Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
#334 How to Re-work Parts or Call Them Scrap
342 Plays2 years ago
TOPICS
- Saunders lost power
- Grimsmo to sell his old Eumach machines
- How to re-work parts or call them scrap
- videos of shotgun barrel manufacturing
- heat treating puck chucks
- Setup sheet tool tracker
- Question to the audience
Recommended
Transcript
Introduction & Business Updates
00:00:00
Speaker
Good morning. Welcome to the business of machining episode number 334. My name is John Saunders. And my name is John Grimsmo. John and I continue to catch up each week on what's going well and not well, I guess, or pushing us together on our manufacturing businesses.
00:00:17
Speaker
Yeah. How's it going? It's pretty steady smooth. Yeah. Things are good on our end flowing. Good. We've got guys coming in and out of vacation through the summer period. Yeah. So there's always a bit of a juggle. He's off all this week. So we either got to prep beforehand so that his work doesn't need to get done or somebody else has to jump in. And it's been good. Team's done really well.
00:00:43
Speaker
We've had similar vacation travel, summer, and it's weird.
Handling Power Outages
00:00:49
Speaker
I've looked at our July reports, good month for sure, but then I also feel like we've had some bigger sales swings and like the rare slow day, like a blood day. And for sure, no big deal. You always just kind of wonder, okay, is it a one-off or is there more to it? And some of it I also just think is summer, like it's, you know,
00:01:12
Speaker
And actually, this is kind of funny, we lost power. First time, geez. This may have been over a week ago at this point, but I don't think we've ever really been out of power. We've had the power flicker on and off before. Really? Like twice maybe, but in eight years. And this was a transformer down the street blew up, caught fire.
00:01:34
Speaker
Yeah, it's too soon. I vividly remember your... That was a year ago now, almost exactly, early August, a year ago.
00:01:44
Speaker
What was funny was this happened a day or two after our heat wave ended. I would have thought it would have happened during the heat wave. The grid is being taxed. Maybe it was still related and that transformer had been taxed and just took another day or two to give up the go. This is one reason why I enjoy talking to you. There's no playbook. What do you do? It was 140.
00:02:13
Speaker
In the afternoon? Correct. So you're like there. Oh, yeah. We always run it. And you got five machines running. And like. Nothing was being tapped. No holes were in there. Tap recovery is really no big deal. But it's always kind of the joke, I think, of like, OK, anybody tapping right now? Yeah. And so the police officer down the street had sort of confirmed, like, hey, transformer blew. I thought, OK, wait a minute. We close up shop at four. That's two hours change.
00:02:41
Speaker
I said to everybody, hey, let's sweep and clean for 20 minutes. We opened up the bay door so you could have some light. Sweep and clean for 20 minutes, and then everybody went home. We gave them an hour on the house, and then the next hour they could come in and make it up later or whatever. It ends up that the power came back on about 315, which, shout out to ADP. I'm really surprised. It's still, it's kind of like, it just kind of felt like one of those summer, like, everybody go home and do a big deal.
00:03:10
Speaker
Yeah, yeah. Were there any weird recoveries on machines? Like you must have programs mid program? I don't know, everybody.
00:03:22
Speaker
I couldn't tell you. Sorry. We leave the Wilhelmin on perpetually because the age of that machine is such that if you leave it off, it seems to cause more problems. Otherwise, most of the other machines are on and off regularly. And obviously, you're in the middle of a program, so you just got to figure out where to repost from.
Challenges in Selling Equipment
00:03:43
Speaker
Everybody apparently did that fine.
00:03:46
Speaker
What else was I going to say? Oh, like little annoying things like our net share and IP servers kind of had to, some TLC the next morning. Um, so we normal networking stuff and good. No big deal.
00:04:01
Speaker
Good to know it's super stable otherwise. We've had it flicker a big handful of times and we've lost it hard, hard the once and then for hours at a time a couple other times. Yes. It's never fun. Yeah, not fun. Agreed. I think the worst is the unknown of when will it be back? Do I come into work the next day? That was brutal. Yeah, exactly. Agreed.
00:04:27
Speaker
I thought that I was like, oh man, could this be a multi-day thing? And I've always heard maybe an old wives' tale.
00:04:36
Speaker
that most United States power grid has dual, like dual fed, dual redundancy. I don't know if that goes down to like every individual house, but it would make sense to think that maybe we're in technically an industrial park. It doesn't really feel like that. But for those that haven't been here, we're on a road that has numerous other factories nearby, one of which is a multi-hundred thousand dollar Kellogg, where these foods like veggie burger plant. So they have freezers and
00:05:03
Speaker
their power bill, I would guess is 100 grand a month. So they've got a different relationship. Actually, it's probably more than that. They have a different relationship with AEP than we do. So that's a good thing to have. Yeah. Excellent. Where I'm not doing a good job, I'm just going to call myself out on this, is I can't seem to sell the machines we're trying to sell. Like the little ones, the Wazer and all that stuff.
00:05:31
Speaker
Yeah. And so, of course, it's inadvertently a little bit of a marketing attempt for our audience. And look, if I was, you know, I've adjusted the prices down, I never intended to be high on the price for sure, but they're also not selling. So that tells you something, either not getting the word out or you're asking too much. But yeah, the big ones would be the PocketNC 5-axis benchtop machine, the Wazer, the 24R, the Torlock router, which is a bigger machine, and then some smaller
00:06:00
Speaker
real small stuff, but those are the big optical comparator. Yeah. So I'm just loath to put them on eBay, but yeah. That'll happen. How about you on the, on the, on the UMAX, not too much progress. We've been
00:06:20
Speaker
We spent a tiny bit of time just cleaning them up and getting them ready for taking pictures and stuff. Got the one turned on, which was kind of cool. I thought we'd unplugged it completely to plug our router to use that power for the router. But it turns out I was able to just add a plug outlet to one of the cables, because it's a 220 three-phase. So I just used a dryer plug kind of thing. And I'm able to unplug the router vacuum pump and plug in the spindle of the U-Mac.
00:06:48
Speaker
in that port and I was able to turn it on and get things figured out. Because yeah, it'd be nice to have a quick video of it functioning to sell. So anyway, we're not super rushed on it, but we're picking
Decision-Making in Business
00:07:02
Speaker
away. Can I push you though? You can try. Pick a date. I don't care if it's December 10th, but if a machine to a company won't resell it for you and get them moved for any amount, then scrap it.
00:07:20
Speaker
could be. There's no scenario where you integrate them into your workflow, right? Like, yeah, and I keep, I keep not that I want to do that, but it does cross my mind. And I just keep going. No. Yeah, you know, yeah. But this is, this is, this is what I want out of you in this conversation is push me, even if it's uncomfortable, like tell me if I'm, if, if you need that nudge, if they're not going to be party, they're taking up real estate, they're taking up space. And it's also like that. Um,
00:07:49
Speaker
There's something to be said about leadership of recognizing what you need to do. And I'm thinking about that now with how we focus in on how we cleaning up the old stuff, getting rid of old material, old, like we have some super old fixture plates that we're working progress where we've iterated beyond that workflow. And yeah, that's some real money, but it's a sunk cost and it's kind of a symbol of the past, like gotta go, gotta go, gotta go. It basically becomes a paperweight.
00:08:19
Speaker
As much emotion as you have behind it kind of thing. I died and my kid was an adult and took over this business. They'd look at that and be like two seconds gone. I wish I didn't have any emotional attachment, but inevitably you think, oh, on a slow day, we could get that back up and do it all the way.
00:08:41
Speaker
needs to go. It's kind of the same thing on our end with say a rework part. Say we make a blade and it needs to be cut again or something so it comes back through the workflow and it goes in the rework bin and oftentimes we're running around too much to spend the time to rework it. It just ends up aging and getting older and older and then I look at it and I'm like that's
00:09:04
Speaker
400 serial numbers ago. It's almost too old to sell. Why is it still here? Well, we haven't gotten around to it yet, either me or the other guys in the team.
00:09:14
Speaker
And on one hand, it really bugs you because you're like, that was a fixable good part. But now it's aged out.
Managing Production Scrap
00:09:21
Speaker
And like with some of our parts, the handles we engraved, you know, January 2022 on it. And I'm not going to sell a knife from almost two years ago. Right. Right. Like it gets weird. Yeah. It's kind of the downside of dating and numbering everything is you kind of want to keep a current flow of product going out the door.
00:09:42
Speaker
But it also shows why did that not get fixed in the moment? Do we need to build in the five minutes it'll take and the pressure to keep those flowing? It's a good part. It's good materials, a lot of money, a lot of time goes into it. And it's not scrapped until it's decided that it's scrapped through time or failure or whatever.
00:10:07
Speaker
No, totally, John. These strike me as the kind of things that are your and my weaknesses as the founders of the company. I'm right there with you. I am no better at this. But again, that outside third party CEO, you hire a sharp
00:10:24
Speaker
whatever, seasoned executive MBA, like been through the ropes, brought companies from X to 5X. They're not going to be burdened with what you and I look at about, oh, man, that garden hose could be useful. This is what happens in my head. That's interesting. But I don't think, I mean, we do have emotional attachment to the parts that we make because we're very proud of them.
00:10:54
Speaker
I still find it very hard to be like, nope, scrap. Just throw it away. Put it out of its misery kind of thing. So it goes into a bin and it's like either a rework bin or an actual like red scrap bin and part of the value in.
00:11:08
Speaker
not making it disappear and keeping it is that it shows like how big is that red bin this month? It gives you that feedback, which is helpful if you analyze it properly. If it just kind of goes in the drawer and keeps going in the drawer, then nobody ever thinks about it. And on that note, one of the things I was thinking about talking this week was our, what do you call it?
00:11:36
Speaker
are okay-ness with creating scrap. Okay. So I'm a very, um, easy going understanding kind of person that makes a lot of mistakes and I make scrap too. And the team does too. And I think that's almost led to a situation where scrap is more okay than it should be. You know what I mean? And, uh, I might need to start pulling back on that because it's starting to bug me. Yeah.
00:12:07
Speaker
I find myself
00:12:12
Speaker
I'm being honest, sometimes it's easier to judge from the outside when others make mistakes. So consciously or subconsciously think like, ah, that's avoidable. And then it's great. I will never venture away from getting involved in making parts here, full stop. Like the company will, it's just not going to happen. I don't care. No one to tell me not to and I enjoy it too much. It's what like, so that's a great thing. That's just, okay, good. I don't like to defend that answer.
00:12:40
Speaker
And it's a good reminder of how human we all are. And, you know, I was on the, Garrett was off last Friday. Vacation was actually great. Gave me a full day on the horizontal to play. And then it was already loaded up to do the night run when I was done. Like that's like full. Yes. Awesome. And, um, and it went great. I guess she's like making any big mistakes, except it just kind of reminded you about, um, the standards to which we hold ourselves on finishes and tolerances. Um,
00:13:11
Speaker
are such that we're using, even on the Akuma horizontal, which is a very good, you ask for it to interpolate a hole, it interpolates the hole pretty darn well, but you're still doing a very minor stock to leaves and adjustment. As you know, spring passes are iterative because after you do the spring pass, it's different. The next time, if you haven't done the pre-spring pass tests,
00:13:32
Speaker
And then the only thing I did where I broke, I broke two end mills, um, back to back, which is $28 each. So in five minutes I've lose $56 in tooling because a slotting 2d contour that I moved over from a, we're doing some gen three accessories that we're getting ready to launch. And, um, I don't know why fusion did it. I'm not blaming fusion. I'm sure it was me, but.
00:13:56
Speaker
the slot was supposed to have the tangential extension and that seemed to have disappeared. So the 3.16 tool was plunging into the part instead of starting a hundredth out app. And I broke one tool and just assumed it was a slotting operation that was getting pushed too hard. Ran it again without stupid bonehead looking like at the cam tool path. Yep. Yep. Yeah. I had a similar thing where I was cutting a fixture and
00:14:21
Speaker
It's doing this weird facing operation on the side of the fixture, like five axes rotated. And I've done it before, but I reprogrammed it to be a different style of toolpath. Instead of a 2D adaptive, it was like, I forget what it is, face or some linear contour, something. Anyway, it broke the tool.
00:14:41
Speaker
And then, but it's such that it was an eighth inch flat end mill, it turned it into a chamfer mill. So the length of the tool was still the same. So tool breakage didn't catch it. So that led to the side of the fixture looking weird, looking finished, but unless you look closer, you're like, that's not done. And then that tool happens to also be used on our Norseman handles. So it started making bad handle pockets.
00:15:08
Speaker
et cetera, et cetera. And that's kind of been on my list. Like that's a fixturing tool. Why is it still used on Norseman? Well, I just, I just screwed myself here. Um, so now that tool is not being used on Norseman anymore. That is now a dedicated fixturing tool, but, uh,
00:15:24
Speaker
Yeah, and then I looked back in CAM, and I was like, OK, what broke the tool? Because I couldn't quite see it on the part, but I went through the toolpath, and I'm like, oh, it's not retracting enough, and it's moving over, and it's plowing into material at rapid until a little bit of clearance plane, and then it feeds into the material. But full rapid, it went over and down into steel. And of course, it's going to break an eighth in gen mill.
00:15:49
Speaker
But it's fun to be able to go through that process and to be able to find the cause of the break and fix it and move on. I enjoy that process. And it's not so much tooling scrap that I'm currently referring to. It's parts, blades, handles, things like that. Is there a particular part or machine or process or person that's
00:16:20
Speaker
It's probably a couple things. It's like all adding up in my head. But take a blade, for example. Blade's probably our most complicated part because there's so many hands involved, so many processes between heat treat and surface grinding and soft milling and hard milling back on the machine and grinding and polishing and sharpening. There's so many ways to scrap a blade. And we have probably several drawers full of scrap blades over the 12 years we've been doing this.
00:16:53
Speaker
I want to get better at avoiding the scrap, like full company. Our standards are so high that if it doesn't meet standards, scrap it and move on. But maybe we're missing some vulnerabilities here. Maybe throughout the team, myself included, we're passing it off as scrap and not reporting an issue that could be fixed kind of thing.
00:17:19
Speaker
Um, yeah, I hear you. I mean, in blades, is it, is it, I don't, we don't
Heat Treatment Challenges
00:17:27
Speaker
want to scrap fish slates because our material cost is, is high as a nominal value and it's high relative to like our overall company. My advice, you know, if we scrap apart, it's, it's, we'll be okay. But it's the same thing. I don't want that moral hazard of thinking like, Oh, you know, every two, every, you know, one or two, we just throw in the scrap bin. Like that's not good.
00:17:46
Speaker
Well, when you really think about it, you think, I'd say a handle, it's a small chunk of titanium, but it's still a, I don't know, $10 chunk of titanium that has $10 of grinding in it, that has $10 of water jet in it, that has how many hours of our time in it that has compounded that one piece of scrap is worth over $100 in hard costs or whatever.
00:18:09
Speaker
And that's no longer just a $10 piece of material, just whatever. I want it to be. So take it. Let's go from the overall scenario to being more data-driven and more specific. So focusing on if you can find out where the most scrap is happening, whether that's a product or a stage, the margin, then
00:18:37
Speaker
and then just look at that individually instead of worrying about solving this overall issue. Let's just say- Well, I'm trying to be purposely vague for the podcast on one day. Yeah, I hear it. Sure, sure. Also, so specifically,
00:18:53
Speaker
because part of me is super obsessed about flatness because I think it's fascinating, but also because we control our own heat treat and we do our own lapping and we do our own surface grinding. Flatness is a real part of our lives. Currently, one of my big issues with yesterday was flatness of parts coming out of heat treat and then going through lapping and lapping not being able to lap through
00:19:22
Speaker
the warp. You treat warp. Yeah. Yeah. And not being able to create a fully cut surface. So it's leaving sections that are not lapped. And there's a lot of finger pointing, I gotta say. Yeah, right, right. And it's very difficult. So I'm trying to find the root cause of all of this. And
00:19:47
Speaker
a lot of it does stem back to the blades are warping through the heat treat and quenching and tempering process. And we've put a lot of effort into that already. But I think there's more effort to put into it to eliminate the warp. Sure. I might challenge a little because so two of the most important lessons I learned in my manufacturing career
00:20:10
Speaker
And both were pivotal. The one was strike mark. This is 10 years ago, about 15 years ago, day job and wanting to tackle the world, wanting to bring this product to market. And no one was willing to tell me or able to tell me, I wasn't able to tell myself, you're trying to do something that's not possible. And I'm not saying that's the scenario here with the blades, but
00:20:37
Speaker
Yes, it's fun to perform miracles. It's fun to create new events in technology. But if the ultimate answer, I'm being a little bit extreme here for the sake of the example. If the example is you just need heat treat to not warp it so that you can lap it, it's like, what if a PhD metallurgist told you you'll never do that, so stop having that be the bar by which you're judging yourself?
00:21:04
Speaker
But then the second one was fixture plates, I don't know, three years ago, two years ago.
00:21:09
Speaker
Our business was over if we just had to keep having them ground. I wasn't going to buy a $340,000 giant grinder and start grinding to steal ourselves for a variety of reasons. And the outsourced grinding just didn't work. It was kind of when it was like, you'd have to just take a step back and I was, this just doesn't work. You're taking two to three months of lead time, huge costs, quality issues. We get plates back. This is a very good grinding shop. We get a plate back.
00:21:35
Speaker
16 by 40 inches, and it'd be within our spec, except the one inch corner would dive down a thou. And, you know, do you reject a full 16 by 40 inch plate because of that? I mean, you could, but if you do, then they're just going to drive the cost up and spend more time grinding them. And what stinks is like, that's the first spot you check with a mic. Yeah, corner, right?
00:22:00
Speaker
even though it probably has no real functional bearing for most end user applications. So I sort of had this internalized moment. It wasn't quite this dramatic as I'm telling it now as a story, but where I realized we're either, our business is either done, like we've got to figure out a different route to go in life, or we need to figure out how to do this in a way that works full stop period. Like don't kid yourself. It has to work. Yeah.
00:22:24
Speaker
in a wonderful way with some really good help from our team. It worked honestly better than I ever expected. And I don't, I don't think I'm up for another one of those challenges in my life. Like that wasn't really, no, I'm serious. Like I didn't, like if you're telling, if your challenge is now like you crystallize needs to revise the process of metal treating and metallurgy and heat treating and quenching to not have, you know, like, Oh, I don't know, John, that's a tall order.
00:22:53
Speaker
Yeah, I think we're close. I think we've done a lot of steps in the right direction and I think we just need to refine and fine tune some of the steps that we've taken to
00:23:06
Speaker
fine tune all of the variables involved because we've noticed through experience. So we have these aluminum water cooled quenching plates that I have in an arbor press that are water cooled and the blade comes out of the oven at 1940 degrees, red hot, goes between these plates and then we put a thousand pounds of pressure on them.
00:23:25
Speaker
I actually created a hydraulic cylinder with a pressure gauge on it so we can tell that it is exactly a thousand pounds pushing on it, which the consistency there between two different people was huge. To gauge that was awesome.
00:23:41
Speaker
that variable eliminated. But we've noticed that sometimes one of the pumps for the water cooled system might fail or the lines will get clogged up because algae tends to grow in this little water system.
00:23:56
Speaker
So if a pump gets clogged up, then say your top plate is not getting cooled and your bottom plate is cooled. So now you have a heat differential between your cooling. So you have a hot top plate, a cold bottom plate, and that's going to create warp because it's sucking the heat from the cold side, not from the hot side. And in those moments, you know, it gets brought up and we're like, the blades are super warped right now. And then we investigate and we're like, Oh, there's a reason for that. So I want to supercharge
00:24:23
Speaker
those processes, I want to probably put a chiller on that, just an aquarium chiller on that line, because the temperature rises for every blade that we heat treat. So it goes from 70 degrees to 72 to 78. So if we put a chiller on it, every blade 70 degrees, it's kind of the current mentality of like, well, the machine's always at 20 degrees Celsius. So that is no longer a variable.
00:24:47
Speaker
Or just increase and get a 30, what was it cycling through a five gallon bucket or something? Yep. Yeah. Change that to a bigger 30 gallon rubber made tote. That'll massively help. And then you could also chill or cool that. Yeah, exactly. I know my opinion doesn't really matter here, but I'm with you on those scientific improvements. And good grief, you've got block filters, rock and roll, like there's a block line. That's all I just, I've told myself in the past,
00:25:18
Speaker
Things like you just said like, oh, I just want to be more excited that we'll figure it out when it's like, what if somebody needed to tell you like, you might get better, but this is now I guess where I'm wrong is you, you have making a number of these correctly and it's only some that are not.
00:25:32
Speaker
Yes, because sometimes they're bang on, and those blades lap perfectly. Sometimes they're fourth thou warped, and we're lapping off less than a thou per side. How does that work? It doesn't work. Yeah, sure. Back when we were trying to use a giant C-frame machine with 32-inch hangover in the spindle and trying to minimize stepover lines,
00:25:56
Speaker
I knew it was a fool's errand to solve that. Like sometimes it would work, sometimes it wouldn't. And if the roll up door got opened or the forklift was being moved or the tool pressure changes from, like it just was.
00:26:07
Speaker
You know. In order to face your fixtures, you mean? So the whole game with making fixture plates without grinding them, machining them is, yeah, your finish matters and the aesthetic, the look, but ultimately you need a parallelism spec that is nuts. Like it's phenomenal. And you cannot see or feel the step over lines. If you have two, three, four microns step over lines, you see them, it looks terrible. You feel it, you figure it out. It's just not something I was ever going to do. Interesting.
00:26:37
Speaker
and we were never going to get there with the seat, which is why we bought the Okumogenos. It wasn't as easy as just buying that machine and hitting the go button. For sure. That was one of those. It was never going to change. I don't know anything about quenching, but
00:26:53
Speaker
Maybe you need to have, you know, like a Adam Demuth style roller bearing die set with those fancy bearings that have all the balls in them. Yeah. Yes. Right. Like the shaft that has a hundred balls right up around it. So we've got this perfect up and down motion and it's controlled. That's also a good idea.
00:27:14
Speaker
And the quenching lines have view ports in them with flow meters. OK, like that stuff. Right. Yeah. And that's not even expensive or hard. I know. And so all of these things are doable, and they would lead to an improvement. And I want to exhaust reasonable improvements in order to at least fine tune our theories behind it. And maybe we get to a point where it's like, OK, realistically. But I do think we can get better and more consistent. I think there's no reason
00:27:43
Speaker
other than the metallurgy of the steel itself, maybe there's variation in blade to blade or sheet to sheet of steel. But I mean, there's a reason we buy very high purity steel that's powdered metallurgy and like, you know, really good stuff. But the kind of the theory of heat treat is that once it's at that malleable state, whatever it's called, you know, at red hot. Austin Martens. Yeah. Yeah. Even I'm not an expert at those things.
00:28:13
Speaker
all the molecules are like all wiggly. And then you put it in the quenching plate and you want to set them in a, uh, as quickly as possible. You want to cool them so that they set in a flat plate. So if that's flat and it's evenly cooled and it's like, there's science behind this. And then even going into the temper oven at 400 degrees, we made these, um, I think they're one inch thick, maybe three quarter inch thick temper plates, like two very thick plates that bolt together and sandwich eight blades at a time.
00:28:43
Speaker
and then we bolt them all down, we torque them and that through the low temperature temper process that also helps to promote straightness and flatness because the way we're doing it before they warp through temper. So we are making improvements but there's still something amiss and got to figure it out.
00:29:05
Speaker
So, and they're bowing, the knife blade is bowing along the long dimension of the blade. Is it only a lake or hill or is it what I call a potato chip where there's multiple undulations? For the most part, only a lake or hill. Like a fairly steady arc. Sometimes they wiggle a little bit and there's various reasons for that. How they're placed in the oven, how they're placed in the quenching plates, the absolute flatness of the quenching plate itself.
00:29:35
Speaker
things like that and they wear over time and we got to check them more often, we got to surface them more often, things like that. Yeah. Have you watched, I think we talked about this like a year ago, the videos of shotgun barrel manufacturers and- We did talk about that. I don't know if I ever did.
00:29:51
Speaker
So look that up. There's more to it than probably what I'm about to say, but in a shotgun barrel, it's easier because you've got a 30-inch long tube that's hollow, so it's easier to impart changes on it, but they have, as soon as you see it, John, you'll immediately go, ooh, yeah, so they have.
00:30:08
Speaker
And it's skilled labor, but it's not skilled labor from an engineering and machining standpoint. It's more skilled labor from a person that just gotten good at using this setup jig. I believe I had two dial indicators on it, and they kind of roll it around and they look at the high spots and then they have a movable arbor press kind of over the barrel, and they can move it to different points. They pull at certain distances and it does a, what is it?
00:30:31
Speaker
beyond plastic deformation. It overbends it to compensate for the three thou or six thou here, because ultimately, if you're just dealing with the potato chip or a lake or hill, create a little jig, pre-bend it a little bit. We do that. It's still not good enough. But yes, we do have our purposes at
00:30:51
Speaker
both by the heat treat cell and also by the lapping station where we can bend out a little warp. But sometimes that makes it into a potato chip. Sometimes that adds crinkle to a otherwise consistent bow. Interesting. Okay. Yeah. And we have a light box set up where with a little flat plate, it's a
00:31:12
Speaker
It's actually an orange vise jaw, one of the bolt-on vise jaws, but it's got a really flat top surface. With a slit through it, and we've got orange LEDs in the background, so you put the blade on top of that flat surface and you can see the light gap basically under the blade. You can tell if it's a hill or a lake. We do that with every blade and we check them. When they put them on the lapping machine, they make sure that it's shaped like a hill.
00:31:35
Speaker
not like a lake first because you want the two outside points touching for the first lapping side, things like that.
00:31:48
Speaker
Okamoto surface grinder fall. So that happens first before to a soft blade. Okay. And then they get machined, then they get heat treated, then they get machined again, then they get lapped. Then they get bevels put on. One idea, you might hate this, but you could have a fork in your workflow. And so the blades that come out good continue on the blades that don't. Let's assume
00:32:14
Speaker
Hill or leg is just a question of which way you flip the nut blade. So you can call them all hills, right? If you want to rotate it over. So, and let's say that the hills are only two, three, four, five, six style. Like they're only a little bit and you can kind of put them in a five or six different buckets. You could build
00:32:30
Speaker
fixtures on your for your Okamoto that have that hill bent built into it so that way when you apply the magnet you're not sucking the hill out and then it's a separate workflow it's more work but it's a consistent process and you could then drop them on the so let's say you've got a three thou hill put it on the three thou hill fixture deck the one side that side is now flat but your parallelism is now gone the pot but just flip it over directly on flag yeah
00:33:00
Speaker
Yeah, it's Angela's used paper as shim sometimes to shim out a couple of thou totally works. The issue with that is
00:33:09
Speaker
Same thing happens in lapping basically. You're correcting a warp by cutting out both sides, which I don't like because now the machine features that were meant to be on a say 0.125 blade are now getting chopped off in the corners or like engravings or tiny little chamfers or something like that. They disappear if this
00:33:33
Speaker
I know you don't like to hear this, then increase your material by three or four thou do this process on everyone. And while you're doing that parallel track, improving your quenching, and if you can get that fixed, then you can abandon this. Great. Yeah, we did increase the surface ground thickness so that there's a bit more room to lap a little while ago, but that hasn't solved everything. And I think, I think the biggest thing is the variability.
00:33:59
Speaker
in work from hatred. Some blades are really close and some blades are not. Then you lap a whole bunch together. Then you have some that work and some that don't. You have to relap some. It gets weird. There's a lot of rework and a lot of double work. I know lapping would be easy if the blades were flat and parallel going in.
00:34:21
Speaker
So I'm trying to find a way to do that, you know? Right. So that's where I'm like, um, solve that without regard to time or cost because it has to get solved. And whether it's the way the idea of like Okamoto in them again or whatever, and then that's going to drive you nuts. It works because you're driving nuts. Then you can improve beyond that, but at least you're now going to almost zero scrap. Yeah.
00:34:47
Speaker
What we're dealing with is now the heat, the puck chuck, because there's a couple of surfaces that we'd like hardened. We've had some really good learning, reading, Googling, viewers emailing in about heat treating, quenching, laser heat treating. It's been edifying, but I've also realized there's so many variabilities.
00:35:07
Speaker
I am of the opinion, there's a chance I'm wrong, but that this product does not have to be heat treated, meaning, or full heat treated, it doesn't need to be a full 45, 50, 55 Rockwell.
Improving Heat Treatment Consistency
00:35:19
Speaker
So we're going to make the first batch out of pre-hard to avoid some of these risks and concerns, because it's not just heat treating, it's the quenching, and then it's the, are we having the blasted, are we having the polished, are we having the rust protective, are we having the Nicholas electroplated or TIN?
00:35:36
Speaker
And those are so many complications, variables, vendors, and frankly, it's things that we're not great at right now at least. So, okay, we're going to parallel track that learning while we also launched the first set of these out of pre hard. And really that there's still a hard bushing. So there's absolutely appropriate hardness where we need it. But it's really just a question of the longevity of the product in terms of if the operators start banging fixtures and do it or dragging such a process. Yeah.
00:36:04
Speaker
Yeah. And so by all means, I care about that. But it's also not, this is uncomfortable for me, because I'd rather than be stable, if you know what I mean, from a design standpoint, but it's also kind of one of those, you know, and there's some truth to this for you, John, don't let perfect get in the way of good, like, you know, is there a problem with a Norseman if it has a three thou taper? I don't know, John. Yes, very much. Well, does everybody agree with that?
00:36:33
Speaker
Let's say yes. I don't want to go into it. But it is. And our tolerances go out of whack quite a bit. If there is any amount of taper or warp, it suffers in the product. But then to combat that, this is a new product you're releasing.
00:36:53
Speaker
And I have to remind myself that for a new product, it doesn't have to be a million percent future-looking, perfect, done forever on the first go. It almost shouldn't be. It's like they say in tech, if you're not embarrassed of your first launch, you launch too late.
00:37:11
Speaker
Yeah, right. They need to work and they do. Oh, John. Oh, this is awesome. We wanted to do a new fixture test idea. It'll hold four of them. So it'll be like a 30 inch long strip. I realized I could make a single one.
00:37:28
Speaker
having to remind myself of my own fixture principles of make a one station fixture first and test it before you go make the floor, no matter how confident you think you are. And so I designed it and then I wanted the 3D print one to just touch it and feel it. And then I realized, wait a minute here, my 3D printed it with 100% infill, including 3D printed ID expansion clamps.
00:37:49
Speaker
John ran great yesterday. I used this in the Okuma. Now, is this going to make 400 parts? No, it may not even make 40, but it was way more solid than I expected. Yeah. And it obviously proves your concept without you having to scrap $20 in steel and time and end mills and breaking and possible failures and things like that. And then only to scrap it and say, I don't like it and make another one.
00:38:14
Speaker
It's made me so happy. We're doing very light cutting on this. It's a finishing thing, but yeah, super big win. So you put a pulse down in the bottom of it?
00:38:25
Speaker
Yeah, so it goes on the punch hook. Yeah, yeah, but I'm just thinking, how's the pull set attached? Is there a nut on the other side? So the pull set has an M6 fastener that goes through, and instead of using 3D printed threads, we just put a hex in the 3D print for a nut. That way you actually have pretty good clamping through the body. Yeah. But I know this doesn't come through on video, but it is solid. Yeah, 100% infill will do that. That's wicked. Yeah, real big win. God, I got to get a bamboo. Sorry.
00:38:55
Speaker
That's OK. I can wait. I'm good. Hey, I want to bring up one thing that I jotted down when you were talking about breaking that tool on the Kern because this kind of goes back to my perpetual gripe that Fusion 360 doesn't have what are called linked libraries, like a master library that tells you tool four is used in these separate files in your project or in your own assigned group or whatever. Wouldn't that be amazing?
00:39:23
Speaker
Well, so we have that through a Python script. That's right. I've mentioned that we put, we post set up sheets into a folder and then the Python script, pretty simple scrapes those for the tool numbers and dumps the tool numbers into a Google sheet. You know, what's funny is I've been doing that manually for the past week or two.
00:39:43
Speaker
As I'm planning for Tools for the Speedio, I'm taking library tools from the current. I'm posting a setup sheet. I'm copying the tools. I'm going into notepad. I'm putting them in a spreadsheet. I completely forgot that you have already automated this. And I like that we did that, still use it. But I also realize there's a little bit of an easier way, which is if those HTML setup sheet files are in their own Windows folder,
00:40:11
Speaker
The Okuma is a strange, it's a panic-ish based control, but it uses, or uses a G116 T4 to do a tool change, different than your typical M6. But regardless, if you're concerned about what programs use Tool 4, you could just do a simple Windows search for either Tool 4 or even the T4 M6 command. It's not necessarily perfect, but in the folder of all, so you just have to post all the setup sheets into a folder and then,
00:40:40
Speaker
search there. Is it foolproof? No, but it's pretty... Are you searching the setup sheet or the code itself? T6M4 kind of thing. Oh, I'm sorry, John. Thank you. I'm searching the posted G code for you. Yeah, I was giving myself... Because it's two actions to create a post and then a setup sheet. And normally I don't create setup sheets unless I need to.
00:41:05
Speaker
although it'd be sweet to link them. It's like every time I post, I also post a setup sheet and put it into this folder. That could be cool. I think you can do that with NT programs, Sean. Yeah.
00:41:16
Speaker
Makes sense. It's a good PSA for anybody out there to know for us because we're running a production environment. We had this issue. I blew up a face mill. That's my fault. I'm sorry. That's the point where I was like, hey, we need to replace this YG1 2.5 inch face mill. I was going to move it to a six flute, but now I need to know which programs use this tool to switch from a four to a six.
00:41:41
Speaker
The setup scripts, she works for that, but I was doing this over the weekend, didn't have a refreshed version of this Python script, and I just searched it that way and it was perfect. Just sort of see, okay, that matches what I thought. Okay, what are next steps on heat treat improvements? So I already had new quenching plates made.
00:42:11
Speaker
I redesigned them, made them thicker, made them better. And I had them made by Easton Bednarik. Bednarik. Bednarik, that's what it is right there. From Mariah Manufacturing, I had to make them. I was like, hey, you want a job? And he's like, yes, I do.
00:42:26
Speaker
I had them made from 7075 aluminum instead of 6061 because the impact strength is like twice. A little bit less cooling suckage heat dissipation, but I think that'll be worth it because we're denting the 6061 once ever so slightly over time. Anyway, I had those made. I've had them for a little while.
00:42:44
Speaker
those just need to be qualified for flatness. We have to check them, maybe flatten them better, make sure they're dead nuts flat, and then buy the plumbing we need and get it installed. It's in the works, and that's happening now, basically. Yeah, I think that'll make a good difference. Cool. How are they? I also want to
00:43:10
Speaker
somehow track progress of this. Like if we CMM to flatness of every blade for a while, we have data before and then we do the plates and then we have data after. So like proof, proof that it works, but might not be required. Might be the extra work to do that. I don't know. I just wonder too, I don't know how you routed the cooling channels through that, but
00:43:32
Speaker
you're what's certain is that the as the plate absorbs heat from the nut knife, it's not heating and cooling itself uniformly, depending on how the coolant lines are routed. So even sticking. I don't know what kind of temperature decisions you'd want to do. But you know, if they were little thermocouple things that could be pushed in
00:43:54
Speaker
you know, literally in half inch increments around the plate, I think that would tell you potentially a lot. You might be you might be screwing yourself by having that. So if you it's, it's heating up where it contacts and I assume that's the middle and then if your cooling channels are halfway out on each side, you're actually doing a really good job of creating a lot of warp and stress and that true. And to that point, I realized as I design the new ones, I've got these zigzag cooling channels on the inside, they look like a maze, basically. So the coolant has to go, you know, throughout
00:44:24
Speaker
And on the new design, I rotated the zigzags 90 degrees so that possibly the current cooling channels are causing a lengthwise warp. And hopefully the new cooling channels, if anything, would cause a short, short ways warp, which is much less of an issue. It's like one inch instead of five inches. Um, so that's another reason which I forgot about as to why I want to try the new plates so badly. How long would it take?
00:44:51
Speaker
So you pull a knife, you pull multiple knives out of the oven and then- Yeah, we do eight at a time. You're putting eight on. One at a time, close the door, quench, and then open the door, grab the next one. But probably within a 10, 15 minute period, we're quenching eight. So are you willing to try one at a time where you allow enough time for the quench plates to completely cool down the ambient between, just isolate this very- Good question.
00:45:21
Speaker
I don't know how that affects the heat treat the blades in the hot oven heat treat one at a time. Oh, fair. Uh, interesting. Yes. If that works, you've, you haven't solved your problem, but you, you, I have a thousand percent confidence that if that works and you will then figure out what you didn't need to do. True. And I hear you. We are tracking blades one to eight. We engrave on the blade one to eight. So we do actually know.
00:45:48
Speaker
the first blade on cold plates versus the eighth blade on hot plates. And I didn't actually know that they were engraving the one to eight on, which I find is fantastic. It's great. But as I was looking through some of the blades yesterday, it's not always the eighth one that's the worst one. But I don't have enough data to really like answer that question. This is where I'm like, part of me is kind of like, I don't care about one to eight, just stop and isolate the
00:46:14
Speaker
Quenching plays totally by letting them, you know, if you run the coolant through it, you quench one and you continue running the coolant through it, I got to think that thing gets back down to ambient within an hour. Yeah, true. It does. I mean, the heat we're pumping into it has an effect on the total temperature of the water.
00:46:32
Speaker
But that's why I kind of want to put a chiller on it. Even just an aquarium chiller, they're a couple hundred bucks. Just put it in a 50 gallon tote that you already had our old cool Qualcomm drum, John. Yeah. Volume of water is way cheaper and easier than trying to regulate a small quantity. This is like, yeah, this is true. Good.
00:46:53
Speaker
Fun stuff. This is fun. Sorry. It is fun. It's stressful sometimes. It's challenging. And I mean, I love a good challenge, but there's, there's a lot of factors to this that I, I'm not loving all of it. No, no, I hear you. I don't mean to, I know I, yeah, I hear you. You'll figure it out.
00:47:12
Speaker
I have a question to the audience if anybody has any updates or tips. We use a fair amount of Heimer tips because we use them in our training classes and students need to use a Heimer. That means they're going to break Heimer tips. We probably break one every three classes. Not the end of the world, but for the part in the training class, it doesn't need to have brand new Heimer tip accuracy.
00:47:38
Speaker
couple tenths or even a thousand, not the end of the world. So we have a little jig where we put in, when the Heimer tip breaks, we have a 3D printed replacement top, we put in pencil lead and then we re-glue in a reused tip. And it actually has worked pretty well, but it's still a fair amount of work to fix the broken ones. And sometimes they do end up having run out. So
00:48:03
Speaker
We had the CAD model and I don't remember why. I think we had one sent out to Shapeways to SLA print it and I don't remember why it didn't work. I don't think it's a great FDM part because the accuracy needed and you don't want stresses that cause dimensional changes and twists. But if anybody has any tips on recognizing this is a compromise on the accuracy of the product,
00:48:25
Speaker
still going to a full 3D printed Heimer tip. It also poses some risk to the Heimer body because if you crash in Z, the 3D printed tip will not shear away like the factory Heimer. Sure, yeah. They have a little upside down cone that causes the ceramic to shear. So when you crash in Z, it self-destructs the tip. Oh, that's cool. It doesn't damage the body. I'm not worried about that. Or if anybody has a form lab that wants to print one for us or for a few months, I don't
00:48:51
Speaker
I don't really want to deal with the cheap ways on this right now. We'd love to learn for us and obviously what we learn we'll share because I think for some folks, if you're just doing quick jobs here and there, you don't necessarily want to throw the expensive Himer tip in. Wasn't it you guys that made replacement Himer tips for super cheap like five years ago?
00:49:09
Speaker
No, that was us. That's what I was mentioning with the pencil lead and you can use the ball tip, but it takes work to set time to set it up and it looks, it does actually look a little janky. Yeah. And Vince was just saying that they're not always proving to be accurate. So now not only do all this setup work and make them, but then you got to test them to see if they even turned out. If we could get an SLA type recipe that worked, you could just print 50 of them. This is not done. That's done. Yeah. Yeah. Interesting.
00:49:39
Speaker
If anybody has the insights, I appreciate it. Otherwise, we're way over and I got to go finish up. I've got some leftover conduit I'm going to throw up on the wall to stretch an existing line over to welcome to the Saunders family and Okamoto.
New Equipment & Upcoming Events
00:49:55
Speaker
I think I got two new machines coming.
00:49:58
Speaker
On Tuesday, so three or four days from when this podcast hits, we have the UMC 350 HD, as well as the locomotive 1224 SA1 semi-automatic, semi-CNC service. Identical to yours, Sean. Sick. You got the fine pole check.
00:50:16
Speaker
Yes, absolutely. Yeah. Yeah. That's what Robin told us for smaller parts. If you're making huge castings, that's different. Yeah. Ours literally might be to the line I'd inspect the same as yours. There aren't that many options. Yeah, exactly. You get the paper filter on the side? Paper filter, but not the magnetic separator.
00:50:36
Speaker
Yeah, we didn't get that either. Paperfield works fine. Yeah, that's why I heard. Awesome. Hey, I'll see you. We're here next week. We're here next week. And then the week after we will have a face to face in person podcast in California at the Autodesk DSI event. So if you guys are fusion enthusiasts, definitely make your way to California.
00:50:58
Speaker
We also have one out of four seats left for our August 24th and 25th 5-axis training class. I think what we've kind of started to do to adjust for training class demand and interest is as the date gets closer, we will bump that price down a little bit, knowing that hopefully that makes it sell. Obviously, it also maybe gets harder for folks to either plan or travel, but we'd like the training classes to be full both from a business model standpoint, but also it's like the class works as a class.
00:51:25
Speaker
So anybody's interested. We have one seat left for our August five access class. Nice Comments see you next week. All right. Yep. Bye