4 hours ago, crazy^millman said:

Mushroom effect is something a programmer has to also think about. I have a project I am currently working on for a process improvement. I have cut from the TOP and now I am cutting from the sides. There is material removed past a certain point but OPTI-Rough still ignores the air cutting with no stock present taking the largest section of material in area above the area it is cutting and carrying that all the way down through the cut motion. Now I have to break it into more stock models and OPTI-Rough sections to avoid the extra 30 minutes of air cutting per part. It is that or create surfaces to avoid which then create other problems with not arcing in and out like it should. 

This is now fixed going into 2023 apparently.  Haven't tested yet with the public beta.  I doubt it is perfect, but certainly looks better.

https://www.youtube.com/watch?v=Gs6XS4Tgmog

45 minutes ago, JB7280 said:

I just can't fathom it moving at 2800ipm without clipping corners, even with look-ahead turned on.

Okuma's without hicut or supernurbs on run really smooth and fast, but boy do they clip corners.... 

2 hours ago, rgrin said:

If a part is going to be a repeat one, then it is time to sit down and work through some cycle time analysis and explore options other than dynamic.

Almost always, if the geometry has any complicated features, or if it is 3d roughing, Dynamic or Opti is going to be better.  The biggest decision that needs to be made will usually be in what to use for a stepover.  Sometimes, a 40-60% or even greater in shallow applications stepover with dynamic motion with conventional speeds and feeds will yield very very good predictable results and will yield more productivity in tool life and process reliability than a conventional toolpath, with only a small cycle time premium.  On paper the MRR might be more with a light stepover with speed bonus, but the extra air cutting and acc/dec slowdowns make it overall slower.  Heavy radial stepovers make for a good balance without a lot of extra air cutting.

20 hours ago, cncappsjames said:

 but in real life can be disastrous if the Engineer and Programmer are not the same person. 

Very rigid and defined processes are needed to make this work, even then it still falls short.  We have had some projects recently that have tried to capture this capability for some product families, it's now better than it was but it is very custom.  This is in NX.

I had a really nice response to this written up. Opened the catalog to look something up and lost it.  Dummy me

Long story short. Seeing as this is my day job...

  If indeed you are running that long tunable adapter.  You will likely need to drop DOC down to .020-.040".  Unless it is 65-45-12 DI and is soft as butter, do not run those at 1200 sfm, treat it like one might steel or use K3 material group numbers, run those D KCK15 inserts at 700-800 SFM, and start around .040" on the feed.  If those don't hold up due to chipping or built up edge, switch to KCPK30 in the GP chip breaker, and run 550 SFM, and .040" on the feed.  Same light depth.  If it is stable, feel free to walk the depth up until it isn't and back it off a bit to regain stability.

Oh and ramping is going to more than double the radial load, so avoid it if you can in this case.  If you can't avoid it due to fixture constraints, you can't...

FYI for the group 7792 inserts should not be run at over 80% of max ap1 listed for it's IC size. When shouldering it can't be avoided at the wall, but it can handle it there due to chip thinning, if you run that deep you will likely blow the corners off the insert at any feed over .015 ipt.  Lower depths of cut at higher feedrates are almost always more productive.

OK, I was able to get it to work by loading op defaults, and with the combination of using the green check.  Thanks for the help!

So I think this is a great toolpath... Used it a few times, it has worked well, tedious, but is easier than doing a 3D contour and doing a toolpath edit.  Trying to use it today and I can't add segments.  I'll add them into the list and then apply the changes close the dialog and regen, and none of the changes come through.  Open it back up and only the first depth and the retract segment is in the list.

GRRRRRRRRR

I would do everything based at B0C0, to shift it in I would just probe starting from a base "0" offset at B0C0,  use CALL OO88 to generate a base offset, and then probe @B-90 C180 and use probing errors from nominal to shift the main work shift in B0C0, then call OO88 from there.  Being on B-90 C180, there isn't any need for rotation center shift info.

If I have it right....

X error will shift Z  (+X translates to -Z) 

Y error will shift Y (+Y translates to -Y)

Z error will shift X (+Z translates to -X)

Best of luck, no need to get fancy.

You can always update the base B0C0 offset and probe again if you don't trust it.  Errors should read close to zero if done properly (won't be perfect as I am assuming it is a cast hole you are probing).

If for some reason you had to probe at a non-orthoganal plane, you could still do the same, but you would have to generate your base X,Y,Z shifts with a little vector math, not difficult, but at the same time would require putting pen to paper for a minute to make sure you get the math correct prior to writing the macro code. (rotation centers still do not come into play in this case as we are dealing with translational errors which are measured when rotation center compensation has already been applied.

Happy to help more if you are still stuck.  Though I will add I am a newb at Okuma macro code.

4 hours ago, mayday said:

Family Farm I grew up there, well I'm almost grown up

We don't grow up, we just grow outward or larger around.

Have you tried using G68.3?  You position the rotary axes and call this then you can shift and rotate the coordinate frame around the new spindle axis, G68.3 X Y Z R.  I used to use this if I was hand coding a single hole and wanted to add pecking at the machine at an odd angle.  I wonder if it does the same side shift thing.

The standard way would be using G43 with G68.2 like you did in the original sample. 

Does the head change angle when you call out G53.1 in your case?

I've chased there error for ages....   Typically though I am moving the rotary and it is the Y axis fluttering back and forth sometimes up to .002mm.  I haven't nailed it down yet, but I have found version to version differences with Mastercam in how it posts out.  I think it has to do with the floating point math and rounding on the backend going from NCI to posted code.  My recommendation would be to create a variant of your post and round your rotary output to two places for now to get some clean code, see if it makes the difference in your finish.  Otherwise, dig into the NCI using excel and run some math's on the vector output to see if the angles resolve correctly in the NCI to see if it is a math error getting from there into the posted code.  If it doesn't exist in the NCI, then it is an issue in the math generating the vector, which depending on the comp method (center or tip) will effect your tangent point position slightly, and is likely a product of poor geometry. 

Just to give an idea of how deep I have dug on this in the past, I have seen a y axis shift issue with simple cylinder pattern surfaces, and no tilt, pointed directly at the center of rotation.

My service contract is up in 6 weeks.  I had fried my i9 a year or so ago and had been operating in limp mode.  Well, two motherboards later I have a fully functioning laptop...  there was a reason I put it off.  Anymore, stability is all I need, but I have been told to maintain a laptop capable of high horsepower programming work.  I MIGHT touch Mastercam once a quarter now for work purposes on my laptop.  Usually anymore I am just coaching customers.  It had been so long since I a fully functioning processor running at full clock speed I forgot how much more productive my work laptop used to be and is again....   Anyway, I have been told by IT that I will be in this rig until the turn of the year unless it fails again between now and then.

Not looking forward to convincing my boss to approve the budget expense for another i9/quadro laptop.  Hopefully I can get 64 gig next time.  I have had this rig since April of 2019.  Minus not being able to overclock for a long while, it's been a good machine.  So far I have had to replace the battery once, speakers once, and mobo twice.  It has only cost me time.

2 hours ago, crazy^millman said:

On the 10th revision of the MT environment.

Sounds like you are going down the right path....  I know I definately wouldn't be asking for custom functionality in that case.  Maybe it would just be easier to do it via manual entry...

Best of luck and good work as usual.

I applaud you Ron for using this tool (advanced drill) and coming up with feedback to alleviate frustrations in applying it to the real world.  God knows barely any of us have time to take the time to do that.  I haven't had the pleasure of using the Advance Drill toolpath yet in a real world application, so I am of no practical use to the situation, just theoretical ;).

Sticking to the theme of improving function of the advanced drill function. Seems to me it would be well suited to create some sort of segment pattern repeat function so you can further customize pecking cycles without a lot of work.  In my mind anyway, it seems you could use this function to create compound pecking routines, that would say peck, peck, clear, peck, peck, clear.  So on so forth.  Giving you the ability to loop certain sections, but able to vary only certain fields based on depth.

I'll quit dreaming.  Back to the real world.

Just an outside view here....  I'm assuming you are dealing with a stubborn customer that won't think outside the box here?  I say that because this is a case where a simple Macro and custom cycle definition in the post would probably be 10x faster to get a nice flexible result.  Course then verify wouldn't function for this tool, and of course it would require some light post editing which the customer probably won't do...  

Thanks for all you do for this community!

Yeah feedmill will be more efficient in that case. 

Per my maths...  I figure with full depth of cut with a 6 flute 1/2 x 1.5 tool, you could get about .85lb/min of material removal.  FWIW, IF, you could go full depth the whole time, I'd figure on about 2 endmills to get the job done.  But given working up to the crest it won't be that nice and you would end up wearing off the end of the tool before wearing out the rest of the tool.   Soo.... yeah.

Not that you had asked what the best way to do this was, but. Maybe post a dummy file and see if the collective brain trust has some goodness to offer that hadn't/hasn't been considered.

Maybe a hybrid stepped approach?  Work it with a feedmill until you could get some good solid full depth steps you could remove with solid tools  Not knowing the geometry of what you are doing makes this a fun guessing game.  Could be way off base on that approach.

1 minute ago, Leon82 said:

That would be a lot of carbide endmills. I need to remove about 100 pounds

What's the material?  I don't think you would need as many tools as you think, unless you are cutting something reallllly crappy.

But yes tool cost per cube removed is very good with feedmills typically, and in the many applications can go toe to toe on time with solids.

I like the concept of Ron's that Colin mentioned.  Has many merits.  Mainly that you maintain a high level of MRR while using the solid carbide tool, and don't wear it out or have poor cutting conditions due to stepping up and not utilizing the entire flute length.  The rest roughing using the feedmill will be a little slow MRR wise, but will be very cost effective and super high level of process security regardless if it is a solid feedmill or inserted.  Using solid feedmills would then give you very favorable l/d ratios to reach pretty deep with small diameter cutters that normal endmills might complain doing.

That's been a question of mine for ages.  I'm curious to see what response you get.  We suggest defining them as a bull mill using the "programming" radius in the book for our tools.  I have never felt comfortable with that though.  I have always been curious.  Me personally I have used a custom definition 99% of the time.  But I don't recall if I was using optirough or not.  I do know that using the custom definition, stock models and verify take way too much horsepower to calculate in comparison to using a bull mill or stock feedmill definition.

At the end of the day, as long as your tool definition creates an extra stock on condition, and your stepover doesn't create islands which overrun your depth of cut capability.  It shouldn't matter much.  You aren't using feed mills for anything other than roughing right?  Chances are you need to go do some rest roughing anyway.  That said, it would be nice for the rest roughing stock to be 100% accurate within the stl creation tolerance and not have extra stock in places you don't expect it due to using a slightly different definition.

Grade 2 is very gummy.  I have only ever seen it in a turning application and it was more difficult to break chips than 6al4v.  IIRC, nearly dead sharp inserts for aluminum with high positive chip breakers worked best.  For milling I would say tools that are sharp and sharper will probably be best.  Like all S type materials, I'd say make sure to give it some chip load and you will be fine.  Other than that I have no more value to offer.

Just pop in some test code, in G18 and see what it does (air cutting without RAH attached).  It's been a decade since I have done it, looking in the manual for machining center, it is possible to just use G18 for Y axis canned cycle.

On 12/21/2021 at 3:39 PM, Aaron Eberhard said:

As it is, I have feelers out to both Western/FTS and Matrix.

Don't rule out a Kennametal Tool Boss.  So long as there is growth opportunity, they are free to the end user.  Sometimes, vendors will put them in for free and insulate the bogey on growth from the end customer.

5 hours ago, cncappsjames said:

I call those tuition payments.

I'm totally going to borrow this in the future.  This is Gold!

I'll go out there and say do exactly what has been said before.  It's all about getting the chips out.  normally this is gundrilling territory being about 40xd.  If you have through coolant you could buy a thru coolant drill and get it done with that.  But that's expensive and likely not on the table.

Good news is, likely you don't have any tolerances to worry about.  If there are, the engineer that designed it needs to be taken out back and beaten.  The function of that hole is a fluid/gas passage.  Just needs to connect the dots.  So as long as it doesn't wander too much, life is good.  

Just my 2 cents, but I would go as far as I could with Jobber and Taper length drills, then walk it over to a drill press or Bridgeport and connect the dots manually as you will be able to feel what's going on, and evacuate the chips as needed.  There is zero reason, other than if you don't have a manual machine to poke it through with, that you can't do this on a manual...

I don't see why not

Run the same tools you would for steel.  Understand the chip control will be easier when drilling and if you ever have to turn or bore it.  IIRC you can run that stuff pretty hot if you want to feeds and speeds wise.  But it's been a decade, so I've slept once or twice.  I think we were running Kennametal Harvi I endmills in that at like 800 SFM and we were burying them.  For milling at the time were running Iscar Tang mills.  Feedmills cut that stuff like a hot knife through butter, possibly better.  There are much better feedmills on the market today.