584,858 active members*
4,307 visitors online*
Register for free
Login
IndustryArena Forum > MetalWorking Machines > CNC "do-it-yourself" > Help with ballpark figures to use for sizing aluminum extrusion for DIY build
Page 1 of 2 12
Results 1 to 20 of 29
  1. #1

    Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    I need help sizing T slot aluminum extrusion for my build. I'm primarily trying to get some ballpark figures on typical real world forces involved to plug into deflection calculators plus any advice you may have would be ideal. I'm not looking for exact force calculations, just some concrete numbers that are good enough to decide between an extra beefy variant of 30mm base profile or medium grade 40mm or to determine that anything less than 80mm would be a waste of money........ At the moment, I'm trying to finalize the type and general layout of my aluminum extrusion frame elements so I can start iterating through the other design decisions. My next step is to select a suitable aluminum profile and redesign my frame to incorporate it.

    Due to pandemic, I have more time at home these days, so I'm attempting to get as much done on this project as I can before other responsibilities require my undivided attention. So, I'm not rushing this, but I am really trying to get at least the linear guide and ballscrews ordered from China sooner than later, and I need to size my machine frame elements before I can determine specific rail lengths (and then revise design based on available parts and cost).

    Sorry about the length of the braindump below, not sure what info you might need for my questions. I am not a mechanical engineer and it has been decades since I took physics and statics classes, so at this point I'm mentally atrophied enough to be a total lay person. On the long term, I'll relearn the math and physics involved, but this ol' brain takes longer on the uptake than it used to, so need assistance getting my frame up to speed. Suggestions on the subject are much appreciated.



    General DIY Build info:
    I'm building this for personal use and for the sake of learning every nitty gritty, PITA detail of making machines like this hands-on, and at the end of the build I'll have a machine that can do an ok job making parts for itself, other machines, projects, or whatever I feel like cutting at the moment (with limitations on material type of course). This project is intended as a concept design that incorporates a bunch of different ideas, more or less an experimental platform that I need to actually work at some point. I have to keep my budget as tight as I can, but not at the sacrifice of practical utility. So, I'm cutting corners where I can in order to spend a little more on having beefier critical elements, etc.

    Machine Type and footprint:
    CNC router with some cost effective attention to more rigid milling requirements, and CO2 laser combo with mostly enclosed optics to protect from swarf/etc. Dual gantries oriented length-wise to accommodate oversized material: one slow-ish for spindle, one optimized for speed and laser engraving. Separate Z axi for spindle and primary work envelope/platform. Yes, it's basically a medium sized cnc router mashed up with a laser engraver, but with an economical attempt to beef up critical parts of the frame a little to be a bit more milling friendly. I'm in an apartment with concrete floors, no garage. I have space for exactly one decently sized machine, thus one of the reasons this build is a spindle/CO2 laser combo instead of separate machines. I have other reasons for the combination, but they're more on the messing around and experimenting side of things. I have looked around at other people's builds and some commercial some commercial machines for ideas and starting points, but AFAIK, I have not seen any DIY combo builds that incorporate a full size spindle and full size CO2 laser instead of a diode, so this is a completely from scratch design.

    Target quality/precision/materials:
    I do not expect to get machinist-quality precision and repeatability. I want picometer precision and repeatability for $ zero dollah, but my slightly more realistic goal is to match or exceed quality of some of the cheap-ish chinese ebay CNC kits like the 3040/6090 kits, NOT the ultra cheapo plastic 3018's. "Materials" target is "good for DIY" precision/repeatability drilling and "ok" milling for aluminum plate up to 20mm thick plus any less demanding materials. I have a one-time project with stone I'd like to try, but stone is not really my target. Sooner or later I'll at least attempt any other material types and depths I can get my hands on, but a machine that can reliably cut ANYthing is waaaay outside my budget, so doing an ok job with 20mm aluminum seems like a decent arbitrary target.

    Oversized plus primary work envelopes:
    The spindle gantry is oriented long-wise to provide oversized work envelope: 50mm Z (barely enough for 1.5" material plus spoil board) and 4' wide plus elbow room, but requires doing a series of jobs and manually shifting material through iteratively. PITA, but hey, no limit to material length if its 4' wide or less. Primary work envelope implemented on a movable platform is 4' x 2' plus elbow room, with Z depth based more on budget than anything else, but figure at least 300mm. Due to laser and other project ideas I want to try out down the road, the main work envelope is on a movable platform and has it's own dedicated Z axis, in addition to the spindle's typical Z axis. When platform is at maximum upward travel, it is flush with the top surface of the frame 50mm under spindle gantry. Tentatively, a vice will be temporarily mounted on the platform when needed.

    Tentative electrical/mechanical bits n pieces:
    I'm in the US, but my apartment has 240VAC outlets available (Shhh!!! Don't tell the landlord!). Stepper motors to start with, might upgrade to a closed loop variant down the road pending additional $. I'm using open source controllers and firmware, but I have enough coding experience that this what I would do anyway regardless of budget. Unfortunately, the TMC5160 stepper drivers I already have are on those mini-pcbs so the maximum motor current is 4 amps per axis. I'm looking at steppers with a brake for both of the Z axi (both spindle and work platform).

    Spindle gantry and axi:
    "Hopefully Genuine" Hiwin linear guide for all axi, probably 25mm for gantry support and 15 or 20mm for everything else depending on price tag. Appx 1500mm gantry span with 50mm clearance. Dual ball screws for moving gantry, 16mm if I can get away with it, otherwise 20mm. Probably ball screw for spindle X axis, but not sure what's required for >1200mm travel, definately bigger than 16mm. Based on the costs I've looked at, I don't think rack and pinion will be an option. I want >= 10mm ballscrew pitch for something less-agonizingly slow, but I don't know what can be economically achieved with a 4 amp/axis (2 amp per stepper for dual ballscrew) stepper motor limit.

    Laser gantry optimized for speed:
    This is mostly TBD. I bought some C-Beam from OpenBuilds that I've been experimenting with. My current thought is to mount gantry with linear guide and use a minimal-load V wheel setup for Laser head X axis. The optics have to be protected and mostly enclosed, so I'm designing my own mirror mounts and laser head, so much of that will be directly integrated into the X carriage itself. Should be extremely low weight to move during laser engraving, just the weight of the plastic carriage and bearings, one mirror w/o mount, one lens w/o nozzle nor head, 3D printed minimum weight low pressure air assist nozzle, plus the drag of the air assist hose, and that's about it...

    Work platform Z axis:
    Lead screws for Z static load and XZ/YZ moment force, and probably minimally sized (12mm? 15, 20mm?) linear guide to deal with X/Y static load and XY moment forces, but I'm mostly throwing out fancy words since I'm not sure the most appropriate way to build and mount a platform like this. How many lead screws for a 4' x 2' platform?!? Well, at least 4. What diameter?!? No idea. Current thought is to use 12mm instead of 8mm for the sole sake of flattening the angle of the 2mm pitch thread a little to reduce backdriving force, but otherwise I have no idea. Feel free to make suggestions.

    Spindle:
    TBD, mostly determined by how much $ I have left over after everything else. Tentative target is a water cooled 2.2kW. I say "water cooled" only because I have the impression that it would be a little quieter than an air cooled spindle, thus less noise I have to dampen for my apartment neighbors' sake. No matter what, it will be a PITA to minimize noise pollution, so need to minimize any source as much as economically possible.

    Laser tube:
    TBD, mostly determined by how much $ I have left over after everything else. I've been considering many different options for this, including a possible dual tube setup with high >=120W and low <60W tubes. At this point, I just need to make sure I accommodate all the different options as far as my machine frame is concerned and remaining budget will determine the rest.

    I have paper sketches and some preliminary CAD done, but nothing worthy of public viewing yet. I need to get the profile selected for my main aluminum extrusion frame elements before I get much more of the overall design out of my head and into Fusion.



    Any help/pointers on T/V slot profile selection, and any real-world numbers to ballpark force estimation to plug into deflection calculators/etc to make De$ign Choice$ would be much appreciated. I freely admit my newb-i-ness, but am quick to point out that I'm working to fix that. Heh.

    Thanks!

  2. #2
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Hi P2N - Extremely optimistic project. Combo machines are compromises towards one of the functions. In this case you are building a Mill not a router due to the mention of 20mm aluminium. You need to forget about construction extrusions and use steel or aluminium structural sections. A 1500mm wide gantry needs to be very stiff to do what you describe. At a guess you will need a steel 200x200x12mm thick square section at least. I suggest you find a commercial machine that does what you expect and this will point the way to dimensions and materials of construction. You will need 16mm plus screws to side step whirling (or whipping). Your process goes like this: take the best paper design you think you have and get it into CAD asap. Ignore all motion items I suggest that 20mm or 25mm square rail will be fine. 2) do structural analysis and aim at 20N/um static stiffness with spindle in the middle of the gantry and full down 3) Once you have the machine at the correct stiffness then you know the weight of things so can do the motion calcs to determine screw sizes and bearing sizes. At this point various bits won't work out as you thought so you then fix all the loose ends and changes of minds and you end up with a working machine on paper Then you cost it fully. Then contemplate the cost and how you could spend your money better. Then after that review A) start again as its too many $$$ or has drifted from your original ideas B) commit to the costs and order everything and start building... budget about 6-12 months to get to this point. Done this several times so know how it goes.

    The shortcut is to find a thread with a machine that you like and follow those thoughts and clues. Saves a bucket load of time and effort. Keep at it, Peter

  3. #3

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Hi Peter, thanks for the feedback!

    A lot of good points, well taken. No argument whatsoever. I have lost count on how many times I've repeated your step A) to get to where I'm currently at. The conclusion I keep repeatedly reaching is that given my limitations and flexibility on target result, my only real way forward is the absurdly ridiculous process of designing in reverse: Do the best I can with what I can get within a minimized budget, selectively add to that budget for critical elements, and then live with the results. But I need to make sure the results are not so horribly bad and unusable that the entire effort ends up a complete waste of money. With the tools and available space I have, I need to stick with T slot profile if at all possible. In my current wack at designing this build in reverse is to design the best frame I can for $1000, and then immediately design another $500 into critical frame elements. In other words, design the frame with light weight aluminum profile and then use high rigidity/higher cost variants of the same profile for the most critical frame elements. It only takes 5 minutes of googling CNC kits to know that you cannot build a good medium sized frame for $1500, but would results from a substandard frame be marginally usable? My expectations are low, but I'm optimistic that an affordable-to-me medium ground exists out there somewhere. I currently have another $1000 set aside for chinese rail and ballscrews. Motors, any electronics I don't already have, spindle, VFD, laser tube, power supplies, etc are somewhat incremental costs compared to frame and can mostly be purchased over time, so I put that in a separate budget category.

    The best I can realistically *hope* for is a frame that is marginally good enough for routing, but attempting to spend more on a few key parts to eek out a little bit better rigidity in specific elements plus some selective mass increase with epoxy granite and/or weights rigidly attached to the frame. I do not expect miracles from adding raw mass, but could reduce the magnitude of error a little bit.

    Right now I'm trying to figure out which aluminum profile is worth the time to do a full design cycle around: there is no way 2020 is good enough and 120mm x 120mm for every frame element will be too expensive for me, thus either one is a waste of time to design around. SO...

    Peter, I read your numbered steps as:
    1) 20mm or 25mm "linear guide rail" will be fine, or were you meaning aluminum or steel square tube for frame elements?
    2) 20 Newtons per "micro" meter static stiffness? Let me make sure I understand this, please correct anything I have wrong. For a given material type/profile with a given span that is rigidly supported at both ends, as long as the deflection from 20 newtons of force acting on the center is less than or equal to 1 micro meter, then that is a good structural element to use for milling purposes.

    Using my tentative gantry span and a variant of Item24's 60x60 aluminum profile as a concrete example (handy deflection calculator at bottom of page):
    Plugging in 20N and 1500mm yields a 0.02mm deflection and is therefore unusable as my gantry?
    https://us-product.item24.com/en/pro...natural-41909/

    But using a thicker variant of 100x100 might be "good enough for milling" for my gantry, although probably out of my price range (NOTE: online calculator rounds to 0.01mm):
    https://us-product.item24.com/en/pro...natural-62456/

    If my understanding of your "step 2)" is correct, then 20N is the ballpark number I was looking for. Thanks! Now to figure out how to apply that to the rest of the frame elements beyond the gantry...

    3) I'm sure I'll eventually have questions on this step, but I've found at least a few articles on ballscrew and lead screw sizing, not 100% sure how to use the equations correctly, but I at least have a starting point on those.



    More info you can probably just ignore that explains a bit on where I'm coming from:

    My primary limitation is living in an apartment without a garage. I have use for a table saw, laser cutter, router table/CNC router, mill, free standing drill press, and any other shop tool you could name. But I do not use any single shop tool enough to merit the footprint within the limited space of my apartment. I already cringe at the full shelf my miter saw takes up when not in use. BUT, I can easily justify a single medium sized frame with marginal rigidity and mass, and use that frame as a base for any shop tool I can get usable results out of, and when I'm not actively using a tool, the machine work platform is automatically an additional workbench. The massive PITA of moving to a house for the sole purpose of having a garage doesn't make sense to me when I already have a perfectly good 8'x6' chunk of floor space to fill with a medium sized multi use frame.

    I really do not want a swiss army knife machine that does a poor job at a lot of different things, but that is all I have space for. That said, one big bonus for my own purposes is using the frame as a platform for experimentation and trying out different ideas. For example, using an encoder to automatically track the relative position of oversized material I manually shift through the machine. I have almost zero long term use for such a feature, but its something I want to try out.

    Within my various limitations, I know ahead of time that anything resembling a "real" mill is flat out not possible. I've lowered my expectations accordingly but don't have the experience to know what is and is not distantly feasible even with degraded quality. If 20mm aluminum plate is not realistic, then I have to lower my target to 10mm, and if not that, then 5mm, etc. Same thing applies to my "CNC router with some cost effective attention to more rigid milling requirements, and CO2 laser combo". But I can't ask experienced CNC'ers questions without stating a material and thickness, thus the arbitrary target of 20mm aluminum plate was born.

  4. #4
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    HiP2N - Its not the actual thickness that is the problem its the depth of cut (DOC) that you expect that is. You can cut any thickness you like if you use 0.1mm DOC but it takes 200 circuits to get it done and you wear out the mill bit because you are only using 0.1mm of it. This is a power available problem and a machine stiffness problem. The more power you have the more the DOC can be but the machine stiffness has to be there to be able to be stable at that DOC. Then there is the large issue of lubrication which Al needs and lasers will hate (mist gets in their eyes). For instance a VMC is about 100N/um stiffness. This means if you push on the tool with 100N (10kgf) then the tool moves 0.001mm (very stiff). Personally I'd forget about construction extrusions they are convenient but expensive per KG and not stiff compared to std extrusions of the same dims. You can build a totally acceptable machine in plywood and aluminium sheet with basic tools. Trying to find your earlier Q's.... linear rail comes in two main types round rail and square rail. Round is cheaper but may work where your headed. Square rail is stiffer but twice the cost. So cost/benefits have to be addressed. Being your first machine I suggest you tone down the expectations and build the "best" machine for the budget.

    I think you will come unstuck with the laser there's many issues with mirrors unless you are going for a direct mount tube which again will require much CAD sorting. 20N at 0.2mm deflection is 20N for 200micro meters (0.02x1000=20 micron or 20um) so 20N per 20um is 1N/um which if you look at the table attached for mills you will see the issue straight away. You need to write down a basic specification of what you want the machine to do (so the target does not change through time) and get into CAD and sort it. Good luck. Peter

    here is a plywood machine example that would do your stuff. Plus there are others in the forum..
    https://www.cnc-holzfraese.de/

  5. #5
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    by the way P2N - When modelling a gantry it is not rigidly supported . Its closer to pin ended. Plus you need to have the load offset to include the Z axis offset so the gantry is in combined torsion and bending. Bolted structures are about 50% efficient so if you "bond" it all together in reality it will be about half the calculated stiffness. If your using fusion I think you can include contact conditions eg a friction connection which will get a closer answer then bonded. Cheers Peter

    ripper – the different CNC – MAXMALI.COM here's a good blog on a build may suit you. Although I still recommend you can do better not using construction extrusions....

  6. #6

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Quote Originally Posted by peteeng View Post
    Personally I'd forget about construction extrusions they are convenient but expensive per KG and not stiff compared to std extrusions of the same dims.
    When you say 'construction' extrusions, are you talking about the type made and sold by the likes of 80/20?

    Thanks

  7. #7
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Hi M880 - Yes 80/20 is an example. There are many brands. They are very convenient but structurally inefficient (see prior attached articles in #4) they have lots of thin edges that vibrate and cutting them square is difficult. There are many many successful extrusion machines however, so depends on what you want.

    An 80x80 construction extrusion weighs 5kg/m yet if you buy a 80x80x4 tube it will weigh 3.5kg and will be the same stiffness (see prior article for calcs). Since these things are sold by weight the tube will be cheaper and do the same job. You then need to figure out how to make holes and join it however but holes or bonding is more efficient then bolting to thin slots.....Regards Peter

    and P2N a big laser needs more travel speed then a router so they usually use belts. So in your spec sheet you need to find out the typical cutting speeds of your laser and router so you design a drive that covers both. A compromise may/will have to happen here.. also you say a mill is not possible. Everything is possible you just have to think bigger & take advantage of the brains trust that this forum is.....Cheers Peter

  8. #8
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Hi P2N - To give you a clue. 16mm al plate is a common size to make a stiff router from. Local list price for this is $463AUD per sq m. If you make a 2mm/16mm/2mm al/mdf/al laminate it costs $90 per sm. Yet is the same bending stiffness as a 16mm solid. Its easy to cut with a jig saw and is much lighter. You can do this yourself and save $$$. You can contact cement the laminate or use epoxy... Lots of opportunities with this approach. If you use 25mm thick MDF in the middle its the same stiffness as a 21mm thick aluminium plate. Now your into cheap very very stiff territory... Leave you to think Peter

  9. #9

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    More great info, thanks! For modelling each frame element and joint, is there a FAQ or something that is along the lines of "For a linear guide (square rail) bearing block, model as a BLAH joint with BLAH derating of stiffness between joined members..." kindof guidance specific for CNC related calculations? I've dug up some of my old textbooks, but it's going to take a while to relearn anything straight through and my old textbooks are more conceptual and not terribly great on the practical application, at least not directly to this build.

    I am not opposed to using square or rectangular aluminum tubing or any of the other aluminum sandwiches, but I really do not think I have enough manual skill to get an entire frame remotely square or rails aligned. With the T slot, I get infinite "do-overs" until I get usable alignment and squareness. But then again, I haven't made anything with metal tube before, so I really don't know. I have 8' x 6' floor space for my build, but that is also just about all the room I have to build the thing too. Perhaps just doing the gantry with tubing might be a good compromise since it has the longest unsupported span anyway. Any rough estimate on how much cheaper tubing is compared to T slot for comparable stiffness?

    Laser engraving and light duty cutting with a shorter focal length is why I'm going with a dual gantry setup, one slower spindle gantry and a second gantry optimized for speed with very low weight. I'm about 95% sure all parts of the spindle side will be ballscrew, probably the spindle Z axis too, but faster laser gantry and carriage will be belt driven. Work platform Z axis has to be lead screw. I've considered a second laser beam path that is routed to the upper aka spindle gantry for long focal length cutting, or mounting a second higher power laser tube on the back of the spindle gantry, but those are decisions for another day. I'm not worried about the laser optics side of things. Don't get me wrong, so many things will be a total PITA, but I think I have a decent strategy to keep the optics fully enclosed when the laser gantry is parked and mostly enclosed during use. Iterating the design of my beam combiner and mirror mounts will be a PITA, so will my laser head, but should result in all the optics being protected and still get alignment after some significant effort. For any dust and coolant misting not dealt with by mostly enclosing the optics... Not sure how much help might come from leaving the air assist and/or exhaust blower running when using the spindle...

    Mill is/is not possible? Well, if I can build something with medium sized frame, the additional complexity of a second gantry and moving Z work platform, within a budget of $2k not including electric bits and pieces, and be able to call it a "mill" with a straight face, I will make no complaints what-so-ever. It does not seem realistic so I've assumed I should let go of that fantasy. But I would be happy about being wrong on that point. The big caveat is my frame needs to accommodate being used as a table saw, drill press, workbench, misc general purpose shop tool, in addition to typical laser and spindle work (but only one of those at any given time...). All of that is pretty straight forward to accomplish after starting with a decent frame, but it also means I have to design and build the thing from scratch (AFAIK).

    My next step appears to be more window shopping at other DIY designs but this time paying more attention to non-T-slot builds. Well, that and figure out/find out how to model the thing so I can make some frame decisions and get a better idea of what is actually im/possible within budget.


    When I graduated, I kept all of my college textbooks, but only a few have survived the many years and many moves. Luckily, one happens to be my old Statics book. I flipped through the thing for the first real time in decades and came across the following illustration. At first I thought it was hilarious, but now it just makes me feel really old. "Yes kids, Err, I mean recent college graduates, that's what computers used to look like. And the computer operator is pretty hip too!"

  10. #10
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Hi P2N - Look for Bamburg - Principles of rapid machine design. It has all the answers. I use sophisticated FEA that includes contacts, friction bolts, sliding etc etc. Fusion 360 has most of this as well. Initially trying to capture all of the real effects will drive you nuts, Just "bond" everything together and use an efficiency of 50%. Your alignment issue will be the same whether you use T slot or build bits yourself. Search for Maximus (or Brevis-HD another of my designs building now) a thread here where I design a machine from scratch to costing. Lots of info in that. $2K USD is $4k AUD what you want to do is not possible here for that amount $4K AUD will build a decent 4x8 router machine at cost... Your requirement for the machine base to be a bench means your rails need to be on the side or underneath. Attached is image of MDF machine with suitable rail design. You will NEED to do everything yourself to get to your budget. So it really needs to be sorted in CAD and sourced and costed before you do anything. It only takes one change to change many things and this will involve lots of $$$ or lots of rework. There are threads that publish complete BOMs and cost search for these you will need the info. Good Luck on your journey Peter

    a std square or rectangular alum ext will be substantially cheaper then a construction extrusion. There will be an al supplier nearby you I'm sure. Don't weld anything, consider bonding everything together this gives you the adjustability you want when aligning stuff. When I studied eng we used punch cards & main frames and slide rules, then 5" floppies, then I had a break from study and when I came back the rooms were filled with PC's, had to relearn everything... I strongly suggest using the laminate as then you can design the size and shape you want. Std extrusions and steel sections limit your possibilities too much.

  11. #11
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Hi P2N - You say you don't have the manual skills 1) Only way to get them is to do it 2) If your absolutely clear that you can't then you have to find a kit, which you won't on your specs so your stuck with doing it. 3) design it and contract all the parts - your now well over budget. So looks like your a) doing it b) forget the whole thing and learn to play golf Peter

  12. #12

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Peter, got copy of Bamburg and started digging through that and the rest of the pile of new-to-me info. Again, thanks!

    It appears that I am doing some version of your latest option "a)" due to process of elimination. If you've seen my back swing and follow through, you'd probably suggest building a log cabin to provide shelter while searching for my golf ball.

    1) Agreed, but I have several factors working against me on that. Last year I built my first decent sized wood project: a 10' x 8' L-shaped floor to ceiling workbench/shelving unit for my Lab/Office. I learned important lessons from that project: my apartment's concrete floors are not even close to level, "square" is a mythical beast closely related to the Loch Ness Monster, storing anything round on a flat surface is simply not going to happen in my Lab/Office, and any decent sized project will require nearly all my available apartment floor space and I have to climb over or under things in order to use the bathroom until I get the project finished. My main problem is poor precision doing manual work. I can measure 20 times and still manage to mess up a cut with a miter saw or drill a screw in crocked and off center. All things considered, I'm happy with my workbench, but the workmanship is not good enough for a machine frame. That is specifically my main worry using anything other than T-Slot, my lack of precision that results in things not being square or level. Maybe I'll do better next project? Not sure.

    Tool-wise: Other than a 6" digital caliper, digital protractor, and a digital dial indicator gauge (resolution to 1um accurate to 30um), all measurement tools I have are on the cheap end of carpentry grade: squares, levels, tape measures, etc. If using materials other than t-slot for my build requires better tools than an economy-grade miter saw and decent 24V Dewalt power tool set, it might be a no-go. I have no problem picking up a cheap-ish bench top drill press if required plus any relatively low cost tools that can fit in a toolbox, although that will start cutting into my available budget.

    2) Yup, one way or another I'm stuck designing and building it from scratch, and I am totally ok with that, but I am worried about poor workmanship for anything I have to cut myself, or drilling holes precisely enough to mount square rails. T slot goes a very long way to mitigate that problem, but the jury is still out on the subject, have to balance budget with likelihood of success and/or usable results. Oh yea, and then there is also that part about getting the design right. Not sure yet which side, my untested (and currently unknown) mechanical design skills or manual workmanship will be the weakest link... Hmmm. Should be interesting finding out.

    3) Yup, contracting ANY parts is not going to happen, but buying t-slot cut to length (+/- 0.5mm) is cheap enough I won't cut that corner (HAH). I haven't looked into it enough to know how much that would cost for other materials. It would be extremely helpful if I have a few frame pieces that are guaranteed square by supplier with edges straight enough to use as reference edges to align everything together.

    You've mentioned the term "bonding" quite a few times and I'm not sure what you mean. I understand "welding", "bolting", "screwing", "gluing", but not sure what "bonding" refers to in this context.

    Punch cards eh? Ok, perhaps I'm not quite as old as I feel. Thanks, I needed that. Hah!

  13. #13
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Hi P2N - Bonding I have used in two contexts 1) In FE work we talk about a "bonded" connection which is a perfect elastic connection 2) Bonding I have also used in terms of adhesives. You can bond joints together vs bolting so you can adjust it, clamp it let it cure DONE.

    So onto your workmanship. Many people build a series of machines vs their "gold standard" dream machine. Even the first lathes and mills were built in a series. So you can build an MDF router that is as good as you can. Then use that to build a better plywood one (MDF is half the price of plywood), then that to build a better aluminium one etc. The learning curve on the first machine is huge but all the bits transfer to the next one easily. Once you have a CNC no matter how poor, it will be better then your hand work. One of the early lathe builders got a grant from the govt (british I think for making cannons) and spent 3 years hand making parts to put together the machine, then used that to build the better one and the better to build the real one. There are examples of this in the forum where they build plywood, then aluminium machines. etc. Look up Gully he's built an MDF machine very nice too and now building a better metal one. Your much better off building something smallish and simple to get you through the curve and into the game then something extremely ambitious and fail or burn out. Heaps of projects here that burn out, dry up or are still born... I've built 4 and they are all missions of blood sweat and tears. Plus there's the 50 others still on the HD that got parked for various reasons. So keep simplifying, KISS and keep at it. But do not buy anything till you sort it on CAD Peter

    if you make a boo boo in extrusions hard to recovery it. If you make a boo boo in plywood , fill it with epoxy and keep going. The repair is better then the timber (nearly)

    https://lecarnetdubois.wordpress.com...tion-dune-cnc/ plywood machine

    If you want something really square the local cabinet maker will put it through his table saw for you....

  14. #14
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Hi P2N - I don't mean to bang on about this but here's a quick note on extrusion vs plywood. If you take an 80x80 construction extrusion vs a 100x100mm solid plywood beam the timber will be 1.33x stiffer and 1.3x heavier. But if weight critical you could make it hollow with no negatives. So now you have a material that you can make any shape or size you want. You can put pockets into it or step ups on it to account for mismatches eg screw drives to rail height deltas etc etc. Only down is change in shape due to humidity but if its indoors it will be fine. Coat it in epoxy or PU or any good resin and your change in shape will be minimal. I selected an F17 ply but you can get stiffer. F17 is off the shelf at the hardware. So I'll leave you with that to contemplate. Off to work on my new bench for the new router... Peter

    Oh yes and wood is 10x or 20X damper then metal win win win...

  15. #15

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Peter, by all means, bang away! It is all good info and covers areas that were not on my radar before. Thanks!

    I'm about 2 pots of coffee through the Bamberg, mostly skipping the equations since it will take a long while for those to sink into my noggin. Also started sketching a much, much more raw fundamental shape for my frame to get into cad. One big drawback to Fusion is the whole cloud credit thing for analysis tools. I really can't complain since the base software is free, but if you are a total newb to the whole thing like myself, I'm going to burn through a great many iterations just attempting to use the software correctly. I considered re-classifying my personal use build as a potential startup company in order to get a free year of SolidWorks since I have the impression I'll have unlimited use of analysis tools for at least a year... Would be nice if they had a hobbyist software license option that was less than a few thousand each and every year... But I digress.

    As far as building a machine to build a machine: I had intended on using my 3D printer to make plastic parts to use as temporary mounting plates for non-load bearing parts and just run the new machine really slow while CNCing (drilling only, no milling) it's own metal parts to replace all the plastic. There are plenty of purely 3D printed DIY builds that prove it is at least a reasonable approach to consider. For temporary gantry risers, either free hand routing MDF or 3DP templates to screw onto the MDF and hand route off of that. That's what I did for a small electronics panel for my current work bench and the results came out well, the wood cuttouts were pretty tight around the panel meters. This result was MUCH better than my questionable squaring and drilling skills. I have a bit more research to do before I return to the question of what build tasks are and are not realistic to manually do myself.

  16. #16
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Hi P2N - Try Freecad it also has free FEA & CAM. Its open software will always be free. Just be aware that if you don't get your machine stiff enough its not a case of "just going slow" to cut aluminium. If the machine is not stiff enough it will wobble around and not drill straight and just be a total frustration. So it has to be UBER rigid from its birth. Peter

    https://www.freecadweb.org/

  17. #17

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Arg. Yea, I was mentally putting motor and pulley mounts/brackets into the precision and repeatability category and neglected their contribution to rigidity... Arg. Will have to rethink how to baby step those parts into existence, well, at least the ones that aren't off the shelf brackets and whatnot. I didn't know freecad had analysis... Will be installing that shortly. Thanks!

    Have you heard of any non-laser/non-3DP builds that have a vertically moving work platform, aka implemented a full Z axis by moving the work piece vertically?

    I kinda doubt one would exist since 99% of the time Z would only be implemented by moving the spindle, but I don't know. There are plenty of examples of moving beds with normal laser machines and some large format 3DP's, but I have not found a vertically moving bed on a machine with higher level of force between the tool head and the bed/work material like a router or mill, I'm sure for good reason. In my crazy CNC router/laser mashup build as *currently* imagined, I'm pretty sure a moving Z work platform would be required, but would also need to be much beefier than you commonly find in normal CO2 lasers and some large format 3DP builds. I have seen one router/mill where the guy could adjust the height of his bed, but that was a somewhat intensive bolting process and in no way was it a full Z axis.

  18. #18
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Hi P2N - No I haven't come across a lifting/dropping table other then in lasers & 3DP's. But in very big (huge type) mills they have lifting gantries which would achieve the same thing. My next two machines a router and a mill will be lifting gantries. I think they tick the boxes for high Z, very good stiffness etc etc. Peter

  19. #19
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    Hi P2N - Here's a moving gantry, lifting gantry mill



    there are 2 small scale lifting gantries in the forum as well. Peter

  20. #20
    Join Date
    Jul 2018
    Posts
    6254

    Re: Help with ballpark figures to use for sizing aluminum extrusion for DIY build

    i P2N - Here's two small scale lifting gantry machines. Peter

Page 1 of 2 12

Similar Threads

  1. aluminum extrusion
    By Gary N. in forum Want To Buy...Need help!
    Replies: 0
    Last Post: 04-07-2018, 04:39 PM
  2. Aluminum T-Slot Build Using Misumi Extrusion
    By Bear5k in forum T-Slot CNC building
    Replies: 55
    Last Post: 10-07-2016, 11:17 PM
  3. Am I in the ballpark? 1/2" aluminum
    By 4EyedTurd in forum Uncategorised MetalWorking Machines
    Replies: 8
    Last Post: 10-24-2013, 04:55 PM
  4. aluminum extrusion
    By georgebarr in forum DIY CNC Router Table Machines
    Replies: 20
    Last Post: 05-14-2013, 09:03 AM
  5. Aluminum extrusion
    By Pysiek in forum RFQ (Request for Quote)
    Replies: 3
    Last Post: 07-21-2012, 05:20 PM

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
  •