Thread: bolt circle size problems

im haveing trouble getting some bolt circles to come out the correct diameter. i have a round part with 3 bolt circle
patterns in the in the face. one is 4, #8-32 on a 3.531 bolt circle. one is 6, #10-32 on a 3.531 bolt circle and the
start angle is 15deg. the third pattern is 4, .328 thru hole on a 3.5 bolt circle and the start angle is 30deg.
the parts are being held in a three jawed chuck bolted to the mill table. each part is indicated to center. the holes are
equidistant from center, the diameter is just small. on the 10-32 and the 8-32 they are around .020 small and on
the .328" they are about .012 small.
the machine is a 3016 fadal with an 88HS controll, and the parts were programed in smartcam.
i went back and calculated all the hole locations by hand and they match what is in the program. i am somewhat
familure with writing G-code by hand and with a G-code chart i can write simple programs, i went through this program
line by line and as near as i can tell it is right.the start angles and angles between holes is right and the holes are
perpendicular to the face of the part with in spec.
the guy who wrote the program and the guy who runs the machine assure me that the program is correct and is not
the problem. i checked the gibs, they are good. i did a quick check of back lash on each axis and the worst i got
was .0013" on the x axis. i did a quick check of repeatability and got .0003- .0005 every time on each axis. im stumped.
what else can i check. could there be something in the program that is not obvious or a paramater or something in the
controll that can cause this. i dont know what to do. im getting ready to have these parts recut these parts and
try again, is there any thing i should try or look out for this next time around. thanks for you help.

How are you measuring the parts to say they are wrong. Even measuring tapped holes on a cmm can be tricky without the right hole plugs.

A few thoughts while I'm trying not to work. Your problems seem like a lot more fun than topping up coolant.

How did you check the backlash? I've seen motor couplings that would slip at high feed or rapid, but give really nice readings
when moving slow. Maybe try running a test part at 25% rapids.

Another thought. Your holding method. Is there something loose in there where the part is wobbling? Is the part
well supported underneath, is it flexing somehow (I don't know what the part looks like).

Have you left the part on the fixture and picked the holes up in place to see if they are where they are supposed to be?

If its a machine problem, a little faky move before each hole to make sure that you are coming at each from the same direction
never hurts.

thanks for your responses. jelrod1, i dont know how they are being measured, they were kicked out by QC, but i
checked them with calipers and i get them within .001 or so of what QC said they were. he may be using calipers as
well, i dont know, i will have to find out. they are in house parts so if that is the measuring standard that they deemed
acceptable then that will be what determines if they are right.
bobw, i checked the backlash by putting an indicator in the spindle and clamping a knee to the table, i then run the
indicator up to the knee in one direction and set a 0 and then clicked back .0001 at a time until the needle moved
.0001. i did this in both directions (+,-) on each axis, and as i said the worst i got was .0013. then i wrote some simple
programs that would start with the indicator set at 0, and then rapid away from the knee about 12" then one program
would feed back at .010ipm to 0, one would feed back at .025ipm to 0, one would feed back at .100ipm to 0. then i had
one that would rapid away and rapid back to 1" from zero and feed to zero at the aforementioned feeds (.01, .025, .100)
i also did these checks in both directions (+,-) on the x and y axis, and every one was .0003-.0005.
about the holding method, the part is sitting right down on the table and the 3 jawed chuck is holding it radially and
keeping it from spinning. as far as the part flexing??? i hadnt thought about that, the part we are drilling into is kind of a
flange that is slightly bigger than the main body of the part but it is about .625 thick so i didnt think it would flex so i
never considered that, i will have a look. i dont think i will be able to get these things set back up and orientated
accurately enough to check them in the machine, but when we run them this next time i will be there to watch and we will
deffinately do that if the first part shows a similar problem. if we have to lie to the machine and/or approach each hole from
the same direction or any other tricks to get these things done we will do it then i guess we will have to have the service
guys come try to figure it out. we have run several other parts since then and we have not had problems so i dont know
what could be wrong.

Watch the program run and see if the machine is going where it is told to