Sorry if I'm coming across as defensive, that's just frustration speaking. I am extremely open minded - if it works I'll use it. I just like to understand the why. Defining a new box stock around the area to Optirough seems to work, and I'll happily continue to use that to get work done. I guess this program just has a very different paradigm for where rest machining is appropriate. It seems like things are more predictable when avoiding it when possible instead of reaching for it first. I suppose that's the lesson I need to take away from this. Edit: Just to be 100% clear - I'm very grateful for this forum and any help that's been offered. It's been incredibly helpful in learning Mastercam. I can't imagine having to learn this thing without being to read though all of these posts.

That's totally possible! I don't allege to have mastered this thing yet ... I have been using other CAM systems for a long time though; this is not my first rodeo :-) However ... This is not really relevant and certainly is not an error. The CAM package doesn't get to pick my machining strategy for me. Could I let it just rough that first, yep. Do I want to in this case? Nope. I actually want to face that to start with and I have a reason for that. The program is supposed to handle rest machining - it should cope with whatever I want to do first and calculate the remaining stock. It seems that if you don't manually define a new stock model for optirough type operations you DO need to use offset, or it doesn't start from the outside. If you do create a new stock model (which it seems you pretty much have to) AND separately create a larger containment boundary than the stock, then yeah, you don't need the offset. If you use a chain around the stock as the containment you NEED the offset or the tool won't leave the outer bounds where you have less stock than the tool diameter. That's what it's for, it works fine. So, what's the correct way to get the system to generate a single full allowed depth cut in one op without defining the cut as a step up? I'll tell you step down on it's own with a depth set to do the outside in one shot sure doesn't do it - it generates only outside cuts and no operations on the pocket. If you don't tell it to step up, it won't pocket the center of the part and do the outside roughing to different depths. I know what step up is intended to actually do and I agree 100% it should work the way you say - problem is that it doesn't generate what I want, which is single depth paths. If there is a better way to rough this in one op than the settings I'm using please share! . Yeah, that's exactly how it should work and how every other rest capable CAM I've used worsk. Problem is that it doesn't actually work all the time here. Sure, some operations calculate remaining stock just fine - OptiRest seems to not be one of them. Sometimes it works, sometimes it doesn't. Try generating an optirest without defining containment - you get a neat little dialog box that says: "A containment boundary is required for rest roughing." Will it auto-generate one, yeah, but if you don't put an offset on the auto-generated version it won't work in all situations because it follows the drive surfaces for the limits, not the stock. I absolutely blame the software for some of this and here's why: I made a simplified test case that worked great (same sizes, less complex internal geometry.) I messed with it a little and it STOPPED WORKING with the same settings that I started with. That's a software problem for sure. I took the same settings that work on the simplified test case and applied them to my real part with freshly created operation - and they don't work. To me that's a software problem as well (or just truly massive ignorance on my part, which is always a possibility. :-) To me rest machining means that the software should calculate the REmaining STock and target it for removal. MCx9 doesn't seem to reliably do that. I don't understand why as I know it CAN reliably create a very good stock model from a list of operations. If I create a stock model from prior ops (as opposed to defining a box stock) optirest sometimes works and sometimes doesn't. If I create a box stock, it seems to work like a champ every single time. I don't consider having to manually define a stock model each time I want to use optirough to be actual rest machining. This part is being run on a 5-axis machine as a 3+2 and I have 5 different optirest operations defined right now. Three of them turned out to require me manually defining a new stock model to get them working. Two of them didn't (and these were ops well down the list) - it just worked out of the box like it you say it's supposed to. Same exact settings, just a different plane. It SHOULD work ALL the time!

I had not tried setting the stock model to a manually defined intermediate for just "one other operation"; that works! Just a simple box stock definition and it's all good - x9 generates a gorgeous high-speed/low-drag tool path that is exactly what I wanted. It really really sucks that we have to do that - it kills the speed of a workflow and makes what should have been a really nice toolpath type more of a pain in the arse to utilize. It just acts like it doesn't know where the stock is anymore, when I damn well know that it does! I'm used to CAM that is a lot more "stock aware" between operations ... the learning curve on x9 is turning out to be kind of harsh. Thanks for the workaround, I owe you a beer!

I have seen that problem when the post processor bombs with an error. We run Vericut and it locks the posted NC file while it's got a simulation running. If I don't reset Vericut prior to posting again from MC, the post bombs out with an error since Vericut has the file locked (i.e. read-only.) After that all sorts of random crap in MC doesn't work right and it will not exit - I have to manually kill it from windows Task Manager. Basically one the post errors out, you are totally screwed and just don't know it yet. This caused me some pretty serious confusion until i figured out what was going on.

So I spent FAR too many hours at home last night jacking around with this. Pretty much every piece of advice on this in this thread and that I've dug out of the archives is correct at one point or another. It's quite an interesting problem, I think I've actually located a bug in MCx9. The behavior is very very inconsistent. I made a really simple contrived example in a brand new project with an simple box buried in a chunk of stock. I found that when I first generate an optirough for that part everything works just great - MC comes in from the outside at full depth for the profile roughing, without doing trochoidal, and helixes in for the pocket. All good. Now when I flip on rest mode (because in real life I want to clean up certain surfaces with a face mill first) things get a little odd. It no longer wants to come in from the outside, instead it ramps down into the stock around the periphery and then whips around the outside without going troichoidal. No exactly what I wanted, but still reasonable. I can understand how it go there. Now for the kicker: if I turn rest mode back OFF, something got stuck! It does not jump back to the way it was working before, but keeps the new style tool path. At this point it's really really sensitive to the containment offset number. If it gets too large, it jumps into full troichoidal, if it's small enough it stays with nice linear paths. That's the part that is really freaking me out: I expect software to exhibit reversible behavior when I flip options on and off, not store some kind of behind the scenes sticky configuration ... Hmmmm .... actually it looks like if I turn rest back OFF, then flip the type of path to "From outside", then turn rest back ON, it generates what I want. It's like they didn't fully disable the selector button even though it's turned off! Argh! :-) Very inconsistent behavior.

Unfortunately is seems that when you have rest mode turned on, the inside/outside radio buttons get grayed out. I've read on other forum threads that in rest mode, MC is supposed to always start from the outside if it has room between the containment region and the stock. They problem is it doesn't seem to be doing that ...

I apologize if the has already been beat to death elsewhere, but I've dug through the archives and I can't find any data that actually can make this work. I've used other CAM packages for years, but am fairly new to MCx9 (SP3) and I'm hoping I'm just missing something stupid here. I'm not able to get x9 to generate an opti-rest tool path that doesn't treat the outside roughing of the part like cleaning out a pocket all the way to the containment boundary. I've defined a stock model (and tried a couple different ways to do it), I've set tool containment with large offsets, and tried all sorts of options. I can't seem to get mastercam to pay attention to where the actual stock is on the outside of the part. Inside he part seems to work great, but outside, not so much. I'm currently using a containment chain around the outside of the stock, but I've tried a manually defined region well outside that as well. Exact same results: outside the part model it cleans out the entire containment region just like it's a pocket, ignoring the fact there is no stock there. Please tell me I am missing something easy here? :-) These are prototype runs, so the cycle time isn't a big deal yet, but I would really like to understand how to make this operate the way it's supposed to.