[PaulRowntree]

Hi Harry;
Engineers Edge is a great site, although I haven't found sections and calculators for C-channels or I-beams.
8x8 seems like a HUGE beam. Would a smaller steel tube be an option? 6"x3"x0.25" or even 6"x2"x0.25" steel would have similar max deflections (but perhaps be less fun to drill and tap...). It would still weigh ~100 lbs, and cost about $100
EDIT : If this is for the gantry, then even 100 lbs seems to be high, and may be inflating your requirements. A router and carriage would weigh 20-30 lbs, and any cutting forces would act against this weight. Lateral forces are similar. So you could probably recalculate for say 50 lbs, and get by with a much smaller (cheaper) main beam.

[OCNC]

So, for a decent, first order of magnitued calculation:
- assume 7 ft beam, just freely supported on the ends - not being tensioned at the ends
- 8 x 8 x 1/4 inch
- 6061 Al
- Just assumed 100 lbs for the beam mass
- Assumed 100 lbs for the load on the beam in the center

First order of magnitude calc - 0.003 inch deflection

I just ran your beam through MDSolids. I used the actual weight of the beam based on a cross sectional area of 7.5 in^2 and an aluminum density of 0.098 lbs/in^3. This gives the beam a self-weight of about 65 lbs or 9.3 lbs per ft. The deflection from self weight alone is 0.0006 in. and with a 50 lb point load added to the middle of the beam (the router and carriage) the deflection increases to 0.0014 in. All calculations were done using the software.

Chris

[OCNC]

Hi Harry;
Engineers Edge is a great site, although I haven't found sections and calculators for C-channels or I-beams.
8x8 seems like a HUGE beam. Would a smaller steel tube be an option? 6"x3"x0.25" or even 6"x2"x0.25" steel would have similar max deflections (but perhaps be less fun to drill and tap...).

6"x2"x0.25" weighs 12.21 lbs/ft so the self weight is a little bit higher than the 8"x8"x0.25" square section. The deflection under the same loading conditions is 0.003" in the 6" direction. In the 2" direction assuming a 50 lb cutting load (no self weight) the deflection is 0.0015". Again the software is doing the calculations.

Chris

[harryn]

Paul and Chris,

Thank you very much for running those numbers for me and the suggestions. I agree, I was a little taken back at the 8 x 8 x 1/4 numbers and size.

That starts to push the weight of this thing well beyond a portable, load it in the back of the min van kind of project. If I am not careful, it could quickly spiral up to 400 - 500 lbs.

If it is not too much bother, would it be possible to run one more calc for me?
- 6 x 6 x 1/4 in Al tube
- Unsupported distance of 5 ft (so 2 ft shorter)

- An extra 1/4 in Al plate running the length of it on the router side (will be used to make a dead flat surface for the rail mount)
- One version of this is to have this Al plate on the bottom of the gantry
- The the other version is to have it on the side of the gantry

- 100 lb load vertical and horizontal load (just for internal consistency if nothing else, and 100 lbs is easy to scale from)

My hand calculation (without the extra 1/4 in plate) is less than 0.001 inches, so I am not sure if I messed up.

I even tried a 4 x 4 x 1/4 inch, and at a 5 ft length, the deflection still looks very low - less than 0.001 inch.

What I am thinking, is to change the gantry to allow use a 6 ft beam (since they are easier to get) and only 5 ft would really be hanging in space.

I guess that cubed length term in the deflection equation really plays hard.

I am using (force ) x (length cubed) / (48) x (modulus = 10,000,000) x (area moment of inertia)

Thanks

Harry

[OCNC]

If it is not too much bother, would it be possible to run one more calc for me?
- 6 x 6 x 1/4 in Al tube
- Unsupported distance of 5 ft (so 2 ft shorter)

- An extra 1/4 in Al plate running the length of it on the router side (will be used to make a dead flat surface for the rail mount)
- One version of this is to have this Al plate on the bottom of the gantry
- The the other version is to have it on the side of the gantry

- 100 lb load vertical and horizontal load (just for internal consistency if nothing else, and 100 lbs is easy to scale from)

My hand calculation (without the extra 1/4 in plate) is less than 0.001 inches, so I am not sure if I messed up.

I even tried a 4 x 4 x 1/4 inch, and at a 5 ft length, the deflection still looks very low - less than 0.001 inch.

Here are some more numbers for the box section without the plate.
(see attached)
The 4x4x0.25 section is showing a 0.006" deflection.
A 1/4" plate will probably conform to any irregularities in the box beam. It's probably not a solution for achieving flatness.

Chris

[harryn]

Hello Chris,

Thank you very much for running those calcs. I realized my math error this morning - when I was calculating the area moment of inertia for the square tubes, I forgot to divide by 12.

You are absolutely right about the 1/4 in Al would just conform to the existing shape, and I didn't explain the purpose or idea very well. The challenges for that part in my mind are:
- How to make a very flat surface for mounting the linear rail (type TBD) ?
- How to make it all adjustable to deal with errors / mismeasurements in being straight / parallel after the build steps

I only have hand tools really, so anything needing serious precision has to be hired out. Some ideas so far:
- Having a shop machine a very flat surface on one face of the square tube or whatever it is all made from. This makes for a flat surface for mounting, but I have not figured out a way to adjust for alignment afterward.

- Pour a layer of self leveling epoxy on one face of the tube to make it very flat. Have a 1/4 in thick piece of Al bar stock twin milled on the faces to make them parallel and a constant thickness. Add some slots for an adjustable position to the gantry beam, and mount the rails to this flat stock.

I will admit that I am worried about the budget for this project, so some of the final approach will come from what I can find a deal on.

[JerryBurks]

The x-rails (Glacern supported 20mm) of my machine are bolted to the outer 80/20 extrusions of the table and it was a concern having a flat and aligned mounting surface (even more for the y and z, with a wood surface).

Therefore I set the rails into a thin film of viscous epoxy (actually I used J-B Weld). I had the rail contact surface first covered with silicone grease so that I can break them out in the future if I ever need to. Since the J-B Weld cures quite slowly this left me a few hours to get the rails aligned better 2/1000" parallel and perpendicular by selective clamping or shimming. After curing I drilled and tapped the mounting holes into the substrate (using the rail holes as pilot) so that the bolts don't deform the rails when tightened.

[harryn]

Jerry - Thanks for that suggestion. I have seen your build, and I am concerned that the workmanship and precision build approach you have taken is well beyond my current abilities and equipment. It is an interesting approach to consider.

I was watching a movie and doodling on graph paper last night. These images are definitely not engineering drawings, but instead just a roughed out concept needing more thought.

The main difference between it and other "fixed" gantry designs is that the entire Y axis moves up / down for the Z motion. The reason for this, is that I would like to be able to make mortise and tennon joints. Since the entire router will be sitting on the floor, this means that I will be flipping the router on its side (pointing toward the x feed end) to make these shapes.

It is still a concept, so we will see what actually happens as the X axis portion develops.

Edit - The images are too large, so I will scale them down and try to post them later.

[harryn]

Trying again with the images.

[harryn]

I could use some advice on bonding square tubes to each other.

I am on a path of using square tube to make the main frame work. The size will either be 3 x 3 or 4 x 4 inch x 1/4 in thick, depending on how the load calculations come out.

I had planned to use Al to minimize weight vs steel, but the total weight is getting high enough that it is less and less likely that I could lift this thing into the back of my mini van anyway.

I don't have welding capability, so I was thinking of using epoxy at the joints. The main question - how do you make a strong L or T joint using square tubing and epoxy ?

I had hoped that the joint itself would be strong enough, but I have doubts, and of course there is the alignment question.

One idea I had was:
- Bond a square piece of wood to the long beam at the joint the size of the tube ID
- Slip the second pipe onto this wood piece
- Apply epoxy, and screw it together using the wood as a common point.

I guess a block of Al could behave similarly.

Is this a common approach ?

Thanks

Harry

[CarveOne]

I would be more inclined to bolt the square tubing together with a combination of 1/4" steel angle and flat bar stock. It's a lot of work to make the parts and drill and tap the holes. It probably wouldn't look very good either. That's why I weld my steel projects.

Epoxy will only hold until it gets a shock load that makes the joint pop apart. A wood insert will be no stiffer than the wood.

CarveOne

[harryn]

Thanks for the comments. I would have to hire out the welding, but that is viable. I have followed a number of build threads over the past few years, and of course welding has the risk of warping the frame, so that is a consideration of course.

The frames built from T slot / 8020 type materials are commonly build with butt joints and similar T connections, so I was just hoping to find a way to mimick this without the expense of actually using T Slot material.

Perhaps I can have some Al blocks milled that match the inside of a square tubing, and screw and glue those to the frame tubing. Those would act sort of like a socketed joint similar to how bicycle frames are sometimes made, except the socket is on the inside instead of the outside.

I wonder what it would cost to have 20 or so 2 x 3.5 x 3.5 inch Al blocks made to match the inside of 4x4x 1/4 tubing ? Maybe even 1 in thick might do it ?

[CarveOne]

I wonder what it would cost to have 20 or so 2 x 3.5 x 3.5 inch Al blocks made to match the inside of 4x4x 1/4 tubing ? Maybe even 1 in thick might do it ?

It may still be cheaper to have it welded. The only way to know is to get quotes on both welding and making that many aluminum plugs.

A knowledgeable weldor can build accurate weldments. It's novice weldors or careless weldors that build crooked and warped assemblies. Precision welding takes time and skill, but it's done everyday.

CarveOne

[harryn]

Last night I started adding up just how many feet of square tube are going to be needed for this project, an approx. cost, and worse yet, the weight.

In rough terms, it will take 10 each 6 ft pieces of 4x4x1/4 square tube, plus rails, motors, drives, etc. Plus the 8x8x1/4 tube.

In steel, this puts the build in the 800 - 1 000 lb range, and in Al, about 1/2 of that. Either one is far to heavy to actually tilt up on its side, or to lift into a minivan without serious effort. The only way this thing is going to be portable, is if I mount wheels on it, and turn it into a trailer.

I was pretty sold on building it with Al, except for the price, then last night I found Al and steel tube for about the same price. ( $100 - 150 each for 6 ft using onlinemetals.) Even worse, the specs on twist and straightness of hot rolled structural tube are fairly - loose. Unless I can find hand picked pieces, the welding warp might not matter anyway.

My son really wants me to build it for heavy duty use so he can at least try some things with steel, so that might drive it in that direction a bit.

I guess it is time to start thinking about the plywood prototype of this thing.

Harry

[PaulRowntree]

Last night I started adding up just how many feet of square tube are going to be needed for this project, an approx. cost, and worse yet, the weight.

In rough terms, it will take 10 each 6 ft pieces of 4x4x1/4 square tube, plus rails, motors, drives, etc. Plus the 8x8x1/4 tube.

In steel, this puts the build in the 800 - 1 000 lb range, and in Al, about 1/2 of that. Either one is far to heavy to actually tilt up on its side, or to lift into a minivan without serious effort. The only way this thing is going to be portable, is if I mount wheels on it, and turn it into a trailer.

I was pretty sold on building it with Al, except for the price, then last night I found Al and steel tube for about the same price. ( $100 - 150 each for 6 ft using onlinemetals.) Even worse, the specs on twist and straightness of hot rolled structural tube are fairly - loose. Unless I can find hand picked pieces, the welding warp might not matter anyway.

My son really wants me to build it for heavy duty use so he can at least try some things with steel, so that might drive it in that direction a bit.

I guess it is time to start thinking about the plywood prototype of this thing.

Harry

Harry, I wonder if this is getting complicated because you are trying to construct a welded frame to high specs. I am working towards a summer rebuild of my 24x48 machine, and am trying to eliminate all wood for the usual reasons. My 4'x6' table design has two 3x3x0.25" tube runners extending the full length of the table, held up by 2.5x2.5x0.25 tube legs welded in place. Diagonal braces are welded from the legs to the middle of the 3x3. These two sides are then connected by bolt-on frames which are cross-braced for racking; my issue of portability is getting the bits home from the shop and thinking of selling the house in the future. This metal work essentially supports the working table, and I expect I can make it flat to within 0.125", easily skimmed off the top spoil sheet. Legs are on adjustable feet to make sure they all reach the floor.
For X motion rails I am going with V-bearings on angle iron, bolted onto 2x6x0.3" channel steel which is bolted to the side of the 3x3 runners. ACME or racks on the inside of the channel. I am planning 0.25" of vertical adjustment available on the channel/rail, so the two channel rails can be tweeked to be parallel in the vertical planes. The angle-iron rails for the V-bearings can move parallel to the plane of the table to make them parallel. If the top of the channel is not flat enough to shim I will use self-leveling epoxy.
Long story short : welded steel table for strength and convenience, bolt-on critical elements for adjustment and transport. It weighs in at ~500 lbs for the table+X rails

[CarveOne]

Look up the local steel fabricator suppliers and ask them for a quote. In this economy they are more competitive and many of them will sell to individuals when they were more selective previously. Tell them what you are building for personal use and they may give you good prices and free local delivery.

Sometimes they have cut-offs that they let go at lower prices. Ask.

CarveOne

[harryn]

Hi Paul,

Thanks for the comments. There are several things that are making this build complicated, of course including my personality.

a) The gantry is nearly 7 ft long to accomodate the material size.
b) I am very space constrained, so the router has to do double duty as "something else" as well.
c) I really wanted it to be portable - stick in the van, so some of the design ideas were driven by that
d) The system and enclosure are all built in one frame
e) I am trying to have it able to cut sheets longer than the X size of the router

It isn't a complicate design, just a rectangular box frame roughly 40 x 40 x 80 inches, with corner posts and 2 extras for the fixed gantry. It all adds up though:
- 4 x 6 ft for the 6 ft length
- 4 x 3 ft for the 3 ft corners
- 2 x 3 ft for 2 braces
- 1 x 7 ft gantry

The price difference, and weight for 3x3 vs 4x4 x 1/4 in square tube isn't that much.

Mostly what you are seeing here is my indecision and rethinking. I have sort of given up on using the hex panel material, not because it isn't great stuff, but because I am concerned about it getting dented, and the challenges of closing out the edges.

C1- Thanks, you are right about hitting up a welding shop. Now that the design is getting more clear, I can almost do that.

The plywood proto isn't to replace the router frame, but rather my way of doing drawings. I call it "Plywood Aided Design" PAD.

[OCNC]

My son really wants me to build it for heavy duty use so he can at least try some things with steel, so that might drive it in that direction a bit.

I don't think you'll be too successful trying to mill steel with a router motor. Were you thinking plasma?

Chris

[FannBlade]

The plywood proto isn't to replace the router frame, but rather my way of doing drawings. I call it "Plywood Aided Design" PAD.

I like that.

Quote of the day!

[harryn]

I don't think you'll be too successful trying to mill steel with a router motor. Were you thinking plasma?

Chris

Hi Chris,

You are right, my router is not steel milling compatible. At least in theory, the router can be swapped out with a different spindle of some kind in the future. Who knows, it might be an air driven grinder with an abrasive wheel. Maybe a drill motor for just some well aligned holes.

My son is 14, so he mostly just wants to be able to make simple things for decoration. If I tell him "no", then his interest in helping goes to zero. If I tell him "maybe" then his interest in helping will be slightly higher.