Hardinge Conquest T42: How to program a sub-spindle to main spindle transfer?

Machine: Hardinge Conquest T42 with Fanuc 18-T (rev B) control
Operator: New to this machine

I tried the manufacturer's example for transferring a part from the sub-spindle to the main spindle as shown in the Hardinge Conquest T42 Programmer's Manual on page 16-19. In this example, they use the G10 P0 command to modify the workshift for both Z (turret) and Y (subspindle), as follows:

G10 P0 Z-1.75
G10 P0 Y-1.75

This shifts the zero of the turret and subspindle to the same offset, which is the zero of the part. Later in the example, the sub-spindle is commanded to G0 Y-0.5n to get a 0.5" grip on the part. Note: The G10 P0 modifies the workshift file, not the G53 through G59 offsets.

After trying this however, the Z workshift offset is modified, but the Y workshift offset is NOT.

And get this... I called Hardinge and spoke with Bob Allington who himself tried this out on one of their machines and got the same result! So it's not my machine, after all.

So, if the manual is wrong, what is the best way to transfer a part from the sub-spindle to the main spindle?

Put another way, how do YOU do it? What have you established as the best practice for this operation on a T42? Do you use G54, G55 for main/sub? Do you program in absolute values? Can you provide an example of gcode that you actually use in production? I can easily program the sub-spindle to transfer, but I am interested to know the best practices for this operation and seek your advice.

My hardinge machines ignore G54-59 workshifts all together. We use the #500 variables in the format G10P0Z#500 thru #5xx. In practice this is the same as setting the G54 etc but it allows workshifts to be permanent in the program and it would then work with the transfer code in the book. The variable are set at the top of the program like so:

#500=-2.4(-2.44)
#501=-27.0(-26.982)
#502=0.
#503=12.(ZSAFEINDEXMAIN)
#504=0.(ZSAFEINDEXSUB)
#505=0.


and then called in the first line of each OP

N1(TURN)
G10P0Z#500
M98P1
M10S1000G97P1T0707
G0X3.25Z.1
G50S3500
G96S400
G1G99X3.05F.008 etc.

We also use a safe index position(M98P1 for the main) that can be modified by the #500 variable for each job.

I just pulled this off a thumb drive.
I don't know what the part was, but the thing sure was long!
But all the better to see how I was dooin' it.

Also note that I like to run the spindles at 100 to pull. I feel like I get more repeatable lengths when clamp/unclamp - when pulling with the sub especially. I don't like to go faster - especially if I have a barfeeder pushing on it too.

I just run the sub up to the part once it's blanked up and figger my Y axis locations at that point.

This is from a T51, but everything but the X value in the last line is likely the same on the 42.

For others trying to make heads/tails, Y is the sub spindle linear axis.


I don't know why WORD lower cased those letters?

N4t202(.156 groove)
m4s100
m56
y3.25m32
m57
g4x1.
M21
g4x1.
Y24.
M22
g4x1.
M56
g4x1.
Y18.
M57
g4x1.
M21
g4x1.
Y18.91
m22
x.6z.584
x.54
g1x.46z.544f3.s800m8
x.375
z.025f10.
G0z.544
g1x.28f2.
Z.064f10.
X.23z.039f3.
X.2f2.
X.2485z.064f3.
X.248z.544f6.
X.55
g0t0x15.z10.


--------------------

Think Snow Eh!
Ox

I see, so you don't even set the Y axis offset. You just rapid in to the absolute value, grab the part, then whip off back to Y24 (long bed.)

For whatever reason, I liked the idea of facing a part, then using that exact length set both the Z zero and Y zero to equal that part face offset, then rapid the Y to -.375 (the depth of my main step chuck.) Without using offsets, I can just move the sub to an absolute offset equal to the sum of the step-chucks + part length - depth of part and be done with it.

I imagined experienced folks would set G54 for the main spindle, and G55 for the sub and simply switch between them, then when a transfer is involved, one would modify the G54's Y-offset to be equal to the part face and rapid in.

In your example, Ox, you keep it simple and just use absolute offsets for the Y.

So...are you guys using Y for the sub spindle axis..parallel to the Z? My hardinge sub machine is an Elite II and they use E as the designation on that machine. Y would be an axis on the turret perpendicular to Z and X.

At least I think that's why I'm confused

I've been told that Hardinge later changed the Y designation to something else... possibly E or B.

Yes, when I speak of Y axis, I'm actually talking about the subspindle which runs parallel to Z.

The Hardinge Programmer's manual shows a workshift offset that sets the part face and Y spindle-nose at the same value:

(The part is 1.75" out from the main spindle nose. All offsets are added to the axis absolute to get a positive offset (to the right) from the main spindle nose.)

G10 P0 Z-1.75
G10 P0 Y-1.75

This second Y workshift tells the subspindle to use the same offset as the part face sticking out of the main spindle. Now you can tell the Y (subspindle) to rapid to Y-1.0 inches and the collet will now take a 1" bite on the part.

This is probably the same as on an Elite, but I'm spelling it out anyway for others not familiar with the T42.

Unfortunately, the T42 machine doesn't actually behave the way they've written it. The Z is offset correctly, but the Y offset is ignored, and so what I get is my subspindle trying to rapid to the left side of my main spindle (i.e. crash.) I have my rapid turned down to zero, my hand is over the feedrate knob, and I run everything in dry-run plus single-step mode so my machine doesn't do anything dumb.

The above workshift solution is small, efficient, and elegant... but it's wrong. The Y offset doesn't actually work.

To make matters more interesting, I use step-chucks, not regular collets. With a collet, it doesn't matter if you're a few thousandths over, but with a step chuck it does. Too many thousands the wrong way and you bump up against the shoulder of the collet which may crush the part, warp your delicately machined emergency step chuck, or pop a torque-limiter (not sure if torque limiter are unique to Hardinge.)

So now I must measure the distance from the spindle nose to the step-chuck shoulder, then measure the face of the part in the sub-spindle to the spindle nose, and subtract that amount from the absolute value of the sub-spindle axis. It's not a show-stopper, but the aforementioned offset solution is much cleaner and convenient.

If this were an elevator, it'd be like answering a skill testing question (and using calipers) every time you want to get to the ground floor. I'd rather just press G.

Ox said:

This is from a T51, but everything but the X value in the last line is likely the same on the 42.

For others trying to make heads/tails, Y is the sub spindle linear axis.

Click to expand...

Not long bed - but rather a 51.

As stated above, Y is linear sub.
Hardinge went to E for that axis when they came out with the Quest line I guess.


---------------------

Think Snow Eh!
Ox