[toyranosaur]

Ya, well, normally I'm pretty good at figuring out how to hold the work, but I'm having a prob now. I'ma new machinist with a Haas VF-3 w/probe and a manual lathe, and am making some small (Under 2") R/C wheels from 6061-T6. They are 2 pieces each: The front screws to the back and Viola-This is necessary because the tyres have a beadlock insert. So, I machine the back sides of the halves-Basically a 2" diameter stepped down halfway to a 1.5" diameter. I then cut some soft Al. jaws on my Anglelock vise and put the smaller diameter in it, and machine the faces. However, no matter what I do, they come out a tad off, about 12 thou: Now when you screw 'em together, the spokes do not line up perfectly, and not only the visual suffers but they'll wobble for sure, killing saleability. (I use the same X & Y zero from the soft jaws for the wheel program, so it should be dead center?) I thought it was jaw pressure moving the movable jaw in once I started gutting the solid wheel half, so just tightened it up as little as possible, to no avail. I've had this prob before & will again, so do any of you more experienced chaps have any thoughts? I have about 50 hours into CAD & prototyping, and have buyers waiting. As a 1-man shop with a Haas to pay for, farming this out is not an option. I know they should be lathe parts, but even if that were possible, my lack of concentricity milling would give me the same prob anyway! Please assist if you have an idea! I can post pics if this would be clearer. Many Thanks, Chris

[christinandavid]

First thing you could do is machine 'V's into the jaws rather than a diameter.

If you are not holding a tight enough tolerance for the location diameter using the lathe you may want to try to mill it instead.

Edit - If you abandon the lathe, rip off the chuck and use it to hold the jobs on the mill.

DP

[kendo]

use the probe to locate the center for the 2nd op

[toyranosaur]

Christinandavid: I am milling the back side diameters now, and, while my jaws are not V'd, I have cutouts at 3,6,9, and 12 o'clock to eliminate straight pressure in X and Y: Thinking is, that it spreads the force out over more area than the 4 concentrated points the V would impose, and increasing surface area/friction so I don't have to clamp the parts so hard. The 3-jaw chuck is on the list, not a bad idea, Thanks! I just wonder if my Korean chuck will be accurate enough.....

Kendo, I considered that one, and it's dang-near foolproof: Only thing there is the added time. If I do 75 sets of wheels at 8 parts per set, and 4 of them now add that time, Yikes! 300 probe cycles! It may turn out to be my best option with what I have however, and perhaps it's wiser just to do it rather than waste my time trying to come up with the perfect holding method. Still open to ideas (Hoping someone comes up with an option that costs nothing and makes perfect parts easily :wee: ) but it's starting to look like it's probe time. Chris

[toyranosaur]

The more I think about it, the more I'm sure that I'll only have to run the probe cycle once: I imagine that, provided I tighten the vise the same amount each time, the parts will be consistently in the same place, obviating the need to run it each time. Should be within a thou each time, no? Sure hope I'm right, it seems to make sense! I'll let you know if it works out that way, Thanks all, Chris

[christinandavid]

If your parts have a through hole, and if you are able to modify the part design slightly, an expanding fixture could be an option. Remember doing some small aluminium parts using this method suggested by a work colleague. It works quite well - a bit like an inverted collet. You could also then make a fixture that holds multiple parts.

DP

[kendo]

Also, keep it mind...
You can cut the cycle time down for the probing routine if you are able to write custom macros for the probe. If you're not currently able to, I'd suggest looking into it, once you learn it, you'll find tons of ways to benefit from it.

Think about it, when you watch the "probe boss" macro cycle that comes with the renishaw package, it starts at the center and probes each of four quadrants, returning to the center each time. You can write a custom cycle that starts at one quadrant and rapids right to the next without ever starting out at the center. You can also make it a one-touch probing as long as your parts are consistent enough.

I cut the cycle time down by more than half with techniques such as this on some weldments that were being probed 30+ times each cycle.

[Caprirs]

I used to run a lot of similar parts on my mill. Even after cutting the soft jaws, I would still have to shift the work offset .003"-.006".

A trick for consistency is to rough the part out leaving ~.01" on all surfaces, then open the vise and re-tighten using a torque wrench. The accuracy of the torque wrench does not matter, only its repeatability so it can be an el cheapo. Obviously, this adds more operator time, but you can get very repeatable results which makes it much easier to dial in. The torque wrench can be set for the lowest setting (~10 ft-lbs) for the finish ops.

[toyranosaur]

Some brilliant stuff here, indeed! I made a test ring and ran a few probe cycles, and dang if it didn't come up over .006 off in Y. I tried it several times trying to close it the same amount each time, and I believe that (Especially with that great torque-wrench idea!) I should not have to run a probe cycle each time. I do need to learn to do macros for it, so I can get a lot more use out of it: You're right, I shall make it a point very soon. I also like the expanding-fixture idea: Thought it won't work for this particular part, I can see beaucoup uses for that approach. Very much thanks for the help and ideas people, your thoughts definitely helped me out on this and even gave me some excellent ideas for future projects! Cheers, Chris

[TheBigJW]

What about a three jaw air chuck with pie jaws.

John