Design Concepts - Page 4

Page 4 of 9 FirstFirst 1234567 ... LastLast
Results 61 to 80 of 175

Thread: Design Concepts

  1. #61
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6252
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Hi Craig - Yes it maybe $2000NZ but the alternatives cost much more than this!! I used mechsoft and for 4-5 axis they want $12000AUD (plus $1500/year maintenance) but in F360 its the $2000AUD extension. Plus you can buy a months worth or a weeks worth if you only use it on and off. I like the ability to go to a client and start it up on their machine and have everything there. Other systems I had to check it in and check it out and if there was a glitch in this it could hang up for days until someone in USA fixed the issue. I'm sure there will be price creep though... Peter



  2. #62
    Member
    Join Date
    Nov 2013
    Posts
    4282
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Hi Peteeng,

    Yes it maybe $2000NZ but the alternatives cost much more than this!!
    Yes I did my homework, and yes Fusion's Machining Extensions are cost competitive relative to other solutions.

    Plus you can buy a months worth or a weeks worth if you only use it on and off.
    I've looked at that, but you pay quite a premium for short term use, but I may not have much of a choice. I cant really afford the full subscription solely for hobby use.
    If I can use it to effect for the business then that would probably swing me from 'occasional use in short term hires vs year long subscription'

    You will no doubt have noticed that there are truckloads of videos on YouTube about Fusion360, but it seems they are all about entry level stuff.....the instructional stuff seem to run out when you get
    to the higher end functionality.

    Craig



  3. #63
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6252
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Autodesk is having a F360 trade show and tech muster next month in Melbourne. I was thinking of going but its a long way for me. I have also done a 4-week webinar series on F360. They seem to be putting a lot of info and time into F360 to develop the user base. Since you are a subscriber contact the F360 autodesk people they are very helpful I've found if you have a particular issue or seeking extra info. BTW basic Simsolid is $11500AUD per year that's up there!! I wanted to do some buckling analysis so needed the next level up and they wanted $50,000 per year for the advanced module. So I had to do that old school... Peter



  4. #64
    Member
    Join Date
    Nov 2013
    Posts
    4282
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Hi Peteeng,

    BTW basic Simsolid is $11500AUD per year that's up there!! I wanted to do some buckling analysis so needed the next level up and they wanted $50,000 per year for the advanced module. So I had to do that old school... Peter
    Holy Snapping Duck S****t, that's getting up there alright. I have used some RF software that is that sort of money, I had access to the universities license and so did not concern me. Once I got
    out in the real world then I enquired about the price,and near died.

    Craig



  5. #65

    Default Re: Design Concepts

    I've been pondering about this for some time now, now that c3 ballscrew bacame a possibility for diy budgets, seems like a good opportunity to tackle cooling of components. Rail cooling can be done by many ways, full or half channels milled on the underside of the inserts, holes inside the inserts(in case they are segmented like mine are) piping in the cast under or next to inserts(inefficient due to low thermal conductivity of uhpc/eg), milled channels right under the rails etc. and all quite cheaply might I add. Rexroth even has channels milled into the rail (as attached), Grob machines use this.



    Rails in this sense are easy, ballscrews remain the problem, active ballscrew cooling is a $$$ no no, so I was wondering and realized we might be able to tackle both components at the same time...cooling the lubrication oil, if we were to cool the oil (cooling grease probably not an option, would be too thick to move) but oil won't necessarily get too thick to move it. As I see it, this type of cooling might be too inefficient given the small amount of oil that gets dispensed and it would probably have to be a continuous stream during machining. Haven't been able to find any papers about this, I'm really bad at using correct keywords.

    What do you all think? Cool the oil, yes? no? maybe?



  6. #66
    Member
    Join Date
    Nov 2013
    Posts
    4282
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Hi,
    what sort of machine are you talking about building??? Cooled rails, cooled ballscrews, well yes I can imagine you might want that for micron or sub micron grade equipment
    or semiconductor processing equipment....but for a hobby machine?? If you are going to the trouble of cooled components then you will spend hundreds of thousands on the rest
    of the machine just to take advantage of it. Is that what you are doing?

    Craig



  7. #67

    Default Re: Design Concepts

    Quote Originally Posted by joeavaerage View Post
    Hi,
    what sort of machine are you talking about building??? Cooled rails, cooled ballscrews, well yes I can imagine you might want that for micron or sub micron grade equipment
    or semiconductor processing equipment....but for a hobby machine?? If you are going to the trouble of cooled components then you will spend hundreds of thousands on the rest
    of the machine just to take advantage of it. Is that what you are doing?

    Craig
    a C3 machine. Think about it, a pipe and a pump, or if cooling the lubrication oil idea, a liquid chiller, all very cheap, not much trouble at all, Granted the room I'll be keeping the machine in is a stable 25C year round, the rails will definitely expand from heat, so will the ballscrew, my reasoning is, implementing this is cost effective and doesn't require much effort, so why not do it?



  8. #68
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6252
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Hi Ard - Every component has its own thermal inertia and will heat up and cool at different rates. Unless you individually cool each component (like the big boys do) you won't have a chance of establishing overall thermal equilibrium. I think you make an insulated enclosure and air condition it. Place TCs around the large masses and areas of interest and map the temps over time to gain an understanding of the temp changes and lags. It takes something like 1 hour per 25mm thick of material to establish temp so you will need to run AC overnight to get everything pulled down to set temp. What accuracy are you aiming at? Every tool room I've been in is air cond as well. Also its change of temp (or a delta temp in the same part) that's the issue not temp itself Peter.

    Hi Ard - its not just about "cooling" its about establishing thermal equilibrium. Just running some "cool" oil thru a gallery may not solve the temp delta issue its a very complex situation. I've done manual heat transfer stuff through heat exchangers and that's a mind bender...This is easy to say but very difficult to understand and implement on a machine. Environmental temp control is the first level of this endeavor.



  9. #69

    Default Re: Design Concepts

    Quote Originally Posted by peteeng View Post
    Hi Ard - Every component has its own thermal inertia and will heat up and cool at different rates. Unless you individually cool each component (like the big boys do) you won't have a chance of establishing overall thermal equilibrium. I think you make an insulated enclosure and air condition it. Place TCs around the large masses and areas of interest and map the temps over time to gain an understanding of the temp changes and lags. It takes something like 1 hour per 25mm thick of material to establish temp so you will need to run AC overnight to get everything pulled down to set temp. What accuracy are you aiming at? Every tool room I've been in is air cond as well. Also its change of temp (or a delta temp in the same part) that's the issue not temp itself Peter.

    Hi Ard - its not just about "cooling" its about establishing thermal equilibrium. Just running some "cool" oil thru a gallery may not solve the temp delta issue its a very complex situation. I've done manual heat transfer stuff through heat exchangers and that's a mind bender...This is easy to say but very difficult to understand and implement on a machine. Environmental temp control is the first level of this endeavor.
    Thanks as always Pete, lots of good insight, future machine room is already AC'ed at constant 25C +- 2C, there is a lag about 30min at peak times (hot summer/ cold winter), the room has 2 vent tubes, 1 is a passive heat exchanger for keeping the air fresh with a bunch of filters and a ceramic heat exchanger and on the other side is a regular ac to heat/cool down. Passive is good at keeping whatever temp is inside but the ac has to create that first if theres a big gap between inside/outside, hence the lag. insulating the enclosure might flatten the delta somewhat, I can see that.

    Every component will heat up and cool at different rates

    I can meter how much oil gets dispensed to each component but thinking about it now, by cooling the oil, the first part that gets exposed to it are the ball bearings/rollers, which might actually introduce more error and backlash to the system than the other way around, since those will shrink, I'd have to wait until they reach a certain temp to start dispensing cooled oil. On the other hand I might not be cooling the rails/ballscrew at all and increasing the delta between rail-carriage, ballnut-ballscrew...

    What accuracy are you aiming at?

    just hoping to take full advantage of C3 ballscrews and roller rails,



  10. #70
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6252
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Hi Ard - C3 is +/- 8um /300mm and 2PI (one turn) tolerance of +/- 6um. Say that 300mm screw changes 1deg in temp (or bed changes +0.5deg and ball changes -0.5deg). It changes 3.6um does that affect what you are trying to do? To have some sort of control your going to have to aim at 0.5deg control or 0.25deg? say 2/8=25% of your geometric tolerance. Its a mission.... Peter

    Attached Thumbnails Attached Thumbnails Design Concepts-c3-jpg   Design Concepts-c3-jpg  


  11. #71
    Member
    Join Date
    Nov 2013
    Posts
    4282
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Hi,
    I also think temperature control is a great way of introducing great complexity with very little gain.

    If you want to improve the performance of your machine use linear glass scales, and good ones, with a servo that uses load sensing, for example a Delta A2 series servo.
    It has a regular encoder, 1,280,000 count per rev, and the drive has a secondary encoder input for 'load sensing' from glass scales for instance and closes the position loop on this
    input.

    Craig



  12. #72
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6252
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    I agree with Craig - If you want to put effort into improving accuracy then scale feedback is a better option. Peter



  13. #73
    Member
    Join Date
    Nov 2013
    Posts
    4282
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Hi,
    Or better still use a laser interferometer, and if you're really keen get an iodine stabilised laser then you could get +- 5nm.....that's not micrometer, but nanometer!. That's how semiconductor equipment
    achieves the accuracy demanded. You'd probably get an iodine stabilised helium/neon laser for $25k, second hand of course.......but they have come down in price. At this rate a century after my death I'll
    be able to afford one!

    Craig



  14. #74
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6252
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Yes Craig - Thats how they correct really big mills these days (by big I mean you could machine your houses gutters with them) they have laser sensors on the spindle and correct in real time, amazing accuracy for 20m long machine but eyewatering cost. Peter



  15. #75
    Member
    Join Date
    Nov 2013
    Posts
    4282
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Hi peteeng,
    I think we are at crossed purposes; yes big machines use lasers for alignment, and that of course relies on a beam of light travelling in a straight line.

    Laser interferometry is a different ballgame altogether. There one uses a coherent beam of light, that is to say EXACTLY one frequency, and then you can measure the difference
    in path length between a reference beam and a measurement beam, both derived from the same coherent source. Using an interferometric detector you can measure easily down to 1/4 wavelength
    of the light beam. If using a helium/neon laser that is 1/4 of 632.8nm or 158nm. If you can improve the 'purity' of the light beam, its called 'coherence length' then you can do better again,
    hence the use of iodine cells to stabilize the laser.

    At University I was able to use a Helium/Neon laser made by HP that used Zeeman Splitting to stabilize the beam, very VERY clever. It was only 100mW output and had a coherence length
    of 7m and was useful for measurement up to about 2m and +-80nm resolution. Note this little laser had a mildly diverging beam, across the room the beam spot was a cm or so.......its not
    collimation, ie how tight the beam is, but how coherent it is, ie spectral purity.

    An iodine stabilized helium/neon laser has a coherence length measured in km's, and that's why with a low noise interferometric detector you can measure down to 5nm.

    Craig



  16. #76
    Member
    Join Date
    Dec 2016
    Location
    Norway
    Posts
    133
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Could a simple solution to your temparture stabilizer be to cool the ball nut mounting block and the end bearing blocks on the screw (maybe one of them is enough) ? Those are stationary parts you can drill some channels and feed with cooling liquid. Same liquid you feed underneath the rails. If you cool the rails - is it necessary to cool the bearing blocks as well? If yes, maybe it is sufficient to have a channel in the mounting plate where the bearing blocks are fastened.

    I have no data to support this, but what I have been reading is that if you have flood cooling then there are 2 things that can have an impact and might be easy to implement: 1) cool the cooling fluid so it has a constant temperature and 2) keep the flood coolant away from structural parts with shields/covers so the machine frame is not affected by temperature change from the coolant.

    And another thing. You allready know this, but buy a spindle (at least if it's a built in spindle motor) that is rated 2x the power you are going to feed it. This way the motor will produce less heat and there will be more mass = lower temperature changes.



  17. #77

    Post Re: Design Concepts

    Quote Originally Posted by Buoyen View Post
    Could a simple solution to your temparture stabilizer be to cool the ball nut mounting block and the end bearing blocks on the screw (maybe one of them is enough) ? Those are stationary parts you can drill some channels and feed with cooling liquid. Same liquid you feed underneath the rails. If you cool the rails - is it necessary to cool the bearing blocks as well? If yes, maybe it is sufficient to have a channel in the mounting plate where the bearing blocks are fastened.

    I have no data to support this, but what I have been reading is that if you have flood cooling then there are 2 things that can have an impact and might be easy to implement: 1) cool the cooling fluid so it has a constant temperature and 2) keep the flood coolant away from structural parts with shields/covers so the machine frame is not affected by temperature change from the coolant.

    And another thing. You allready know this, but buy a spindle (at least if it's a built in spindle motor) that is rated 2x the power you are going to feed it. This way the motor will produce less heat and there will be more mass = lower temperature changes.
    this paper says those are very good ideas! https://sci-hub.hkvisa.net/10.1007/s12541-013-0312-7
    a good alternative and finally something that's been tested that isn't a hollow ballscrew, good one!



  18. #78

    Default Re: Design Concepts

    Quote Originally Posted by joeavaerage View Post
    Hi,
    I also think temperature control is a great way of introducing great complexity with very little gain.

    If you want to improve the performance of your machine use linear glass scales, and good ones, with a servo that uses load sensing, for example a Delta A2 series servo.
    It has a regular encoder, 1,280,000 count per rev, and the drive has a secondary encoder input for 'load sensing' from glass scales for instance and closes the position loop on this
    input.

    Craig
    I will be using 1um scales, is load sensing a general term or it depends on the company? I have to make sure the drives I get have secondary encoder input, thanks



  19. #79
    Member
    Join Date
    Dec 2016
    Location
    Norway
    Posts
    133
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Cool. Didn't read it all, but it seems there is a significant difference from an uncompensated system.



  20. #80
    Member
    Join Date
    Nov 2013
    Posts
    4282
    Downloads
    0
    Uploads
    0

    Default Re: Design Concepts

    Hi,

    I will be using 1um scales, is load sensing a general term or it depends on the company? I have to make sure the drives I get have secondary encoder input, thanks
    Load sensing is a general term. Different manufacturers have their own terminology that tends to create, by design, the notion that they and they alone are doing this. Leaving the market hype
    aside any sensor like linear scales, but equally could be an LVDT (often used for optical stages) or laser interferometry that senses the actual position of an axis and is used to close the position loop
    is 'load sensing'.

    All the top tier servo manufacturers have models that do this. I mention the Delta A2 series as an example purely because of my familiarity with Delta's product line, but feel sure that Yaskawa and all the rest
    have models that do the same thing.

    I use Delta B2 series servos, which are in effect the entry level servo with a 160,000 count per rev encoder but DOES NOT have load sensing. A 750W Delta B2 servo kit (servo, drive and cables) cost $438USD from
    my favored Chinese supplier whereas a 750W Delta A2 kit costs $636 from the same supplier. They command a premium.....but then so do the linear scales to match them! In my case I find the the more humble B2 series
    servos are more than adequate for my machine.....linear scales would be nice but I don't need to chase that last micron....my machine is more accurate than I am!

    https://www.fasttobuy.com/delta-750w...ew_p27770.html

    To be honest I think good linear scales and load sensing servos are a far better use of your resources than temperature compensation.

    Craig



Page 4 of 9 FirstFirst 1234567 ... 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 for manufacturing industry. The site is 100% free to join and use, so join today!

Follow us on


Our Brands

Design Concepts

Design Concepts