I was going to say the same thing. It all depends on your material and what you your operation requirements are. Onesy twosy parts, sure it will be fine, but in a production setting or where you are looking for the fastest and most secure process, pre-drilling will likely (but not always) be best.

Kennametal has HARVI™ III UJDE endmills with those lengths, in stock. 1.25" Dia, 6" flute length, 6 flute, 9" OAL many radius options They can be pricy, but they have never let me down, and I can't say that I have always had a fair shot for them to succeed. Regardless of what tool you end up buying, for that length, even for finishing you had better have a lot of machine and workholding behind it to run it if you plan to take that all in one depth. That's a lot of contact points!

Well I'm not down there anymore so, I'm just in an advisement role to our guys that can drop in on site. So I don't know. I guess at this point this is more for my knowledge than anything at this point. I think the simplest way would be to just time the cut and see if it makes sense. They are currently going F1.0 so it should take whatever the diameter is in minutes or double that.

Title says it all. So feeding in X (diameter), if you want say a .010 chip load @ 1000 rpm feeding in X, but are programming in G94, what would the G94 feedrate be? F10. or F20. because it is only moving half the distance? Or does the machine compensate for this? That is today's mind bender. I have a customer that is struggling with a finish bore on a half round (think fish mouthed tube). I can almost guarantee he is still underfeeding it, so he is just rubbing his inserts to death, hence why he gets whitness marks after only a few parts.

As said before you are going to need to limit it a little bit with some containment boundaries. Make a sample file if you can with just the isolated surfaces needed and post it up. We would be happy to help. Generally speaking I have always just tweaked the parameters to get what I want. Tedious and slow, but it usually gets the job done, not fully understanding what each setting does really slows down the programming process... More lately I have been using the module works pencil in the multi-axis suite, seems to work a bit better with less effort most of the time. Not knowing what your surface inputs look like. If you are simply cleaning up a sharp or filleted corner between non organic surface shapes, or something of the like, you can actually pretty easily create geometry and drive a ball endmill tangent using contour. I actually prefer that method when I can because you end up having 100% control of the end result.

Or you could look at it as the 7 P's Proper Planning and Preparation Prevent Pi$$ Poor Performance

Define your cutter as a bull nose with the same end diameter and it should get rid of the overlap. Do not use this definition for anything other than dynamic milling or facing/pocketing operations. 3d toolpath should not use this definition. For 3D you can still and likely should use a bull nose as the definition, but you will want to program with a smaller corner radius value (usually provided in the tooling manufacturer catalog) to prevent profile overcutting.

Yes you can on a Fanuc. It is well documented in the IF statement section of the fanuc 31i manual. Just keep in mind you need to nest the brackets. WHILE[[#1EQ0]OR[#2EQ1]] blah blah blah Also, if this isn't possible on other controls you can always process the or logic statement previously and bring in using a single variable test.

Machines that behave like this are a blight on any machine tool builder. All it takes is one to ruin a reputation with that customer and surrounding local customers the owner drinks coffee with, or any place the employees eventually migrate to. Poof nobody will buy and Integrex again, regardless of their actual merits. I have heard of issues with multi-function machines not holding tolerances throughout the day, but IMHO usually it was due to unreasonable expectations, or very heavy production with lots of roughing and once again having very a unreasonable expectation of small thermal growth issues, with wild swings of temperature in the machine. I'm guessing, that machine on a bad day should hold .002" without any comp changes to account for thermal drift. Many times, the average techs out in the field are not properly equipped or trained to diagnose or correct the underlying issue causing these problems, and won't call into Mazak service to get some additional help due to pride, or previous lack of help on simpler issues. I would find it very hard to believe the machine was moving that much mechanically, unless it was a large machine without influence from the thermal comp software. If it was moving that much without thermal comp helping, it likely had lack of ball screw tension on that axis, or less likely either a lack of preload in the milling spindle, or a partially broken casting somewhere in the stack up above the z axis that causes asymmetric thermal growth. Either way, very likely it was due to the major crashes the old boss had.

Honestly, I think all machines with any combination of rotary axes should have TCP, TWP and a slew of high speed lookahead / smoothing / fairing options at this point. With the way CAM systems can generate code and with the benefits that combining the technologies together can accomplish, people are leaving lots of time on the table, whether it be in run time, setup time, fixture build time, or just in scrap. It's pretty amazing that more machines aren't equipped.

I'm with James, I'd be making sure they had enough such that they are not limited in any way. IMHO, compensating full fourth would be the only real reason one would "need" to do it. But having TCP and subsequently having TWP, would certainly make the machine far more capable with less effort. A side benefit is that the programs become more portable from one size machine to another from the same builder or if they have taken the time to breakout machine specific functions into sub routines. Alternate Rotation offsets and calculation routines become a thing of the past, and now you aren't restricted to running on centerline for full fourth toolpaths. You can not probe a feature and start cutting in relation it, however it lands on the pallet, so long as mentioned before you don't have a tooling reach, travel, or fixture interference. I have always wanted full blown TCP/TWP on a horizontal. Oh the time that could have been saved over the years......

This is what I struggled with. It can be fixed, I just don't remember right now how we did it, annnnnd my emails have been purged back a year.... so don't have any record of the conversation back and forth with tech support..... Guess I need to start archiving emails.

Zeke, See attached, hopefully this gives you a leg up. Once you start using this toolpath you likely won't use dynamic with axis sub. If I get a chance in the next few weeks I will look into what I have done in the past to get axis sub working right. My memory isn't having a lightbulb moment right now on what was wrong with the sample you provided, but I have had the same troubles in the past and was able to get it working right after some quick discussions back and forth with CNC. IIRC it might have been machine definition related. Happy toolpathing. AXIS_SUB_MXPOCKET.mcam

For the un-initiated, this information is Gold. It took me a long time to realize similar numbers was were things typically run best. Sometime I will end up using a smaller tolerance, but that point spacing just runs good in most cases and isn't overwhelming to really any modern control data wise. Also with the MW swarf toolpath. .0008 vs .001 can make a world of difference in the actual toolpath result depending on the input geometry quality. I sometimes have to sit here and tweak the tolerance by .0001 or so at a time and end up with completely different toolpaths. A .0001 difference can be the difference between a super smooth swarf path or a total abortion that you won't even be willing to run in verify....

Well, out with it! Don't let my perceptions muddy the waters if it isn't actually what is happening out there.

Well unfortunately, they only know what they knew when they hired in, or what some guy they had working for them decided to educate them on. I'm not aware of much for formal training out there that isn't just bare bones basics of the machines. When it comes to advanced features, the only way the AE's learn about them is if they get a project that they need to use those features on. Which in lay the problem, who knew to spec that option in the first place unless it was standard on the machine in the first place....... As for what mode to run your machine in. You will either have to do some testing, or see if you can get a hold of a Mori Apps engineer and see if they can get you some Mori specific documentation on the option. I would guess you would want to use G05.1 Q1 R#, but you might also have other options like nano-smoothing which would add another wrinkle into things but would be well worth using in many cases. James would be the pro around here with those options. However Matsuura has custom codes to turn all that on and off such that it is done properly and in the right order, fully utilizing the capability of the control and machine. IIRC/IMHO, Matsuura is the only builder using Fanuc controls that does this particularly well and consistently across their product line. I'm guessing if it isn't happening already the other builders are starting to catch up on this, but I haven't been on a new machine in a while so I wouldn't know firsthand.

yup, just replace the models with the like components from you machines, adjust the travel limits and voila you have a perfectly partially functional sim you can use to check clearances, don't use it to actually verify what you think you will have for posted code as they won't match up perfectly, especially during null toolchanges, or if your post has lots of fancy pants features for getting from op to op.

I think that was just a comment based on seeing what we saw, and commenting around a generality in the MTB world where they don't train or setup their customers to their best potential right from the get go. If that was just a simple sample program that you generated and not what you actually use day to day, disregard. But on a serious note set your misc int/real defaults to a value that gets you the output you normally use, IIRC that's done in the control def, I don't think kit can be done in the op defaults but I may be wrong. I'm so guilty of not having these setup properly, like ALL THE TIME. But then again, I am hardly ever on the same machine twice inside the same year...

My pleasure Colin, near as I can see it, given that we will have reasonable geographic proximity soon, I see it coming full circle in some way shape or form, eventually.... But as always no expectations implied or otherwise. I'm always for helping out a friend in need if I can.

Use the old parallel or new raster surface toolpath's, take the shape you want and create a flat boundary surface, use boundary as containment. Then add a 2D Contour around the perimeter to clean it up. Define the tool as a small ball end. Nick

I'm glad to hear you are OK. I just can't seem to comprehend the general misfortune that seems to follow you around sometimes. I find the ways of the world to be terribly unfair at times. Especially more so to some that don't deserve the "extra attention". Makes one wonder what they have done wrong, when in reality it likely isn't even the slightest bit their fault when talking in terms of Karma.... If I had made my move north I'd be happy to come help move stuff. But I have my hands full getting ready to move from Texas back to New England in two weeks. Wish you guys the best in getting settled again. Thankfully all physical possessions are replaceable. Drunk drivers suck.

That setting has burned me so many times...... I will forever have that checked unless somehow I only do the same thing over and over and cease to care about optimizing my feeds and speeds per operation and cut width/depth. Which will likely be never...

Not a problem! Used to all the time!

You don't have to for a standard one axis G68 coordinate rotation, at least I don't recall you had to with the old 16 controls anyways. On on 31i controls an use IJK to specify which axis you are going to rotate about and it will give you a format axis error if you don't. Assuming Z rotation, you would be looking at: G68 X0 Y0 Z0 I0 J0 K1 R2.

Look up info on the roll over function. Should be under parameters, 1007 and 1008.