Page 1 of 2 12 LastLast
Results 1 to 12 of 19

Thread: Choosing a precise benchtop mill

  1. #1

    Default Choosing a precise benchtop mill

    Hi everyone,
    I apologize in advance for the long post.
    I'm a mechanical engineer and I'm in the market for a small bench top CNC mill so I can produce prototypes of an invention I've come up with. Due to my relative lack of CNC experience, I'm hoping the forum could help me choose the best mill for my needs. Here are my desires/requirements in a nutshell:

    - I'll be machining aluminum and engineering plastics and possibly some PCB routing. The largest parts will be roughly 150mm x 65mm x 35mm. Features range from large 20mm deep pockets in the aluminum to sub-millimeter features and possibly some micro drilling. The small features lead me to believe I need a high speed spindle, but I worry about my ability to mill the pockets with such a spindle.

    - I have tolerance requirements that seem to be significantly tighter than most bench top machines are capable of. My current design calls for dimensions of +/- 0.01mm with a surface flatness/perpendicularity of +/- 0.005mm. I came up with these requirements based on some engineering analysis, but ultimately I think it will take some trial and error to know what my real tolerance requirements are. So ideally a machine that could handle this tolerance or a bit better.

    - I'd like a machine that can reasonably fit in an apartment or a bedroom, so I'm looking for 110V power and a manageable size/weight. I'd like to spend less than $15k. I would prefer a used machine.

    Here are the machines I've considered:
    EMCO Concept 55 Mill: accuracy/repeatability doesn't quite meet my needs, size/weight at the upper limit of portability, slow spindle speed. However, I like the "mini-VMC" features of the machine like an ATC and a full enclosure.

    Sherline/TAIG: Love the price and portability, but I don't see how these will meet my tolerance requirements.

    Minitech: Seems to have the best accuracy/repeatability of the bench top machines, but I don't know if it will fit my budget. Also uncertain about the proper spindle choice (NSK?). Seems hard to find a used machine.

    Tormach: Like the features of the machine, but it's probably too big and heavy for me, and I don't think it can meet my tolerance requirements.

    MDA Precision/Wabeco: Not entirely clear if it can meet my accuracy requirements, and it's pushing the limits of portability and budget.

    Haas Office Mill: Ok, this is obviously way beyond what I'm talking about, but it is the machine I dream of. I'm sure it could do everything I need, but the price, size, and power requirements would be tough to manage. It's also a bit overkill for just some R&D on small parts. However, the ultimate goal would be to produce the product commercially, and the Haas would probably be most up to that task.

    So far it seems like minitech is the best bet, but the price is just so hard to stomach. Are there any other machines I should consider? Are my requirements unrealistic? I've considered outsourcing to a prototyping shop, but the number of design iterations I'm anticipating will make that a fairly expensive proposition (I've gotten some quotes). Also, I've been wanting to own a CNC mill (and learn to use it) for almost 10 years now. I feel like an engineer without a machine tool is like a musician without an instrument.

    Thanks for any help you can provide!

    Similar Threads:


  2. #2
    Registered
    Join Date
    Apr 2005
    Location
    Canada
    Posts
    389
    Downloads
    0
    Uploads
    0

    Default

    IMO maintaining that kind of accuracy takes a lot of knowledge and tweaking even with a very good machine.

    I'd honestly recommend you outsource the job. It will take a lot of time to learn how to use a mill even if you manage to get everything you need under 15K (unlikely once you add on tooling, and measuring equipment capable of ensuring you are within tolerance).



  3. #3

    Default

    I mean this with respect and I'm not making a pass at you:
    Do the tolerances really have to be that tight?

    A lot of people subscribe to the "if you're going to make something, you might as well make it perfect" mindset, but there's also the "good enough is good enough" mindset. I don't know your design or its true tolerance requirements, but if +/- .01mm isn't actually needed then this purchase and learning process will be a lot easier on you.



  4. #4

    Default

    Hey guys,
    I totally understand that my requirements are ridiculously tight. The parts I create will be moving/sliding against each other, and performance of the design is dependent on minimizing gaps between the parts. So I've analyzed the performance impact of the gaps and came up with this tolerance requirement. I've considered designing the parts to break-in/wear-in to minimize the gaps, but my wear analysis suggests an unacceptably long break-in period.
    Anyways, I could go on and on about the details, but the requirements represent the best estimate I can come up with before actually making a prototype.

    Are my requirements unrealistic for even a full-size VMC, or is this mostly going to just be an issue for a small bench top machine? I've looked at the accuracy/repeatability specs on a number of larger machines, and I'm led to believe that it would be producible on a larger machine without too much trouble. I imagine some of the complicating factors (thermal expansion, etc) will be the same, however. My employer likes to keep us engineers far away from the manufacturing side of things, so my understanding of the issues involved in precision machining is probably a bit lacking.



  5. #5

    Default

    I very much doubt $15K will be anywhere near to enough to buy a machine capable of such tight tolerance. Big, very expensive VMCs cannot do it. And, any such machine would HAVE to be kept in a tightly temperature-controlled environment to hold such tight tolerance, as thermal expansion of the machine itself exceeds your tolerance. It would also likely hold that tolerance over only a VERY small range of travel, like an inch or two, and that would require screw-mapping, so there's the whole issue of calibration to deal with. And, whoever is operating the machine would have to be a highly experienced, and talented, master machinist - not something you're going to learn to do in a few weeks.

    As an engineer, I've always felt that, except in extremely rare cases, designs that REQUIRE such tight tolerances are poor designs. You should be able to design so that you get the performance you want without having to have parts made to ridiculous tolerances. If you need tight clearances, then provide adjustability, or use scraping or lapping to take off the last few tenths, rather than trying to do it all in the machining.

    Regards,
    Ray L.



  6. #6
    Registered
    Join Date
    Jul 2006
    Location
    USA
    Posts
    855
    Downloads
    0
    Uploads
    0

    Default

    most tight tolerance parts and brought close to size then ground to final.
    Very tight tolerances even in the hands of a highly skilled machinist are hard to hold. break a tool and you get to start all over. Variances in tool grinding will take you out of tolerance.

    Also remember this, a machine is just a tool. The knowledge to utilize the tool is what is important. Even with cheap tools a skilled person can get great results. its all about knowing the limitations of the tool and working around them.

    Another thing to remember is 90% of the machines this size are constructed of aluminum. With having to hold high and tight tolerances even the temperature must be controlled. I am not talking about ambient temps. I am talking about the thermal expansion of the frame of the machine. the long a machine runs, the warmer it gets due to heat transfer from the motors to the frame.

    I have worked with a client that required high precision. We ended up using liquid "HEATING" to maintain machine temps higher then that which it would see from normal run time usage.



  7. #7
    Registered
    Join Date
    Dec 2006
    Location
    USA
    Posts
    839
    Downloads
    0
    Uploads
    0

    Default

    Yes without knowing anymore about what your really trying to achive I expect that there will need to be second operations performed to achive this. The machine will not need to come anywhere close this good, and the second operation will do the fitting ( hand scraping, grinding,ect,ect).


    The .01mm can be achived through a good grinding setup, .005mm might take something a little better. It all depends on exactly what the parts are. It may be even easier through some type of ajustment, but I would expect the parts still needing hand scraped or ground to achive the flatness.

    Your boss also needs to learn how comunication and cooperation can really enhance the working flow. This right here is a very good example of what has lead to crappy products, and even companies going down because of the products. We have a lot with in our reach now which has not been there before, best use it, and get someone within the company to help with what is needed here, or bring someone in that can help. A product with this level of performance surely can bring a price which will allow extra guidance. All though your guys on the other side may know the anwser all ready and when they see the part description back up and ask for a redesign, or simple changes.

    Dont get me worng, I admire your willingness to research and try and find anwsers on the quick and easy. But without knowing more about the product its hard to even give guidance on building such a part, and I probably couldnt if it envolved more than such easy operations as hand scraping or grinding.


    If more could be given about the part there is a section on down in the forum where there are those that can give good guidance on such needs. Finding out what it takes to build the part would be good in giving guidlines on what type/types machines, or if a machine is even needed. Heck it may even be possible to cast the part and use the simple second operations I meantioned to finish it, who knows without more info.

    One thing is for sure, your not going to get a machine to spit out parts at this level for the funds meantioned. And even if one is bought at the high price commanded it will take much more than the price of the machine to achive it. Heck it will cost more than this to even have the machines/tools to measure the part for this level of accuracy (in house).

    But dont give up, I have seen auto machinist hit .01mm with a hone in a small corner shop. And there have even been guys that have aligned such quality parts at home, or hand scraped to this level of fit.. So the key is knowing what is being built, or atleast, a very good description of the machanics for someone to go by. And then posting in the section for getting such info. I believe its "General Material Machining Solutions" or " General Metal Working Discussion" or maybe "General Buisness Pactices and Pricing"( might be best one).


    There may be another section if you look, but again finding what it takes to build such parts would be a good start. Or you may find a way to add ajustment to achive the level needed. Good Luck.

    Jess

    GOD Bless, and prayers for all.


  8. #8
    Registered Bwana Don's Avatar
    Join Date
    Apr 2009
    Location
    Detroit-USA
    Posts
    105
    Downloads
    0
    Uploads
    0

    Default Tolerance

    You need a grinder to hold those tolerances. Way beyond a mill for .0002" flatness and .0004" size. You can get a grinder for cheap and still have money for a little mill.

    Use a little cheap mill for roughing, then follow up with another process (grinding), or redesign the product for bigger tolerances. I know easier said than done. lol

    I agree about an engineer needing a machine tool.

    Don

    Still working in the "D".


  9. #9

    Default

    You could do it on a high end machine, stress relieve the material, and have temperature control. We have a few machines that are very small that would hold that no problem, but your looking at 200K for a machine with a work envelope of 5.9"X5.9".



  10. #10

    Default

    For flat tolerances, get yourself a nice flat piece of granite, and wrap it with some fine wet silicon carbide sandpaper... Then it's nothing but elbow grease and time...



  11. #11

    Default

    You say you want to produce prototypes.
    This implies there will not be many. Based on this assumption, I say you can do it, by conventional machining (no grinding) with a good quality repeatable machine tool. You'll just have to 'sneak up' on the dimensions. This is why you need repeatability even more than accuracy. It's difficult to say for sure because we have no idea what your parts look like, but cutting it undersize, measuring, and cutting again until the part is within spec is a perfectly acceptable method for making prototypes.
    If you want to do this as a production process, that's different. My company does this on parts requiring 0.01mm tolerances and the machines are customised VMCs costing hundreds of thousands each. They all have on-machine probing and parts are also inspected off-line using CMMs. But for prototypes, I've made the same parts to the same tolerances on a 10k CNC converted Wabeco.

    LongRat
    www.fulloption.co.uk


  12. #12
    Registered
    Join Date
    Jan 2012
    Location
    keniduh
    Posts
    17
    Downloads
    0
    Uploads
    0

    Default

    You also need to consider cost of CAD/CAM in case your home setup doesnt have it yet. Depending on what you choose easily a few thousand there or more or less.

    What is your absolute maximum budget for this project?

    With a sherline or a taig, I think one can get +/- 0.03mm without too much problem given the cartesian axis are calibrated to good accuracy and have good repeatability. The +/-0.03mm here comes from the spindle runoff actually. These machines have some backlash on the axis's that cannot be eliminated with these machines short of putting ballscrews for which you simply cant retrofit off the shelf or even if so it wont be done cheaply, but software backlash compensation works miracles for their case. BTW, the minitech is essentially kinda 'sherline with ballscrews', but many many times the price.

    The thing with machining precision is that it is very much a function of the skill of the operator as much as the capabilities of the machine. There are lots of skills, machining strategies and techniques and other factors to consider, that a good machinist can use to improve machining precision. As for focusing entirely on the machine capabilities, other than high positioning accuracy and resolution, you also have to consider spindle runoff which for most spindles normally in the +/-0.02mm range. That is before you put your tool in. With the tool holder and the tool, tool length, the actual runoff at the cutting edge can easily be more than this. Like everyone says, a quality micromachining CNC mill runs in the hundreds of thousands of dollars. Also don't think you can plug a tool in and let it go. You absolutely have to measure to make sure the tool is spinning to achieve the rated or acceptable runoff. ALso, not all cutting tools are made the same. They may not have perfectly axially centered cutting edges which can add to the error if you just take for granted what the manufacturer says. If you say you want precision machining results, now you really have to measure everything and every single step.

    Other factors to consider is whether the general tolerance needs to be this high, and the size feature you intend to create. This dictates what size tools you need. With a smaller tool, there is longer runtime, and this affects device choice. If you need 50um cut-in features, you will need <50um end mill which is expensive, but moreimportantly if its everywhere the runtime takes forever. Also since you are into precision machining, you need a precision technique of measuring actual cutting tool diameter both statically and dynamically (which includes effect of runoff). You may need those expensive laser tool diameter gauges such as those offered by renishaw. A cheap alternative is to mill a hole in the intended material and then precision measure the hole diameter.

    If you intend to make a large flat surface by means of rotational milling, I dont think that you can easily achieve 5um surface roughness deviation easily, since the classic machining path marks have this level of surface roughness in general, and I know of little mills or their operators capable of leaving machine mark free surfaces using a mill. YOu definately need to grind those surfaces to precision. However depending on the geometry, maybe you cant even have access in there to grind them using conventional means.

    Maybe you should look into other ways of manufacturing. You dont have to do this with a CNC rotary mill. Maybe explore EDM, or even lithography techniques that are conventionally used for high precision and small minimum size feature parts. Developing rapidly is sterolithography techniques using lasers to control sintering and these parts have some amazing precision too, but again by the technology they are often pretty poor on surface roughness and perhaps suffer from issues of porosity which may be of detriment to your prototype.

    Ultimately I'd agree with others. You may want to redesign so that you can achieve your end goal given your financial resources to make it happen. You also may want to consult professionals in manufacturing to show them your actual design (under heavy NDA of course but also with a hefty payment for their consultation I'd imagine), e.g. manufacturing/industrial engineers (i know youre a mech but they are different), machinists, etc - so they can come up with a more concrete plan for you.



Page 1 of 2 12 LastLast

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
  •  


About CNCzone.com

    We are the largest and most active discussion forum from DIY CNC Machines to the Cad/Cam software to run them. The site is 100% free to join and use, so join today!

Follow us on

Facebook Dribbble RSS Feed