[iMisspell]

Quick backround... been running, setting-up and milded programing for cnc lathes for around 8-9 years in a production shop. A week ago i was tossed on a TOYODA FA550-II Horizontal Machining Center. Aside from simple stuff on a Bridgeport and Milling slots and bolt-hole pattens on a HITACHI-SEIKI HI-CELL 23 i have just about no Mill experience.

So my question is
How would you go about writing a "ramp on" path to face a 7.5in flange with a 3.5in bore through it.

Heres how they have it set up now but its creating a crap finish.
The finish only has to fall between 93-125 but can not have any draw or ramping marks on it.

Z.0 = finish depth.
X.0 = center line of flange
Y6. = clear rapid point above part.
Using a 4" face mill (EDIT: miss-labled as fly cutter before, sorry) with six inserts on Gray Cast Iron.

Orginal code...

G0 Y5.5 X-3.25 S600
Z.25
G1 Z.0 F25.5
Y.0 F30. (feed onto part)
G3 X-3.25 Y0. I3.25 J0. F30. (rough)
G3 X-3.25 Y0. I3.25 J0. F30. S1000 (spring pass)
G0 Z.75
G53 ....

Heres how i butched it to get a pretty good finish.

G0 Y5.5 X-3.25 S600
Z.25
G1 Z.005 F15.
Y.0 F30. (dont like this, want to ramp on with a G3)
G3 X-3.25 Y0. I3.25 J0. (rough)
 G0 X-5.75 Z.032 (dont like all this eather, want to ramp off)
 G1 Z0. S1750 F47.5
 X-3.25
G3 X-3.25 Y0. I3.25 J0. (finish)
 G1 X-3. Y-1.5 Z.001 (dont like this, want to ramp off cleanly)
G0 Z.75
G53 ....

I was trying to ramp on the part but keept gettin an alarm abuot radios is wrong, think it said it was too big or somthing...

G0 X.0 Y6. Z.25 S600 (rapid above and to center line of flange)
G1 Z.005
G3 X-3.25 Y.0 I.0 J-6. F30. (Ramp on but get an ALARM)
G3 X-3.25 Y0. I3.25 J0. (rough)

I dont have a full understanding of the J, I and K values for a G3 right now which seams to be my problem.

Think i see when your making a full circle X goes with I and Y goes with J and ill guess Z goes with K but im lost with partial circle.

Guess my bottom line question is how do you make a partial circle with a G3 to ramp on and off a part ?

Thanks for any help or advice.
- joe

_

[ajl6549]

Try this :

G0 X6. Y0
Z0
G3 X0 Y2. I-4.4923 J-3.4768 F?? (RAMP ON)
G3 J-2. (RGH)
G3 J-2. (SPRING)
G3 X-6. Y0 I-1.5077 J-5.4768 (RAMP OFF)
G0 Z1.


Also you may be able substitute R5.6805 for the I and J in the ramps if your machine has the radius opption.


A.J.L.

[Geof]

This will feed you in at Z0.05, do a half circle down to Z0.0, do a full circle at Z0.0 then lift up to Z0.05 over another half circle.

It is possible your alarms where because you did not change the sing on the I value. The first half circle uses I3.25 because the center is to the right (+) of the start point; the full circle and the last half circle use I-3.25 because for these the center is to the left (-) of the start point.

I tested this little program on my Haas and it does run.

edit: This gives you the same number of full circles, two, as your original programe so it should run in the same time but give you a better finish.

%
O00000
T1 M06
M03 S600
G54 G00 Y5.5 X-3.25 Z0.25
G01 Z0.05 F25.5
Y0. F30. (FEED ONTO PART)
G03 X3.25 Y0. I3.25 J0. Z0. F30. (RAMP DOWN TO Z0.0 IN HALF A CIRCLE)
G03 X3.25 Y0. I-3.25 J0. F30. (FULL CIRCLE AT Z0.0)
G03 X-3.25 Y0. I-3.25 J0. Z0.05 F30. (RAMP UP TO Z0.05 IN HALF A CIRCLE)
G00 Z0.75
M30
%

[ajl6549]

This will feed you in at Z0.05, do a half circle down to Z0.0, do a full circle at Z0.0 then lift up to Z0.05 over another half circle.

Geof,
It may not be safe to assume an FA-550 toyoda has the ability to helical interpolate. So use caution. It'll either alarm out, or position Z, then do the circles.

[iMisspell]

Thanks guys, im at home now (work 2nd shift) but will try both methods once i get to work and then post back the results once i get back home (no net access at work).

Have a few questions now.
With A.J.L's code, how where the I and J coordinates achived on line...
G3 X0 Y2. I-4.4923 J-3.4768 F?? (RAMP ON)
A.J.L. looks like you coded that by hand where Geof looks like a CAM.

And with this said...

It is possible your alarms where because you did not change the sing on the I value. The first half circle uses I3.25 because the center is to the right (+) of the start point; the full circle and the last half circle use I-3.25 because for these the center is to the left (-) of the start point.

The below should work ?
G0 X.0 Y6. Z.25
G1 Z.01
G3 X-3.25 Y.0 I-3.25 J-6.

That shuold start feeding from the center line above the part and stop feeding on the middle left side of the part doing a quarter circle at an odd radius (more of an egg shapped radius) ?
On a lathe we had do create a parametric radius and had to do it coordinate by coordinate in G1 cause a G2/3 would only create a true radius. Does that logic not apply to a Mill ?

Geof,
It may not be safe to assume an FA-550 toyoda has the ability to helical interpolate. So use caution. It'll either alarm out, or position Z, then do the circles.

Machine/control is capable of this. Whill playing around with a ramp off path last night i tried...
G3 X-3.25 Y0. I3.25 J0. Z.005
but i was not too keen on making a full circle to ramp off the part (it was a third pass)... being a prodution shop and only on the mahine for a few days, boss might not like the cycle time. That was why i opted to do the...
G1 X-3. Y-1.5 Z.001 just to get the cutter off the part before rapideing and not doing a full circle.
Thanks for the warning thouhg.


Thanks again guys.

_

[Geof]

You are correct about wasting time cutting air; my program does that for half a circle. You could ramp down over a quarter of a circle; do a full circle at your F30. another circle at say F40. as a spring cut then ramp off over another quarter circle at say F60. to give two full circles and an overall cycle time close to the original.

Regarding the I and X values it does depend on where you are relative to the work zero but in your example your work zero seems to be at the center of the flange so if X is - I will be + and vice versa.

ajl's approach is quite nifty but not easy to figure out at first glance. I would call it spiralling in and it is a good option if your machine cannot handle helical moves.

Incidentally my stuff was hand coded.

[ajl6549]

Trig... or if you have AutoCad maybe this dwg. file will help. Yes, hand coded. Also helical would be the way to go finish wise. But we all have to compete with the Chinese. However if you have access to a face mill with a large wiper insert you could feed much faster. I'm using Iscar 16mill's I get an 80rms finish at 1100 rpm and .140 ipr (154 ipm).

[iMisspell]

in your example your work zero seems to be at the center of the flange

Correct and thanks for the explanation.

Incidentally my stuff was hand coded.

Opps... no offence intended. When ever i see the same feed, line after line and double digit G and M commands (rather then G1 or M1) i always asume its a CAM program.

However if you have access to a face mill with a large wiper insert you could feed much faster.

Sorry, that is what we're using, a 4" face mill with 6 inserts... my Milling lingo is not the best
As of now i cant go and change tooling, for the reason below.

I hear you about the speed but that got me into trouble friday, took 10mins off a 41 min program and the first shift guy felt like i was stepping on his toes, being i was only on the machine for a few days. Maybe some where down the line the two of us will work out the cycle times alittle better.


_

[Geof]

....When ever i see the same feed, line after line and double digit G and M commands (rather then G1 or M1) i always asume its a CAM program.

Repetition comes from me copy-pasting in Notepad and then just changing coordinates in the copied line. Double digits come partly from habit and partly because when I taught myself G-code I did it double digit to reduce my error rate from typos. I know that doesn't seem intuitive, typing more to reduce errors but it worked. Now I rarely write anything from scratch I nearly always copy paste and change numbers.

Absolutely no offence taken, just amusement at being mistaken for CAM ; go and pull up my posts in other threads and you will see why.

...I hear you about the speed but that got me into trouble friday, took 10mins off a 41 min program and the first shift guy felt like i was stepping on his toes, being i was only on the machine for a few days...

Naughty, naughty :nono: :nono: :nono: Not allowed to do that! Reminds me about the time I started a job in Scotland. Manual machines and production work with a bonus for exceeding a certain number of parts per day. I got into deep doo doo for making bonus on my second day. Some of those Scottish guys were a lot bigger than me and I had to back off so I moved on after two weeks.

Actually 25% off a cycle time is pretty good.

[ajl6549]

I now, working 2nd shift is pain but if you get a chance check-out Iscars website and look at the 16mill by the way don't worry about that dayshifter were not that bad :cheers:

[iMisspell]

... production work with a bonus for exceeding a certain number of parts per day....

Bonus would be nice... We dont get a thing for chopping times, union shop with the petty politics which go along with it. On a personal level i just cant (or should i say have a hard time) seeing somthing which i know can run better and just let it sit there un-touched.

I now, working 2nd shift is pain but if you get a chance check-out Iscars website and look at the 16mill by the way don't worry about that dayshifter were not that bad :cheers:

This what your talking about ?
Looks interesting, but all these round tools with more then one insert look neat to me right now
As time goes by and if they keep me on the machine, ill do some research and maybe suggest some new tools, gotta get the programing down first. And for the record, day-timers... slack :stickpoke cant undersand how little can get done with all the "bosses" walking around been working nights for over ten years now and never understood that. Its all good between me and my partner, hes a good guy.

Anyway...
Was able to play alittle, but like i said, its a production environment so i couldnt spend too much time (alittle each night).

With what you posted AJL i had to change the ramp off because it alarmed out, aside from that it worked good.
Liked how it ramped on the part, but (and i should have said this before) how we are holding the part it would be best to feed from the top down rather then from the side. When i have more time im gonna juggle your numbers around and feed from the top. I think the day-guy will like that also if the G3 J-2. lines dont through him off (hes been looking at the same program(s) for the last two - three years, i cant go and turn his world compleatly up-side-down, he's a good guy and we get along good).

Heres what i used and it took 00:01:13 from M1 to M1

G0 X6. Y0 S600
Z0
G3 X0 Y2. I-4.4923 J-3.4768 F30. (RAMP ON)
G3 J-2. (RGH)
G3 J-2. S1500 (SPRING)
G3 X-6. Y0 R5.6805 (RAMP OFF)
G0 Z1.

With what you posted Geof, it ran with no problem, i did change the Z.05 to Z.02
The op time on that was 00:01:34

AJL's gave alittle better finish (was able to kick up the spindle speed for the spring pass) but left ramping marks, Geof's looked good and better then the Orginal, but in the end and as ugly as it is, the following worked the best, nice finish and no ramping marks.

Full clip from the op. if any one is intrested.

N016 G0 B90. 
N017 G54.1 P02 
N018 S600 M03
N019 G00 X-3.25 Y5.5 M08 
N020 G43 H02 Z0.75
N021 Z0.375 
N022 G1 Z.005 F25. (LEAVE FOR SKIM PASS)
N023 Y0. F30.
N024 G3 X-3.25 Y0. I3.25 J0. F30. (ROUGH)
     G0 X-6. Z.032 (GET OFF PART)
     G1 Z0. S1750 F47.5 (FINISH SPEED, FEED & DEPTH)
     X-3.25 (FEED BACK ON PART)
N025 G3 X-3.25 Y0. I3.25 J0. (FINISH/SPRING)
     G1 X-3. Y-1. Z.001 (SLIDE OFF PART WITH NO MARKS)
N026 G0 Z0.75
N027 G53 Z0. M09

You guys have been a big help, thanks, even thouhg i was not able to use your ideas in full here, they will be combinded and used down the road.

Thanks again.

_