What's new
What's new

Steel / Stainless roughers for older machines with flood coolant and traditional paths?

I do miss when Mori Seiki actually meant something... these DMG Mori's are garbage maybe they've gotten better since I've been in the woods but I doubt it.

To be fair though the OP was asking about optimizing an ancient grey beard machine... 7000rpms is like model T ferd stuff lol, thing probably can't feed fast enough to take advantage of the high feed insert cutters anyway.
i'm all waxing poetic about a time when rapid was 300ipm and folks thought that was fast lol...
Regardless of the machine, HSM paths optimize tool engagement, and removes full slotting in the same operation.
So you can still run 2500 rpm, 20 IPM if you like with a 25%-50% step over, but the machining strategy will be better.
Doesnt mean because your using a HSM path your screaming around banging around, and if your using something like Volumill even if your machine has shit acc/dec the code has slow ease in/out feed rates optimized.
 
To be fair though the OP was asking about optimizing an ancient grey beard machine... 7000rpms is like model T ferd stuff lol, thing probably can't feed fast enough to take advantage of the high feed insert cutters anyway.

Now Now, I can get up to 500 imp feed but it starts cutting sharp corners above 40ipm. Trying to figure out the accel/decel settings but it's hard to get much useful help on an old 520meldas control. Thing is, this machine is old but practically new as is resided in a high tech shop just before they shut down in the .com crash then moved to a college where it barely ran as no one knew how to program it. Still has most of its paint on the inside and is damned near minty, hold tolerances perfectly and leaves great finishes.

I can and have used HSM paths on this machine and an identical one at the previous workplace to decent effect. Reason I am sticking to the old school paths as mentioned is partly functional regarding limited memory, but also because I do not yet personally own a software that can generate the Adaptive paths, nor am I in a position to purchase one just yet. Fusion is out due to the internet requirement.

I'm talking with a local tooling supplier right now, see what they suggest. Going to look at powdered metal options from YG and Minicut plus others. The machine can handle the larger tooling and I have the holders so I'm looking forward to experimenting. Their applications guy said that one of their customers un in northern Ontario tried another PM cutter and swapped to it from carbide due to performance/cost. Curious to find out what it was.

Thanks for the input so far and I will report back with what is offered. I love watching the HSM paths but there is some joy to be had watching a good rougher slot full depth nearly silently through material.
 
Looks pretty quick.

yeah get conFusion360 for cheap, on a isolated network(VLAN) rig, will help also.

Also, I have never tried them, but a lot of people buy them, Lakeshore carbide has carbide tooling at HSS Co prices.

When I tried powdered metal it sucked. I was pumped to try it when it was new, disappointed.
 
Last edited:
Now Now, I can get up to 500 imp feed but it starts cutting sharp corners above 40ipm. Trying to figure out the accel/decel settings but it's hard to get much useful help on an old 520meldas control. Thing is, this machine is old but practically new as is resided in a high tech shop just before they shut down in the .com crash then moved to a college where it barely ran as no one knew how to program it. Still has most of its paint on the inside and is damned near minty, hold tolerances perfectly and leaves great finishes.

I can and have used HSM paths on this machine and an identical one at the previous workplace to decent effect. Reason I am sticking to the old school paths as mentioned is partly functional regarding limited memory, but also because I do not yet personally own a software that can generate the Adaptive paths, nor am I in a position to purchase one just yet. Fusion is out due to the internet requirement.

I'm talking with a local tooling supplier right now, see what they suggest. Going to look at powdered metal options from YG and Minicut plus others. The machine can handle the larger tooling and I have the holders so I'm looking forward to experimenting. Their applications guy said that one of their customers un in northern Ontario tried another PM cutter and swapped to it from carbide due to performance/cost. Curious to find out what it was.

Thanks for the input so far and I will report back with what is offered. I love watching the HSM paths but there is some joy to be had watching a good rougher slot full depth nearly silently through material.
have you considered canned cycles? (I'm not familiar with meldas, but most machines can do sub routines) then you just have one "tool path" and a series of M99/m98z's
 
have you considered canned cycles? (I'm not familiar with meldas, but most machines can do sub routines) then you just have one "tool path" and a series of M99/m98z's
Absolutely, that is my primary method and the older versions of cam software excelled at it. When using early 2000's Mastercam or similar era Gibbs and my first job, all the code I was used to seeing was subroutines by default. I love having full programs that are only a few hundred lines, easy to read and to edit. Unfortunately Fusion, and other like Esprit seem to lack this functionality (i.e huge code), not sure if newer versions of Gibbs or MC still spit out subroutines by default but if they do then I will keep them in consideration in the future.

I tried the fusion posts fusion offers for the older Fanuc controllers which are almost identical output to Meldas 520, but they are still nowhere near as efficient as the older cam softwares were. I think it is a trend that programs are way less optimized now that we have the computing power to just brute force our way through the volume of code. I hate to have to drip feed basically every program if I were to use fusion. Been there done that and that is why I am here asking.

Houdini:

I'm dissappointed to hear about your experience with the PM cutters. I'm anxious to see what the sales/solutions guy comes back with, he is usually pretty good and I will report back. I a used to doing mostly aluminum, mild steel and 4140 so never really needed much in the way of special cutters so this is a bit of a new adventure for me as I get more stainless and exotics coming in from some new customers.

That FV-800 in the video looks like the identical model to mine, and it can run faster than shown there. Looks like the sad example in the video has its guts hanging out from the head cover being removed and has been beat to hell. These damn things are nearly indestructible though, owner tried to kill one at the first shop by drilling 3/4" holes in steel bars at 60 to 100 ipm and 3000rpm (Guhring carbide drill). Load needle bounced from 0 and the off the stop pin in the red 1000's of times a day, all day for about 3 years before we gave up trying... Spindle was still whisper quiet.

Enjoying this discussion so far and look forward to digging up more suggestions.

Anyone have experience with International Minicuts? I have heard great things about their aluminum cutters but nothing about the 995's for exotics.
 
Gibbscam has a new interface but it's still the same, they added stuff but didn't get rid of anything.
drill subs, machining subs, All that is selectable still.
I went from 2006, 2012, 2022-2024
 
think the last shop I was employed at was "attempting" to use fusion, I was not impressed... especially since before the merger I was using Gibbs and writing most of the programs for the old shop. Its a long story, but its also one of the reasons I've spent the last 7 years in the woods.

Dollars to donuts I'd bet someone elses money that the newer software can spit out sub routines too, I mean I was once told you can't program threads, tapping or single point on Gibbs or Mastercam... which is weird cause I did it daily without issue (or surfcam... remember surfcam...)
 
think the last shop I was employed at was "attempting" to use fusion, I was not impressed... especially since before the merger I was using Gibbs and writing most of the programs for the old shop. Its a long story, but its also one of the reasons I've spent the last 7 years in the woods.

Dollars to donuts I'd bet someone elses money that the newer software can spit out sub routines too, I mean I was once told you can't program threads, tapping or single point on Gibbs or Mastercam... which is weird cause I did it daily without issue (or surfcam... remember surfcam...)
Yeah whom ever said you couldn't do 'anything' in Gibbs was just a low end programmer. We had a guy who still liked his 2D Smurfcam at home. :D
 
Ironically the same guy that tells everyone he taught me Gibbs... is the one that couldn't figure out threading... insisted you had to hand program it... they also told me that you had to program xy radiuses in absolute and by quadrant on a multi axis lathe... cause like g17 I guess never occurred to them? i.e. g91g3x.1y.1r.1; x-.1y.1;x-.1y-.1;x.1y-.1 g1g90x0y0 etc... FOR EVERY DANG HOLE grrrrrrr given me a little ptsd lol (I don't remember if you had to use I J or if a modal R worked? but I wanna say they used IJ cause they insisted on hand editing each hole to size, I managed to avoid running these setups somehow lol)
 
Ironically the same guy that tells everyone he taught me Gibbs... is the one that couldn't figure out threading... insisted you had to hand program it... they also told me that you had to program xy radiuses in absolute and by quadrant on a multi axis lathe... cause like g17 I guess never occurred to them? i.e. g91g3x.1y.1r.1; x-.1y.1;x-.1y-.1;x.1y-.1 g1g90x0y0 etc... FOR EVERY DANG HOLE grrrrrrr given me a little ptsd lol (I don't remember if you had to use I J or if a modal R worked? but I wanna say they used IJ cause they insisted on hand editing each hole to size, I managed to avoid running these setups somehow lol)
That is just brutal... I'm so glad I never had to suffer that very much other than when I got the chance to "enlighten" some guys when I started running a Toshiba TUE vertical lathe with a Fanuc control. They were programming every profile pass by hand, like every DOC, and using something dumb like a g32 to do it, can't quite remember. I had lots of time during runs so I pulled out the manuals, drafted up some generic facing, profiling and boring canned cycle programs using G70, G71 and G72 and showed a couple of guys how to use them. They loved it, just program the start point, final path, fill in the blanks and hit go.

I bought a motorcycle off of a guy who started working there about 6 years after I left and he was like:" Oh, you're mister canned cycle, I heard about you! We still used those all the time, pisses off the programmer because we have the parts running before he ever has a program for us." Gave me a good chuckle, their programmer was a total dink.

I loved the lathe programming on Gibbs 2006, put out canned cycles just fine which were easy to tweak at the controller for threading and being able to adjust the start and end of paths along the line by dragging and dropping is something I sorely miss in other soft wares. I never really liked the gibbs tile system though, found it hard to keep track of ops in long programs. Not an issue on the lathe but in my small bit of experience in mill programming, was not really a fan.

Always painful to listen to morons talk about what they have no clue about. Unfortunately they are usually louder than those who actually know things and their parroted BS spreads too quickly to keep tabs on it...
 
I never really liked the gibbs tile system though, found it hard to keep track of ops in long programs. Not an issue on the lathe but in my small bit of experience in mill programming, was not really a fan.
I used to machine injection molds with Gibbs, talk about a lot of process tiles.
The newer version they are actually manageable now, color coding, collapsing based on groups or custom groups.
A lot better than before, searching through 70 processes got brutal :D

Wish they added dimension to it, instead of linear, more like a node based software. Like this.
1709648402293.png
 
Last edited:
That is just brutal... I'm so glad I never had to suffer that very much other than when I got the chance to "enlighten" some guys when I started running a Toshiba TUE vertical lathe with a Fanuc control. They were programming every profile pass by hand, like every DOC, and using something dumb like a g32 to do it, can't quite remember. I had lots of time during runs so I pulled out the manuals, drafted up some generic facing, profiling and boring canned cycle programs using G70, G71 and G72 and showed a couple of guys how to use them. They loved it, just program the start point, final path, fill in the blanks and hit go.

I bought a motorcycle off of a guy who started working there about 6 years after I left and he was like:" Oh, you're mister canned cycle, I heard about you! We still used those all the time, pisses off the programmer because we have the parts running before he ever has a program for us." Gave me a good chuckle, their programmer was a total dink.

I loved the lathe programming on Gibbs 2006, put out canned cycles just fine which were easy to tweak at the controller for threading and being able to adjust the start and end of paths along the line by dragging and dropping is something I sorely miss in other soft wares. I never really liked the gibbs tile system though, found it hard to keep track of ops in long programs. Not an issue on the lathe but in my small bit of experience in mill programming, was not really a fan.

Always painful to listen to morons talk about what they have no clue about. Unfortunately they are usually louder than those who actually know things and their parroted BS spreads too quickly to keep tabs on it...
not to derail this entirely..

I once made some billet angled license plate holders for my motorpsycho... hand programmed it while running 2 other machines, gathered most of the tooling I needed 1st break, had it set up during 1/2hr lunch, ran 2 of them out of scrap, including the hemstitching where it met the fender... copy and paste on a haas is skookum lol, roughed with a 2" face mill, finished with radiused end mill, and a square cnr E/M... spot,drill,counterbore 4 holes, just basic tool paths no crazy angles to plot out...

I made the mistake of leaving one of them on the table, so when the day shift supervisor came in, he refused to believe I could program a hog out with 5? tools and set it up on my breaks... bruv... they have security cameras watch em...

I would regularly reprogram his Dog doo programs and have them bought off often before the first part was finished running... some how I didn't get fired then lol.

Moral of this story is this dingus later went on to become the teacher at a tech school... he was fired after 1.5 years.

I'll try and get pictures later today of the license plate frame (thats held up well after 20 years lol) but I gots to get chooglin back to the woods for now.
 
Thanks for the input.

I'll certainly be looking to plumb in an airline for steel roughing but will be running flood otherwise.

As for drip feeding adaptive paths, that is am option i tried but I have had buffer overrun problems due to the nature of adaptive toolpaths using point to point which overwhelms the read ahead. I can drop the tolerance to help some but even then the machine accel/decel may cause gouging without a lot of extra material left.

Another point to add is that I am still running an older cam software without adaptive and am not in a position to upgrade yet. I keep everything offline for security so Fusion is out of the equation as well, though I do use it at my day job. Just trying to keep things as seperate as possible due to previous complaints by another coworker accusing me of skimming jobs from students.

I have never had much luck roughing stainless with regular or variable flute endmills, always get edge failure from heat buildup and/or recutting chips. With flood, through tool or air blast, always a short tool life. That is the reason I use serrateds to reduce cutting edge force and have small, easily controlled chips to evacuate. Been working well for me so far but want to try some different geometries.

I have been using mostly 3/8" and 1/2" cutters but figured if the machine has the rigidity in the column with the box ways, i could try the larger cutters and actually make the load meter move above 20% before breaking tools😁.

My parts thicknesses vary wildly. Currently roughing out a stainless pocket 10mm deep, 70mm wide by 280mm long but other jobs are 4 inches deep profiled and pocketed. Of course I choose cutters based on stickout, longer is wider or tighter helix or relieved. I'm not a total noob, but I have only been doing this for 19 years so I'm still learning, haha.
find an inserted end mill. there are great ones around like R390 sandvik. $$$ but worth it, and feed it hard and fast, over 10 thou per tooth.
 
Haven't been active for 7? years, Modern machines are "better" in some respects, though the shop I mentioned having drive motor issues was with Mori DMG mills that are about 12 year old (they were shiny new when I quit working for them, I've stayed in touch though) Both have bad x motors, or drive problems.
You're missing the boat here.

Let's say that the Mori shop billed at $25/hr more than you do, because they are hauling ass and "beating up their machines". I'd bet that the deficit is a LOT more than that, but we'll use $25/hr for this example: That is $576,000 more they billed than you over 12 years.

I'm pretty sure they can afford to replace some drives or thrust washers or spindles or whatever boogie man keeps you up at night, and keep on making more money than you.
 
You're missing the boat here.

Let's say that the Mori shop billed at $25/hr more than you do, because they are hauling ass and "beating up their machines". I'd bet that the deficit is a LOT more than that, but we'll use $25/hr for this example: That is $576,000 more they billed than you over 12 years.

I'm pretty sure they can afford to replace some drives or thrust washers or spindles or whatever boogie man keeps you up at night, and keep on making more money than you.
This type of thing is what the old dudes said when I told the owner I read a paper on HSM machining, after I hand wrote a part with HSM style machining specs.
The owner wanted me to start re-writing anything in steel that was production.
The old manager was like, your gonna ruin the spindle, ballscrews.......
Now it's the norm!
They're like Dorito's crunch all you want we'll make more.($$$)
 
Last edited:
They're like Dorito's crunch all you want we'll make more.($$$)
Really depends on your situation in life. Small shop, doing prototypes, short runs, mom and pop, bought a nice mill and want to keep it until they kick, there's no need to beat stuff up. It doesn't give you anything, you'll spend more time waiting for UPS to arrive than you'll save on the parts and there's no more parts to make today anyhow, have to draw up that thing Joe sent and write a program anyhow.

Now, if you have 40,000 in process and it's a decent-size biz with everything ready for the next 50,000 of something else and the tax benefits matter as much as the machine itself, that's a totally different matter.

So, you get guys arguing over what's right but diff'rent strokes ...

I'm also a believer that these hsm toolpaths ? I swear some of them are slower than the old-fashioned way. A lot of them spend half their time driving around in the air, cutting nothing. I know they ain't supposed to but sure seems like it. Just drop a 1" cutter and chomp through there on a machine with square ways and a real spindle, I bet a lot of the time it'd be quicker. Not always, just sometimes.

Once again, diff'rent strokes ...
 








 
Back
Top