Student organization looking to build CNC milling machine for school donation

[peteeng]

Hi 308 - If you have an ATC its best to have a tool setter, there are many flow on effects from seemingly simple design wants. KISS says no ATC to me for a training machine.

I used to train people to use welding robots and any machine training needs to be really, really simple in the beginning. This is not only for the single person that can use/train a mill at any point in time but the que of people behind that will get frustrated because someone assumed the tool was set then crashed the machine, not pretty. Future people can develop the ATC if needed. ATC is really only useful for production machines.

Gerry has answered the shopbot Q and here's the excerpt from Bamberg. Its unwise to have rails "facing" each other, can be very difficult to adjust. Better to be on the same plane (and same structural element) for your purposes. Peter

Seems you have access to machinery and skills to fit. So now you need lots of time. You learn more from your failures then your successes so sit tight and buckle up. Your going to get a lot of flack here and from where-ever you are. Be prepared to be scarred for life!!

[Daedalus308]

Hi 308 - If you have an ATC its best to have a tool setter, there are many flow on effects from seemingly simple design wants. KISS says no ATC to me for a training machine.

I used to train people to use welding robots and any machine training needs to be really really simple in the beginning.

Gerry has answered the shopbot Q and here's the excerpt from Bamberg. Its unwise to have rails "facing" each other, can be very difficult to adjust. Better to be on the same plane (and same structural element) for your purposes. Peter

Thanks for the quick reply! i see it now! My thought process was to do more like the design on the rightmost side of the chaart, but mirror the gantry so across the X axis so its doubly supported. My intent was not to get two small gantries rather than 1 large one, but rather 2 of the largest we could readily obtain and fit within a given footprint/price. If you believe that is still an unwise direction, I will take that heavily into consideration and will (immediately upon figuring out how) do some FEA on it to help me understand it even better!

Thanks again for the assistance!

[peteeng]

Hi 308 - No its not doubly supported. The second beam is structurally redundant (as they are not actually connected, this thought bubble is an illusion) and makes it overly complex from many angles.. One correctly sized beam is best.

Your model seems to indicate that you are going down the welded tube rabbit hole. Since you have access to lathes and mills I strongly suggest you go down the path of billet machined elements (Al or steel) . They will be as stiff, easier to assemble and disassemble and align and size. There is a lot more design freedom in this approach then a tubular one. There are lots of examples in this forum if you dig around. And they achieve rigid mill stiffness. Once you weld something you have a very very large bucket of worms to deal with. Peter

[Daedalus308]

Hi 308 - No its not doubly supported. The second beam is structurally redundant (as they are not actually connected, this thought bubble is an illusion) and makes it overly complex from many angles.. One correctly sized beam is best.

Your model seems to indicate that you are going down the welded tube rabbit hole. Since you have access to lathes and mills I strongly suggest you go down the path of billet machined elements (Al or steel) . They will be as stiff, easier to assemble and disassemble and align and size. There is a lot more design freedom in this approach then a tubular one. There are lots of examples in this forum if you dig around. And they achieve rigid mill stiffness. Once you weld something you have a very very large bucket of worms to deal with. Peter

I'm sorry, i guess im having some trouble understanding how it wouldn't be doubly supported. Would not an extra set of linear rails and trucks count as a connection?

As far as welded tube, we intended to weld brackets to the ends, and connect each tube/bracket combo to each other with bolts. I would love to use solid steel, but solid steel of the same size as these tubes is extremely expensive (not to mention would require a crane to hoist into place). Unless you meant using a lot of smaller sized steel billet stock

[peteeng]

Hi - You are assuming that two separate beams share a load equally in this arrangement, they do not. Plus the bearings are rigid and they do not allow a twinned set of bearings to load share either. You will have to figure this out for yourself I suppose. Read prior post with mind experiment in it. If you take the web out of an I beam it does not hold up the house, that is what you are doing. The bearings are not a connection in the sense of being able to transfer moments and shear across the gap its a "rigid" bearing so can only transfer the direct load due to the deflection of the lead beam which is tiny. Peter

Plus to be equal to the HAAS you are going to need 800kgf at the tool which your bearings are mounted on thin pieces of metal which will deflect. Lots of work to do... The HAAS requires a crane as well...

I'm sure a sponsor will get you the material if you ask around, plate aluminium is the go easy to machine and light... design the machine and the material will come, stay away from welding.

[Daedalus308]

Hi - You are assuming that two separate beams share a load equally in this arrangement, they do not. Plus the bearings are rigid and they do not allow a twinned set of bearings to load share either. You will have to figure this out for yourself I suppose. Read prior post with mind experiment in it. If you take the web out of an I beam it does not hold up the house, that is what you are doing. The bearings are not a connection in the sense of being able to transfer moments and shear across the gap its a "rigid" bearing so can only transfer the direct load due to the deflection of the lead beam which is tiny. Peter

Plus to be equal to the HAAS you are going to need 800kgf at the tool which your bearings are mounted on thin pieces of metal which will deflect. Lots of work to do... The HAAS requires a crane as well...

I'm sure a sponsor will get you the material if you ask around, plate aluminium is the go easy to machine and light... design the machine and the material will come, stay away from welding.

Alright sounds good! I will try to wrap my head around the concept. when you say plate aluminum, are you meaning to create structure with them similarly to as you have designed for maximus?

[peteeng]

Hi 308 - Maximus is a laser cut and bent machine. In your case I'd pick say 1/2"-5/8" aluminium plate and try to make every component from that. To begin with anyway. I've had a dig around the forum and can't find some of the machines I'm thinking about. Gerry maybe able to point at a couple. There is a bonded aluminium machine and a very nice one that did video blogs. So your gantry would be a big box, edge bolted together from the 1/2" plate. The fwd and back sides could be from one piece so the "corners" are one piece like an arch. Webs would go across these. This is how aircraft and race cars were made for many years prior to composites so don't think its low tech. Ultimately once you get it assembled and bedded in you could reassemble it and epoxy set it together. Many machines are done this way now. Or use high strength loctite same theory. 1/2" or 5/8" will be easy to manage, easy to mill to shape and finish. 6061-T6 is a common grade. 5053-H32 is common here as well. In states you have a better selection of plate. This also lets your students get their hands dirty as you can rough the plate using electric hand saws and hand routers, watch their fingers. I'll keep looking for some examples. Peter Thinking about HAAS stiffness you probably will be in the 1" to 2" think range of stuff. The engineering will figure that out.

look at thread mactec54 cnc router build for a friend

This is a very good example of how you can do your machine. Yours will be scaled up a bit and does not need the pocketing and webbing to KISS. If you can get going on the tormach you can make the parts more elegant. get busy a long road to travel.

[peteeng]

Hi 308 - Here's the hit list then you'll have to do some work.

1) talk to HAAS and find out how stiff your target machine is. Say its 50N/0.0001mm at the tool
2) I made a 250x250x12mm square Aluminium gantry and put 50,000N on it. It deflected 1.12mm so it's in the ball park
3) Now you have to design a machine around those parameters. I'd go 300x300mm to be safe but your team will do some FEA to figure that out more.

cheers Peter

[peteeng]

Hi 308 - You mention Maximus' construction method. This maybe a good idea. You can have the main parts laser cut and bent if needed. 12mm or 16mm bends will be interesting... Fusion has sheet metal? Then finish machine the parts. Find a laser bender shop near you and get a sponsor going. Lasering the parts would be quite cheap for them, they would be buying material at a good rate and you can have a sausage sizzle to raise funds for the material or get the material supplier to chuck it in.

You'll have to put your persuasion hat on. No emails, do your research and cold call or find the right person to speak to. Local people are generally happy to help... but think outside of your area, the world is small. We are ??? miles apart and having fun....But right now its a design exercise so you need to pick appropriate technology for the job and this I think is a good way to start. Your using a good technology but still lots of manual work for your minions. How many in the team? Peter

On second thought laser profiling is the go, bending maybe difficult. But you say you have access to a Tormach pcnc, happy days!! Maybe speak to Tormach they have a great student program too. Find out the Tormach stiffness. If you have to downgrade your target work towards the Tormach then you have a example machine at your fingertips. Peter

[peteeng]

Hi 308 - some examples of billet construction.

https://jeremyyoungdesign.com/2015/1...-building/amp/

I think you will find this more satisfactory then a fabricated mill. Plus do not use construction extrusions they are convenient but not efficient or effective for your job. Your team will have to learn a little about bolted connections. Look at friction, thread design, preload, snug bolting vs friction bolting etc. Being a machinist you are familiar with dowels? Cheers Peter

[Daedalus308]

Peter, I'm going to try to address all your past 4 responses in one reply so here goes:

reply 1: building by plate is a great idea! that will allow for lots of easier customization! Now, I've read that aluminum is only about 1/3 the stiffness of most steels, do you think we'd gain appreciably by switching from aluminum to steel (of course FEA will be done on both to see the final numbers, I'm tasking a member with learning fusion's FEA suite right now). Autodesk has given us access to their generative design suite, although I've not had time to learn it yet, hopefully that allows us to generate some great strength increasing internal frame members. When you say "The fwd and back sides could be from one piece so the "corners" are one piece like an arch. Webs would go across these." could you explain that more? I'm having trouble visualizing this.

reply 2: thanks for the ballpark numbers and "hit list" I'll try to get the ball rolling with Haas tomorrow!

reply 3: do you think laser/waterjet sheet metal is the way to go, or thick plate machined via the same method? I haven't gotten all the way through the Maximus thread yet, so i'm not yet fully up to speed on the construction. you've given me so much reading material! I'll be doing my best for persuasion and calling! we have something like 12 members currently, hoping to get more come start of the fall semester. most of them are unfortunately swamped with work/internships right now so productivity is certainly limited.

reply 4: thanks for the link and the images! they give me some great ideas. I'll be taking a look a the bolted connections stuff asap. I am mildly familiar with dowels. Ik their concept and use well, I just don't have much experience actually using them and knowing their tolerances and fits.

[peteeng]

Hi 308:
1) Aluminium will produce a lighter and stiffer structure if you are not limited in geometric size. Which you are not. ie if you had to fit something through a hole or had a particular package size then steel may win. But I think you will find aluminium will do what you want here and be lighter and stiffer. Plus aluminium is easier to machine which is a plus here. But do the FEA to prove it. The generative design stuff will produce some weird geometry but will give you ideas. Your building a bridge it's up to you now, to get some geometry going.

I use steel because its cheaper. Its cheaper to cut stainless steel then aluminium on a laser as aluminium cuts much slower. I'd prefer Al. I use SS and AL so I don't have to paint it. Paint sounds simple but it has created issues in the last production run so I'm eliminating that problem.

2) Put Tormach on your list as well. They have an interest as you have a PCNC?


12 team members is a lot of labour hours available but a project management nightmare. You will need to brush up on some project and people management skills. I use https://www.ganttproject.biz/ for this sort of thing.

If you look at any performance oriented construction they started in steel, went to aluminium then to composites take planes, bikes, cars thats their developmental path. So forget about steel think about aluminium and composites.

Now another thing since your now convinced about a billet machine. Machines need prototyping. There's a lot to a machine to sort, wiring, dust control, assemble logic. You can prototype in plywood like formply. That's easy with hand tools routers and planes to make the same shape as you intend in metal. You still use the rails, screws and everything and it will function very well but it takes experience to design something in CAD that covers most things that will happen in the real world. It's easy to bang a hole thru plywood so you can wire something properly or get access to a bolt head that's buried. or screw something onto the machine like a junction box... Then once you sort the functional stuff out you can commit to metal knowing you have solved the issues. With new machines we always do things in small steps so we don't waste a lot of time and money. Perhaps the way to go is to 3D print in plastic for the appearance prototype. I worked in the auto industry and we always had to do an appearance prototype to show management before we could move to a functional one. Your on your way now. Peter

plywood machine attached not a great example. GER21 makes great timber machines have a look at his stuff.

Oh by the way you say strength increase, routers and mills are stiffness dominant you could make it out of very low strength material and it would not matter. You need to think of stiffness. Every thing in your machine is about stiffness not strength. This will be an issue in generative design as it works in the strength/stress domain usually. I have not done generative in the stiffness domain so will be interesting to see how that goes. I'm not sure how the FE code can make decisions based in stiffness. Its easy to make stress decisions, global stiffness is difficult to quantify/evaluate at a mesh level.

Lets call it a low-function prototype. I'm sure it will cut aluminium just will flex a bit. So stepping stones:
1) CAD design and develop a machine specification
2) simulation to prove you meet the specification
3) costing of machine, get sponsors and patrons on board
3) commitment and production of low function prototype. Review of LF Prototype
4) commitment to production of Machine No1, whats the project or machine called? Its not real until it has a name.

[Daedalus308]

1)We may be limited by space... I need to get another meeting together with the head of the department to make sure but That is some good information to know!
2) you're right! I will put them on my list!

thanks for the management advice, you're right though, it has been very difficult keeping tasks straight. I will look into that software whenever I get the chance.

Thanks also for the prototyping advice! you're right, I have certainly made mistakes in the past by not make prototypes before sinking time and effort into a final product that ended up failing. you make a great point about stiffness vs strength. I'll see what autodesk has to say about that in their generative design suite! I'll take a loook at Ger21's stuff and I'll try to think of a name

[peteeng]

Hi 308 - A common term these days is a Minimum Functional Product or prototype MVP this is a prototype that captures the minimum functional requirement of the product sounds simple. But its best to take many small steps towards your target vs overstepping and failing. Failing is fine also if its a small fail. They also say test often fail fast, learn move on. So you need to boil the machine down to basics that's why you should consider not doing an ATC. If you had a bare mechanical machine sitting on your desk today you still would have a long journey to get it moving. So come up with a plan and execute it. Up to you now. Peter

[Daedalus308]

Hi 308 - A common term these days is a Minimum Functional Product or prototype MVP this is a prototype that captures the minimum functional requirement of the product sounds simple. But its best to take many small steps towards your target vs overstepping and failing. Failing is fine also if its a small fail. They also say test often fail fast, learn move on. So you need to boil the machine down to basics that's why you should consider not doing an ATC. If you had a bare mechanical machine sitting on your desk today you still would have a long journey to get it moving. So come up with a plan and execute it. Up to you now. Peter

Agreed on the ATC! i just wanted it to be future compatible if student wanted to do so after my time on this machine is done

[peteeng]

Found a machine close to what you want on the forum. Look for Linux_fan new machine build gantry mill. Peter its a moving gantry but just nail it down and move the table.

[Daedalus308]

Found a machine close to what you want on the forum. Look for Linux_fan new machine build gantry mill. Peter its a moving gantry but just nail it down and move the table.

Thanks for finding that for me! I'll be taking the time to read through all you've sent me!

[peteeng]

Hi 308 - Happy to answer your questions and please continue your research. Many of your questions will be self answering if you develop an objective and scope for the machine. A machine needs a reason to exist as I've mentioned before. If it does not have a strong reason or utility then it will be forgotten. So a bullet point list of its objectives is needed. Ideally it has a specialist use as well, makes it even better for the Uni or someone You say you have 12 people plus in the team and they are doing internships. One of those companies will have something special to do or difficult to solve put the word out. You mentioned tool steel in the beginning, was this a just a wish? or was something in mind. Tool steel is a tough call for a beginner machine but let's not put up any barriers...Peter

[peteeng]

Hi 308 - Just to discuss setting your machine. Use a wicking loctite like attached. This means you will not need to pull down the machine once aligned. The idea is not to "glue" the machine together but to take up the empty space between joints. The parts are machined and unless you scrape the surfaces (a skilled job) then lap them they will sit on the high spots and crush these over time resulting in a loose machine. By filling up the space this process is slowed or stopped. A wicking adhesive is run along the edges and capillary action draws it into the joint then sets. Cheers Peter

[Daedalus308]

Hi 308 - Happy to answer your questions and please continue your research. Many of your questions will be self answering if you develop an objective and scope for the machine. A machine needs a reason to exist as I've mentioned before. If it does not have a strong reason or utility then it will be forgotten. So a bullet point list of its objectives is needed. Ideally it has a specialist use as well, makes it even better for the Uni or someone You say you have 12 people plus in the team and they are doing internships. One of those companies will have something special to do or difficult to solve put the word out. You mentioned tool steel in the beginning, was this a just a wish? or was something in mind. Tool steel is a tough call for a beginner machine but let's not put up any barriers...Peter

Hi 308 - Just to discuss setting your machine. Use a wicking loctite like attached. This means you will not need to pull down the machine once aligned. The idea is not to "glue" the machine together but to take up the empty space between joints. The parts are machined and unless you scrape the surfaces (a skilled job) then lap them they will sit on the high spots and crush these over time resulting in a loose machine. By filling up the space this process is slowed or stopped. A wicking adhesive is run along the edges and capillary action draws it into the joint then sets. Cheers Peter

Tool steel was simply a goal as I know it is very difficult to achieve this with non professional machines. I will keep an eye out to see if I can find any ways to make this machine stand out and have special functions, rather than just being a working machine. Thanks for the advice on the wicking adhesive! do you think it would be beneficial to spray on of the connecting surfaces with a mold release so that the machine components can be disassembled for maintenance or moving if necessary?

Thanks,
Daedalus