Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
#266 - Achievements, Automation, and After-Hours CNC Machining
253 Plays4 years ago
TOPICS:
- Okuma Horizontal: After Hours Machining Dream - Why We Bought a Horizontal Video - Fusion 360: Changing Tools at Master Library Level? - Renishaw Probing, Okuma Probing, and Macro Variables - Fixturing, Solving Tolerance Issues, Post Modifications - Multiple Tolerance Ranges - Metrology Tools - Update on Grimsmo's Swiss Coolant/Clarified Butter
Recommended
Transcript
Introduction to CNC Podcast
00:00:00
Speaker
Good morning. Welcome to the business of machining episode number 266. My name is John Saunders. My name is John Grimsmo. This is the podcast that talks about amazing milestones, steps, achievements, things in the world of CNC machining as it relates to our story at Saunders Machine Works and yours at Grimsmo Knives. You got something you want to talk about?
Achievements with Horizontal CNC
00:00:24
Speaker
The amount of energy, amazingness, awesomeness of the past week with the horizontal is difficult to, it's just amazing. And you know some of this because you have the Kern, and it's really about, to me, this is about automation more so than it is the fourth axis versus the fifth axis. Right. Yeah, it's about palatizing and scheduling and tools. But John, it's amazing.
00:00:54
Speaker
You know, and we're only still scratching the surface. Yeah. But you know, we did that. We put that video out and I'm really glad we did that. This video on kind of why we bought the horizontal, I filmed it spur of the moment right after I, it was a really fun
00:01:11
Speaker
I kind of felt like a chase or a game. I decided I want this machine. We got a quote from Gossager. They started giving me options in pricing, and it's tough because some things takes a day for them to go figure out if that's available, and they say, well, that would be Japan, so that would push the whole machine back.
00:01:27
Speaker
four months or six months or that's not, you know, that's, we could install it, add that later. It's not this immediate back and forth where you can just sit down and bang it out. And then the bad news was machines weren't gonna be hitting the ports until May, you know, so not even five weeks from today when we're already making chips on it. And then they found one, again, I'm very cynical. Like to this day, I don't know how they, quote unquote, found one. One explanation is it might've been a machine that was allocated to a different
00:01:57
Speaker
Okuma dealer in the United States. There's three, I think it's Gosker Hartwig and Morris, but I don't genuinely believe it was like a sales tactics because I heard from enough other people and other machine tool builders that these machines just weren't in stock.
Benefits of Horizontal CNC
00:02:10
Speaker
So they were like, hey, we actually do have one that we could
00:02:14
Speaker
earmark for you. And we bagged out the deal in like an hour. And so I recorded that video kind of saying why we bought it. And you know, the big thing was automation and multiple setups. And it was really a personal moment of pride, edge precision. He's a really good YouTube channel. And he's a accomplished
00:02:34
Speaker
guy runs a shop, I think in Louisiana doing big work, Horizontal's, Bill Turns. His comments are wonderful over the years. He's made some really nice things that reinforce what we're doing well, or he said some things like, hey, you should consider this. One of those guys I think of like a Tom Lipton or Robin Renzetti, where you very much want to listen when they offer some advice. I loved
00:02:53
Speaker
And he was like, you're buying a horizontal for the exact right reasons. It's the ability to have a huge tool matrix. It's the ability to have all these jobs set up. It's not to overproduce, but it's to run what you need when you want it. And I kind of knew all these things. But until you do it and feel it and taste it, that's where I'm at with the current. And I totally am with you. John, yes. So what I thought of was we were buying a machine and a spindle. And what I'm realizing is you can kind of make this
Optimizing Machine Scheduling
00:03:20
Speaker
You can make this argument that it's like three machines because first off, I realized we're never going to run it during the day, meaning the production work it has to do for Saunders Machine Works products is going to start at four o'clock every day. That means all of a sudden, this machine that's more capable because it's a horizontal, it has a fourth axis, it has this big tool matrix, it has six
00:03:42
Speaker
pallets that have tombstones with two, four or six or even more sides. All that work can happen outside time, evening hours. So now we have this machine totally available all day to do whatever we want. Right now, I'm learning. What's great is I'm learning and I'm tweaking things without worrying about, okay, I'm going to go back and run some reversibles because I've got a meeting for two hours. We're already doing that. We're making some of the own fixtures for it.
00:04:09
Speaker
because it's just so easy to do it. We're tweaking things in the machine. It's just it's absolutely wonderful. And the way it's going to free up other spindles already, which gives us those spindles back to use for other stuff is just it's the biggest force multiplier I've ever been through. And it's so exciting. That's incredible. And it's it's hard to it's hard to wrap your head around it until you've seen it.
00:04:39
Speaker
Yeah, until you do it. Yeah. And it takes time to implement and do well as well. But yeah, it wouldn't have worked with a two palette horizontal, completely different zip code, completely different, not even the same. I'm going to have an own one, but I know already. Similarly, having the small matrix, thank God we didn't.
Tool Management Challenges
00:05:01
Speaker
I mean, or the built in 60 to 100 tool. Yeah.
00:05:05
Speaker
Yeah. Because you're just open now. You're free. You can have as many tools. You'll never remove a tool from that machine. And I know Lockwood made the joke. He's like, yeah, sometimes we'll look and there'll be like six quarter inch end mills in there because we just keep adding more and forget that they're in there. But I mean, a good tool, life management or a good tool.
00:05:24
Speaker
You know management process process. Yeah Saves for that and yeah on the current I almost never remove tools I just keep adding more and I'm very cautious and conscious about what tools go in and what tools get used for what and I have both a spreadsheet and the tool library infusion to track all this stuff and
00:05:45
Speaker
Yeah. Yeah. Your current wins because of how easily visible the tool change system is. Ours has windows, but it's around back. It's kind of dark back there. So we actually bought like an LED 110 light. I'm going to literally hack it into the, actually going to hack it into the shop light so that when the shop lights turn on, it turns on a light inside the APC. That's the one.
00:06:08
Speaker
flaw in my opinion in Fusion, I'd love to see them fix, which is there isn't a way to only use tools out of a master library. So let's say tool 60 is a quarter inch end mill. And in this case, I really care about the gauge length and the holder because of how we're using the Okuma collision avoidance system and just general fixturing. Well,
00:06:29
Speaker
that tool 60 can get pulled into four, five, 20 different part programs, some maybe one-off fixtures, some maybe repeat production. Well, if I want to change that tool at a master level, there's no way to do that right now. Correct. And that's not okay. Yep. Because it speaks to the same issue. How do you keep a clean-ish library? Really, we're
00:06:52
Speaker
or just doing it because of tribal knowledge amongst me, and then soon to be the few people that will be involved in that machine. Yeah, what I do is I have, I kind of, I set up the tool, I set up the gauge length, I kind of guess at my speeds and feeds, or sometimes I'll fine tune them over time, but then I just kind of leave it. And I have multiple programs pulling from that same master tool library, like at my current
00:07:15
Speaker
But how? Tool library. But there's not feedback. So it pulls from the current tool library once, and now it's in each thing. And if I want to change the speeds and feeds of that tool, I have to open up both the library tool and each individual file if I care that much. I'm being nipping here. It doesn't pull. It duplicates or copies. That's the issue. I want it to stay linked. Yeah. Pull would be sweet.
00:07:42
Speaker
If you're not careful, it could negatively affect a whole bunch of programs that you thought were good. If you take your quarter inch end mill and you shrink up the gauge length, it might not clear on that one clamp that needs that gauge length, so you'd have to be very careful.
00:07:59
Speaker
You're right, but that doesn't mean it's still not the right approach. It's similar to what they've done. I actually didn't catch this at first.
Probing and Verification Macros
00:08:08
Speaker
They've done a wonderful change on every operation with speeds and feeds where the operation like a 2D contour says, is your speeds and feeds the custom or default? Because you know how if you go change a retract feed rate to 20 inches a minute from 12?
00:08:25
Speaker
It now changes it to the default and you can, or custom, and you can realize, no, no, no, I want to change all that for all those tools. And it kind of behaves like that within one file, which is great. You want that for a tool. So that way if tools, if tool 60, you open it in a part file.
00:08:40
Speaker
it could sort of say, hey, Tool60 is on a custom setting. Do you want to refer back to the default, the default being the master library? The new default or something, or you open the file after two months and it says, oh, these tools have changed in the master library. Do you want to update the speeds and feeds or something? Yeah, I understand that it would be a tricky shift in the way Fusion works and the way programmers use it, but it would still be sweet.
00:09:08
Speaker
Yeah. Yeah. And in your example, you're right, it's risky, but you'd actually crash either way. Yeah, probably. Like if you shorten up a gauge length, you don't realize like it's, you'd be better off shortening up and having the software be say, hey, say there's a collision. Well, sometimes I will open up a program from a year ago and post it and run it. You know, like it worked then.
00:09:33
Speaker
And if the tool gauge length or something was changed drastically and it didn't warn me and I just posted it without doing my due diligence and simulating and like watching it, it could be a problem. We've all done it. Yeah, exactly. The other thing that's amazing is I spent
00:09:52
Speaker
I've spent more time probing and writing probing macros and verification than I have anything else on that machine. What's funny about that is it has the same probe as all of your other machines, similar macro structure, I guess, to all of the other machines. But how are you utilizing it so much differently now than on a VM3?
Innovating Fixture Design
00:10:14
Speaker
So not actually really true. It's the same Renishaw, except there's Renishaw probing and there's Okuma probing. What?
00:10:21
Speaker
And.
00:10:24
Speaker
So that's one difference. The other difference is the way we're building our fixtures, which also speaks to the fact that we have a fourth axis. So we're now able to rotate apart and we built fixtures that have pockets cut away so that we can pivot and reach an otherwise occluded feature, like our jaws. This is the best thing ever. We used to make our mod vice steel top jaws in three ops on a vertical and four at a time. And the op three was a pain in the butt, but it was necessary to get
00:10:54
Speaker
like we really care about this half inch nominal dimension, both nominal and parallelism. And we got it to where we'd mic them and hold, I don't know what the tolerance is on the sheet offhand, but legitimately within for sure plus or minus three tenths, probably closer than that. And there were a couple of tricks we figured out to how to get that to be repeatable, but it still required this op three and it was limiting and it was honestly a pain in the butt to set it up.
00:11:23
Speaker
Now we do it in two ops, 16 at a time. We're probing each one before and adjusting as necessary for minor variances. The parts I pulled off have all been within one tenth parallelism.
00:11:39
Speaker
Nominal can actually vary a smidge. I got to work on that like a smidge meaning one or two tenths, but two or three tenths. But the parallelism has been perfect kind of as you would expect because of the way we're doing it. But like it's different to write a macro in theory and then to put it into practice. And I actually this morning I did. So I'm on like cloud time. This has been awesome. This morning I fixed a fixture. We realized I was getting some movement and it bothered me like big time bothered me.
00:12:05
Speaker
Why are these you know, why is a chamfer moving a little wise is pro part moving a little bit? What's going on? And I realized we're using uni forces pushing the part up against a rail and the rail was secured down with I don't know eight or nine cap screws But the cap screws were on the smaller side never even thought about this when I was designing the fixture and in the
00:12:29
Speaker
Uniforces plus the part we're deforming the bar, we're flexing the bolt on rail. So I'm remachining that this morning. It just finished. I didn't remount it because I had no reason to rush. That should fix it because I know I could exactly see exactly what happened.
00:12:48
Speaker
Yes. You put an indicator on it, flex it in there, and you get, oh, wow. Okay. Got it. That's awesome. Yeah, it's been great. There's so many little things I still don't know how to do. I don't know the M code for the wash down system, like the internal enclosure. Yeah, the cattle curtains kind of thing. Yes, thank you. Yeah. And the last thing I made a note of is I've been shocked. I've made eight
00:13:12
Speaker
like actual legitimate post mods already, which I kind of didn't expect to have to do. And I love it. It's fun. This is a fusion default Okuma post. It is. I was surprised that the fusion default Okuma post had
00:13:31
Speaker
It appeared to have built-in provisions for rotaries because you had the drop-down options, but they were effectively written but not developed. Regardless of what you chose, it didn't change anything. That was the initial thing I had to figure out. That was the only real post-flaw, if you will.
00:13:52
Speaker
The second one would be that it wasn't rewriting the rotary axis,
Post Modifications in Fusion
00:13:57
Speaker
in this case B. It wasn't reposting a B0 when you, or B whatever, when you changed coordinate systems, which is obviously a fatal flaw. And so that one was not that hard to figure out. Although I felt a little bit like first time I modified a post where I thought, I want to make sure I'm not stepping over my knowledge limit here because that could be a
00:14:22
Speaker
crash type of situation. So I did it. I checked it. I always write comments to see where my post mod went through. And then I shot it over to Lawrence to be like, Hey, this, I did this. This is what happened. This is how I fixed it. This is what the results are. Uh, you, you good with how I did it. And he basically confirmed that made me feel good. So thank you, Lawrence.
00:14:39
Speaker
When you say you make comments, you obviously make comments in the post itself, in the post file. Do you comment things that end up in the G code? Absolutely. In this case, it's reposting a B0 move after it does that, it posts a comment, JWS made this B0 move explicitly. That way, every time it shows up for now, it's a reminder that it's my post hack. I love it.
00:15:04
Speaker
I can delete it later or if Fusion revs or develops the Kuma post and I can merge mine, I'll delete my mod there. Yeah, that's awesome. Yeah, I did that on the current too. I wanted on every tool change, I want the X and Y to park in the corner, the far corner and I want the
00:15:23
Speaker
Sometimes the B and C home, sometimes they stay down. But definitely on purpose, I'm putting the X and Y to the corner because that's out of the way of the B and C and the big tombstone and everything. And so I watched videos on Instagram of other people, other currents doing tool changes. And I'm like, that looks weird. Oh, because they're not going to the corner. They're just tool changing.
00:15:44
Speaker
from the operation in midair. Yeah. And I'm like, OK, I think it's faster, but I like my way better because it's safer. And it gives me that like mental checkpoint of like, oh, it's doing a tool change. It's going over there. Yes, everything is right in the world. You know, that's the other amazing thing is I feel no pressure to push this machine because again, we've got enough.
00:16:09
Speaker
capacity on that machine. It's going to be so productive running at night that I don't know the exact time, but there's no way we're even close to having it run from 5 p.m. or 4 p.m. until 7 a.m. It's going to finish halfway through that. It has really- It's like 12 hours of free time right there.
00:16:29
Speaker
Oh, yeah. So it means I don't need to worry about speeds and feeds as much yet, or cranking it. And I can go slower to prove out my fixtures. I care about tool changes, because I always think of tool changes as costing money, like ultimately, you know, you don't want to change a tool every time. But it also means I'm not worried yet about collapsing. Fusion patterning makes me a little bit nervous. Like I don't want to face mill on B
00:16:52
Speaker
like what is it, combine tools to save operations setting. I don't need to do that yet because it's fine if I do eight parts, finish them and then use the same tool later on eight parts on the other side. I don't care about that.
00:17:05
Speaker
And those are the things you fine tune as you go. For now, as you're doing, do it safe, do it smart, set up good practices, good tool library. Figure out the things that will last for the next five years of operating this machine that now is the time to get in the habit and the good habits of doing those things. And you can make it faster and sexier and cooler over time. When it finishes machining the top jaws,
00:17:30
Speaker
it then has the option of going back in, re-measuring every one, but we don't have to re-measure the backside because I already measured it. So I'm pulling that part of the variable value back in. I only have to measure the top side. Then I'm doing the simple math of computing subject to a tolerance range. Then if any of those 16
00:17:53
Speaker
are outside of the tolerance, it throws up a information only message to the operator to go look through the table of all the values. You can actually say which one was out, but I don't really care about that because if more than one was out, it gets noisy. It's more just like, look, if they all pass, there's no message. If any one of them failed, you just open up a common variables table and scroll through to see what happened. Is it pausing the machine at this point waiting for you to look at that or is it a keep going and
00:18:19
Speaker
Right now, we've chosen not to. We can change that later. Yeah, I've got a few of those pop-up warnings on the Kern. This tool is coming up one of the grinding wheels. It's going to expire very soon, so replace it today sometime, things like that. But that's awesome.
00:18:40
Speaker
The other thing I'm like, I'm on a high right now. The other thing is you it's easier to program out of fusion, I guess any cam, but out of fusion because I don't have to post.
00:18:52
Speaker
every setup or cam operation under one file. I can post op1 as 1001.nc, op2 as 1001.nc, op3 as 1003.nc. In our case, op3 does have some hand code in it because of the way we're doing Okuma probing. I don't want to, but it doesn't ever change. That's the nice thing. Op3 is good to go forever.
00:19:17
Speaker
You can schedule programs so easily in Akuma? Yes. Yeah, exactly. Yes. John's is amazing. Oh my gosh. Yeah, that's what I do too. Okay, my question is, how are you implementing probing and operations? Is it all with Infusion? Is it all posted cleanly? Or do you have handwritten probing files that you schedule in between ops?
00:19:39
Speaker
The only probing thing I've done yet is the probing on the top jaws. And what we did was wrote a sub-program that is straight Okuma probing, not Renishaw. So we're not using the 9863s, 9811s, all of that. We're using the Okuma version. And then in my, and so that program,
00:20:02
Speaker
is a sub that is, I call it alias, they call it something else, but is now aliased to a gcode. So when we call g11x, I forget what it is, we then also pass through x, y, z, h, and p values that we made up is what those mean. And then it runs that subroutine pursuant to those values we passed through. So to probe
00:20:25
Speaker
and update 16 parts, that same line is updated 16 times with minor variances in those values, parameters you're passing through, but then that program is done. I don't envision ever having to change that over the coming years on our machine. Yep. That's awesome. Yeah, I've created some probing subroutines that are very flexible.
00:20:46
Speaker
So you change a couple of variables and it changes the whole way it works. Like you can probe the inside of a bore or the outside of a bore or, you know, a square or whatever. And I just, it's a massive handwritten file with kind of all of the probing options and you tweak two variables at the top of it and you can do anything. Yes.
00:21:03
Speaker
I did post, because the next program we're going to move over has a fusion probing in it. And I posted it the other night from home, just playing around with the stock fusion Okuma post, or the one that we've modified, but we haven't modified any of the probing stuff. And it looks like it works. It's posting the Renishaw equivalents, but it was
00:21:26
Speaker
appear to post them correctly, but I will need to get that set up on the machine and test it to confirm.
Learning the New CNC Machine
00:21:33
Speaker
If it doesn't work, I'll be able to rewrite it. Of course, but yeah, that's when you put your surgeon gloves on and say, I'm testing probing, leave me alone. Yes. Oh, it's been great though. That's fantastic.
00:21:45
Speaker
How has it, um, so the training guys are gone. They've been gone for a week. Yep. So you've had the keys, you know? Yep. Um, has anybody else driven the machine? No. Okay. Um, my real question is how is your, you training and you playing integrating with the rest of your work life? Like you just spending most of your time doing that and
00:22:11
Speaker
Yes. That's a good question. You're going to bring me off my high now if I'm being honest, but I knew rather than trying to do two things as well as I think I can, which is usually not so well, I became a surgeon.
00:22:31
Speaker
And so all of last week I was in training. I took a couple of calls. We'd break for 20 minutes and I'd do a couple of emails, that type of thing. But I was present. I was there. And that's hard because, well, I was excited, but there are other things to do. Look, it wasn't that bad. I shouldn't say that. We've done a legitimate good job at distributing stuff within the team and the show goes on. Nevertheless, things fall into my lap. For sure.
00:22:58
Speaker
And then, but what's great now is, um, yesterday today or Monday, Tuesday, a little bit. I, um, I'm still a surgeon. I'm still focused on the Akuma, but now it's running a 20 minute program and I can go do whatever I want. And so it started to transition back, but.
00:23:15
Speaker
Humble unsolicited advice to anybody listening is, I used to think there was more merit to multitasking, jumping around.
Focus vs Multitasking
00:23:23
Speaker
While Fusion was generating a toolpath, I'd alt-tab and check something else in an email, because I thought that was more, I thought I could get more stuff done. And I've really, I think I've talked about this before, I've done a hard stop on all that. I'm just like, I'm here, I'm present. And if something's doing something for 20, if it's probing for 20 seconds or tool changing or whatever, I just, I'm present.
00:23:42
Speaker
That's fantastic. Yeah, I'm trying to be very similar. I'm trying to notice the effort of task switching.
00:23:50
Speaker
in your brain, when you're trying to jump between two different things, it just doesn't work for me. That said, I'm not going to watch a cam generation op calculate for three minutes or something. So I try to fill in some of those two-minute time periods. Sometimes I'll just jump on Instagram for catch up on what my buddies are doing for something like that. But as opposed to trying to be present somewhere else for a very, very short period of time, I'm not
00:24:17
Speaker
good at that. So to be hyper analytical, yeah. I think, honestly, a bathroom break, stretch, Instagram, all that stuff, good to go. I think what I don't want to do is do something where I want to bring my best foot forward, which honestly can even be as simple as answering a sales question or checking off a to-do listing. If you don't want to be someone that's high-stressed and flustered, then don't. It's okay.
00:24:46
Speaker
there's different periods of time where I'm okay answering an email within two minutes. Like I can bang that out, you know, like being productive in a very short period of time. But you know, it's case by case, be aware. I haven't worked on the weekends in any material amount of time in honestly, a long time. And it's great. I love it as as, you know, family and I got other things to enjoy life. That's okay. You know, I think that's that's age old question of like,
00:25:11
Speaker
You know, should you be grinding, grinding, grinding, or is success mean you don't have to work on the weekends? You know, I don't know the answer to that because I still, I resonate a lot with hard work.
00:25:21
Speaker
Well, I was raised that way and I still like that and on people I look up to are still that way So I think that's what it all comes back to anyway, I worked and I appreciate my wife She was like you go do what you need to do. I worked from like 7 30 until 8 p.m 7 a.m. Or so till 7 p.m Saturday and Sunday I couldn't like six it was
00:25:45
Speaker
Awesome. Alone in the shop, quiet, cranking. The training guy left Thursday afternoon. I actually wanted him to leave a little early because I wanted me to struggle on my own. And now I could still call him or text him if I got in a pinch. I only had to call him on one thing. And so Thursday, Friday was like training wheels, steep learning curve. Saturday, Sunday, coming into Monday, we're going. Yes.
00:26:13
Speaker
That's it's so good to be able to spend that time. Yeah, I've definitely come in on Saturdays and all right. I got a couple hours. I got two things on my to do list and let's just.
00:26:22
Speaker
Focus. Get it done. There's so many other things you could be doing. You could be doing emails. You could be ordering stuff. No, no, those are those are midweek things. I'm here for a reason, you know. Yeah. Yeah, I did that. I closed like fresh desks. I didn't close my email, but like on the week, I was like, it's OK. You don't know what needs an answer on a Saturday. And I'm getting much more comfortable at leaving things for a little bit.
00:26:48
Speaker
I used to feel so guilty about doing that all of the time. But I'm like, no, I'm here to focus. I've got my critical tasks. I'm going to do each one as necessary.
00:26:59
Speaker
Julie came up with a new scheduling parameter, if you will, which is awesome. It's awesome because the outcome is awesome and it's awesome because I had nothing to do with it. And it's a little bit counter to my own wishes in a weird way. We quote a lead time on stuff on our website that's out of stock. I think it's seven to 10 business days, which is, you know, it's a decent lead time. I don't love it. I'll put it that way.
New Scheduling Strategy
00:27:24
Speaker
But what we used to do was kind of reprioritize that to massively exceed expectations. And look, of course, we still want to exceed expectations. But the reality is we've told somebody knew it was out of stock before they purchased it. They knew the lead time. So what Julie has now done is changed the scheduling. It's not an algorithm. That's an exaggeration. But it's just kind of like formula.
00:27:47
Speaker
of how to work things in. So now the team knows what needs made and when, which means we can continue running other stuff, which is so weird to keep working on inventory when you have a sale item out such that the sold item still gets made, finished, QC'd and shipped within the standard lead time and ahead of time, hopefully. You know what I mean? But it's great because the truth is we don't need to be mucking up production
00:28:17
Speaker
to make something on day two, and we told them it's actually three times that long, if you know what I mean. Meaning it's okay to, kind of what you said, it's okay to leave things. It's like, years ago I was talking, when I got my Nakamura lathe, and the training guy was in, kind of teach me about it, and he used to be a machinist at a production shop, and I was talking about tolerances, and I was like, yeah, I want this to be one eight zero zero.
00:28:46
Speaker
You know, 180 chow. Bang on. And he's like, well, what's your tolerance range? And I'm like, I want it to be bang on. There is none, yeah. And he's like, well, he told me something I didn't really know, being a self-taught machinist. He's like, when you have a tolerance range, like plus 1, minus 1, he goes, as a machinist, I will use that whole tolerance range as necessary.
Tolerance Ranges in Manufacturing
00:29:04
Speaker
Yeah, yeah. And I'm like, that feels so weird, because I want 1,800. Yeah. But it's the same with time and with scheduling and with
00:29:16
Speaker
planning, there's a tolerance range. And maybe it's not a big thing to use that tolerance range. It's a good question. As a non-engineer and quote unquote self-taught machinist, I'd love to hear input from listeners on this. But I've kind of wondered if it would ever be considered acceptable. Because it kind of feels taboo, but I love it, of having two tolerance ranges. So take something that's
00:29:46
Speaker
half an inch nominal. Say the tolerance range is minus two tenths plus three tenths or something, but call that the target range, meaning a statistical significant number of parts in an R that range, but then have a secondary range that say plus or minus four tenths of like, look, that's the outer limit where it would change between
00:30:10
Speaker
Rejection and like we won't ship it or you have the right to return it if we goofed on something because the reality is Sometimes in fixture plates are also a good example of this No joke, honestly, we don't really have this issue But we have in the past sometimes had to make a decision Do we rework a plate do we kind of ignore it where we'll have a plate that's plus or minus three tenths across 30 inches but then the corner inch of
00:30:38
Speaker
is seven tenths out completely, for all intents and purposes, completely irrelevant. How do you handle that? And I know how you handle it because ultimately you handle it kind of at the business level of like, we still stand behind it, we make a note, it's on a QC sheet, but like, do you see what I'm saying? Like having kind of the nominal target, the target range, but then the outside limit. Yeah. And it's not that you're selling two products with two different tolerance ranges.
00:31:07
Speaker
Because some companies will do that. Like when you buy gauge bin, there's XX and ZZ and tighter tolerances. And I've heard that when Iskar makes their collets, like an ER 16 collet, they have their UP line, their ultimate precision line or something like that, which as far as I'm understanding is they're all made the same way, but these ones are just inspected way, way, way, way, way better. So they're like proven to be perfect or rejected otherwise.
00:31:37
Speaker
I see what you mean though. I guess what you're saying, that could end up being kind of the two different lines of product. I guess what I'm saying is basically, let's use bigger numbers for the sake of it. Plus or minus one thou is our published tolerance range period. But the reality is 80% of our products are actually plus or minus four tenths.
00:31:57
Speaker
Yeah, it's like plus or minus one thou is our tolerance range. If something happens, we'll still pass it within that range, but shoot for nominal. That's what I tell my guys. We'll shoot for nominal or shoot for a much narrower range. Yeah, whatever the middle is, the goal. Yeah.
00:32:15
Speaker
Yeah, and I guess we do that here, you know, if it's nearing the upper end of the tolerance, we're like, I'm not happy about it. But I know it'll work. Do better, you know, like, yeah, it's a good example, though, of like the difference between a machinist who's working within the tolerance range, totally acceptable, totally fine. If it's upper or lower, it doesn't matter. But I want my guys to feel like it should be this, you know,
00:32:41
Speaker
That's right. That's to me like I've always felt what does it mean to be made in Germany or made in Japan or made in the USA or Canada, you know And sometimes it's like looks it might be the same piece of cast iron in the machine But it's like the culture the decisions the preferences.
Metrology and Future Tools
00:32:56
Speaker
Yep
00:32:57
Speaker
Prints are there for a reason. You can take up that tolerance, but as a manufacturer where you own the process, it's nice to know, no, no, no, we're hitting this actually. Well, it's called full Grimsmo for a reason. And you and I are both in industries, you probably more so, where our customers have micrometers.
00:33:18
Speaker
Yeah, not all of my customers, but many of my customers are machinists or industry people or whatever. And I've seen people post on Instagram of them like, like, you know, the mid to toy, the super duper mic, digital mic that measures to like five millionths of an inch or whatever. It's like a thousand dollars or more. Somebody posted a picture with that in their hand and measuring a saga. And they're like, not bad. Not bad. We.
00:33:42
Speaker
were getting weird measurements on a block or a part. And I was like, this is weird. We were kind of starting to test and we ended up going down this big rabbit hole of getting values that were different between probing and between a Mitsutoyu quantum mic. So it's not the one you mentioned, but it's a... Yeah, yeah. I've got a quantum mic. I love it. Love it too. I could use the other ones and I'm like, no, these are gross.
00:34:08
Speaker
Yeah, love the quantum mic. PSA as a metrology novice, mics and lots of tools aren't as accurate as you think across their full range. So if you're trying to measure, there's a range of variants on the screw itself of the mics. If you're trying to measure
00:34:25
Speaker
a one inch part with a zero to one mic, you really need to zero it off of a one inch master, not zero it off the ample shut and open one inch because there's a significant, yeah, there could be a pretty significant amount of range there relative to what the
00:34:40
Speaker
mic itself can do if you zero it off of that. So we have a Sarah block, which is the ceramic gauge block, we have a deltronics gauge pins. And I really like that workflow and I trust that process a lot. But then we also have our mid to toy you digital height gauge, which is a five figure
00:34:58
Speaker
you know, poor man, CMM, like it's no joke. Um, and we went down this rabbit hole of trying to figure out what was what and how we got it. And then it ultimately, it was kind of funny cause I feel, I feel like for the first time ever now I'm being, just being a copycat, but I truly, I trust that it's the most wholesome source of desire ever. I'm like, we need to get a CMM. Yup.
00:35:22
Speaker
We don't yet. We've got some ways of going more with samples, but it's in your future. Yeah, absolutely. It'll be good for you. It'll be really good for you guys when you're ready for it. Yeah, I keep thinking more and more about ours. I'm so excited for it to come in. What's the latest? Fine tuning on financing. I've got a phone call today, just nailing that down. A part of me is sick of waiting and is wondering if I should just pay it off and then deal with financing, and financing can pay me back.
00:35:52
Speaker
You know, yeah. Well, they, are they willing to do that? I don't, I don't know. Yeah. But, um, yeah, I'll talk to financing guys today, but, uh, and then I also think I should buy a speedio like right now. Yeah, done. Yeah, exactly. So I'm gonna talk to financing about that today as well. Yeah. Um, and then on top of that, I'm also thinking hard about a second Swiss.
00:36:19
Speaker
Well, hold on. Can we interrupt you? I've had more than one demand for an update on the oil info. In the first Swiss? The only Swiss. Yeah, exactly. Still foaming. Still a problem. Yeah, we replaced all of the hoses going into it. We took the transfer pump apart. We've talked to L&S so much we don't want to talk to them anymore.
00:36:44
Speaker
Blazer took a sample of the oil and analyzed it, and they said there's 11% whey oil in this oil, but the oil that's left, the 89% is good, like Chex Good, Blazer oil. So the whey oil that's in there, it could be transferred from the bar feeder carryover. It does seem like a lot, and this is cool, it's like a month old. So I don't know about that, but basically they said the oil should be fine.
Swiss CNC Oil Foaming Challenges
00:37:13
Speaker
It doesn't have a lot of air in it, and it's chemically fine, normal. But what it's supposed to be, it's not like a bad batch or something. So my next call is our dealer, Elliott Matsura, where we bought it from. I emailed them last night, and I was like, I need somebody to come here, and who knows what they're doing? Who knows hydraulic systems inside and out? Because we need to sit down with somebody and tell them everything we've done. They need to look at it. I need the guy that's going to be like, well, there's your problem.
00:37:41
Speaker
Right. Someone must have seen this. Exactly. So I'm there now. I talked to them last night. I'll get one of their guys in hopefully this week and figure it out. No, that said, we are making parts. The machine's running, just not at full capacity. We're editing our programs to not use the high pressure.
00:37:59
Speaker
or only use it for like a drilling op or something like that. Right. Whereas before I would use it for 90 percent of everything because it's amazing. And we never had this foaming before. So it's it's frustrating, but we are still making parts just at a slower capacity, which is slowing down our production. And it's really we're really starting to feel it.
00:38:18
Speaker
And Pierre will spend a week, especially in the beginning of this, not making parts because he's digging into the whole system. He's got it all apart. So it's frustrating. So I don't know if it's the oil or the pressure system itself or something. I watched it from a distance. He was trying the high pressure. And it's aerating the oil. It's like foam coming out.
00:38:42
Speaker
Yeah. So either it's sucking in air on the suction side of the pump. Um, I can't see there being a leak in the pressure side. Cause otherwise it would leak out, not just suck in air. Okay. That's our theory. Um, so maybe it's, there's a cracked fitting that we've, we've looked at them all. We've tightened everything, but maybe there's somewhere that it's sucking in air. You got to fall. I mean, I don't know how hard it was. I mean, Occam's razor.
00:39:10
Speaker
to me says, like you said, crack or something, which means empty the system and fog it. Or it's actually going to be cheaper just to buy another thing of oil from Blosser versus a service tech. And the thing is, oil doesn't really go bad. So if it's not the oil, you just leave it for the next time you need it, or the next Swiss or something. But even if a chemical lab tells me, we're good, I would still swap the oil. Yeah.
00:39:37
Speaker
a whole drum that wasn't open, but I think Pierre put about a third of it into the tank, so we only have two-thirds of the drum right now, which by itself is not enough for a clean fill, so we'd need another... Does it have to be... Well, first of all, I would use a different barrel.
00:39:52
Speaker
What's at $2,000 for a barrel? $1,200 or so. Yeah. I mean, I don't know what your service cost, but it's probably going to be that much just to have somebody come out and start taking tools out of their toolbox. And again, it's $2,500 that you're just spending sooner than you want, because it won't go bad, realistically. Yeah, assuming it's, if it's the oil, then we're talking about a batch to batch difference. It's the same brand, same model that we were using here for.
00:40:20
Speaker
You ran it for years too, right? Yes. With probably two fills of oil, maybe the same batch.
00:40:31
Speaker
And we did an oil swap a month ago with a full tank clean out, maybe something up bumped or broken or something. Which is when it started. Which is when it started. Yeah, John, totally. Look, what we all know is true across the world is from microchips to green paint, there's been huge stress on quirky parts of supply chains, which tells you
00:40:54
Speaker
The reality is it could very likely be the case that a substitute was made, or maybe somebody wasn't even made aware of the fact that a downstream ingredient was changed or compromised. It changed the oil. You would have to make sure it's from a different batch. On the label of the drum, it says the batch number and the date to manufacture.
00:41:15
Speaker
Didn't you keep the old stuff or no? We don't have... No, we don't. Got it. We did put some quality chem stuff in a while ago and it stunk so bad though. It was just disgusting smelling. I thought the shock was on fire when I smelled it. Yeah, that's what I'm wondering right, John. Maybe we do. You don't care about, you could take it right back out. But that answers your question.
00:41:44
Speaker
I'll talk to the guys about it. So if you run it without the, like you were saying, where you're not using the high pressure as much, is it still able to be run unattended lights out?
00:41:57
Speaker
Are you nervous all night? Not so much nervous. Just if it does start to foam, then the high pressure pump will alarm out. And that just beeps all night and does nothing. So I think some of the ops piers run into the night a bunch of hours. But I mean, we used to run, and it would still be running in the morning when we come in. So we need to get back to there.
00:42:19
Speaker
Yeah, having just come out of this week of making all these post-mods part of me is like, oh man, it would be so easy to make a post-mod that puts a 10 seconds dwell between each operation. That's what he did. Yep. Okay. It was a minute dwell, which helped. Sure. It's way slower, but who cares? It's still going to run that marathon overnight. Exactly. Yeah. He did that about a week or two ago. And yeah, that helps. Is it variable pressure?
00:42:43
Speaker
It's supposed to be able to, but it's not. I forget if it's 1500 or 1800 PSI out the nozzle. That part of it's never really worked, I guess. The variability? Yeah. I don't know how to make it work.
00:42:59
Speaker
I think you can turn it down and make it 500 PSI, but yeah. What if it was actually always running at 800? You didn't know it, and somebody how, when you change it, it bumped up to 1500, which all of a sudden would have foamed all along. Well, the gauge reads 1500. Got it. Yeah, the output gauge. Yeah, but I'm hoping to all ideas at this point.
00:43:20
Speaker
Yeah, we got to get this fixed. So I had getting a second Swiss, I mean, trying to sell that to my team right now is a little difficult because they're like, that just sounds like more problems. Oh, that's funny. But in reality, we have 14 or more parts, individual parts that we're making on this machine. We're doing setups changes all of the time. A second Swiss does have a lot of merit.
00:43:44
Speaker
in what we do. And we want to grow our production. And I want to be able to spread different components across two different machines.
Expanding Swiss CNC Capacity
00:43:50
Speaker
So that setup changes are reduced incredibly. And then you have two spindles running at the same time.
00:43:57
Speaker
Well, maybe that's a good, we should probably wrap up because it's been 45 minutes. Maybe it's a good segue to start next week talking about what role a Willman would play and whether, whether somebody should create a false demand on you that you're not allowed to buy a second Swiss until you do get either get the woman working or sell it. If it's, if it's not going to work. Yep. No, I think I'm going to spend a good portion of tomorrow, um, putting that Willman back together. Okay. Let's talk about it next week. Looking forward to that.
00:44:26
Speaker
Cool. What do you have today? Talk to the bank about financing. I don't know. Stuff. Good. Swap that oil out. Walk on the Wilhelmin. Yeah, that too. Sorry. I'm the one who gets to tell you what to do. Yeah, exactly. Encouraging. Yeah, we'll get it all fixed. What about you? What are you up to?
00:44:49
Speaker
We were remachining the fixture that was flexing. If I do that, I'll move to one of our parts. I'm going to move from making one at a time to two at a time. So shouldn't be an issue, but that's the classic do it is different than knowing how to do it. I'm going to do that as a pattern of multiple offsets. So it'll be
00:45:10
Speaker
Akuma uses G15 instead of G54, so G15H10H11. Right now, it's H15, so in Fusion, I'll just check the additional work offsets, H16, so it should be the exact same toolpath just in a new coordinate system, so it really shouldn't be an issue.
Managing Fixture Offsets
00:45:25
Speaker
I had to remind myself when we started this, just make one at a time. That'll be great. And then you can build out from there. And that has been a way to keep. We've actually been, those are customer parts. They're going out, which is great.
00:45:41
Speaker
And it meant we didn't have to do what I initially thought, which was that I was going to bring that machine online with just test parts while we kept running these parts on the other machines for a while. And it was a little bit awkward because I did feel some pressure to be like, oh man, I got to make these parts on this machine. And it went great. And those machines have already, the other machines have already been retooled to run other parts, which is just oxygen. It's awesome.
00:46:04
Speaker
That's amazing. Okay, last question. Is the fixture offset, does it tell you your B rotation and the side of the tombstone and everything? So you run the exact same program in a different offset and it'll do to the different faces of the part of the tombstone? It will or would, I'm not, yes, short answer, yes. Is that how it works pretty much?
00:46:25
Speaker
So take a four-sided tombstone, and let's say you had the exact same part at B0, B90, B180, B270. The way you would program it would be, and let's say it was offsets H10, H11, H12, H13. In the OSP, the control work coordinate systems, you would set your XYZ datums and the B datum. So the B for the first one would be Z to the B for the second one would be 90, and then the same program would work.
00:46:53
Speaker
Okay. Or you could post it from Fusion and Fusion would post B90, B180. Correct. In a weird way, I don't think we're really going to do that. I guess we would basically to answer your question, I'm doing some fourth axis work right now, but we're not using tool orientation in Fusion. Really? Because we're actually picking up the datum from the other offset.
00:47:21
Speaker
I'm sure at some point we will. A lot of ways to skin the cat. Interesting. Cool, man. Awesome. I'll see you next week. All right. Take care. Bye.