Announcement

Collapse
No announcement yet.

My personal findings on going to CNC from manual

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • My personal findings on going to CNC from manual

    I've always manually machined by "feel", never paid attention to the numbers, as a result, I am finding my manual machining experience to be almost completely useless in the CNC world. There is no feel, just an odd sound before catastrophe. The capabilities are amazing, but the learning curve is literally paid in the cost of broken endmills. I've broken probably about 10 1/8 flat endmills, 1 1/8th ball endmill, and 1 1/4" 3 flute endmill.
    Good news is that I am getting much better results now. Still need to play with chamfering for a smooth shiny surface, and ramping down movements.

    On another note, I have a need for a test bar with exact known dimensions, within .0001 for calibrating my CNC mill. Something that is 9" long by 3" wide. Stacking my 123 blocks is what I've been doing, suppose they are accurate enough? I mic the individual blocks and they are accurate enough. Stack them and use my digital calipers or height gauge shows .003 over. So I am having trust issues with my metrology. The micrometer is a 2-3" Mitutoyo that I acquired used which I am assuming is accurate. It is to the standard it came with.

  • #2
    [QUOTE=RB211;n1901463]I've always manually machined by "feel", never paid attention to the numbers, as a result, I am finding my manual machining experience to be almost completely useless in the CNC world. There is no feel, just an odd sound before catastrophe. The capabilities are amazing, but the learning curve is literally paid in the cost of broken endmills. I've broken probably about 10 1/8 flat endmills, 1 1/8th ball endmill, and 1 1/4" 3 flute endmill./quote]

    There is a cost to education. School, study, experience, time, broken tools, etc. Its why somebody who can do things has value.

    Good news is that I am getting much better results now. Still need to play with chamfering for a smooth shiny surface, and ramping down movements.
    Because you are paying for your education and its working.

    On another note, I have a need for a test bar with exact known dimensions, within .0001 for calibrating my CNC mill. Something that is 9" long by 3" wide. Stacking my 123 blocks is what I've been doing, suppose they are accurate enough? I mic the individual blocks and they are accurate enough. Stack them and use my digital calipers or height gauge shows .003 over. So I am having trust issues with my metrology. The micrometer is a 2-3" Mitutoyo that I acquired used which I am assuming is accurate. It is to the standard it came with.
    There are a number of reasons your stackups could be off. 123 blocks don't usually have the ability to ring together like gage blocks, but putting them together like they do could reduce some error. .003 seems like a lot. The first thing I would do I think is check them on a surface plate. Look for slight swells or ridges around the holes, or along the chamfers. Maybe stack them up in different ways and see if the error or location of the error changes. Consistency to .0001 in every metric may be going down a rabbit hole. I'm sure you can find those who will say "well my...", but that is a hard spec to meet. Just holding a block in your hand for a while can make that much difference from one measurement to the next.

    I only have a couple granite tools, but I've found them to be pretty good They are not 100% immune to the issues of other metrology tools, but they are pretty good and less susceptible to many. That may be something to consider. Alternatively look for a 246 block or some other precision block not subject to error stacking.

    Another thing to consider is that your table may not be perfect. Some years ago I had a conversation with John Herzog (maker of Taig) about making some long tables for Taig mills, and he mentioned that there is some growth during anodizing. More near the ends. I don't know what you are chasing, but its something to think about.
    *** I always wanted a welding stinger that looked like the north end of a south bound chicken. Often my welds look like somebody pointed the wrong end of a chicken at the joint and squeezed until something came out. Might as well look the part.

    Comment


    • #3
      It sounds like you should make a test bar very slightly oversize and take it to a good machine shop with a drawing and get them to grind it to size. Then make a nice box for your master cnc standard.

      Comment


      • #4
        You have too many variables.
        If you want accurate lengths, you need Micrometer Standards or other professional standards ( mine were used on a Jig Borer)

        https://www.ebay.com/itm/MITUTOYO-NO...UAAOSwqphfbSB6

        They are available
        The problem with your 1,2,3 blocks is flatness and parallelism .
        In "stack up" all errors are amplified
        Rich
        Green Bay, WI

        Comment


        • #5
          Originally posted by RB211 View Post
          I've always manually machined by "feel", never paid attention to the numbers, as a result, I am finding my manual machining experience to be almost completely useless in the CNC world. There is no feel, just an odd sound before catastrophe. The capabilities are amazing, but the learning curve is literally paid in the cost of broken endmills. I've broken probably about 10 1/8 flat endmills, 1 1/8th ball endmill, and 1 1/4" 3 flute endmill.
          Does the machine have feed override controls?

          If so begin with slow feeds and override them upwards while it is running in a cut, when you achieve a good feed rate jot it down then change the program for the next part.
          Every CNC lathe or mill that I have ever used had simple up/down buttons for feed override. This is a Bridgeport 2 axis lathe control, in the lower left corner of the screen are up and down arrow buttons marked FD OVR, these override the programmed feed rates on the fly.

          As a side note the reason that I took the picture is, one Monday morning I turned it on and these were the axis positions displayed, this machine has 15" of X travel and 36" of Z travel.
          X -27" Z -63" would place the tool roughly 25" past the left end of the spindle and 12" past the back of the enclosure.
          In other words "in thin air" (-:
          Last edited by Bented; 09-25-2020, 06:14 PM.

          Comment


          • #6
            It does have feed overrides, was using them.
            246 blocks sound like a good idea.
            So more evidence to the stackup being over .003, the last part I made was off by .003... Instead of 9", I could just do it with a single 123 block, which the Z and Y were done with.

            Comment


            • #7
              Second question, are you using flood coolant? This is an important consideration especially when milling aluminum.

              Comment


              • #8
                Originally posted by Bented View Post
                X -27" Z -63" would place the tool roughly 25" past the left end of the spindle and 12" past the back of the enclosure.
                In other words "in thin air" (-:
                That poor machine has probably been in the same spot for years. It was asking you to move it so it could get a better view through the window.

                Comment


                • #9
                  Only broke 10 small endmills (1/8") ? Your doing good ! I broke so many in the early learning days ! Still today, those pesky small endmills are VERY unforgiving ! Recutting chips will break them, runout will break them, entering a corner fast will break them, you name it ! I also found that the various feed/speed charts for small carbide endmills vary all over the place ! On 1/8 and smaller endmills, the chipload is pretty small, typically under .001 for them to survive.

                  What sort of chipload were you running when you broke them? What do you use to get the chips out of the way? What about coolant of some form so the chips don't weld to the endmill?

                  Comment


                  • #10
                    The best view through a window is from the outside lookin in (-:

                    Comment


                    • #11
                      http://www.pmtnow.com/nano
                      Try one of these endmills, .0002" dia.
                      Beaver County Alberta Canada

                      Comment


                      • #12
                        I am using mist coolant

                        Comment


                        • #13
                          Originally posted by Bented View Post
                          The best view through a window is from the outside lookin in (-:
                          Not in Aviation!

                          Comment


                          • #14
                            Ok, I cheated, I changed my revs per inch based on version 6 on the X. Version 7 came out perfect!
                            Click image for larger version

Name:	20200925_214045.jpg
Views:	244
Size:	1.05 MB
ID:	1901569

                            Comment


                            • #15
                              What software are you using to do your spindle rpm, chip load, DOC?
                              Not a CNC programmer or operator (still in class at Haas U.), but does your control handle corners properly?
                              Read that some control software approaches corner cuts differently, fewer stress on the cutter.

                              Comment

                              Working...
                              X