584,800 active members*
4,543 visitors online*
Register for free
Login
IndustryArena Forum > MetalWorking Machines > Drilling- and Milling Machines > Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CAD
Page 1 of 2 12
Results 1 to 20 of 30
  1. #1
    Join Date
    Mar 2019
    Posts
    28

    Lightbulb Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CAD

    Hello, I'm going through a very extensive iteration of ALL the suitable CNC Router designs for a machine that can mill steel with a decent material removal rate.
    I plan to document the project with a YouTube video series (Williams Garage Engineering, not much worth watching yet) as well as articles spread all over the internet haha.


    I'm currently simulating different gantry beams to find the best design relative to stiffness, vibration dampening properties, and manufacturability.

    I'd very much like for you guys to suggest more gantry beam designs that I can put on my comparison chart. The criteria for the beams is that they should be easy to
    machine and build, therefore I'm only allowing the use of aluminium unless you can come up with a design that doesn't require machining on the beam (like steel tube
    shafts, I have already designed this though). Hollow cavities like the inside of a square aluminium profile can be
    filled with epoxy granite (it's basically "stiff" cement with superb vibration dampening properties).

    I'm thinking about adding a C-beam to the comparison chart even though I know it's not going to be stronger than some of the other designs, but it would for sure
    be good proof about the fact that there's better alternatives than using a C-beam as a gantry.


    I've linked two pdf files with all the designs that I have currently simulated, so take a look and see if we need more designs to simulate!

    Looking forward to seeing your design proposals =)

  2. #2
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hello William - Welcome to the hornets nest. Your in very early days in your research and capability. You need to study commercial mills of the size you propose. You are not talking about a router...I think you will find them to be much more substantial then you think. Personally and I have put a lot of vibration engineering analysis in these threads filling steel tubes with FG is only adding mass not much damping via material damping. I looked at your chart and you note CF as 400GPa stiffness. This is a high modulus fibre std modulus is 200GPa and HM fibres will be extremely expensive. Plus it will translate into a laminate of around 180GPa max (at say 1550kg/m3) by the time you build it at 55% fibre volume and have enough cross ply to provide torsional stability. With std CF you get around an 80GPa laminate with 210GPa fibre. It is quite difficult simulating the dynamic response (damping) of a multimaterial multipart structure unless you have specialist software. Doing a modal analysis is quite easy in most std FE programs.

    stiffness and rigidity - go big very very big and thick. Thin sections vibrate
    damping - exceptional vague area even for the experts. Aluminium and steel are very live, damping ratios in the order of 0.05% so need lots of mass to mitigate
    manufacturability - all things are buildable, depends on your budget, available resources and the desire to do something
    MMR - you need to specify the MMR you want then look at commercial machines to see what they do. Do not think they weigh 2 tonnes or 15 tonnes for no reason. KE=mV2 so the faster you go the forces are exponential and the more mass you need to oppose the inertial forces F=Ma and then you suddenly need tonnes of mass so the machine does not skip across the floor. Been there done that with robots...
    Beam sections - will need to be closed sections to maximise torsional stability and local stiffness, your current approach will not have the torsional rigidity to get what you want. Current research indicates an outer shape of a square with an interposed circular section within to be the ultimate shape. The square outer for bending stiffness the circular centre for torsion of which there will be heaps required where your going
    machine stiffness - initially you will need to decide on a static stiffness for the machine typical VMC's are in the order of 100micrometer per N stiffness plus 100um/N at the tool. very small hobby mills maybe 10um/N but their MMR is tiny. Your FE model will overestimate stiffness by about 50% as a built structure. Part by part will be Ok...
    easy to build - if you have a commercial requirement for the machine then buy a mill, where you are going is not easy, be prepared for pain. You have about 2 years to go before you will be capable of completing this project. You think you can build a machine in a few months then a year goes by and another... unless you are employed full time to get this done (and you have reasonable mechanical and electrical engineering experience) then a year maybe...

    You need to specify your cutting forces, rapid speeds and various mechanical electrical requirements (and then validate them via commercial benchmarks or the knowledge in the forum) before you do too much more FE stuff. Plus a build budget, ultimately the $$$ will count. Do lots of searches for mills in this forum you have lots of reading to do.

    Oh people here can't help you much with gantry advice because you have not told us the size of machine you need. Gantry types depend on whether your making watch parts or airbus wings. So a spec is important please...

    Unless you know what the target is you can't shot at it, so a written spec is very important. My day job is designing and building machinery so I do not say these things lightly... Good luck and have fun Peter

    By the way it is very difficult to guess at what is optimum in a structure. If you are using fusion360 or similar they have form finding algorithms which will produce near optimal geometry in one hit, no need to guess.

  3. #3
    Join Date
    Mar 2019
    Posts
    28

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hello Pete, and thank you for the warm welcome to the hornet's nest. First of all, I'd like to express my gratitude for your great reply, just wow. You opened my eyes to the deeper levels of machine design parameters and other potential designs, now there's much more to take into account when continuing the design iterations... sigh.

    Carbon Fibre. Thank you for pointing out the fact that 400 GPa is a bit much, I did a lot of research of weather or not the results would be realistic as I figured fibres would be hard to simulate. I'll just scrap those beams and list carbon as something that's not going to be an option for the design iterations. I'm however really intrigued by the fact that carbon in some load cases can be soo much stronger than steel... I guess this is something that would have to be built in order to be measured and compared (might do this in a couple of years when I might own or have access to a tensile tester).

    I will look into doing a couple of modal analysis test's on the winning and easiest to build beams to ensure that everything looks good. This is a field in which I know next to nothing about, so I'll probably spend the rest of the day reading up on this topic until I know enough to correct my designs if needed.

    Regarding stiffness and rigidity, the length of the beam plays a very big role. I forgot to tell you that the beams will have a total length of 600mm. I have not given much thought "directly" to the torsional strength of the beams, but the load on the gantries have been a remote force of 1400N applied on the square surface of both 20x20mm rails. The load is positioned 90mm below the lowest point on the beam itself, which is where the force will come from when the Z-axis is fully extended. This will account for the torsional load. All this information will also be added to the simulation "charts". Regarding the ultimate gantry beam shape, it's very interesting with the square that has a circle inside! I'll simulate this after reading a bit more about torsional strength... lot's of reading today haha.

    The manufacturability of the beams is just to indicate how easy they are to build. Great feedback, I'll point out what the "manufacturability" refers to on the comparisons as well.

    Regarding MMR (material removal rate): The aim of the "design research" in this project is to find the strongest router configuration that is easy to build. That means that I don't really any specific number in mind, I'd just like to find the most rigid design that can be built with relatively common materials. Ohh well... didn't really think about the fact that the acceleration is exponential... sigh (again). It has always been my plan to cast a very solid epoxy granite base, so I guess I'll make it a little bigger than I originally planned. Another thing, I'll probably end up welding a steel frame together for the router to stand on, so that will likely also decrease the chance of the machine walking itself around in the shop lol.

    The router is meant to be a universal all-around machine that can eat aluminium and perhaps just be able to lick a tiny bit of steel (instead of eating it). I've come to realize that it's okay if can only take small cut's in steel. I plan to build another router someday that has a "fixed" Z-axis and a gantry that can move up and down... this router would of course also have a much lower rpm spindle, but I'll talk more about this in another post if that project becomes real.

    One last thing: The aim of the projects is also to myth bust a lot of bad designs. I've seen loads of poorly designed CNC routers and I'd like to provide people that are new to CAD and machine design a nice overview and understanding of why their router should be built one way over the other. This way they can compare multiple designs that are seemingly equally good, by looking at my comparison charts (and perhaps also watch my videos/read my articles).

    Best regards
    William

  4. #4
    Join Date
    Mar 2019
    Posts
    28

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA)

    Okay so I've been testing the impact of filling 'Beam 5V' with tubes, rods and some square steel profiles, but none of them seems to make the beam deflect less. They're actually increasing the deflection, and this doesn't really make sense to me. The winner beam that I've been filling with different shapes consist of a square aluminium profile filled with epoxy granite. Since steel is stiffer than epoxy granite, the beam should get stiffer, but it doesn't. All the simulations are made in Autodesk Fusion 360.

    Right now I've done a comparison between two beams, let's call the beam A and beam B. Both beams have the same dimensions.
    Beam A: Made of solid stainless steel, and has a deflection of 0,0018 mm.
    Beam B: Aluminium square with filled epoxy, has a deflection of 0,0018 mm.
    Take a look at the picture for reference. You can also see the material properties there

    Now when steel is stronger and stiffer than both epoxy granite and aluminium, how can it then be true that they have the same deflection?

    I see 3 potential answers to this question.
    1. There's some magical and incredible engineering concept that I don't know about.
    2. I have input the wrong parameters to my custom epoxy granite material that I added into Fusion 360 myself.
    3. Fusion 360 is not having a good time computing and simulation a beam that consists of multiple materials.

    Can you help me figure out what is going on? - So that I can continue simulation my way to a great CNC router design hehe.

    Alright, thanks in advance and I hope we get this sorted.
    Attached Thumbnails Attached Thumbnails Is something wrong with epoxy granite.jpg  

  5. #5
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hello William - Are you a graduate or student engineer? Just give me some background on yourself so I know at what level to reply. I assume you want to be a machine designer and you are a Maker. Firstly do not mix up strength and stiffness they are totally different things. Check that F360 has "coupled" the geometry together. I don't use F360 so not sure how it is managing this. We talk about "bonded" surfaces or solids. You have back tracked from steel to aluminum. Just because you take "light" cuts does not mean a machine that can cut aluminium can cut steel. This is a rigidity issue. You have to specify what you want to do and then the result is self evident. You need to work on your specifications and requirements first. In three posts you have changed the machine rigidity requirement by a factor of 100x. You will chase your tail and burn up 100's of hrs if you do this too often. Designing to cut steel is just as easy to design to cut foam. You just have to understand the stiffness, accuracy and power requirements. Its all here in this forum or in members heads just ask a clear question. Search for Maximus I design a machine from scratch to cost it covers lots of stuff.

    I also encourage you to aim high. This is not an easy project for a newbee. You have CAD and computation resources but you do not to have a deeper understanding of stuff vs just plugging geometry into a FE program. eg strength vs stiffness I encourage you to go composites as this is where machine design is going. Cast Iron in advanced machines is not the ticket. So two paths 1) design a small starter machine in MDF. It will have the same electrical and mechanical issues a metal machine will have but is far quicker and easier to develop. 2) Then design the big machine. By then you will have suppliers, software, electronics sort of sorted. 2) design the dream machine in composites. Its a design exercise say 400 hours of work and at the end you can decide if you will build it. For every machine I build I have designed 2 or 3 which never made it off the HD.

    back to your posts:
    1400N or 140kgf is about right for cutting steel with small bits and hand feed rates. Plunging with a drill can go to 250kgf. Get a bath scale and measure the load on a drill press you will be surprised.
    Strength in a machine is not needed. materials are way stronger then the service loads can create. The exception is a crash. But you can't design for a crash.
    Moving a gantry up and down is complex and then you have poor gantry clearance. Maybe a good idea for a printer or a laser but high gantries are for high objects which means they have deep insides and a vertically moving gantry does not solve that problem. You actually have to not, have moving parts. The least moving parts to solve your problem the better.

    Machines are designed to do specific things. If its general purpose then its hard to design to. This is way GP machines have evolved into particular shapes. If it can be done "better" I'm sure the machine companies would they are filled with engineers who want to change everything.

    A router that eats aluminium is a mill. I think you need to forget about routers. Myth busting bad designs, I'd steer clear of that at the moment in your career position. You have already stated things that destroy your credibility as a myth buster. Your journey will be interesting enough without trying to disrespect other peoples work. Read up on Karma as well its just as good as FEA...

    Now to your FE problem. Since the deflection is exactly the same in both results you need to check your models. Something is going on. Garbage in garbage out is the credo. You need to understand second moment of inertia and how geometry works. Rigidity is the combination of inertia and material stiffness. When multi materials are combined they are summed by rigidity. R= sum of EI1+EI2+EI3 etc see attached. The outer skin of structures do 90% of the work thats why we make tubes. I looked at a 100x100 steel tube and then looked at the same al tube filled with EG. The steel is clearly stiffer. Steel has a very high stiffness. Check and recheck and do some hand calcs to validate the FE. Peter

  6. #6
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hello William - this is where your headed - https://dspace.mit.edu/handle/1721.1/88839 Bambergs thesis is a watershed work in my view. He put a few years into this so pay attention...Now in summary of your intent:

    1) You are heading for a fixed gantry moving table machine,. a) because its small b) because this is the stiffest least moving parts configuration. Unless it fails your footprint spec. Or a high rail design...
    2) you need to select the material steel, aluminium, EG, fibre composite I suggest you stay in one media. Become expert in one material more efficient at the moment
    3) Steel is old tech step forward to today prefer the future
    4) Aluminium, billet machined and bolted very good if you have access to a mill (but that's a circular problem, why build a mill if you have one? usual reason you use a small machine to build a big machine) . If no mill available expensive to contract the parts
    5) EG - the EG Thread here has over 2 million views so very popular. Totally doable excellent material easy to do in your garage. Does need some finish machining but that's reasonable
    6) Composite CF or fibreglass doable in your garage learn about composite infusion, need a router to make MDF moulds or use simple shapes such as the EG people do... EG best E=35GPa CF E=70-90GPa so well ahead there
    7) Other - titanium printed will be very common soon. Printing costs very high but interesting design project. Most high end machine companies have metal printing machines now and printed parts are out there. They are even qualified to fly primary structures so tech is mature
    8) CF printed parts is doable now
    9) Tetrium-S or Tetrium-A do a search

    Pick your point of difference and develop the scope, specification and requirements... Peter

  7. #7
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    G'Day William - I'll chuck something else in then leave you alone. Although steel and aluminium are live and make good machines they can be brought up to date by laminating them. "Glare" is an aluminium laminate that is used extensively in the aircraft industry. If you have a laser cutter nearby and their rates are reasonable or you are happy to cut parts yourself (small mill, router or jig saw) then you can make parts from steel or Alum that are damp by laminating them. A stack of say 2mm aluminium epoxied together will be stiff and quite damp. This can be modelled once you sort out your multi material modelling. But I expect you can neglect the epoxy and be quite close.

    https://en.wikipedia.org/wiki/GLARE
    https://www.compositesworld.com/blog...gence-of-glare

    The issue with going CF for instance is that you will call up your supplier and say whats the stiffness of this laminate and they will say "don't know" "that depends on XYZ" etc etc. So really hard to design with that info. However ST and AL really well known absolute figures. Composites have the opportunity for any shape you need, consolidation of parts, bonding of many parts, so many opportunities. Old tech has a habit of getting reinvented so depends on what you want to do. have fun Peter

    You will hit the usual issues with metal laminates however, bending allowances block shapes etc. Up to you... The aerospace like it because it has superior fracture toughness and fatigue resistance properties and cheaper then CF... a machine does not need to have these but these materials do have a lot of visco elastic damping something we need lots of.

  8. #8
    Join Date
    Mar 2019
    Posts
    28

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Okay, a little something about me: I graduated last summer from a gymnasium, and I'd like to be an engineer (not sure which kind yet but something that relates to consumer products or metal machines) or an industrial designer. I'd probably end up getting an education as an engineer. In my father's garage I have an Emco Compact 8 lathe and a nice medium-duty mill drill. The shop as it is right now is only set up for manual machining and I'd like to use my equipment to build a CNC machine. In this year I've also tried my luck as a freelancer that offer product design services (only taking on projects that I know I have the experince to make a great design). It worked out pretty great, I got a thing going where I was working for an electronics engineer and learned a lot from him, but the corona situation kind stopped our work together, so I'll likely end up starting my education this year instead of next year.
    You can use technical terms, I'd like to learn them, I will just google what they mean.

    You are absolutely right about getting some hard constraints on the machine's rigidity and capability. The thing is, that I don't care how stiff the machine is, just that it is as stiff as I can make it without having the gantry weigh more than 30-38 kgs, while still being relatively easy to build. Easy to build can mean different things depending on who you ask, but for me, it means that one doesn't have to do unnecessary complex work to add a few extra percent of stiffness to a given part. Alright... with all these vague constraints to the design of the machine I don't think I should call this research or anything like that because it's not. You've made me realize that I should wait with publishing something about machine design and "research", so instead this project of mine will get to be a follow-along kind of thing.

    You're right about having the dream machine in composites, this is definitely where the machine is going. From the start, I imagined having an epoxy filled base, but I have today come to realize that there are many great alternatives to epoxy granite. In the last couple of months, I've read many good things about epoxy granite and had, therefore, got so caught up in that specific material that I forgot to look at the alternatives. Not much comes up on google when searching for "epoxy granite alternatives" but today I've been doing a lot of reading on filamentary composites, and googling that instead, gives a lot better results for the available materials to use instead of epoxy granite. I ended up reading about composites and a few articles about vibration dampening because of your machine design/build thread "Designing new Router called Maximus" - what a great forum thread btw. Another thing, I do have a grasp on the concepts of strength and stiffness, it's just now that you told me that the words have a very different meaning in engineering, that I've come to realize that they don't sort of mean the same thing (I'll use the terms correctly now, thanks).

    About the CAD simulation problem, it's definitely fusion 360 doing something funky, thank you for leading me in the right direction to solve the problem.

    Funny thing: I actually measured the force of the table on my mill with a scale when cranking the handwheels with about the same force that a crank them when cutting steel... I broke the scale to pieces by accident hehe (only 20 kg scale).
    All my simulations have been with a force of 140 N because I've heard on the forums that that's what it takes to cut steel on a small scale.

    Umm about the myth-busting thing, I didn't really mean it in a way that would disrespect other people's work. I might have used the wrong words to describe what I meant. What I meant by myth-busting was actually just to find out what simple shapes in thick aluminium plates or profiles are best suited for a CNC gantry. Before this project, I had an idea of which shapes would be the strongest for a moving gantry CNC machine, but I didn't know for sure.

    I actually thought that "router" meant "moving gantry frame" in the CNC world. Now I've looked up the differences between router and mill, and I found out they refer to machines' ability to cut materials. Thank you for correcting me, you're right, I'm absolutely designing a mill.

    Pete, I'm still reading your Maximus build thread and I'm constantly getting distracted by the articles that you refer to, it's really great stuff and a have a lot of questions that I need answers for (I'll create threads for those because they deserve their own threads).

    When I read these two words, filamentary composite, in an article today I got a great idea. Imagine having a 3D printed shell and then fill it with a machine base filamentary composite. I know it can be slow to print big parts on the printer, but the fact that you can have a relatively complex part like a tailstock on the lathe and then be able to build its body at home without the need for a metal foundry, it's pretty damn amazing. Metal pieces, sleeves and threaded inserts for the tailstock could "simply" be aligned in the tailstock as the composite cures. Then it becomes a question of positional accuracy instead of having to machine all the faces and groves on the tailstock on big and expensive machines.

    I'm not that great with engineering formulas yet. Rigidity R = sum of EI1+EI2+EI3. What does EI mean? E = MODULUS OF ELASTICITY?

  9. #9
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hi William - Yes E is the stiffness or modulus of the material. It is a material property. I is the second moment of inertia of the shape. It is a geometric property. The bigger the section the more inertia it has. The combination of the geometry and the modulus gives its rigidity (R) = E x I in this way you can compare a steel beam to a timber beam. If the beam is a "composite" then you have to sum all the parts. Filament or fibre composites for the purpose of analysis we "homogenise" ie its an average of the fibres and matrix. This can be estimated or tested for.

    So to your mill. One requirement is that you can build all the parts using the tools you have, another is that you can lift any part by yourself. Add researching getting an engineering education. Look up your technical college on line uni etc etc. You need to get into a qualifying path... long term it will be the best thing for you. So I call the Project Design Rules the terms of engagement. These need to be written down and referred to. Designers suffer project drift and design inertia along the way so you need a framework to guide you.

    You are in whats called "The fuzzy front end FFE" of the project, its chaotic and poorly directed. That's fine you will get direction and more firmness in a short time.

    So
    Requirements
    being able to lift parts
    make parts in own workshop
    add more here

    Specifications
    X=600mm Y=500mm Z=500mm ??
    footprint 2000mm x 100mm
    etc
    X velocity = ??

    Project Objective?
    xxxxx??? add stuff here

    The idea with these is not to define an outcome but to define whats needed. The outcome will show up in the work.

    Now to your idea about printing "armatures" then filling them. Its good, its done but you don't have a 3D printer. So you need to buy one or contract one. But I suggest you work on the specs and requirements a bit before you get to that. I've been working with composites for about 35 years so can help a little with that.

    When you write something here its set in concrete nearly forever so be clear on what your saying and check before you hit the post button! Good Luck and happy to help.. Peter

  10. #10
    Join Date
    Mar 2019
    Posts
    28

    Red face Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hi Peter, thank you for helping out in general and explaining the formula for rigidity, this is very helpful and valuable information. =)

    I've been thinking about the requirements and specifications of the mill and decided to sleep on it. I've written down some great requirements and constraints today, that I can refer to when designing the machine.

    Requirements
    All parts should be able to be lifted by two people relatively safe (around 80 kg).
    Should be buildable with my own equipment (mill, lathe, 3D printer, bench grinder, chop saw, drill press)
    The machine must have decent vibration dampening properties.
    The machine should not be complex to build. (filamentary composites are allowed)
    Compact design
    Must contain dust and chips
    Benchtop machine (for steel frame tables)

    Specifications
    Max deflection: 0.015mm 90mm below gantry at the spindle, with a gantry length of 600mm.
    Part accuracy: 0.02mm
    Gantry clearance: 120mm
    Part size: 400x800mm
    Max gantry weight: 45 kgs (Including Z-axis and spindle)
    Max total weight: 360 kg

    I will create a design/build log for this project this weekend (I'll post a link to it in this thread)

    As mentioned I've been reading a lot about composites lately, and yesterday I began reading the first 10 pages of the epoxy granite thread on the forum, there's still something like 260 pages to go hehe.
    I have questions, and I should be able to find an answer to one of them by reading all the posts in the EG thread, but if you have the answer at the top of your head I'd very much like to hear it.
    - Is there an alternative composite to EG that is stiffer but still has relatively good vibration dampening properties? I know there are many answers to this question, but if you know of a composite that cost about the same as epoxy granite and has about the same stiffness of aluminium (or greater) then I'd be very happy if you could tell me what it's called. It doesn't have to be machinable or anything. So far I've heard that one could either mix the composite with materials such as boron or chopped carbon fibres. Ideally, the composite mix should cost the same as what it would cost to buy the solid aluminium for the equivalent volume.

    Just a quick side note. I do actually have a 3D printer, haha, but it only has a build volume of 200x200x300, so not much potential for printing moulds in the size that I require. I have also designed my own 3D printer last summer. I has an airtight enclosure with a fan in the bottom that controls its temperature. I've attached a picture of the 3D printer for anyone's interest.

  11. #11
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hi William,
    Good to see the Scope coming along makes life much easier.
    1) When you spec a deflection you have to also spec a load. Machines and machine elements (in the metric world, where are you?) are specified by N/um. If you look at bearings, ball screws, couplings and data sheets they will have the part stiffness in them. So your machine will be speced the same 50N/um or 10 or whatever you need
    2) Part accuracy comes from the machine accuracy so you need to spec the machine tolerance. Tolerance, accuracy and precision have specific meanings in engineering speak
    3) remove fibre composites from the requirements. This is an outcome and we don't want to pre-empt an outcome (yet)
    4) Re "better" material then EG - there are 1000's stay away from boron although very stiff its very poisonous. One attraction for home builders in EG is cost you can get the sand at the beach for free.
    5) So leave the material in the requirements as "composite" and we will work on that.

    Homework
    As materials performance increase so does the cost so a very important addition to your list is budget for the build. $1000USD $5000, $10000, $100000. This will immediately filter (boron, HM CF, PBO, diamonds, metal printing etc) unless you have the cheque book.

    In the scope you may want to add External Resources
    machinists
    composite fabricators nearby
    sponsors - material money labour food
    material suppliers, hardware stores, software you will need these
    epoxy suppliers and composite manufacture suppliers - vacuum bags, mould wax etc
    sand and gravel suppliers, look up sand blasting suppliers

    you will need a vacuum pump

    We will get to the candidate materials shortly but we have to get the project foundations cast and cured before we blast off. We are still in the FFE Peter

    Re Material cost
    If you are using aluminium as the cost baseline then it will be difficult to beat. Extrusions here in Oz cost say $7/kg up. Epoxy starts at say $18/kg up. That's one reason EG is popular because sand is free to cheap. You do need to settle a budget $$$ as this will answer many Q's. CF bought as roll product is about $150AUD /kg so in your 100kg of machine you will have 150*50= $7500AUD of CF which I expect will blow your budget. CF is the best material I believe but we also have a mass and cost issue to resolve. If your in europe there are recycled CF coming to market which are quite cheap. More on materials later once you nail down some stuff. Steel and AL are hard to beat stiffness/cost so other benefits need to be factored in to make it worthwhile.

  12. #12
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hi William - Machine tolerance is important as this is directly proportional to machine cost. A way of looking at this is to understand the International Tolerance Grade system.

    https://www.engineersedge.com/international_tol.htm

    This will give you a basis for selecting machine parts and to place your machine into a tolerance/quality space. To achieve a part accuracy of X you usually need a machine accuracy of 0.1xX So in your case the part accuracy of 0.02mm means the machine has to be able to work at 0.002mm Peter

  13. #13
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hi William - Heres a 800x375 mill just to see what it looks like. Its a manual mill so not as stiff as a CNC VMC but you can see the weight and bulk of it. Its $12,000 AUD mass produced in taiwan. So you need to come up with a budget for your machine.

    https://www.machineryhouse.com.au/M607D

    In your spec you say the gantry clearance is 120mm. This will disappear if you use a vice, which for metals is usual. So this needs to change. This is why mills have a "knee" so you can adjust for the job height but the machine stiffness stays relatively the same. The Z on this machine is 425mm. It weighs 1.3 tonnes.

    Not saying that this is what you need to do, I suggest you find a commercial benchmark machine that you are trying to better. Are you still cutting steel? I didn't;t see a material in the requirements...

    Back to composites - 3D printing is increasingly being used for making moulds so maybe your printer can do that for you. They would be segmented so you can build any size mould you like. Glue the segments together. You will need a plastic that has a Heat deflection temp (HDT) of over 110C as when the composite exotherms they get quite hot especially with thick sections... You don't want moulds going soft and changing shape as the part cures...Peter

    or you can print the "plug or master" then take a mould off that then mould the part. To be figured out later...

    datron maybe a good avatar or benchmark machine... definitely a high std machine

  14. #14
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hi William - I'm going to point you down the steel fibre direction. Steel is the same stiffness as CF but CF is 1800kg/m3 and steel is 7800kg/m3. So if you can pack it correctly you acheive the same stiffness as CF yet its 7800/1800= 4.3x heavier. Since you are unlikely to get over 50% fibre volume ratio its effective density will be 7800/2= 3900kg/m3 so its denser then aluminium (2700kg/m3) but maybe close to its stiffness. So shotcrete uses steel fibres, get the shortest you can, 20mm is a common size prefer shorter. or Find a machine shop that makes lots of swarf that you can get for free, find out wholesale supplier of steel wool, it packs quite well. Now to get a good pack you will have to press it or vacuum press it. So I expect steel fibre composite to be heavier then AL which is good, not ultra light but we do need some mass, yet its stiff and its cheap and it can be post machined or you can mould in inserts or studs.

    Being an epoxy composite it will be very damp same dampness as EG as the interface is the same and the matrix is the same...

    Grit blasters sell steel shot this packs well and maybe used in conjunction with fibre. If you find the right machine shop doing turning then you may find a free source of fibre. Mill swarf or chip would be good. Swarf is very hard due to its heat treatment and working during cutting so its hard to post drill but its free..... Lots to digest good luck... Peter can also use aluminium chip not as stiff but post machines easily.

  15. #15
    Join Date
    Mar 2019
    Posts
    28

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hi Pete - I only got to reply to message #7 before, because I forgot to refresh the page the next day. Peter this project is really warming up my heart and love of engineering. I'd really like to thank you for helping out.

    Now first off... the machines budget for the frame only will be around 1200$. This does not include transmission elements like linear rails and ball screws. The rest of the parts (motors, spindles, driver, controller, transmission) is something that I'm hoping that I might be able to get sponsored. I'm planning on making some youtube videoes (a lot better than my current ones) and a build series on my channel as well in order to get some parts and components for the machine sponsored. When I get to the point of the actual build with the complete CAD model I'm hoping the channel will have enough videoes for companies to sponsor and support the project. It's alright if I can't get any sponsors and if the youtube channel doesn't become what I'd like it to be, then I'll just save money up for the parts myself ;-). Do you know of other ways to get funding for open source projects such as this one?

    Alright, you mentioned that the machine would have to be a lot more accurate than the parts need to be. I'll edit my requirement to a positioning accuracy of 0,005 mm then, as I'm not sure if I can make it more precise than that with the equipment I own. I'd very much like to design the machine in a way so that the accuracy of the machine doesn't rely on the accuracy of one's machinery. This is why I originally decided to build the machine out of milled flat bars, then the builder wouldn't even be constrained by the working envelope of his machines. - Now I really want to make this machine in composites because of the materials high specific stiffness (just learned this word in a scientific article, this is awesome). Now the problem with doing composite castings at home is that they would have to be milled flat by myself, which contradicts my original plan for being able to mount milled flat bars. I've spent the last two days being concerned with the fact that aluminium would add vibrations, but I've just now come to realize that the vibrations dampening properties likely wouldn't change much if I used 5-10 mm plates of aluminium for mounting linear rails where they are needed. If I screw a very flat piece of aluminium onto an uneven surface they would, of course, bend, so the aluminium will have to be cast in place as a piece of the casting. An alternative to using milled aluminium flat bars would be the use of epoxy leveling to create the mounting surfaces of the linear rails, as this process also can be used for making a machine that is bigger than one's capacity to machine parts flat with personal equipment. I'd like to avoid using epoxy resin although it would completely exclude the need for aluminium. People on the forum doesn't seem to have good luck with epoxy resin and I have this feeling the flat milled aluminium will be more precise. I will make this a requirement as well "use the flat milled stock for mounting linear rails", I also like the idea that one can drill holes for the rails before they become a joined piece of the casting as a big and heavy casting is a little awkward to move around and machine/drill and thread compared to aluminium plates/flat bars.

    Added Requirements
    + Use flat milled stock as mounting surfaces for the linear rails
    + Total budget: $4400 AUD

    You have given me a lot to think about and given me many directions to go in, so I need to find a better "scope" for this project - I need to select my materials. Ideally, I would have a high-performance composite material (within the budget, if this is even possible) for the gantry with great specific stiffness and then use "cheap" EG for the base. Now I'm a little concerned about what potential materials can be used for the Z-axis as the machine. Originally it was intended to be made in aluminium. Now I've come to realize that aluminium has bad dampening properties and I'm thinking that it either has to be made of stress relieved cast iron or composites. Filamentary composites usually have less stiffness relative to its volume compared to metals which potentially could make the z-axis assembly take up more space, which then would increase the leverage of the whole thing, as it would be further away from the gantry. This problem could potentially be solved with some clever shapes of the composite material, but it probably won't come close to the simplicity of a thick piece of cast iron.
    Now I'm in a bit of a dilemma here because using cast iron kind of defeats the purpose of the extreme vibration dampening that the whole machine would have in the gantry and the base, so I'd like to figure something out here that doesn't evolve having the z-axis frame weighing a lot more compared to a solution consisting of the use of composites. This brings me over to something like GLARE or something similar and perhaps a bit simpler. Glueing pieces of aluminium together as you suggested sounds like a good solution - It also sounds like a lot of fun, this is really something I'd like to try, although an uneven spread of epoxy could make the material a little uneven. One could simply use a press to press the slaps of aluminium together using a flat milled piece of material (like a flat milled piece of excess stock from the materials used to build the machine) so that the pressure on the z-axis plate would be even.

    Regarding the machines rigidity and ability to cut materials I've decided to let the amount of deflection dictate what kind of materials the machine will be able to cut. I could try and find some data on what is needed to make a specific cut. I just found out my deflection specification was 93 N/?m. I will round this up to 100 N/?m.

    Already now I can say that the base of the machine will be made of epoxy granite. Now the next step for me is to decide which composite I will use for the machine base and then find a material to use for the z-axis.
    Added specifications:
    + EG base
    + 100 N/?m deflection
    + Positional accuracy: 0.01 mm (I might be able to get it down to 0.005 by tightening/loosening the linear rail screws if I use an straight edge and an indicator).
    + tolerance: ±0,015 mm (this should hopefully be achievable with fine finishing passes)

    I do have a question that I've been thinking a lot about. If one were to cast an aluminium plate into a composite, should the plate then have screws though it for the composite could "grip" the piece better? Let's just say it's a piece that with the following dimensions 120x600x8mm.

    As I've already made a few design iterations in CAD back when I intended to use aluminium, I've included a picture of the current z-axis design - just for kicks. - it will be redesigned when I have selected my materials.

    I'm going to do my homework and read more of the great articles that you guys recommend in your build thread https://www.cnczone.com/forums/austr...76822-cad.html.
    Peter, thank you for helping me, you have already added great value to the material selection stage of the project as well as potential designs.

    Best regards
    William
    Attached Thumbnails Attached Thumbnails z-axis.png  

  16. #16
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hi William -
    1) So write out your new scope. version 16.01 section A
    2) I suggest you stay with one material seems you like laminated aluminium or EG so you need to do a "trade study" or SWAT analysis to establish which direction to go. So pick a gantry for instance and design one using both methods that has the same stiffness. Then cost each and look at the manufacturing logistics and see which one is best
    3) I suggest you look at steel fibres as the "EG" especially if you can get swarf or steel wool at zero/low cost, Plus you need to define what you mean by EG? its different for different people, depends on what they can get...
    4) 100N/micrometre is a tough gig. But aim high...
    5) Set an overall machine budget at the moment. The best path is to set a value that you can fund yourself. We don't know the cost of the machine yet so the budget is a bit academic. But a budget of $4400AUD probably won't get you a 0.005mm accuracy machine. Maybe 0.01mm is better. This will need some tuning. For instance you could spend $1000x4 easily on high class ballscrews to achieve this.
    6) You don't need to select a material yet it will come out in the wash, keep working on the scope the more detailed the more direction it will give you. It will show up clashes faster then doing CAD models or supplier searches. We are looking for a doable reasonable path
    7) Budget ballscrews are class C7 which are described as transport screws (not positioning screws) C7 is +/-50um per 300mm so that means your accuracy is say 0.050mm per 300mm so in your 600mm axis you could be 0.1mm out. So you need to get familiar with the machine part tolerances
    8) I suggest the first machine is a learner machine and $$$ do count, but we need to pick a start that is scaleable to the dream machine. So we design the dream machine then cut it back to the learner machine (or lets call it a prototype)
    9) Then we need a project name this helps in visualising and realising the project. Everything has to have a name...

    But always aim high... Peter

    Find your local composites supplier. They will have a glass veil usually 30gm/m2 its about 0.25mm thick. This will be the interleave in your Al laminate. This will provide an even bondline thickness and stop the AL coming togther creating a zero thickness bondline.... You can use your vacuum cleaner as a vacuum clamp although I recommend you find a vacuum pump. You probably will cook the cleaner eventually using it this way. Peter There are 60gsm glass weaves which maybe better as well. handling veil is a bit fiddly. So try to find a very light weave. For that matter you could use a very light cotton from the haberdashery...

    if you are going laminated metal then steel may win. Its stiffness/$$$ will be better then AL I suspect. And since your laminating making hollow parts is easy so weight is designable... plus bonding steel is easier than bonding AL, will come out in the manufacturing study...

  17. #17
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hi William - re your money question. I can't see people sponsoring a machine that does not exist and that to some extent is a high risk machine done by a learner. So my strategy would be to design the dream machine to the best of your ability, disregard cost. In the scope this will be your research bit. "Dream" will use new tech etc and solve a particular problem. Your prototype is the springboard. If you have a prototype that demonstrates it works, plus this puts some of your skin in the game then you have a chance at sponsors.

    Sponsors want something out of the sonsorship so I suggest you look for a company that needs a machine to solve a problem,. This may not be a mill it maybe a cloth cutter or a chocolate printer. If you have a composite fabricator nearby and they need a cloth cutter then lets design a cloth cutter. This will give you work and solve their problem. Most laminators don't have the skill set to graduate to cloth cutting. A huge area in composites is automatic tape laying (ATL). So your CNC platform could become a tape layer. There's probably more money to be made in tape laying then milling into the future. Mills as such will turn into finishing machines that finish 3D printed parts. No more hogging needed very soon in your career. In your career machine shops will change dramatically.

    So find your composite supplier(s) and local composite fabricator(s) I think you will need both. Peter

  18. #18
    Join Date
    Mar 2019
    Posts
    28

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hi Peter
    1) Good idea, I already have all the requirements and specifications written down, but it would be great for me to expand a bit on what my personal goals would be and such.
    2) I will limit myself to two materials
    4) Yes, I'm aiming high for this one, it's my dream mill after all.
    5) + 7) I forgot all about the ball screws, I was mostly accounting for the flatness of the stock aluminium. If the ball screws accuracy deceases linearly over a given length, then we should be able to account for some of it in software. For example travel length x scaling factor. This could get us a tiny bit closer to the goal. Let's edit the positional accuracy, it will be 0.03 mm instead, for the whole working envelope of the machine approx 900mm. I've just visited BST Automation on Alibaba and found that they sell 20mm machined ball screws for a price of $100 AUD! - I've sent them a message asking for the accuracy of the ball screw.
    6) Good idea peter, I will do some comparisons of the gantry tomorrow. Epoxy granite (35 GPa) vs. carbon fiber epoxy (60-70 GPa).
    8) That's a great approach, we will let the design be my dream mill and the I'll figure out a way to build something cheap as a learning machine.
    9) Yes, a name haha. The very first thing I did was create a folder called Will Mill in my cad program. It's pretty funny because my name is William (obviously)... 'Will Mill'.

    3)
    I've been lurking around on the internet looking for data, and on Wikipedia I found a chart that compares basalt fibers to other various fibers. It states that basalt fibers has a young's modulus of 85-87 GPa and that large carbon fibers has 228 GPa. When comparing the prices on Alibaba you would find that carbon fibers is at leas 10-20 times more expensive than basalt.

    Basalt fiber strands ($1 AUD/kg: https://www.alibaba.com/product-deta...6f932c14dS5lHe

    Carbon fiber strands($10 AUD/kg): https://www.alibaba.com/product-deta...52bd1dectf9rK0
    Carbon fiber strands($5-25 AUD/kg): https://www.alibaba.com/product-deta...65f21759TZPlK0

    Do you know if this type of ruff carbon swarf/strands can be used for composites?
    I'll have to do more research on how to create carbon epoxy but I do have a question. When casting composites consisting of strands and swarf how do one obtain a low amount of epoxy? Do you just mix it together, or is a press used to compress it all together? You mentioned infusion with a vacuum pump. Does this mean that I "simply" have to put a vacuum bag around the casting and then use the vacuum pump? I'd be very happy if you could describe how to go from fibers/swarf to solid composite in major and simple steps as I don't want to take up all your time ;-) - I will buy a vacuum pump hehe.

    Very interesting suggestion with the tape laying machine Peter. I came across one of these today on YouTube, this was the first time a saw one (it looked very expensive). But cloth cutting machines and automatic tape laying machines are so different from what I'm trying to do, right? They don't require the vibration dampening and rigidity that I'm trying to achieve. But I thing there's more to it right? Did you have something else in mind Peter? - this project can still go in every direction, as we still can edit the entire scope. I will try and get in touch with some local composite suppliers and fabricators here in Denmark.

    - William
    Attached Thumbnails Attached Thumbnails Fiber comparison.png  

  19. #19
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hi William - Your up early must be excited.

    Don't limit materials yet. Materials will come out of the analysis and needs
    1) basalt fibres are interesting but will cost I buy them here for about $45 AUD per kg.
    2) Where your going you will need to learn a bit about infusion. This entails stacking a mould with dry fibres, powder or whatever you have, then evacuating that space then use the vacuum to draw in resin to fill the void. In this way you use atmospheric pressure 10T/sqm to compress the stack and you have no air to inhibit the fill. There are many videos on infusion on the net
    3) They require extreme damping! They are faster then mills so need to be lighter and damper. Again pointing at composites
    4) What you describe with the screws is called mapping. Way over your pay level at the moment. This means you need a co-ordinate measuring machine, run the screw, map its actual travel then write a correction code to correct the travel. Lets just say whatever the WillMill needs you will get and use C7 for the WillMill Junior.
    5) I suggest you find your local composite shops and ask them that any CF they are going to put in the garbage bin to put it in a bin for you. These will be small off cuts but they will add up to a lot of CF over time.


    5) long fibres vs short fibres - short fibres are becoming more and more used. You can pour them into a mould and infuse it. This is much faster then cutting and draping fabric. The cheapest recycled fibre I found is here
    https://en.procotex.com/index.php

    6) same if you find steel or al swarf just wash it and use it. No different to sand but stiffer and free.
    7) I see you collecting CF scrap to make the "skins" then fill will swarf then infuse. Now you have a CF structure and a heavy core perfect for machines.
    8) I'm assuming that this machine is a one off for you, If you intend to productionise the machine then the approach has to be totally different. Peter

    Its your machine I'll just keep chucking stuff at you until you are overloaded or put up a little flag if a fatal error is about to happen...

    I built my first XYZ gantry robot some 35 years ago from 6061-T6 aluminium... seems its time to change.the metal hasn't

  20. #20
    Join Date
    Jul 2018
    Posts
    6248

    Re: Router Design Research - Help me find more GANTRY DESIGNS to SIMULATE (FEA) in CA

    Hi William - when doing a study use the form finding in F360 . It will make a shape that is optimum for the load case.

    infusion video I've infused timber, CF, glass, rocks anything that's porous and you can put in a bag or mould. Look up stabilizing wood. Even human body parts. Look up plastination anatomy.

    cheers Peter

    https://en.wikipedia.org/wiki/Plastination

Page 1 of 2 12

Similar Threads

  1. 16M Length 5 Axis Gantry Router Design - NEED YOUR OPINION
    By omnicnc in forum CNC Wood Router Project Log
    Replies: 5
    Last Post: 12-19-2019, 02:16 PM
  2. New CNC router design - Gantry
    By yngndrw in forum Linear and Rotary Motion
    Replies: 10
    Last Post: 03-10-2019, 01:14 AM
  3. Hybrid Design - Fixed Gantry Router for High Speed Aluminum Milling - Metal Nibbler 5
    By Bob La Londe in forum Uncategorised MetalWorking Machines
    Replies: 7
    Last Post: 04-10-2014, 05:16 PM
  4. Questions about DIY gantry designs
    By Flyn Hawaiian in forum DIY CNC Router Table Machines
    Replies: 9
    Last Post: 11-28-2011, 03:30 AM
  5. ?? 4 fix gantry designs
    By eloid in forum DIY CNC Router Table Machines
    Replies: 1
    Last Post: 06-19-2009, 01:19 AM

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •