Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
#268 - The Beauty of Kinematic Structure and the Blade is the Ham!
201 Plays3 years ago
TOPICS:
- Process Capability Index (CPK) - Six Sigma? - Batch Production in the Auto Industry - 3D Printed Haas Parts Catcher & Grimsmo Knives Fixture Loader - Heat Treat Sandwich & Stacking Blades - Vendor Relationships: MAKE SURE YOU HAVE DETAILED PRINTS - Okuma Post Modifications: Almost a Catastrophe - GAME CHANGER: The "Compare To" Feature in Visual Studio Code - Shop Computers & Remote Desktop - Lights Out Production & Air Compressor Ball Valve Solution
Recommended
Transcript
Introduction to Episode 268
00:00:00
Speaker
Good morning. Welcome to the Business of Machining, episode number 268. My name is John Saunders. And my name is John Grimsmough.
00:00:08
Speaker
And this is the podcast where John and I talk about everything we do as it relates to manufacturing. And today we're going to chime in on a really cool topic.
Understanding Process Capability Index (CPK)
00:00:18
Speaker
One of the things I think we both knew a lot about without actually realizing it, which is process capability index, also known as C, abbreviated as CPK. Do you know about this? I've read about it in books.
00:00:33
Speaker
Do you want to take a stab at explaining it? Not really. Go for it. I'm probably going to do a horrible job to anybody who's actually legitimately like a manufacturing engineer and works in this realm, but it's the coolest thing because it's one of the many listener emails that we had in in regards to this idea of like, I was asking kind of a different question, which was,
00:00:55
Speaker
two sets of tolerances. Like, hey, look, we need to have this band because we're not going to reject a 40 inch fixture plate if there's a one inch in the far corner that is 1.2 thou out and the rest of it is 5 tenths out. You know what I mean?
00:01:14
Speaker
The answer is this beauty of CPK. There's a great Wikipedia article on it. Basically, one of the emails then explained it in the context of car parts and manufacturing mating components like pistons and blocks or cylinders.
00:01:37
Speaker
and this idea that they require, if they were the supplier or the vendor, but they require a certain percentage of parts to be in within the bell curve of the tolerance. So you think about a piston in a sleeve fit, that's a pretty tight, you care a lot, I guess, actually I don't know how much you care about that because a piston rings, but nevertheless, let's assume you care about it being a really tight fit.
00:02:03
Speaker
It's not that you have a plus or minus X tolerance. It's that every one you want to be, when you manufacture a thousand of them, you want 990 of them to be plus or minus one or two, but a couple of them could be a little more out, but they require their manufacturers to submit these reports. I think it's 100% inspection where they show, okay, this is the distribution of range of tolerances. I just think
00:02:30
Speaker
I don't think about Saunders and our fixture plates as much for this because it's such a one-off. We care about each one on its own and they aren't mating. I guess it's because they're not mating in the sense that your knives are, but I think about your knives.
Automotive Production Processes and 3D Printing Integration
00:02:42
Speaker
I think this is such an interesting way of looking at the batch processing of production. Then he goes on to explain this idea that then we could do very simple probabilities that show if you happen to have
00:02:56
Speaker
One of the statistical odds of having one cylinder that was within tolerance, but on the small side and one piston that was within tolerance, but happened to be on the high side is infinitesimal, even though that could exist, which is also funny because it's almost like we're acknowledging that there could be effectively a known manufacturing tolerance defect that could result in a problem, but it's so low, this is the best way to contemplate it, right?
00:03:25
Speaker
Yeah it's like on the surface you don't want to even think about accepting that possibility you know that that 0.001% chance that somebody's gonna get the big one and the small one and they're not gonna go together but the reality of production is like
00:03:41
Speaker
You got to work with what you got. Yeah, exactly. And I always admire big auto because the first vendor that we found was a laser and water jet shop in Rhode Island to help us with strike mark. And this guy, I think he actually went to MIT. He was rough around the edges and great guy, like helped us a ton.
00:04:01
Speaker
And he was kind of talking about like, man, you want to figure out how to make something and make it cheap and make it right. You know, call it Detroit, you know, they'll spend $10 million to save a nickel. And so, um, I think, I think some of that is kind of lost on my generation, our generation and even the younger generation where we, I want to take care a little bit less about cars and tinkering and you know, this EVs change things and that's a whole other conversation, but man, a lot of respect for the amount of process that goes into bulk
00:04:31
Speaker
automotive production. Yeah, it's beyond our world. You and I spend insane amounts of time dialing in our processes and our businesses and things like that. Imagine that with a billion dollar budget and 300 people to do that. Yeah, exactly. Basically, the equipment that you want, hard stop. Yes.
00:04:59
Speaker
I mean, that's, that's how we do it on a small scale. You know, I buy the equipment we need and, you know, I talk very closely with the rest of the team. Like, what do you need? What do you want? What, you know, what's going to get you to the goal? What's, uh, what's going to remove those bottlenecks and barriers so that we can do our jobs better and faster and easier. And as a business, I love that we get to spend that money and buy the toys to help us do the job better. Um, as opposed to trying to pinch every penny and be like, no, no, you know, it's out of our budget, yada, yada.
00:05:27
Speaker
Yeah, yeah. What budget? Let's work on this. It's awesome because some of the stuff just is expensive. We've been really happy with the Okuma Genos and how it's helping us with the fixture plates, but we're still buying different face mills. We're buying different wiper inserts. We're just tweaking. It's totally good, but we're tweaking.
00:05:47
Speaker
That's all frankly expensive, but then we wanted a quick QC gauge to check something kind of differently or new And so Ed just sent me a file and he's like, hey, will you print this one convenient? So I 3d printed this look like a candy bar with you know three pegs on the bottom in a hole and two through three tapped holes and I'm gonna ask just printed for him and I go over the other day and I see it and it's an indicator holder on three points that tests something and
Blade Tempering and Heat Treating Innovations
00:06:15
Speaker
doesn't matter that it's 3d printed the way it's being used. And it's just like, Oh, I love that stuff. So he asked you to 3d print something that you had no idea what it was actually for. And you're just like, Yeah, sure. Here you go. Yeah, yeah. I love printing now. I didn't I know, I hadn't printed, I don't even know two years ago, I'd never really used one. And now I'm like, I mean, I probably print
00:06:37
Speaker
So we're between five and 15 things a week. Yeah, nice. I go through phases. I don't touch the printer for a month or two. And then I go nuts. And I'm like, Oh, yeah. And then I print 20 things. Yeah, we use it all the time. We printed a new actually this one's legit.
00:06:53
Speaker
parts catcher tray for the Haas, which I will share that because that's working out. Grant did a great job with this new design. I actually need to do the video on all the 3D inserts on the Haas machines that have really helped with chip evac. And then what else was I just printing? I honestly forget, but yeah, it's been awesome. It makes me want to stand up on my, you know,
00:07:17
Speaker
podium and just scream at everybody who's running a manufacturing company that is like, oh, I thought 3D printing was a toy. No, it's not. There are so many- You should absolutely do that. Yeah. Well, so like unloading the reversible inserts off of our horizontal, the beauty of a horizontal as a kinematic structure that involves one less loop, although we can come back to that. That's not as true as I thought, but this idea that you've got better chip evacuation because the parts are
00:07:45
Speaker
Perpendicular to earth yeah around is is great for machining horrible for loading and unloading yeah, I can imagine so i've been thinking about this little uh
00:07:57
Speaker
Did you ever build model airplanes as a kid that you'd snap the plastic out of the lab of parts? I was thinking about making one of those that had just a couple of magnets on it that are rubber coated. Then when you went to unload the parts, you just draped this thing over it that held all the parts down temporarily, unscrew them all, and then you can just lift them all off at once. Nice.
00:08:22
Speaker
You could even put magnets to capture the parts as they come off, depending on the part and the size. Yeah, exactly. Nice. I like it. Yeah. Sorry, I've been talking too much here. No, it's good. Actually, I've got a design and 3D print. When we're heat treating the blades, when the blades go into the low temperature temper oven at 400 degrees,
00:08:45
Speaker
We've got these two thick sandwich plates. Imagine you're making a ham sandwich and the blade is the ham and these two pieces of bread, of A2 steel that are like half inch thick, I think, that sandwich the blade together. I think we stack eight blades together, so you got eight slices of ham in the sandwich.
00:09:04
Speaker
And then there's six bolts around the outside that tighten down. And this sandwiches the blades together and keeps them straighter through the tempering process to help them be straighter afterwards so they don't warp so much. Interesting. It seems to be working. It cuts the pre-warp in half after tempering. Oh, wow. Which is great. So if there's a thou of warp, it'll be half a thou after this process. So that's good.
00:09:31
Speaker
But we just have like stockin' head cap screws with nuts on the back, six of them. And you need an allen key and you need a wrench on the other side, like a box of wrench to tighten them. So the guys are like, what if we 3D print something with six allen keys, like literally an L-shaped allen key where the short leg is sticking up?
00:09:49
Speaker
and it's just the Allen key is captured in the print. Yes. Then the hex is sticking up and you just lay your fixture on top of it and it captures all the six bolts and then you just tighten the top. I'm like, yeah, done. Let's do that. There's no printed plastic that will last at that tempo, right? No, it's just for tightening the screws. Oh, interesting. Okay. A loading fixture.
00:10:14
Speaker
How much is it require a lot of force or are you just basically more than hand tight? I don't know. Because I was thinking about the bicycle wheel toggles, like the cam toggles, make it two of us. Good question. The heat cycling does make things weird. We have to use nickel bolts or some sort of alloy so they don't stretch over time going through all those heat cycles, something like that. Interesting.
00:10:41
Speaker
When you first were talking about this, I was thinking of your Arbor Press quench plates, but this is obviously totally different. This is just after that process, but yeah. You're stacking them eight on top of each other or they're all adjacent? Yes, like on top of each other. They're all lined up together. Interesting. The blades aren't our profile though, right? Correct.
00:11:10
Speaker
Huh. You stack them on top of each other. And there's some theoretical, like, if they're warped in different ways, are they sharing? Or if they're all coming to the left, does it help? I don't know. But it seems to be working so far.
00:11:25
Speaker
I'm asking, not challenging. Absolutely. Why not lay eight of them out like you would lay out cookies on a cookie sheet and then put the A2 sandwich above and below it? That way, each knife has touching A2 on each side that can be hardened and ground and all that. Imagine laying out eight blades. It would take up 10 by 20 inches of size and our heat treat oven is not that big.
00:11:51
Speaker
Got it. Totallyness that. I'm assuming at this point you have a walk-in heat treat oven. Yeah, I wish. I've looked into them. I actually did a deep dive on vacuum heat treat ovens about a month or two ago. And for $300,000, they're amazing. But don't you have those? Were they Paragon or Evenheats or something? Evenheats, yeah. We've got three of them. But I thought they were small enough for a 12-inch pizza. I thought they were pretty big. Barely. I think they're
00:12:20
Speaker
12 inches squared and one of them is much smaller than we have, but yeah, still. Okay. Interesting. I remember, this is actually kind of a good, funny story. When I was trying to make strike mark work and I had this idea of leaving my job, setting up a shop in Brooklyn with a tormach. I didn't own a tormach yet, but I knew I was going to buy one and we had to make these parts that were these mounts that held up the target
00:12:45
Speaker
There was one on the left, one on the right, and then a similar one in the middle that held the DC gear motor. The two side mounts had a giant hole through them. In hindsight, I don't actually know what that hole was for. I think it was probably superfluous, but it wasn't like we were really trying to reduce weight. My thought was, this is such a dumb idea to stack. These are all half-inch.
00:13:08
Speaker
thick aluminum, probably four inches by five inches profile with some features in them and so forth. My thought was instead of machining them one at a time, was to drill the hole first and then slide the hole over like a piece of pipe that would line them all up. And then on the Tormach, I could come through and do like 20
00:13:25
Speaker
top ones on the top side where I'd be drilling and tapping the holes from the top. It just doesn't work like that. You can't stack parts left and right on a CNC machine up against each other and expect to hold any tolerance and so forth. Is that the difference between an engineer and a machinist where the engineer is like, yeah, this will be great. Just do it like this. And the machinist starts laughing. Oh, I was neither at the time. Why not? Yeah, same theory. Yeah. Anyway.
00:13:52
Speaker
Interesting. So what caused that this is all driven by the lapping time? Kind of. Yeah. Like lapping being our end process for this, this thing, the flatter the blades come in, the easier the lapping is. So they're not taking out a taco into, you know, cause if you have a curve, lappings only going to cut the high point and low end points and the flatter it goes in the happier everybody is. So we're, we're trying all these theories to make the blades flatter and flatter and flatter and flatter.
00:14:21
Speaker
and trying to reduce the hand bending of them in an arbor press right before the lapping machine, because sometimes that just creates a squiggle, not so much taking out the cup, you know? So it's tricky. Your parallelism is... It's great. Perfect, perfect. Parallelism is surface ground soft, so they're probably parallel within a tenth. Yeah. So you have either a lake or a hill,
00:14:49
Speaker
you don't want it to turn into a potato chip. Correct. Yeah. Have you watched How They Make Shotgun Barrels? No. Did you mention that a while ago? I feel like I probably did because I feel like that's something I would say. How They Make Shotgun. I'm going to have to look into that. There's a YouTube video on one of the factories. I don't remember whether it's Beretta or Remington or maybe both. It's probably not going to help you because it's effectively
00:15:14
Speaker
Exactly. It sounds like what you're already doing, but they have this jig with two rollers. And then they have an indicator that quickly shows them where it's high. And then this guy with like 72 years of tribal experience of firsthand knowledge knows exactly where to put the point of the press and like push down and over flex it just a hair and like, you know, a shotgun barrel is quite pliable compared to a relatively shortened the knife. And it's the same way we do it. Yep. Yeah.
00:15:41
Speaker
except we don't have a roller, you know, with a high point on it, we have a gauge block or a it's actually an orange vice jaw from orange. But it's flat. It's really flat. And I made a 3d printed light box with LEDs in it so that we can see you put the blade on top of the the
00:16:03
Speaker
vice jaw, the flat piece of metal. And you can see a light gap in between. So you can see where the high points and low points are. And you can see when the lights getting thinner and thinner. And then that's cool presses right beside. And because we used to literally like slap it on a handheld vice jaw, hold it up to the light, right and then down again to the arbor press, bend it, hold it up to the light again.
00:16:24
Speaker
up and down a bunch of times and I started looking at that going, that's dumb. Okay, we need a light source right there. So about a year, year and a half ago, I had my buddy Martin kind of, I just gave him the idea and I was like, here's what, just here's idea. Go nuts, 3D print something. I'll pay you for it. And it's great. We got three of them around the shop now. They're fantastic. When you bend them with the arbor press, is it free? Like, is it, is the blade back stopped or is it just free bending?
00:16:53
Speaker
Um, basically free bending. I'd be curious if you could find, like, let's say there's a thou. Yeah. Like how much to bend it a hill. You probably want, I'm making this up a seven thou Lake backstop.
Detailed Service Drawings and Outsourcing Challenges
00:17:08
Speaker
That way you push up against a hard stop to do the over flex and then have it come back. I think it's too variable for that. Okay. Um, and they go different ways. Some blades go this way. Some blades go that way. Cause
00:17:20
Speaker
I think a lot happens in heat treat. If they're on the left side of the oven, they're different than if they're on the right side of the oven, maybe. The way they're quenched between those aluminum water-cooled plates, there's a lot of factors involved. Honestly, I'm outside of both of these processes. I'm just kind of fly on the wall, hearing stories and watching every now and then, trying to help.
00:17:45
Speaker
It takes some operator skill to be able to use the arbor press and not create a taco, but to create a flatter object. Yeah. Yeah, for sure. Guys are good at it, but it's tricky. You heat treat quench and then a temper.
00:18:04
Speaker
Yeah, so heat treat to 1930 degrees for a little bit, 20 minutes or something, and then quench in the water-cold plates to room temperature, and then...
00:18:18
Speaker
I think you temper once, then cryo and liquid nitrogen, then temper again. Oh, yeah. I forgot about the liquid. I guess, does the temper act as a form of stress relief? Yeah, exactly. Okay. The blades are much harder. We're shooting for 60 Rockwell, 61 something. I think right out of main heat treat, they're like 64 and brittle.
00:18:40
Speaker
So you want to stress relieve them in the tempering oven at 400 degrees to just like let everything chill out and all the molecules kind of like find their set. And the tempered temperature will affect the final hardness. So if you temper hotter or temper colder, you can change your hardness by five or 10 points. That's cool. That's awesome. So that's how you pick your, it's like you heat treat in the hot oven for the steel and you temper for the hardness, like the exact hardness you want.
00:19:08
Speaker
That's cool. We do two full temper cycles at two hours each. Oh, wow. Plus a while in cryo. And we do this every single day. So the guys have a, you know, okay, Dave's coming in early, so he turns on the oven. And then it's hot by the time sky gets in and the sky processes all the blades. And then even sometimes guys like I leave early, Eric, can you pull the blades out, you know, at exactly five o'clock? Yeah.
00:19:34
Speaker
That's awesome. It's fun. It's good. It's a good process. I'm really glad we do it ourselves, even if it does add a lot of challenges to the process, but I feel like we're all learning and getting smarter and.
00:19:45
Speaker
have control. When you send blades off to heat trees like we used to do back in the day, you have no control and they come back all discolored and sometimes warped and like, what are you going to do about it? Okay, I'll just work with them. Now we have that feedback loop instant daily of even if it sucks like, oh, the blades were really warped that time. Yeah, sorry.
00:20:08
Speaker
Dennis made a good point a while back on WhatsApp about the importance of vendor service drawings, which is probably obvious to anybody who's done this a lot, but as a self-made entrepreneur, you don't always realize if you're setting a thing out for galvanizing or anodizing or heat treating, you actually have a print that has a lot of specifics for the rules to follow. Here's where to rack it. Here's where to mask it. Here's what you can and can't do.
00:20:36
Speaker
You know, like I suppose there's probably heat treated out there that will use vacuum bags, maybe even to this, you know, don't put two knives in one vacuum bag, like all that stuff. Right. Yeah. Most of them have vacuum ovens, which is a whole different kind of process than we do theoretically better.
00:20:54
Speaker
But yeah, if you don't have specifications and a print, we found this with a lot of our vendors, they're just going to do it their normal way and not ask any questions. And like who says their normal way is acceptable to us, like nobody knows. So the problem is when you're a new,
00:21:12
Speaker
manufacturer like a new Saunders or a new Grimsmo and you're dealing with a vendor for the first time and you don't know, you don't have the experience to know like, okay, when I get water jet parts, make the tab in a good location. I as a knife maker know what a good location is. You as a water jet person have no idea, so you put it anywhere, you don't care. And then I get a bunch of parts with a tab in a really dumb location that always cuts us or causes a problem or it doesn't fit in the fixtures or something.
00:21:41
Speaker
The, uh, a long time kind of friend of the channel, a guy named Steve was up at the shop this week and saying hi. And we were actually talking about you and knife making and all that stuff. And we were all kind of wondering, why don't you have a water jet yet? Yeah, I don't want one.
00:21:56
Speaker
It's $150,000 that is extremely messy, extremely loud and extremely breaks all the time. It needs its own room and it's super messy.
Water Jet Machines: To Own or Outsource?
00:22:11
Speaker
I mean, I know there's enclosed ones and there's exceptions to this rule, but everything I've learned about it, I'm like, no, not interested. I have precision equipment. I don't need all that junk floating in the air. Even your $10 million to spend, like we're investing in Grizzly and I's, we're building you a new facility, you wouldn't want to vertically integrate that? That's a good question. Maybe.
00:22:32
Speaker
I mean, if it's a simple process that's done reliably, then yeah. Yeah, I don't know. Nothing on the water jet. The water jet doesn't touch the customer part. Like it's all that credit. I mean, credit stuff is gone by that time, so it doesn't matter, right? Yep.
00:22:47
Speaker
Yeah, waterjet is probably one of the easier, easiest to do well in our experience with our vendor right now. Got it. It's cheap, easy, and our titanium supplier also has his own waterjet. So as we're buying sheets of titanium, he cuts it up for us, which is kind of tightly integrated. It's great. So we just, more rask handles, please. And he's like, okay,
00:23:14
Speaker
I was always thinking about your blades. I guess I forgot that the- Well, we actually have two water jet suppliers. One in the States that we buy titanium from and he cuts them, and then we buy our sheets of steel from Sweden, get them shipped on a pallet here. So we're storing the sheets here. They're like two foot by three foot, I think.
00:23:38
Speaker
And then we have a water jet company down the street, literally five minutes away. That's awesome. So I load the sheets in my trunk. I'm like, okay, guys, today we're doing 15 sheets, put them in my trunk, and then I drive them over and unload them, and then a couple days later they're done.
00:23:55
Speaker
here's an outside processor thing. Some shops don't care. And this is a soft steel, even though it's a stainless steel, it's still rust like crazy if you're not careful with it. Water jet's a wet process with just plain old dirty water. And first batch of blades we got back from them, the entire box was brown rust.
00:24:14
Speaker
And like, so you cut them on the water jet machine, you take them off, you throw them in a box, maybe hose them off with your hose in the water jet machine, throw them in a box and wait for two weeks to call your person. Yeah. Sprinkle some salt in there and wait for them to do it. Exactly. So like, holy cow. So there's a lot of back and forth, especially with Angelo and them, like Angelo flipped when we got that box. This was two, three years ago.
Anodizing Challenges and Vendor Communication
00:24:39
Speaker
And so we're like, okay, we need like coming off the water jet machine, they need to be dipped in something in some sort of slushing oil that prevents this. So we worked with them, we found gallons of this stuff that we had them buy. And so now the blades come back oily, which is good because they don't rust. But now we have to remove that oil with varsol before they go into the surface grinder. And like it's, it's adding process. But yeah, I don't know.
00:25:05
Speaker
Why not build a VCI paper-coated tray that has each blade held vertically so that now you have a count of numbers of them? It's one of those things where when your vendor has a purpose-built crate, there's an implicitly more amount of care, even if it's that second shift thing or whatever. Great idea. I don't know. The oil, I don't know. It works, but it's annoying. We haven't revisited the process. We're like, just keep doing it.
00:25:34
Speaker
Yeah, get rid of the oil. Yeah, that's an interesting idea. We had a batch of... Our processes are just coming off so great. We're doing these bi-monthly runs of anodizing on a schedule with like... Nice. I think I mentioned this, right? It's going great. And then all of a sudden, I think there were 13 plates in an order and one of them was
00:25:59
Speaker
out of spec. It was undersized on the bores. Anodizing is a self-terminating process. It's not possible to add 4,000 because the anode and cathode contact is lost as the anodizing builds up because anodizing is an insulator.
00:26:18
Speaker
So it's kind of a perplexing thing. And I'm not like an anodizing expert, but the colder your tank is, or where the anode and cathodes are, or where the plate is in the tank can affect all this. But it's still super weird that out of all these batches, one is materially different. And so
00:26:36
Speaker
Honestly, that points the finger at us because it's a lot easier to miss machine something, frankly, than it is to miss anodizing. We're still talking tens here, but it's notable across all those plates that we make. Wouldn't you be able to see the growth? Because it would grow equally on all surfaces. Your thickness would be fourth out bigger, your boards, your threads, everything would be bigger or smaller, right? If it's anodizing.
00:27:02
Speaker
Yes, correct. We have our Lex QC system and we just have our own QC process with the pins and checks and so forth. It was my opinion, I was like, we didn't goof on this. It happened to be one plate that were... There were three identical plates to it, so we ran those together. It's just not the case that we're going to be out four or five tenths
00:27:24
Speaker
especially compared to those two. And so I had a good call with our vendor, and he was nice, but he was just like, look, you know, kind of a not us. He's like, we ran these together as well. They're in the same tank. We have your process, blah, blah, blah. But then, but then I was like, well, you know, we've made you a go no go gauge. It's two deltronics prints in a 3D printed case that gives you guys instructions. I was like, did your guys run the go no go?
00:27:50
Speaker
And then it was kind of a, Oh, it looks like maybe we didn't. And then, uh, and then it was awesome. Like two days. And I was like, look, well, you know, it is what it is. Um, we need you guys to make sure you run the no go at the end, because even if it's a fail, I want to know that it, I want you guys to know it failed before you go run a bunch more plates to figure out what happened.
00:28:10
Speaker
And maybe we have you guys strip it and rerun it, although that's never great. So two days later, I get this lengthy email from him and he was kind of like, Hey, I'm unpacking a whole can of worms.
00:28:23
Speaker
And it was really cool. It's a good comment about like, treat people well. And, you know, I didn't even ask him to take off the invoice. And I was just like, look, we got to figure this out together. Not that I'm like some saint, but just so you know, we're in this together. Let's figure this out. Let's work on this. And I never got ticked or upset and all that. And then so
00:28:43
Speaker
He basically, and I don't remember the details, but they had photos of it from their process cameras of all their tanks. And then he's like, so we had a second shift operator who was misfiling out reports and loading things in.
00:28:57
Speaker
the wrong way and this and that, and it wasn't running the batch, it was said it was running. It's pretty like, oh, you got to get your house order, but on the flip side, it's like, okay, great. That explains everything. That's awesome. For him, it sucks to start that process because you're like, no, we're on top of things. We're pretty good. We watch our guys.
00:29:18
Speaker
and you don't want to dig deeper, but then you start scratching the surface like it sounds like he did, and he found some more stuff
Post-Processor Modifications and Safety Concerns
00:29:24
Speaker
and more stuff. And then it starts to get interesting. Like, as long as as a business owner, you get like curious, like, I just want to find the cause and effect here and like fix it. I'm not trying to point fingers or blame everybody or fire anybody or whatever. It's just like, I just, this is getting interesting. This is getting fun. You know, like, holy cow, let's fix this. And it's such a Dale Carnegie thing about like, look,
00:29:47
Speaker
It's really hard, the human psychology, it's really hard to actually genuinely be mad at someone when they're honest. Interesting. And it's that awkward thing of even like, hey, I don't want to come to your party tonight. I don't feel well. It's like, dude, you feel fine. Like, I don't care if you don't come, but like, don't tell me the lie of that. And it's like, you can just be honest about stuff. And it doesn't mean you need to have a, you know, no, uh, what to call inner monologue. We don't have to say everything every time, but
00:30:13
Speaker
When he was digging his heels in a little bit, I'm like, well, it wasn't us. I'm trying to figure out how it could have been us. I left it alone because it could have been us. I didn't want to make him dig his heels further in. Then man, when they're honest, it's just such a great way to...
00:30:33
Speaker
keep that going. Can I share a potentially big booboo? I almost botched the horizontal. That's a big sentence there.
00:30:50
Speaker
I don't know how to say this. I was really upset on Sunday. More upset as a machinist, manufacturing guy, person who loves what I do, person who cares a lot about the inevitable public outcome. If I had really crashed it, I'm not going to not share that somehow, which would stink. In the spirit of being honest, it's exactly what folks may have thought, which is it was a post-mod by me.
00:31:20
Speaker
But not really. The reason I wanted to share it is honestly more as a PSA. To cut to the chase, I was spending Sunday morning at home trying to modify our Okuma post to handle tool load monitoring as an option property, which is going to be, I will get it figured out. I've taken a back step away from this for now.
00:31:45
Speaker
Option properties are a way to effectively use your post plus machines configuration infusion to allow every single CAM operation to have its own definable load limit. It's a new tab in each property, which is so freaking cool. The way you can program it is you can have an overriding default of some value, which is fine for most things because I don't care about tool load limits on engraving or something. The roughing and the facing and drilling and all that.
00:32:14
Speaker
I couldn't get it to work the way I wanted to. Toward the end of the morning on Sunday, I thought, okay, well, this isn't happening like I wanted to. No big deal. Let's call it a day. I reverted back to my original post. I did that because I had already forked off a copy that was sandboxed.
00:32:35
Speaker
pasted that back in and I was totally back to the original post code. And what happened, this is frustrating, is when you go to like manual and see your post that pops up a box or window, and you can choose your posts, you can choose the location, and then you can choose a bunch of post properties. Well, in the process of having been testing a bunch of different post things, it lost
00:33:05
Speaker
the settings for the Y axis and it was kind of gone and I didn't realize Y axis as the B axis rotation axis. And so even though I went back to my original post, that setting had been wiped and so the next set of code I posted out didn't have the axis rotation in it.
00:33:31
Speaker
even though, again, I had gone back. To throw myself under the bus, I shouldn't have even been modifying the post that it thought was... I honestly thought I would get it figured out, and so I didn't think about taking the post offline on a computer with no ethernet, literally totally off, because it was the process of modifying that post that caused Fusion to lose that setting somehow in the...
00:33:58
Speaker
because to get deeper on it has to do with the fact that the Kuma post has its own post built machine config and I needed to use the
00:34:07
Speaker
fusion machine config, not the post machine config to allow this to work. As the process of commenting that out and going to the fusion one and then going back to it, it broke that link. Look, it's on me at the end of the day, but anyone out there who spent a lot of time modifying your posts, you're not going through and reanalyzing every single line of code when you go back to your original. It's just not
00:34:31
Speaker
Yeah, because you trust your original and you're like, I know I just undid my changes, so I'm good to go. I have done the same thing on the Kern, like maybe not exactly, but certainly post tweaks, complete tweaks,
00:34:47
Speaker
go back to normal, assume it's fine and it's not either break a tool or yeah, I've, I've bumped the current a few times. Oh really? Yeah. Not in a while, but in the beginning for sure. And yeah, certainly the, uh, put your hands over your head and be like, Oh no, a couple of those, but yeah, just wasn't, wasn't X that was my,
00:35:12
Speaker
That was where I, it wasn't arrogance, but that's where I was just wrong was I went back to it. I posted code and I wasn't even in the like, okay, John, you're in the exploratory phase. Let's see how this goes. I was just like, oh, we're back to good normal rock and roll.
00:35:29
Speaker
Um, I'm not like running. I'm still taking it super easy on that machine, but it just caught me off guard. Yeah. No, I've done that. I've done exactly that. Yeah. Back to normal. Good to go full send. Uh, yeah. So, so what happened? What, what was the outcome? So the outcome was that I, uh,
00:35:52
Speaker
was rerunning one operation and reposted just that operation. I had jogged a tool over to look to see if I liked or the condition it was in like, cause the edges weren't all that. And then, and this is where it's, I got lucky, but it stinks. When you jog over with the door open, it's super slow. So I turned the rapids up to a hundred percent to jog it over hand.
00:36:18
Speaker
But then I always turn them back down to like 10, 15% when I run the machine. Well, a lot of times, and I've really have changed my behavior based on the Gossicker training on how I use those jog feed handles. It's great. They're wonderful. So I usually start the machine, the cycle at a slightly higher feed rate to let it start getting into position. Then I turn it way down.
00:36:41
Speaker
Yeah, just being honest, it's kind of hard to say this. It was still at 100%, and I went to hit cycle start, and it basically did a transition move, but without the B rotation. And I saw it right away, and I luckily, I didn't even hit cycle start, or E-stop, or pause, because my hand was already on the feed rate, so I just whipped it to zero, and it stopped before it crashed.
00:37:09
Speaker
Okay. Amazing, but super, super embarrassing and crummy. And I saw it. It's funny. I saw it. I knew it. I figured out what happened and then I wanted to fix it and move on and get back up on the horse. And I was so upset that I realized I just needed to go home. Like I just had to walk away from the machine. Don't worry about it. Like you got lucky, but man, it was,
00:37:38
Speaker
Tail between my legs, man. Yeah. No, I've done that a bunch of times. And like you said, sometimes I don't – I assume everything's going to be back to normal, back to fine. So I run it and I can't think of any examples. But a lot of times when I'm testing and tweaking and it rapids to the wrong spot and I've got my hand on the feed rate, I crank it down to zero, I'm like, what just happened? Why is it – holy cow, this is way wrong.
00:38:03
Speaker
And then I spend a while mad at myself not understanding what's happening. When you know the solution, that's one thing. But when you have no idea what's wrong, then that's a very confusing time. And then the whole machine is dead until you fix that, until you fix that post-mod and understand what's happening. And I'm so big on understanding what I did wrong.
00:38:29
Speaker
that I will not move forward until I know the solution. So actually, this week's video, which comes out today for you and me, but it'll be two days past for the audience, is on mods we've done to the Okuma post. And I learned from CJ. Thank you, CJ. Do you know about this in Visual Studio Code, the Compare to Files option?
00:38:55
Speaker
Oh yeah, all day. Dude, that is a game changer. What I also now realize is when I go back to my original post, I do keep backup copies of good code for some of our production files now. I could compare those two to see what's different. That would have saved my butt for sure. Yep, yep. That's cool. So watch that video. Folks are listening who want to do post stuff because woo.
00:39:20
Speaker
Yeah, it's potentially catastrophic, but also gives you so much control and so much amazingness that it's worth it. I just felt like life isn't fair. I felt like, man, that's not fair. I wasn't being reckless. I wasn't being Wild West. I wish you could have seen how judicious I was with
00:39:45
Speaker
the posts and copying things off and all the comments and going back to it and referring to it. I was in a good state.
Remote Desktop Solutions in Manufacturing
00:39:52
Speaker
That's what it felt unfair. This wasn't the whole like, oh, I forgot to measure a tool or I forgot to set a coordinate system. Man. You simply found a flaw in your process.
00:40:02
Speaker
Yeah. And that's a good thing. It's like the anodizer found a flaw in the way his staff was working on stuff. Yeah. Once you're out of the, the, um, I screwed up mentality. Okay. How do I fix this? Okay. How do we not do this again? You know, yes, you tighten up the process so that, uh, stuff like that doesn't catch you by surprise anymore. Can I change the subject? Yeah. What's this? What's up with this new computer?
00:40:29
Speaker
Um, so the Fraser, um, my marketing director is, you know, super computer nerd, and he's got, um, a side business at home doing a lot of video, um, like weddings, like he'll go film weddings. So he does a lot of photo editing, video editing, things like that. So he's been upgrading computers over the years. He built this sweet PC, but a year or two ago, like, you know, full out graphics card. Um, and then he bought a really nice Mac.
00:40:57
Speaker
Mac book pro mac pro i don't know i don't know anything about max are you about like the nice mac video editing program computer for here at work. And he's like i love mac okay they came out with this new laptop i'm gonna buy the laptop for home and sell my my my rig my pc. I overheard him talking about this to some of the other guys at work and i'm like sold.
00:41:19
Speaker
Yeah. So I bought his old PC and he put it up in my office and it's sweet. It looks like the thing that that's somebody's hand me down. It looks PC. Yeah, it's full spec. I'm looking at it right now. It's got a GeForce RTX 2080.
00:41:37
Speaker
which I think is good. Two monitors, nice tons of RAM, 64 gigs, I think. I have no idea. I just said I'll take it. I trust you. That's my desktop at work now. I've been working off of laptops for five years exclusively and I love it. But to have a home base at work will be nice.
00:42:04
Speaker
Dennis, we're trying to get that film, tour film video with him. Oh, yeah. And he does something that I've heard about, but never really seen somebody actually implement as a true workflow, which is that all every, well, I'm not sure if every like brother, but most all of their machines have a cheapo cube computer at them with monitor. Okay. And that gives them access to like their ERP, but then they use remote desktop to remote back to their
00:42:34
Speaker
beefy office computer that has real horsepower and has the licensing and stuff you get for like NX and all that. I use Google Remote Desktop all the time, probably every day, but I don't use it to work. I use it to get something fixed or tweaked or scanned or something weird like that. I don't use it to actually sit there and program parts and they do. Yeah, I do sometimes because it'll take like
00:43:01
Speaker
30 seconds to close fusion on one computer and open it on another computer and open all your side like it's already open on the work laptop so I'll remote into it and then you know do my edits there but it's it's a little laggy and it's a little it's not something I would want to do for fun I'll do it if it's there and I have to and make a quick tweak but speaking of posting like
00:43:22
Speaker
Sometimes I'll be sitting at home at my kitchen table where I do all my work. I know I need to make a tweak, like re-output this finishing pass and repost it. I can do that from my kitchen table, remote, post the code, upload it to the current and know that it's going to run that night.
00:43:42
Speaker
99% of the time, I'm confident that it's going to work perfectly. So I've done that a bunch of times, especially if it's a very known tweak. I'm like, I know what I'm doing. This is fine. Nothing's going to go wrong. I haven't changed the post in weeks or months. This is going to work. So I'm building that nerve of being able to pull it off successfully, and then I have.
Shop Updates and Future Equipment Plans
00:44:03
Speaker
So if the Kern was totally idle, are you able to load a program and hit cycle start? I cannot hit cycle start and I cannot stop it. I can't adjust the feed rate. I can schedule different programs. I can load different programs. I don't think I can stop it or feed rate it or anything like that.
00:44:29
Speaker
But if it's running its palette cycle, which is just a list of programs you can basically tack on, that's the trick. Yeah, exactly. I can reschedule and I do it all the time, anytime, anywhere. I can do it for my phone.
00:44:43
Speaker
That makes me want to add a single dummy program that has a 70,000 second dwell. Yes. That way it's always running and if I could just bolt something else, I guess you couldn't do it because it would start to finish that first dwell program. You know what I mean? The trick is mentally knowing what's loaded and what's available and what you can schedule next.
00:45:04
Speaker
Yeah, we have an M0 program. If we don't want to stop this cycle counter basically, if we're at 27 hours and we're like, yeah, let's go for 80, we'll put an M0 afterwards, not to calculate time, but just to keep it running, pause the cycle time and then continue.
00:45:21
Speaker
So, if we need a few minutes to load the fixture or something like that, or we're not sure what to run next or whatever, well, it'll stay in the cycle with an M0. But you do have to hit cycle start to continue for that. Oh, got it. Although, I could totally do what you did or what you said and put like a four-hour dwell program in there that you could tweak. How do you not? How do you get it out? You can't get out of a dwell. Unless it loops a million times and you can control the looping amount.
00:45:51
Speaker
Could you break the loop? Probably. It doesn't matter. I say this because I love you. Don't get too hung up on stats. If you ever make a mistake, you'll be like, oh, I was just trying to shoot for an ego number here. We do that a tiny bit, but not a lot. We don't run 24-7 lights out like you do, but the horizontal has run almost every night in the last 10 days probably. That's huge.
00:46:21
Speaker
No, John, I could talk about it. It's amazing. But it ends up finishing up probably before midnight most nights. And so I don't want our compressors on overnight yet. And in case in point, we had a blow off overpressure valve get stuck open when I came in. And I came in at like six to swap parts out. And I was like, oh my God, what's that noise? And I'm walking up to the shop door.
00:46:46
Speaker
And that's kind of why like, man, you're doing like an air dump for six hours. And that's just, you know, hey, a piece of a speck of dirt dust or something caught in there and in or something. Anyway, I had these $70 110 volt
00:47:05
Speaker
stepper ball valves. So it's not a solid because I hate solenoids when you're trying to like hold them open or closed for long periods of time because they generate heat and they fail and blah, blah, blah. I wanted a ball valve like you rotate with your hand that was just motor controlled. So it just uses a motor to open it, you use the motor to close it. Well, they make them.
00:47:23
Speaker
And so i bought one actually probably a year ago and just sat there and it's great except it was built to be normally closed so when it's actually really cool has a pretty beefy capacitor in it so if you pull power to it even though you've pulled the power away it has enough
00:47:40
Speaker
juice in the cap to Complete one last closed cycle really so it fails to close But I wanted to fail to open because worse I'd rather have the air fail open than fail closed So I actually was able to just break open the housing Rotate the motor mount 90 degrees Okay, and then I relabeled the has a little indicator that says whether to open or close, but now it's weird because turning it on Powering it closes the ball valve. So it's kind of weird cuz yeah
00:48:10
Speaker
I'm calling, so we hooked it up to an Alexa with a smart plug, and so we call it Fioc, what do I call it? Fioc, not break, the air kill or something. It doesn't matter. So now if you open up Alexa anywhere in the world, you can hit this button and you turn on the air block, Fioc blocker. That's what it's called.
00:48:33
Speaker
Um, and it's great because when you turn it on, it blocks the air from coming out of the compressor. And again, if you have any concerns or even if somebody's at the shop and they're like, ah, this isn't working, you just unplug it from the wall. It just opens itself up. Nice. Okay. Yeah. I always wondered how this servo controlled solenoid valves work. I'm glad to hear it's so far successful. Yeah. It's fun. Oh, sorry. We're going over. Sorry, Phil.
00:49:04
Speaker
trying to wrap up the Wilhelmin install, things are good. I got the umbilical cord and transfer plate all pretty much installed. It's super close. Awesome. Super close. What else? What else? Trying to wrap up a speedio order and the Zeiss order as well. It's taking longer than I hoped, but whatever.
00:49:27
Speaker
Yeah, otherwise, good production. We had a new guy start two days ago, Grayson. Oh, congrats. So let's shop assistant role. So Angela's been walking him around and showing him all the ropes, and super excited for that. That's great. Yeah.
00:49:42
Speaker
We talked a lot last week about the coolant, that ring around the pump and all that. Has that been fixed, replaced, tested? We are trying to purchase a replacement, and our supplier is confused as to what the replacement is supposed to entail. So there's been some back and forth. You should just buy a whole new pump. Yeah, but we just need the seal kit. Yeah, but I don't know if the seal kit has all the parts you need. Yeah, but here's a picture of all the parts we need, and here's the manual that we found, and this is what we need. Just give me that. Yeah, but I'm not sure.
00:50:11
Speaker
Yeah, got it. So that's the answer. And then we're going back and forth on that. A last update was we might just have to buy a new pump and that would solve the problem. And it's clearly the problem. So I'm happy we found it. Yeah. I just got to get it solved. Yeah, got it. That's good. Yeah. Cool. All right, man. I'll see you next week. Have a great day. Bye.