Parameter change max rapid feed

anybody ever change max rapid feed rate using parameter change ? bigger machine tends to bounce back and forth with servos and hydraulic pressure when stopping from max rapid feed rates. i am considering slowing rapids with parameter change.
.
since it can always be changed back i would be interested if it helps at times. sure i can always turn rapids to 50% but this way even if another operator got rapids at 100% cnc would still not go above certain max feed even in rapid mode

It can be done, but oscillation as shown in your picture is more often a sign of another issue with the machine. Dropping the rapids may "seem" to fix the trouble, but really it is just masking it.

Servo tuning could possibly help this. Has the machine behaved like this always or did this just start recently?

Does the machine have scale feedback?

What control?

Age?

Overall condition?

Vancbiker said:

It can be done, but oscillation as shown in your picture is more often a sign of another issue with the machine. Dropping the rapids may "seem" to fix the trouble, but really it is just masking it.

Servo tuning could possibly help this. Has the machine behaved like this always or did this just start recently?

Does the machine have scale feedback?

What control?

Age?

Overall condition?

Click to expand...

.
Fanuc 15 and yes the servos are adjusted periodically as well as the hydraulic spool values cleaned periodically and it helps. recently Z axis hydraulic cylinder counter weight was found leaking and tie rods were loose. the repair is expected to help. still servo oscillation was seen of day one, every day after and 23 years later servo oscillation still there at times. when it gets to certain levels we stop and call maintenance. the borderline oscillation is still annoying
.
i am just looking at servo screen for gantry mill. the W columns one is master and other slaved to it. they tend to bounce back and forth trying to match positions. when Y axis moved the weight shifts and the W columns bounce. machine acts like on springs.
.
i often use G4 delays for example X axis moving 10 ton table stops from 600 ipm feed even a handle move of .010" you can see with indicator it bounce back and forth .0005" for a few seconds. bigger cutter it is not a problem but with a horizontal head attachment and short carbide drills and back and forth side movement tends to want to snap the drill off. sudden tool failure has stopped by slowing feeds and speeds so drill not near limits and have more safety factor to resist machine bouncing better

I have been wondering about this as well. The problem I am trying to mask is worn ball screws on a machine I don't want to spend 8k on to fix. Is it as simple as changing 3 lines of parameters or is it more involved? Control is a Yasnac J50. The machine has 1400 ipm rapids with no over rides so the only other way to slow them down is to use G1 but I have a maximum feed of 200 ipm.

my machine has parameter 1420 i believe it says Rapid F and numbers appear to me in centimeters so when multiplied by .3937 comes out to inches
.
if i change Y axis max feed, the weight shifting and the W columns going up and down should be less and faster at stop bouncing. even the X axis you cannot have a 10 ton part on a 10 ton table stop instantly and not have bouncing back and forth .0005" for a second or 2. if you visualize a carbide drill in a horizontal head the bouncing side ways would obviously tend to want to snap small carbide drills.
.
a sudden tool failure and say 24 holes needing rework can add 10 hours to machining job. slowing rapid feed so it takes 30 minutes longer but save 10 hours rework over 50 parts it is worth it. as the saying goes sometimes to go faster you have to go slower

Yes, parameter 1420 sets the max rapid rate.

Are you able/willing to say what brand the machine is? Big machines present bigger challenges for servo tuning. The dynamics of moving the workpieces precisely when the weight varies by tons is tough.

Master/slave axes also pose a problem on a machine that age. The servo and control technology of the day not really being up to it.

theres an easier way.....one is I never rapid faster than 25% on the dial it wears the machine out 10X faster than the savings on the parts....and if I need to go faster without changing the parameters or the % dial I program G01 x5.0 F75.0 and then F5.0 back into cut...I make a G01 move to right before touching part....part might be at 4.5 so fast feed it to 4.6 and then feed regular the rest of way...same thing on out cut...cut to 4.6 F5.0 and then say x8.0 F75.0.....Y40.0.....x-8.0......y0.0....x-4.6....then up the feed again.....x0.0 F5.0....that other jazz I use it to go up and around the clamps or something and never go into G00...hope that helps

PS....unless you don't need parts every 7 seconds don't even go 50% on the machine....and it will hold its repeatability for 40 years

I lowered it on both my Mazak lathe and my Brother TC. Different reason though. Instead of a BANG! I now get a bang. After a crash and some reading on here where some guys recommended it I figured it was safer.

Tom I guess what you say makes sense. If everyone (management) is on board with the idea then giving it a try could have you looking like the hero that reduced/eliminated a nagging problem. But I before I took it upon myself to slow their whole operation down by a parameter change and then walking off I'd want to have someone's blessing.

Actually running the machine at 25/50% rapid speeds does fix the oscillation issues your visually seeing on the part?

Apparently it does is why your asking?

Brent

Johnny Larue said:

theres an easier way.....one is I never rapid faster than 25% on the dial it wears the machine out 10X faster than the savings on the parts....and if I need to go faster without changing the parameters or the % dial I program G01 x5.0 F75.0 and then F5.0 back into cut...I make a G01 move to right before touching part....part might be at 4.5 so fast feed it to 4.6 and then feed regular the rest of way...same thing on out cut...cut to 4.6 F5.0 and then say x8.0 F75.0.....Y40.0.....x-8.0......y0.0....x-4.6....then up the feed again.....x0.0 F5.0....that other jazz I use it to go up and around the clamps or something and never go into G00...hope that helps

PS....unless you don't need parts every 7 seconds don't even go 50% on the machine....and it will hold its repeatability for 40 years

Click to expand...

You are an idiot.

To my mind, at first glance, that kind of oscillation is much more a symptom of accel/decel than raw speed. On the Fanucs I have worked with there are lots of parameters for tweaking the accel/decel curves, and they are separated for rapid and cutting moves so you have a good level of control over how each axis behaves.

However, a simple test of increasing the rapid decel time period would be a single parameter change that would make the axis come to a stop more gradually.

This is an elevating rail and ram machine yes? Going by memory from other photos you've posted...

As Vancbiker already stated, master-slave servos on a long beam complicate matters a lot, as do hydraulic counterbalances. It could be that the servo tuning is all just too aggressive for the mechanics of the machine.

The oscilation is a servo tuning issue, you need to retune the machine but with it loaded, not empty as it left the factory. Not so relevant on a small VMC, but when you strap a couple of tons to it, you need different gain and damping parameters set.

IMHO you don't want to mess with rapid rates, the speeds not the issue, its the de-acceleration - acceleration thats your problem area. Equally its not the fast moves that causes the wear its the hard high speeds starts and stops. If your happy to lose the time, reduce the accelerations, but leave the feed rate alone. That will take the loading off the drives and ball screws.

The oscilation is a servo tuning issue, you need to retune the machine but with it loaded, not empty as it left the factory. Not so relevant on a small VMC, but when you strap a couple of tons to it, you need different gain and damping parameters set.

IMHO you don't want to mess with rapid rates, the speeds not the issue, its the de-acceleration - acceleration thats your problem area. Equally its not the fast moves that causes the wear its the hard high speeds starts and stops. If your happy to lose the time, reduce the accelerations, but leave the feed rate alone. That will take the loading off the drives and ball screws as well as reducing the excitation of the machine components into oscilations.

I see what John is saying if you use high feed instead of switching over to G0 rapid the control will make a smoother transition because it doesn't have to cancel the milling cycle to go back to rapid. Whether or not that's going to resolve the issue who knows it helps prevent dwelling for sure though. I assume that material is some sort of cast iron or aluminum. Judging by the swirls maybe adding a small radius to the tip of the flutes of your tool maybe help reduce that marking/bouncing. If we could watch as this happened I'm sure a better conclusion could be drawn

Whysosharp said:

I see what John is saying if you use high feed instead of switching over to G0 rapid the control will make a smoother transition because it doesn't have to cancel the milling cycle to go back to rapid. Whether or not that's going to resolve the issue who knows it helps prevent dwelling for sure though. I assume that material is some sort of cast iron or aluminum. Judging by the swirls maybe adding a small radius to the tip of the flutes of your tool maybe help reduce that marking/bouncing. If we could watch as this happened I'm sure a better conclusion could be drawn

Click to expand...

The post you are referring to, along with everything else he has ever posted, is unfathomably stupid. Trying to extract any reason from it is an exercise in futility.

The problem the OP is fighting is servo oscillation or maybe low frequency vibration of the machine frame. Doing anything that will increase the jerkiness of motion (like using high feed G1 moves instead of G0) will only make the problem worse. Probably much worse.

Johnny Larue said:

theres an easier way.....one is I never rapid faster than 25% on the dial it wears the machine out 10X faster

Click to expand...

LMAO you can't be serious

adama said:

The oscilation is a servo tuning issue, you need to retune the machine but with it loaded, not empty as it left the factory. Not so relevant on a small VMC, but when you strap a couple of tons to it, you need different gain and damping parameters set.

IMHO you don't want to mess with rapid rates, the speeds not the issue, its the de-acceleration - acceleration thats your problem area. Equally its not the fast moves that causes the wear its the hard high speeds starts and stops. If your happy to lose the time, reduce the accelerations, but leave the feed rate alone. That will take the loading off the drives and ball screws.

Click to expand...

.
.
Toyoda BN-25a with Fanuc 15 control
picture of servo oscillation is a extreme case where either ball screw servos needed tuning or in picture shown the hydraulic spool servo valve on hydraulic counterweight was dirty and fluttering. when hydraulic valve cleaned i have seen it stop. many causes for oscillation and often it is barely noticeable. i can see it every time on a tool change as it moves in Y along gantry beam the W column servos are bouncing back and forth continuously adjusting for weight shifting. amounts are often 5 to 10 times normal. when i run 50% rapids usually finish is better and more accurate on finish passes. i also use G4 P20000 for 20 second delay and i give M code to lock W and Z axis when trying for .0002" flatness tolerances.
.
but when drilling with short carbide drills with the horizontal head at 100% rapids i have occasionally gotten sudden tool failure where it does not happen with vertical head drilling. my belief is the oscillation even if +/-.001" is putting side stress on short carbide drills.
.
i can change parameter to slow rapids and can change back after a day or a week.
.
X axis is a concern too. the table with part can easily weigh 10 to 20 tons. in handle mode moving .010 i often see with .0001 indicator it going back and forth .0005" for a few seconds. i have had to when doing boring bar test cut and it moves 10 feet over to allow operator to measure bore i put G4 delay when it moves 10 feet back in X over bored hole. if it is still oscillating in X if gives a slotting hole even if only .0002 to .0005" giving delay stops that. trouble is with other operator with a program with no programmed delays and rapids at 100% can have problems.
.
same with G83 peck drilling with a long over 16" drills a full peck can be rough on drill when reentering hole at high rapid. when rapids reduced we never have problems pecking. but some operators forget to reduce rapids. i am thinking going from 472 to 400 ipm rapid max speed is something to try out. depending on how it works i can always change again next week

Gregomarwick. I can see that most of his posts aren't really let's say helpful necessarily. All I'm saying is if you stay in G1 instead of flip flopping and use a faster federate maybe one that is 25% of your full rapid the finish would be better regardless. We have a similar problem on our Kia KV25 drill tap machine and the solution was exactly the above

@ DMF_TomB...

In an earlier post I asked if you know whether the machine has scales but did not see an answer.

When a machine has scales and some lost motion (backlash) in the mechanical components (screw, nut, thrust bearings, gibs, couplings, screw torsion/wind-up) there is a big potential to get low frequency servo oscillation. There are parameters that can reduce that behavior but they should only be employed after one has eliminated as many of the causes of mechanical lost motion.

Whysosharp said:

......All I'm saying is if you stay in G1 instead of flip flopping and use a faster federate maybe one that is 25% of your full rapid the finish would be better regardless. We have a similar problem on our Kia KV25 drill tap machine and the solution was exactly the above

Click to expand...

While that technique may help you make acceptable parts and I understand the need for that, you are really just working around an underlying problem. The underlying problem may not be resolvable, but properly designed and operating servo motion control and machinery does need this type of work around.