Torsion box router with a 4th axis.

[TheWire]

Don't epoxy the whole bed. Waste of $$$


You can just cut it flat once you have your linear rails flat / adjusted.

Pippin,

I have access to free epoxy so the cost isn't an issue. I also thought it would protect the torsion box from mist coolant or lubricant if I was routing aluminum. But I shouldn't use something just because its free. I also had not considered cutting the surface of my torsion box when I put 200 screws into it. I was thinking I would line up the rails using an indicator sitting on the epoxy bed. Obviously, the rails need to be parallel, but after that, I now see how milling the table might be easier and cut down on the need for adjustability that I was reluctant to let go of.

I wonder if its possible to trim down the epoxy bed with the router?

Thank you for your help!

Jerry

[peteeng]

Hi Jerry - Do you know what second moment of inertia is? There are beam calculators on the net that's probably the way to go for a layman. So if you have a beam calculator put 1000N as the centre load and then ratio the number.

https://mechanicalc.com/reference/beam-analysis

See attached. Your current beam in bending is about 8N/um. But its the torsion component that will bite you and that's hard because using hand calcs you have to account for the aspect ratio of the beam etc etc. easier to do it with FE for me anyway. Your CAD may have an FE module? cheers Peter

From analysing a few of these things I look at it like this. 1) Pick a target machine static stiffness say 21N/um. 1/3 of this will be connections, 1/3 will be Z axis and 1/3 will be Z axis. ie 7N/um is the component stiffness target then all of these are inefficient say 50% so you end up at ~10N/um which is a stiff machine, if you get 5N/um for your first your doing well. So the gantry has to be, in this example at least 7N/um but this considers the tool in the bottom position so the gantry is bending and torqueing. Torqueing is the current ???

In several other threads the conclusion for an aluminium cutting machine has been a steel gantry 200x200x10mm thick at least. Not saying you have to do this but here's the calc for an equivalent timber beam. Works out 300x300 solid. You got one direction close. By the way this does not have to be to the ends, it just has to be the middle maybe 75% of the gantry. So maybe just beef up the back of the middle third? More thought bubbles... This is interesting to me because my next machine is going to be plywood....

[TheWire]

Pete,

The info regarding multiple component stiffness is very helpful. It turned some more lights on for me. I plan on having a raised bed about the same as the height of the walls for most of my aluminum work and to build the Z-axis to raise to the point the bottom of the spindle collet is even with the bottom of the gantry beam. So hopefully I will have less torsion on the beam when I need the precision on metal. The reason I made the walls a bit shorter than needed was to allow the Z-Axis, when raised above wall height, to traverse the full length of the gantry beam to give me a wider cutting area. I probably would use this mostly on routing larger wood signs < 25mm thick that are neat the top of the wall.

Steel tube 8x6x0.375 or ~200mm x 150mm x 9.5mm wall is > $11USD/inch from the local metal supplier. That's about $517! for 45" of it. I can look at local salvage yards. I believe steel tube sides are rarely flat. It looks like people either have them milled flat on one side or pour epoxy to create a level surface.

I also have 20' of 4" Schedule 40 pipe. I thought about building a tight box around 2 pipes the length of the gantry beam stacked vertically and filling in the gaps between the pipes and box with epoxy. But this seems like a solution which would be very hard to estimate its performance ahead of time.

In trying to keep the cost down, I'd like to try the laminated plywood. I will look at making it as thick as possible. If the beam doesn't exceed 7N/um of stiffness, I can add more plywood laminations or the aluminum skins. I'm willing to step back and replace the gantry beam if it doesn't perform as needed. Or if I find a different spot for the router, the depth of the beam wouldn't matter as much. Understanding the impact of each lamination and aluminum skins may be helpful for the DIY's here.

Thank you for your help!
Jerry

[peteeng]

Hi Jerry - Personally I'd keep going with the ply. Its stiff, easy to modify, very damp and low cost. Steel tubes need milling flat and they are very "live". Pay attention to laminating everything flat or get access to a thicknesser and good sander and the LVL or ply approach is a sound one. If your comfy laminating Al on each side once you get the front flat then you double your stiffness. It must be kept symmetric so AL must be both sides!

Make all the decisions now, its easy to say I'll pull the machine down and add Al latter or do that later but once a machine is built, working and square you will find it very difficult within yourself to pull it down. I have a benchtop machine needing paint (for over 2 years) vs its rust and it takes a few days to pull it down and rebuild so I just call it Rusty...I start building my new machine base today, very excited...Peter

[peteeng]

Hi Jerry - My new workshop is about to get its roof on in the next couple of days. Unlike yourself with restricted space I'll then have room to build some big machines. Looking fwd to getting some big space... Peter

[TheWire]

Hi Jerry - My new workshop is about to get its roof on in the next couple of days. Unlike yourself with restricted space I'll then have room to build some big machines. Looking fwd to getting some big space... Peter

Nice shop & steel frame. Let me know if you have any 45" pieces of steel beam left over
I have a 24' x 48' barn with a 16' x 48' enclosed work area in the barn's 2nd floor. But it's not insulated or heated. Wisconsin gets hot in the summers and -20 degF (-30 degC) in the winter. I'm worried about those temperature cycles on my lathe, mill and soon to be router. So I'm reworking my in house shop to be primarily for machine tools. It's only about 14' x 7'. I could see myself eventually building a 1.5' x 1.5' high rigidity router for aluminum in my shop and a full sheet router in my barn. But the 24"x36" I'm building now will hopefully give me a feel for both wood and metal routers.

[pippin88]

To reinforce what Peter is saying, torsion boxes are NOT stiffer than a solid of the same dimension.

What people are often getting confused with is:
If you take the same amount of material (mass/weight) and make a solid or make a torsion box (or other large section), the torsion box will be stiffer.

Stiffness depends mostly on the material furthest from the neutral axis. Material at the center of a beam carries very little of the load and does not contribute to stiffness much.



If the epoxy is free, go for it. It will seal the wood very well. You can mill / route epoxy fine. Wear a respirator when sanding or machining epoxy.

[Zorbit]

Confusion between strength and stiffness seems to be very common. Throw in "toughness", "damp" and a load of terms ending with "modulus" and it's not surprising there are so many wobbly machines built.

I'll have a search for a clear guide to mechanical design, if anyone already has a link I'd love to see it.

Here's a good one to start with:

https://www.linearmotiontips.com/mec...nd-deflection/

[peteeng]

Hi Zorbit - for mill/router design Bamburgs thesis is the bees knees, have you seen it?. Its quite technical for the layman. Do you mean a description of the meaning of all the words? or a step 1 do this step 2 do this if you need this go to step 6 if your get this (fatal error go back to step 1 etc? ) I've been designing machines for +40 years and still need to learn stuff....

https://my.mech.utah.edu/~bamberg/re...e%20Design.pdf
Plus Slocums publications. His course notes are on line somewhere and they are worth a look. https://www.surgibox.com/team-blog/p...der-slocum-phd

Hi Jerry - very nice barn you need to take advantage of that space somehow... Peter

[Zorbit]

I've been lucky to work with a chartered engineer for the last few years, it's been great to have him put the maths to things that I knew from experience, and to learn where I've been going wrong.

A real eye-opener for me recently is the 3d printed parts designed by automatically removing pointless material, wonderful organic shapes that I would never have dreamed of.

I wonder what a gantry designed by this process would look like ? Very much like a birds bone I imagine.

https://fineartamerica.com/featured/...n-iverson.html

[peteeng]

Hi Zorbit and lurkers - I use Inventor and like fusion360 it has a "generative design" module, You put in a blank of material, loads, restraint's, define what bits you don't want removed and hit the button. It figures out a "best" design. I've done this a couple of times (even with a 400 tonne trailer) . I put the results in a thread called biomimicry I think.... I'll loook it up. Peter

see here ----> https://www.cnczone.com/forums/uncat...iomimicry.html

Ideally I think a moving gantry would be carbon fibre intermediate modulus nearly circular with struts/bosses out to the rails. Internally it would be filled with 400kg/m4 PU foam. So it would in effect look like the birds bone.... The machine I'm making now is intended to make the moulds for my next next machine which will be 100% composite. Looking fwd to threading that.... Peter

Hi Jerry - Yes bits of steel are showing up but they don't fit in an envelope...

[peteeng]

Hi Jerry - The "free" epoxy what is it? Do you have a data sheet? There are 1000's of epoxies and the one you can get may not be suitable for pouring or sealing or levelling. Peter

[TheWire]

Hi Jerry - The "free" epoxy what is it? Do you have a data sheet? There are 1000's of epoxies and the one you can get may not be suitable for pouring or sealing or levelling. Peter

Pete,
The epoxy is ProMarine self-leveling table top epoxy. https://promarinesupplies.com/ I have used it with success on several tables that are still like new after 20 years. I plan to use it on the bed with a lot of surface area, but I would need to research its load bearing capacity if I was going to bed the rails with it. Which I'm not planning to do. Once I laminate the plywood gantry beam, if needed, I may pour a coat on it if the rail mounting surface needs any leveling. Then I would glue the aluminum skin to that.

I also began covering the sides of the torsion box with plastic laminate. It will dress it up and provide more moisture resistance. Plus, I will pour the epoxy bed right up to the sides and that should seal against the laminate.

What are your thoughts on aluminum skin thickness for the gantry beam. 0.125" or ~3mm was brought up, but there are numerous aluminum skin high strength plywoods with a lot thinner skin.

[peteeng]

Hi Jerry - Good your experienced with the epoxy. I chose 1/8" as this doubled the stiffness of the beam with the calculated stiffness. I wouldn't go thicker, prefer to make the timber bigger and stay with 1/8". But if you use 2x LVL I suppose 1/4" gets you more stiffness.

Some of the laminated ply with aluminium: the Al is a very soft grade which won't take bolt loads very well.

Here's some scrap for you? Peter

[Zorbit]

You guys probably know this, you can get a good bond between aluminium and epoxy by sanding through a thin layer of wet resin, the epoxy can then bond before the aluminium forms a protective oxide layer.

[TheWire]

You guys probably know this, you can get a good bond between aluminium and epoxy by sanding through a thin layer of wet resin, the epoxy can then bond before the aluminium forms a protective oxide layer.

I had not heard or thought of that. Thank you. Do you sand through the resin with the hardener in it? I guess I would want to contact the company to ensure the epoxy I'm using would adhere well to aluminum.

[TheWire]

Hi Jerry - Good your experienced with the epoxy. I chose 1/8" as this doubled the stiffness of the beam with the calculated stiffness. I wouldn't go thicker, prefer to make the timber bigger and stay with 1/8". But if you use 2x LVL I suppose 1/4" gets you more stiffness.

Some of the laminated ply with aluminium: the Al is a very soft grade which won't take bolt loads very well.

Here's some scrap for you? Peter

Its been an interesting learning experience with the gantry beam. A couple weeks ago I would have thought steel from a building frame would have been stiff enough for a gantry. Then I thought 1 LVL would be fine. Then 2 LVLs. Then I likely will have to laminate aluminum on to it.......

[Zorbit]

I had not heard or thought of that. Thank you. Do you sand through the resin with the hardener in it? I guess I would want to contact the company to ensure the epoxy I'm using would adhere well to aluminum.

Mix epoxy as normal, spread a thin layer on the aluminium and sand with coarse grit. The epoxy prevents the oxide forming and makes a strong bond. If you're interested in epoxy and wood construction the Gougeon brothers book is excellent and free: https://www.westsystem.com/wp-conten...k-061205-1.pdf

[peteeng]

Hi Jerry - Your first options would have worked, maybe not up to what you think. Re: aluminium bonding with epoxy. The type of epoxy you describe is quite thin. If you laminate with it and clamp it you can create a zero thickness bondline that obviously has no strength. This sort of connection is perfect for contact cement but since you have epoxy.
1) I'd firstly do a small test piece to understand how the epoxy soaks into the timber or not and to time its gel time. You need to know that it does not leave a dry surface and when the epoxy gets tacky. You can also use a "primer" coat on the timber so it seals and allows a consistent epoxy film when you do the join. I'd do a primer coat on the timber.
2) scotchbrite is good for wet sanding the AL surface but any fine wet and dry paper is OK (240 or 320G). And yes you apply the mixed epoxy to the surface (use a V spreader). Wear gloves!! I suggest butyl gloves latex tends to tear
3) If you want to get technical you should use a glass fibre tissue or some very light cotton cloth to create a consistent bondline thickness. Glass tissue is 0.25mm thick. The perfect bondline thickness for structural components is 0.25 to 0.5mm thick. You can achieve this without a scrim by applying the epoxy to the timber, allow to get tacky then apply the wet AL to the surface. The tacky epoxy then can't be squeezed out of the bondline. Do a small scale test to get to know the timeline

In this application the bondline strength required is tiny so I'm sure if you dived in and did it it any old way it will be fine. Peter

To get a little more technical the most common reason epoxy and aluminium bonds fail is surface moisture. Epoxy does not like water. If you have ever TIG welded you will see the moisture coming out of the metal. So a waft with the hot air gun (or propane torch) before you coat the metal is a good idea but this can change the gel time. Automotive and aerospace now use flame (plasma, corona) treatment to prime surfaces prior to bonding. And laser prep have you seen laser cleaning of rusty surfaces aaaaaaaaAMAZING!! I expect Dakota is dry air however compared to where I live. The deserts in USA are full of aerospace component makers for the dry air and cheap land.

re laminating plywood and LVL - I'd prime coat these as well before laminating the result is much more consistent.

[peteeng]

Hi Jerry - I said Dakota but your in Wisconsin sorry about that. My US geography is zero. Peter