1 Attachment(s)
Re: New Machine Build - Massive CNC router for timber components
Hi All - There is some discussion about welding vs bolting vs bonding so I thought I'd put some numbers on paper. I've kept it simple. If we take a 50mm (2") wide mild steel bar 6mm (1/4") thick and weld it what do the other connections have to look like to support the same tensile load? Interesting thought. So attached is the calcs, but the summary is the welded joint yields at 75kN so this is the target for the others. If we look up the shear strength of class 10.9 bolts an M12 has a shear strength in shank of 74kN, a 5/8" grade 8 has 84kN. So a single snug fit bolt is same strength as the weld.
If we try to do this with a friction fit we need 7xM12 Class10.9 bolts to hold the 75kN. This is why aircraft use snug bolt or rivet fits to minimise the number required. Plus its why old riveted structures used so many as they were very low strength rivets and they used conservative approaches to the design.
Bonded connection. If we use a toughened epoxy or a urethane adhesive we can get 30MPa shear strength glue. There are 40MPa adhesives around as well. 30Mpa gives us a 50mm wide connection for the same strength as the weld. This is why aircraft have steadily moved into bonded connections to minimise the number of holes to make, ream and fit fasteners to. Now machine structures are stiffness driven not strength driven so it would seem that relatively small bonding areas or joint geometry is required to achieve adequate strength connections. Plus a bonded joint is visco elastic so is damper then a welded joint.
Obviously we can't butt joint an adhesive yet (we probably can with a big fillet) but if we revisit how riveted structures are made out of angles and flats we can bond these easily and use a few bolts for alignment and fixturing while the bondo sets. Then there is no heat and it will be the same geometry as fixtured when cured. Cheers Peter S
https://www.prosetepoxy.com/standard...bly-adhesives/ 30MPa epoxy to steel
Re: New Machine Build - Massive CNC router for timber components
Yes, the honeymoon phase is definitely the most exciting part. Even the first glitch free run seems to pale in comparison. So George, try to make sure you don't let that glow interfere with the business decision(s) you need to make.
I agree with Peter that you should make a business case for a new machine but also include any used machine(s) you may find. Just make sure you add a bit for the extra maintenance you are likely to encounter.
Peter, thanks for the comparison of weld vs bolt vs adhesive. It is good that you used the shear strength of the mild steel rather than the weld steel since the weld material is at least twice as strong. I would like to comment on the bolt comparison; try to picture having to put 7 of those m12 bolts into that small area.
Rick
Re: New Machine Build - Massive CNC router for timber components
With bonding, I would think Cleavage and Peel forces would be the most problematic?
Re: New Machine Build - Massive CNC router for timber components
Quote:
Originally Posted by
peteeng
Hi All - There is some discussion about welding vs bolting vs bonding so I thought I'd put some numbers on paper. I've kept it simple. If we take a 50mm (2") wide mild steel bar 6mm (1/4") thick and weld it what do the other connections have to look like to support the same tensile load? Interesting thought. So attached is the calcs, but the summary is the welded joint yields at 75kN so this is the target for the others. If we look up the shear strength of class 10.9 bolts an M12 has a shear strength in shank of 74kN, a 5/8" grade 8 has 84kN. So a single snug fit bolt is same strength as the weld.
If we try to do this with a friction fit we need 7xM12 Class10.9 bolts to hold the 75kN. This is why aircraft use snug bolt or rivet fits to minimise the number required. Plus its why old riveted structures used so many as they were very low strength rivets and they used conservative approaches to the design.
Bonded connection. If we use a toughened epoxy or a urethane adhesive we can get 30MPa shear strength glue. There are 40MPa adhesives around as well. 30Mpa gives us a 50mm wide connection for the same strength as the weld. This is why aircraft have steadily moved into bonded connections to minimise the number of holes to make, ream and fit fasteners to. Now machine structures are stiffness driven not strength driven so it would seem that relatively small bonding areas or joint geometry is required to achieve adequate strength connections. Plus a bonded joint is visco elastic so is damper then a welded joint.
Obviously we can't butt joint an adhesive yet (we probably can with a big fillet) but if we revisit how riveted structures are made out of angles and flats we can bond these easily and use a few bolts for alignment and fixturing while the bondo sets. Then there is no heat and it will be the same geometry as fixtured when cured. Cheers Peter S
https://www.prosetepoxy.com/standard...bly-adhesives/ 30MPa epoxy to steel
No stress from not welding this is great, but how do you stress relieve the steel parts once it is all glued and bolted together, all the steel parts have stress from it's manufacture, just gluing and bolting a frame together does not get rid of this built in stress, so when you machine the linear rail and ballscrew mounting points the steel glued and bolted parts will move all over the place
1 Attachment(s)
Re: New Machine Build - Massive CNC router for timber components
Hi Gerry - Peeling and cleavage need to be accounted for in the joint design. Toughened adhesives these days cope with these very well. If the connection is symmetric then these issues are minimised. Also many bonded connections use bondlines that are too thin which makes it easy to peel due to the stress concentration at the edge.. The supplier should be able to provide guidelines on ideal bondline thicknesses. A good connection design should have a thick edge. So the bondline maybe say 0.5-1.0mm but at the edge its 3mm thick via a chamfer or some edge feature. this allows the edge to stretch further.
Mactec - I think the plan would be to machine everything first so any of those effects could be accounted for. I agree that rolled sections have internal stress that does relieve during machining so this has to be accounted for. But transport and SR of a small object is easier then a full frame.
Lots of commentary is based on failure but a machine frame or component is stiffness dominant. Nowhere near failure!! So strains are small, loads are small. So things like 100% welding is not needed, bonding could be done with low strength economical material not high cost fancy extra toughened adhesive. There would be a few single component adhesives that would work at the local hardware. Just putting forward some ideas on the subject. Hobbyists and Makers need a better frame build method then welded steel or AL construction extrusions. They don't have cranes, access to stress relief, large milling machines for correcting frames etc. So we need to explore these options.
I've attached a doc on adhesive joints for those interested. It's a Hexcel doc so a bit technical but it covers many aspects of joints well. Peter
1 Attachment(s)
Re: New Machine Build - Massive CNC router for timber components
Hi ftech/Rick - Yes packaging sizes for some connections is the worry. But making a snug fit bolt is easy if you use epoxy to set the hole. In this way you set up your frame with std clearance holes or slightly oversize holes. Use two bolts or 3 in each side to provide geometric stiffness as well as strength, in this case three smaller bolts vs one big bolt and then adjust frame and snug up the bolts until happy. Then undo each bolt and place some epoxy into the hole then tighten up to correct torque. Now you have a snug fit connection plus some adherence in the epoxy. If the joint is something you want to come apart then wax the components prior to assemble. Or use a soldering iron and heat the bolt to undo in future.
Another approach is to use solder for the connections. There are many soft solders in the 50-100MPa strength range that melt at 250-300C. This is easy to achieve using a cheap propane torch. So after you have your important parts machined straight you could assemble the frame and solder the steel SHS bits together quite easily and have at least half the strength of the welded frame yet be very stiff as solder is metal. This means you can use butt joints for the connections, Butted hollow sections create stiff joints. Bolted connections tend to be planar and will have a soft direction. If the section is thick which for a machine frame gives it weight then I'd expect the joint to be near as strong as the welded joint if say you are running small welds to prevent distortion. But it would be clearly strong enough for the application. Peter
George - Have you found a stress relief company nearby? This will be an important part of your build!!
Re: New Machine Build - Massive CNC router for timber components
Even if you're going to be building a machine from scratch that suits your exact needs, I would encourage you to buy a used CNC router.
Get your hands on one now, because you can use it immediately to help get yourself up to speed on how to use CNC in your business, train staff and prototype your processes.
Consider it part of the build cost of the new machine and consider that money well spent. A deficient machine will inform you in a thousand ways about what you want in your purpose built one. You can even modify it as you go along.
At the end of the process you can sell it and recoup some of your costs, or you may find it useful for secondary operations. I use my "sheet" style machine to flatten slabs, cut mortices and tennons in large stock. Operations that I no longer have to use dedicated machines for.
It's faster and more accurate than traditional woodworking machines, obviously, but there is a huge space between traditional tools and a state of the art timber cutting CNC. A space that is often overlooked. Do you use patterns? You could be cutting those on the CNC.
If it won't cut your long parts you'll still be able to cut your short pieces. It also might be easy enough to come up with a system that would allow you to index pieces longer than the machine bed. You could pack the bed and cut multiple parts per cycle.
If your machine build has any delays, the temporary machine will boost your productivity in the interim.
Re: New Machine Build - Massive CNC router for timber components
I have to agree with post #43 if the topic is looked at from a business point of view.A used machine would get you running in a matter of a week or two and you could start earning money with it.The mental exercise of designing a machine for work that you believe will come along in the future can continue while you are producing parts for the present customers and if ,or when , the big jobs show up you will be a long way ahead of starting from scratch and the money earned in the meantime will be on hand.
In fact it might even be a good time to review the market and see if there are any 5 axis machines on offer as they can tilt and thus get around bigger jobs than a 3 axis machine with a very long tool.The type of accuracy needed for joinery is not as demanding as for aerospace work and a machine that has been slipping out of tight tolerance work could have years of life in a joinery environment.
1 Attachment(s)
Re: New Machine Build - Massive CNC router for timber components
Hi All - I've been looking for the place that the "rivet" image I placed earlier came from. I've finally found it!! It's the Newark Penn Station. Looks a grand structure. Arches in two directions lovely. Got to have a riveted machine frame next time....Peter