is to minimize that deflection (make it stiff) EG is poor choice of material.

1. The objective here is to build a solid, 1- 12ft CNC frame (or table, or gantry). Leveled, precision surface for linear rails (up to 12ft long).

2. You can use ANY material, but...

3. You can't use a metal saw, milling machine, welding machine or drill press.

What would you do..?

[digits]

Thanks Andrew - I had a feeling it couldn't be that simple!

TBH it is a bit embarrassing doing all this learning so publicly, but I really do appreciate being set straight by those in the know

:cheers:

[Mcgyver]

1. The objective here is to build a solid, 1- 12ft CNC frame (or table, or gantry). Leveled, precision surface for linear rails (up to 12ft long).

2. You can use ANY material, but...

3. You can't use a metal saw, milling machine, welding machine or drill press.

What would you do..?

The objective is challenging (but also needs 'solid' quantified), creating large, rigid, accurate structures is not easy. who knows, lightning may strike, but my current opinion is that I'd loop back and question the assumptions on why you can't acquire some of those resources or look at outsourcing, just because no clear way strikes me to meet your challenge.

A solution that occurred to me is to embed steel structures (tube, small I beam etc) into the EG castings underneath the 80/20.

[Geof]

I was out of the loop for a day and three pages of posts went up! I have not read them in detail because I am working with a 4" by 6" screen and old eyes

Mcgyver you are correct about the differences in elastic modulus (E) but you appear to be overlooking that stiffness also depends on the cross sectional area, more specifically the value of I the moment of inertia which depends on area and shape. If you have a much larger cross sectional area for the low E material I will be larger so for the same load the deflection will be the same as with large E and small cross section.

This is partly what I was talking about in a different thread that Igalla mentions, when I was saying you could distort a straight frame when attaching a bent linear rail rather than pulling the rail straight. Both parts have the same E but if the I for the linear rail is greater the frame will deflect more than the rail.

Regarding holding your 8020 straight for casting the embedment around I was thinking that this could be fastened to a wooden substructure and aligned on this. Maybe wood would not be stiff enough so possibly the extrusion could be fastened to a couple of lengths of steel tube. It is relatively easy to drill and tap 1/8" wall tube by hand. Straighten the extrusion along the tube using the tight wire method then align the two tubes. Build the mold around this, pour the EC and unbolt the tube.

[walter]

I'd loop back and question the assumptions on why you can't acquire some of those resources or look at outsourcing.

Steel tubing is a bad idea. $2k for a frame + another $2k for heat treatment and grinding. Not to mention the welding and trucking this thing around! (And how do you get the level under 12ft of linear ways?)

I guess we wanted to "pour" the entire thing like the home buiders do. (yeah..I know...)

8020 came in later, to save money and give the E/G structure a spine.

[greybeard]

Walter - Two 12' long brick walls onto the concrete floor with cross walls if necessary.
Epoxy/sand along the surface, then lay on a 12' length of rails to each. When aligned, pour a filet along them to give full support.

John

[Geof]

....
8020 came in later, to save money and give the E/G structure a spine.

In my mind it is the other way around; the E/G provides an exoskeleton for the 8020, and forms the table while the 8020 provides that attachment points on the E/G

[walter]

Greybeard,

Yeah, pouring epoxy to get the perfect level- that was the original idea.

BTW, I appreciate all the comments- keep them coming!


EDIT:

Geof,
what are your thoughts on this 8020/ EG combo? Do we have something here or not?
So many people would benefit from this! Think of all the mini router frames, small tables, gantry type molds, single piece milling machines, etc.

[digits]

Regarding holding your 8020 straight for casting the embedment around I was thinking that this could be fastened to a wooden substructure and aligned on this. Maybe wood would not be stiff enough so possibly the extrusion could be fastened to a couple of lengths of steel tube. It is relatively easy to drill and tap 1/8" wall tube by hand. Straighten the extrusion along the tube using the tight wire method then align the two tubes. Build the mold around this, pour the EC and unbolt the tube.

Why not just bolt 80/20 to 80/20 as the designers intended? There's infinite adjustability because of the slots, and I think I've come up with a simple way to get good parallelism between two rails - you bolt down the 'master' extrusion on top of a pair of cross-rails, using a set-square to ensure squareness. Then you bolt parallels across the cross rails using the T-slot, so that they butt up against the extrusion and eachother, giving you exact multiples of 6" or whatever across the cross rails. If you then line up the 'slave' extrusion so that it butts up against the end of the last parallel, the two exrusions should now be properly parallel. You then bolt that rail down and remove all the parallels.

[digits]

In my mind it is the other way around; the E/G provides an exoskeleton for the 8020, and forms the table while the 8020 provides that attachment points on the E/G

Yes, this is what I had meant when I first suggested 80/20 - it's there for precision alignment and adjustability without fancy tools. The E/G is the actual base material.

[Mcgyver]

Steel tubing is a bad idea. $2k for a frame + another $2k for heat treatment and grinding. Not to mention the welding and trucking this thing around! (And how do you get the level under 12ft of linear ways?)
.

i meant kind of like reinforcing rod in the EG - include a small i beam in the mold maybe? might not even be have to be attached, just buried like rebar its not shear and that ends that going to stop deflection, its along its length, like a beam

re the 4k, i don't understand your point. are you saying that to create a EG structure the delivers the same deflection under the same load is going to be substantially less? maybe, I don't' know, but there are hints that it won't. the elasticity suggests it'll be a challenge. another is the machine tool industry has not abandon cast iron or fabrication filled with EG in favour of straight EG. lets say the density of steel and EG is similar (i don't know), you now have (to achieve the same deflection) a machine weighing roughly 6x what the steel one did, what does it sit on?

if you are talking building a concrete or brick wall, you need a foundation etc and that adds cost. as per John's idea which has merit (just don't move). Here machine accuracy will be tied to the movement of the concrete wall, That may or may not be a acceptable, but why bother with the EG at that point, what does it accomplish?

the stiffness question is or should be what deflection is the acceptable limit for a given load, and what would it cost to create a section in steel and what would it cost to do so EG without steel, each within the acceptable deflection limit. I'm not sure the steel at 4k (at 4k it would be a beast) is the losing pony in that race, but really the point is, you can't assume who the winner is without engineering and costing it.

Geof, I know that section matters, i mentioned somewhere section and that you need a lot more of a less rigid material. its not that EG couldn't be stiff, more that it was being assumed to be so and if its stiffness you want, abandoning steel in favour a material with a fraction of the rigidity is a decision that should have some compelling other merits to justify the switch. EG may be answer, I'm just challenging some the conclusions

[greybeard]

.... as per John's idea which has merit (just don't move). Here machine accuracy will be tied to the movement of the concrete wall, That may or may not be a acceptable, but why bother with the EG at that point, what does it accomplish?

Could you build the walls on wheels ? (chair)

Some time back someone pointed out the ageing problems with new concrete, so I was thinking that if you used old concrete slabs (a chance to get rid of that old patio) and epoxied them together.... ?

John

[lgalla]

My 4200lb.paperweight is designed with 4"X4"x3/8 wide flange I beams,which is totally self supporting.The E/G was only intended to damp resonance and provide a surface plate for accurate rail mounting and gantry orentation.If E/G adds to the structural integrity that is an added plus.Sorry I am pretty dumb and get lost on the youngs molastity thing.[modulus]
If a tube or structure is filled with E/G besides vibratory damping it should "stiffen"the structure to compressive and twisting loads.Good example,we had hollow aluinium tubes 3"dia .016 wall.Step on one it crushes,fill with urethane foam,you can drive your car over it.The calcs on youngs molasity:cheers: are they based on the same crossection?This would result in E/G looking bad.2or3 in of E/G should look better than steel or castiron.
I am going to take a big risk here of Geof hitting me on the head with a chair again.If you bump a total E/G machine it will break.You guys are probably picturing an industructable material.Not so.I f you hit E/G with a hammer its gunna break.
Thats my 2:cheers: worth
Geof I will do it 4 U(chair)

[lgalla]

I would like to welcome some good friends and members who have helped me in the past to our E/G community.Man these guys are hot.They really know their stuff.
What we are doing here is similar to a big corporation.
Our thread starter,Walter is the President or chairman.He says to our staff "develop new techonolgy based on E/G"
We have on staff 2or 3 master machinist's,A couple of engineers,A strength of materials guy,A guy with a huge machine build experience,bearing experts.
welding experts,mathematicians,physist,NASA engineers,the list is to big,hope I didn't leave anyone out.And me backyard epoxy guy.No corporation could afford the research we are doing.Infact I wouldn't be surprised if large machine builders are following our progress.(group)
Larry
Scarry stuff for to-morrow's post.(flame2) (flame2) Title Removing entrapped air with a blow torch.(flame2) (flame2)
Geof will you stop the chair thing?I know I deserve it.(chair) (chair) (chair)

[Geof]

Things are progressing too fast for me to keep up!

Regarding question by Igalla about Young's Modulus: it is a property of the material and the value of Young's Modulus does not depend on the area or shape of the material. But the stiffness of a beam or shaft that is carrying a load depends both on Young's Modulus and the cross-sectional area and shape. A thicker beam with a low Young's Modulus can be just as stiff as a thinner beam with a high Young's Modulus.

digits; the reason I suggested using something like a steel tube and securing the extrusion straight for embedment is because I have never seen an extrusion that is straight to within a few thou per foot but I think with a bit of effort it could be adjusted straight

[walter]

Mcgyver, check out message #63. This builder mentioned vibration damping but also rigidity.

[walter]

Larry, I hope I got the link right.

http://www.moldmakingtechnology.com/...es/040401.html

The picure shows a large milling machine base.

[Aramon]

http://www.epucret.de/maschinenbau_anwendungen.html

Taak a look at more designs...

[Mcgyver]

Mcgyver, check out message #63. This builder mentioned vibration damping but also rigidity.

....and?.... Of course you can make a rigid machine with EG. There are optics machine that machine parts to a millionth of an inch, extremely small cutting force and they make it massive out of EG so there is no vibration. the combo of it being so massive and the cutting forces being so small, it has has minuscule deflection. this puppy is rigid - so what? it might have also cost 10x what steel would have. none of that has any relevance to the questions of comparing steel +EG or just EG for a 12' router - you don't know how massive you'll have to make it in EG to have comparable deflection, and you don't know the cost of making it in EG to have comparable deflection.

Do you agree that these questions need to be answered before its assumed its going to be cheaper building EG than buying a welder or outsourcing a steel fabrication? that's my point and is where this started

[walter]

Mcgyver, I am very disappointed in you. I've seen the movies. You use to build some cool stuff using an old sock and a toothpaste. And here you are, talking down epoxy and granite...