[mflux_gamblej]

25% is slower than sh* I only use that when I am scared / un-vetted part or probing routine

50% is what I always use. and I have older DC motor and inch pitch ballscrew machines so its 400 ipm stated rapid max rate.

100% sometimes I feel that I should run 100% rapids. Especially when I am making a lot of the same part (i.e. greater than 10 of them?) I get thinking to myself as I'm standing there watching like the 3rd or so part get made, and I wonder if 100% rapids would help to make it faster. It almost always takes off a pretty useless reduction in cycle time.

here's whats wrong with 100% rapids. I usually save about 30 seconds on a program and I supose most of my programs have about 6 tool changes on average, maybe up to 10. it seems that no matter what program I'm running, I save 30 seconds off the cycle time for running 100% rapids. so basically on a 5 minute program it is 10% off the cycle time. does that matter at all? maybe it does if you're running so many parts that its a whole 10 hour day of running the same thing over and over.

For me I spend more time setting up and programming than making parts on most jobs. typical qty is 20 to 30 pieces per order including all line items so it may be 3 part types on that order. that accounts for about

0.5 of my orders... another
0.4 are qty 30 ea per part type (these are my fav orders), and
0.05 big orders like 500 pieces for very little money per piece, and
0.05 1 or 2 off crazy parts that cost like $2k ea. or even more

so maybe on the big volumes it matters? usually on those i run a whole bunch of mitee bites and the machine just doesnt rapid around much anyhow - actually even less because there are very few tool changes per part on a pallet of mitee bites and no reason to rapid around because the arrays are dense.


Why I see no point in running 100% rapids on my old fadal at 400 ipm:

(newer ones with even better rapids would be even more pointless I think)

1.) 30 seconds saved for running 100% vs. 50% per part on a 5 to 10 minute part with 5 or so tool changes.
2.) the spindle takes a long time to start and stop. when on 100% rapids a lot of the time my spindle isn't up to speed before it enters the cut, and for sure it is never off and stopped by the time the z axis is back to the tool changer for a tool change, so you end up just waiting anyhow on both ends as I have the setp wait for spindle speed parameter set on.
3.) the coolant is never on fast enough - due to this problem I program a 3400 millisecond which is like G4 P3400 or something like that after every tool change to allow time for the flood coolant to be flowing. therefore on 100% rapids i get to the part, then wait a long time.
4.) it only serves to heat up your machine's ways and make parts drift around tolerance wise
5.) it only adds wear and tare to the machine

there's just no point at all because everything else is slower than the rapids. not to mention that the tool changer is glacial.

i guess the point of all this *****ing is that 100% rapids never help, and i feel like it just takes life off of my machines. it pisses me right off that the machine cannot cycle faster - there should be ways to help move faster. any suggestions?



some things I have thought of so far:

- make the tool changer motor faster (i dont have servo turret or side mount)
- get the air to discharge faster into the slide on the tool change and put some buffer springs or something at each end so it will move faster in and out
- mess with the spindle drive's acceleration parameters - you see modern machines reverse in like 10 miliseconds its crazy fast. my fadals get to 5000 rpm in like 3 to 4 seconds on to steady speed, or steady speed to off.
- put a check valve on the coolant line so it cannot drain back to the sump when pump is off
- put a faster pump on the coolant line


also, i dont need a lesson in effecient programming, no g-code tricks or order of operations will speed up the processes I'm talking about above, unless its some kind of crazy servo gain g-code and i already have those written into my post

[jbranting]

The check valve on the coolant line is a good idea, I find myself hitting slide hold all the time to let the coolant catch up.

[schuter]

I saw a YouTube video of a Fadal where the tool carousel didn't stop at each tool position. So it would spin from tool 3 to tool 10 in one solid motion. Much quicker than my Fadal that pauses at each position. I could never figure out why his was different...but if a guy could upgrade it would make tool changes much quicker.

[mflux_gamblej]

I saw a YouTube video of a Fadal where the tool carousel didn't stop at each tool position. So it would spin from tool 3 to tool 10 in one solid motion. Much quicker than my Fadal that pauses at each position. I could never figure out why his was different...but if a guy could upgrade it would make tool changes much quicker.

the continuous selection tool changer you mention is the 'servo turret' option. this consists of a servo motor, an axis control card, servo amp, encoder, and so on. while it is not strictly off the table as an option, it is likely very expensive to install this option built up from used parts. my statement here isn't complete. i did read about this in the fadal manuals at one point and have had a glance at the wiring diagrams for it too. i don't know everything about it, but it sounds like it is very over-done.

in other words you may be able to build a servo turret option yourself from modern components such as a motor drive, matching servo motor, and a microcontroller to fake the fadal machine out into thinking that it is running the geneva gear style tool changer by providing the relevant signals.

i like the idea of building this style of toolchanger up from new parts rather than trying to acquire all the gear needed to do it with the old fadal used parts. one problem is that the tool changer is quite heavy and likely would want a gear box or powerful motor to drive it. this rules out simple setups like stepper motors (probably anyhow?)

[mflux_gamblej]

The check valve on the coolant line is a good idea, I find myself hitting slide hold all the time to let the coolant catch up.

hey on this note i bought a new 1/4 hp coolant pump and it shipped with a check valve. ha. will let u know if its worth a damn (i actually don't know if my machines have a check valve already come to think of it)

[mflux_gamblej]

hey on this note i bought a new 1/4 hp coolant pump and it shipped with a check valve. ha. will let u know if its worth a damn (i actually don't know if my machines have a check valve already come to think of it)

FYI a check valve is just great - decreased coolant flowing fully time from 2.7 seconds, to 1.0 seconds. great upgrade - would cost ya $20 at home depot i think - don't get a spring loaded one.

[ThatguyDave]

You got me. Every fadal I've seen, 30 tool umbrella changes included have a Geneva wheel indexer so incremental movement of the carousel is a guaranteed thing. It would take to a servo drive setup to attain the fluid motion you're describing.

[ThatguyDave]

Yea. Its also worth mentioning that even with DC fadals, rapid speeds increased as the years went on. My older Fadals are slow enough VMC40, VMC15xt, ect that they tend to live on 100% rapids. My 96 4020 only sits at 100% if I'm rushing out lots of repeat work sort of like your scenario.

[mflux_gamblej]

You got me. Every fadal I've seen, 30 tool umbrella changes included have a Geneva wheel indexer so incremental movement of the carousel is a guaranteed thing. It would take to a servo drive setup to attain the fluid motion you're describing.

hey about the geneva indexer - i think you can simply run the indexing motor faster. you'd need a programmable motor controller for this most likely with encoders so that you wouldn't overshoot. if you watch some of the haas umbrella style geneva tool changers they index about 3x the speed of my fadals. I think this is possilbe to do without using a purely 'servo' style changer, i.e. you can keep the geneva indexer but just find a way to run it fast while keeping it under control. this may have a benefit with regards to diy'ing an upgrade - you can likely use the existing motor and gears. if / when i find time to do this I'd make it up as a producable product and could make many units if others wanted one. i feel like i'll not find time to do it though. such is the way of the machine shop world.

by contrast if you want to go full servo control i think it is a bigger project even if it may be conceptually cleaner.

[HalfRhoVSquared]

hey about the geneva indexer - i think you can simply run the indexing motor faster. you'd need a programmable motor controller for this most likely with encoders so that you wouldn't overshoot. if you watch some of the haas umbrella style geneva tool changers they index about 3x the speed of my fadals. I think this is possilbe to do without using a purely 'servo' style changer, i.e. you can keep the geneva indexer but just find a way to run it fast while keeping it under control. this may have a benefit with regards to diy'ing an upgrade - you can likely use the existing motor and gears. if / when i find time to do this I'd make it up as a producable product and could make many units if others wanted one. i feel like i'll not find time to do it though. such is the way of the machine shop world.

by contrast if you want to go full servo control i think it is a bigger project even if it may be conceptually cleaner.

Come up with an upgrade kit that reduces the toolchange time by 50% and you would sell the holy snot out of them. I'm just learning about Fadals but the toolchange is why they have a reputation of being slow.