Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
#394 DIY 3d printed machine tools with concrete!
3.4k Plays1 month ago
TOPICS:
- DIY 3d printed machine tools with concrete!
- Flextures and insane tolerances
- Machining soft blades in the Kern is slow, but better
- Patterning stress
- Hardmlling tool life?
- Chatter machine monitoring
Recommended
Transcript
Passion for Machining
00:00:01
John S
Good morning, welcome to the business of the machining episode 394. My name is John Saunders.
00:00:07
johngrimsmo
And my name is John Grimsmo.
00:00:09
John S
And John and I just purely love this idea that you can buy a machine that has screws on it, ball screws and guide ways and put a sharp cutting tool in it and end up doing some pretty cool things with those tools out of metal, titanium steel, aluminum, full stop, yes.
00:00:27
johngrimsmo
Yep, I am subscribed for all of that.
00:00:28
John S
Yeah. Yeah, right. um Yeah.
Innovative Machine Tools: Chris Borge's Approach
00:00:35
John S
I have a menagerie of random topics.
00:00:37
johngrimsmo
Yeah, me too.
00:00:37
John S
The first one was, I've thrown up throwing them out in that WhatsApp group, but there's a guy, I'll look him up while I'm talking, on YouTube, who is doing 3D printed machine tools.
00:00:49
johngrimsmo
Yes.
00:00:50
John S
Okay, so the obviously disclaimer is, um his name's Chris Borg, B-O-R-G-E, and that's his channel. um And he made a lathe, he made a precision like a watchmaker's drill press, which has a really cool, like, almost like a pedestal style.
00:01:03
johngrimsmo
who
00:01:04
John S
um your depth of the hole is is is by pushing down on a pedestal that lifts up the whole platter into the drill bit um but he's um doing these 3D prints with a concrete filler which as much as that seems sort of barbaric anyone who's read any info about machine tools knows that there's mineral granite castings we've done epoxy tombstones here like this is actually not ah far off something that I think is far more interesting than even he's letting on to.
00:01:34
johngrimsmo
Yes, and you you sent that to the WhatsApp group like a minute before I had to go out the door and I saw it and I i looked at the video. like I opened it and I looked at it and I was like, holy crap, save for later. like I'm gonna watch this later.
00:01:45
John S
Yeah.
00:01:46
johngrimsmo
And I haven't watched it later yet, but I got the gist of what he's doing and it put my mind in a frenzy. I was like, that is incredible. To be able to 3D print basically a shell and then pour concrete or epoxy granite or whatever the whatever the hot thing is nowadays,
00:01:55
John S
Yeah.
00:02:02
John S
Yeah.
00:02:04
johngrimsmo
to create you know plenty of rigidity for whatever small thing you're doing is phenomenally interesting.
00:02:11
John S
Well, I think I've said it here. I'm saying it again. And it's just an idea. Somebody can poach the idea. It's just an idea.
CNC Micro Grinder Retirement Project
00:02:20
John S
It's a fair game. But you know in my retirement slash free time, which yeah whatever whenever that's happening, it's like, OK.
00:02:27
John S
um go back and or now now that you've been through 20 years of this like you now know what you know you've got these well we've got this equipment and these processes and you've got guys like silos garage Adam Adam ah the machinist and and Robin I keep mentioning them sorry but like you know they're helping pull the rest of us up and learn more what I want to do now is go build a CNC precision micro grinder
00:02:35
johngrimsmo
Yeah.
00:02:52
John S
you know, like ah ah the size of an old drill, like a Derek's drill sharpener, something that's like the size of a laser printer. But, you know, it may only have a three inch area, but you could put a grinding wheel on there and do precision CNC grinding on it.
00:03:05
johngrimsmo
And all you need to do is take off a tent at a time and it's fine. Like who cares how long it takes? That's not the point.
00:03:10
John S
Yeah, right.
00:03:13
johngrimsmo
Interesting.
00:03:13
John S
It's gonna, yeah it's gonna be like a 30 times spark out option because, yeah, no, but, ah yeah.
00:03:16
johngrimsmo
Yeah, like who cares? I want auto dressing. I want a Raspberry Pi controller. I want, yeah, it's cool.
00:03:22
John S
Yeah, it's it's really fun.
00:03:25
johngrimsmo
Yep. And like, as I was driving to work, I was thinking about that video and possible applications for it. And I'm like, okay, first of all, I'm not, I'm not here to build tools.
00:03:36
johngrimsmo
Like this is a major distraction.
00:03:38
John S
Unsubscribe.
Mod Vice Rework Process
00:03:40
johngrimsmo
Yeah, exactly. But I don't know how can it be used?
00:03:44
John S
No, you're right. Well, look, how it can be used is we have built processes around our mod vices where when we fail, we try to fail tolerances to where they can be reworked. And the rework works well as for our Okamoto grinder. I think when I got that, I was like, I don't want to do this. And now I realize it's actually really simple. Like we can regrind 16 jaws at a time so you can work through I'd like to think we don't have that much rework, but you know look you know we we will we told tight tolerances and so if it's a 10th out, we just put in the rework bin and we grind them, no big deal. Well, that's taking up a $80,000, 10,000 pound grinder to take a 10th off a fixed dimension. like That could probably be done on a little like,
00:04:28
John S
3D printed Dremel-ish style. I'm sure there's limitations I'm not thinking of here, but even as so coincidentally, Silo, ah who's, the is that his name's the mirror diamond turning lathe guy, Silo's Garage.
00:04:35
johngrimsmo
Sure.
00:04:43
johngrimsmo
Uh, I forget his first name or his actual name.
00:04:46
John S
Cirrus or something. I apologize if he's listening.
00:04:47
johngrimsmo
Yeah.
00:04:50
John S
He talked about Spindle. design and spindle error and motion and how if you're doing spark out on something like a grinder, you can ignore certain aspects of of that.
00:05:00
johngrimsmo
Hmm.
00:05:01
John S
Um, so the juices were flowing.
00:05:03
johngrimsmo
Cool. I haven't seen his latest video yet, but I think it's, it's opened in a new tab for my weekend viewing.
00:05:11
johngrimsmo
Cool.
00:05:11
John S
Speaking of that, can you give, I, I, I'm not sure when I'm going to have an hour and a half to watch your latest video. Everybody else seems to love it. So I'm on the anomaly.
00:05:19
johngrimsmo
Yep.
Field Clip Development Journey
00:05:20
johngrimsmo
Yeah, exactly. You know, Ryan was editing it here and he's like, I've done a rough cut and it's like close to two hours. I'm like, good. Um, yeah, long story short, it's, it's the final development stage of the field clip where I'm basically just testing and tuning for an hour and a half of edited footage.
00:05:38
John S
Yeah.
00:05:39
johngrimsmo
Um, going through a lot of the theory, a lot of the talks I did the, um, Adam, the machinist yellow posted note cutouts.
00:05:48
John S
No way.
00:05:49
johngrimsmo
Yeah, totally.
00:05:49
John S
Oh, like the paper, the construction paper.
00:05:50
johngrimsmo
the paper things to be able to visualize a dovetail, a taper fit, things like that.
00:05:51
John S
Yes.
00:05:56
johngrimsmo
And I think I did a bad job at it, but it it kind of got the point across. It's my first time. um
00:06:02
John S
and think you're they're I think they're supposed to sort of look bad. You know what I mean?
00:06:05
johngrimsmo
Yeah. Yeah. I think I kept switching between a two-dimensional object and then using a two-dimensional piece of paper as a three-dimensional, um, explanation.
00:06:15
johngrimsmo
And I think that's where I got a little stuck, but whatever. Um, but yeah, the video is, it starts with the clip having wiggle and not making me happy. And then it ends with the clip having no wiggle and making me very happy.
00:06:27
John S
Good.
00:06:28
johngrimsmo
What's up?
00:06:28
John S
That should be on the hiccups.
00:06:29
johngrimsmo
So it's just a big, uh, a big struggle with me in the shop while I try to figure this out kind of video.
00:06:29
John S
That's great.
00:06:36
John S
Yeah.
00:06:36
johngrimsmo
And ah feel free to skip through it as much as you want, but you don't have to watch every second of it, you know?
00:06:42
John S
Well, but that's the thing. I actually ah but obviously me goes out saying we're friends and but on a personal level and a professional level. And I enjoy seeing what you do. but I sat down at breakfast this morning and thinking, Oh good. Here's a 14 minute knife making Tuesday.
00:06:53
John S
I can bang through.
00:06:54
johngrimsmo
Right.
00:06:54
John S
I'm like, well, not not getting through that one right now.
00:06:56
johngrimsmo
Yeah, exactly, exactly.
Mechanical Designs: Flexure Clips
00:06:59
johngrimsmo
Well, yeah, people seem to enjoy it, which is good.
00:07:01
John S
So you're happy with, I saw the intro where you talked about there's guide. There's a, I would call it a flexure, like kind of a solid metal thing that's thinned out such that it can a flex probably a terrible description and then yeah.
00:07:13
johngrimsmo
in the clip, right? Yeah, I'm calling them ears. There's two little ears that spring load in and hook on these features on the handle. So as the thing slides in, they they compress and then they come out and then they expand into the to lock it into place kind of thing.
00:07:21
John S
Okay. Yeah.
00:07:28
johngrimsmo
um But the trick I think we talked about last week is that taper wedge, where a five degree wedge goes into another five degree, you know, female wedge and they stop, they come to a dead stop pretty much.
00:07:33
John S
Okay.
00:07:40
johngrimsmo
um Which has been phenomenal.
00:07:41
John S
and the Okay, so you have two wedges going to each other, but there still has to be something that's stopping the wedge from backing out. Is that the flex?
00:07:49
johngrimsmo
That's the ears. Yeah.
00:07:50
John S
You know what's funny is in some respects, this is very analogous to the dual taper contact of the puchuck.
00:07:57
johngrimsmo
Yeah.
00:07:58
John S
Yeah, I mean, one doesn't work without the other. um ours is Ours is fully rigid.
00:08:00
johngrimsmo
Exactly.
00:08:03
John S
The idea is that the mating part will flex a little bit to the extent that you have a couple of tenths of deviation. But yeah, I mean, it's totally, and we've been doing more testing on the mechanical one, which is like, I wanna like scream
Metrology Setup Innovations
00:08:16
John S
at the top of the mountain.
00:08:16
John S
I love this thing so much. And Alex and and Grant and everybody's done such an awesome job on it. Like when you, forget what Alex did, Yeah, we built a, sorry, I'm robbing, stealing the conversation here.
00:08:28
johngrimsmo
No.
00:08:29
John S
We built a new metrology set up for measuring pull side diameters along the taper, which Alex also did a great job of.
00:08:38
johngrimsmo
Oh.
00:08:39
John S
I should honestly get a, I don't know if I want to do a video on that, maybe an Instagram thing or something, but.
00:08:45
johngrimsmo
Five minute YouTube video would be amazing on that.
00:08:48
John S
Yeah, it's pretty ingenious buying off the shelf components to to avoid angle errors and basically measuring along the taper for diameter.
00:08:48
johngrimsmo
Just talk about it. Yeah.
00:08:55
John S
Kind of like, think of it like thread wires to measure thread cutting, but except for big, big pull studs, you know, one inch pull studs and doing this whole thing on a fixture plate, on a granite plate.
00:08:55
johngrimsmo
Yeah.
00:09:08
John S
So we found we intentionally turned to two pull studs at the extreme ranges of our low high, which is a very tight tolerance band, but nevertheless, there's tolerance variation to make sure if the pull stud is quote unquote too big, then that bigger diameter pull stud won't sit all the way down in the puck chuck taper, which means the Z datums may not be touching or could have an issue and so forth.
00:09:35
johngrimsmo
And you won't have dual contact at that point.
00:09:37
John S
Bing, bingo.
00:09:38
johngrimsmo
Yeah.
00:09:38
John S
And basically we did some testing around all that and we're good. Like we're good. It's awesome. I mean, it's immediate full lockup and lockup in a good way, like securing it.
00:09:51
johngrimsmo
Yeah, yeah, yeah.
00:09:51
John S
And then it pops out like so cool.
00:09:54
johngrimsmo
Nice. Yeah, I
Engraving Techniques: Macro Code on Blades
00:09:55
johngrimsmo
can see. I mean, in a way, i'm I'm creating a dual contact where one is the two tapers that have to mate. They have to they have to go till they bottom out pretty much.
00:10:05
johngrimsmo
But the second feature is the outside of the clip. I want it to match visually with the outside of the handle. And there's you maybe a thou of wiggle room but before you start to feel it.
00:10:11
John S
yeah
00:10:16
John S
Yeah.
00:10:17
johngrimsmo
So I'm trying to hold ah ah a fairly really tight taper tolerance and a fairly tight outside radius tolerance. And so everything kind of matters.
00:10:27
johngrimsmo
So I'm using the same tool to finish all features, and it kind of you learn these things, right?
00:10:29
John S
Bingo. Yeah.
00:10:31
johngrimsmo
You don't know.
00:10:31
John S
Yeah. Yeah.
00:10:33
johngrimsmo
and Yeah. And then I also had to create a measurable feature of this thing to be able to have, it's all tapers and weird and organic shapes on the outside and all of what do you measure?
00:10:42
John S
Uh-huh.
00:10:43
johngrimsmo
So I created a straight section that I can make. And then I'd put that to a nominal tolerance of 1995 and that's my zero, you know, plus or minus a few tents, but that's one of the last, my goal is this, you measure this one feature and if it's within range, everything else fits.
00:10:47
John S
Yeah, yeah.
00:10:50
John S
Yep.
00:11:00
John S
Yeah.
00:11:01
johngrimsmo
That's so funny. We're both kind of going through that process right now.
00:11:06
John S
When like pulling in some of my own. ah Lessons I've learned the hard way it's like okay it's okay if you do things slower right now do them in the way that you know is going to be exceptionally good and you can consolidate the process later so.
00:11:18
johngrimsmo
Speaking my language.
00:11:19
John S
we're. milling the puck and the puck base and all the other parts. Then we're heat treating in-house right now. We'll we'll start sitting this out. Actually, we're building batches right now to be sent out. And then we're grinding. We're basically grinding three times. I think we can eventually go to grind once, but don't worry about it. Grinding three times lets us break it down. We can do one thing at a time, do it really well.
00:11:43
John S
And then the hard milling happens at the very end, which establishes all the final criticals. It's one tool, and it's, you know, one tool that does all these features, and it is in the Rego fix holder, it's
00:11:59
John S
What was I going to say about that? It's hard to say. I know it's working great. but The reason my grandma laughed at my monitor was angled when we started this call today. And it's because the and NS tool, which is really the Micron tool rep was just in here.
00:12:13
John S
We just met him for the first time and we were shown in the shop and the sample parts.
00:12:15
johngrimsmo
Nice.
00:12:19
John S
what hes Oh, he just, he just, he complimented the shop, which is always nice to hear.
00:12:23
johngrimsmo
Yeah.
00:12:24
John S
Sorry. Yeah.
00:12:25
johngrimsmo
Especially people who who get boots on the ground in a lot of different companies, in a lot of different places, you're like, it feels good to know.
00:12:28
John S
Right. Yeah.
00:12:31
johngrimsmo
like
00:12:32
John S
When he, we gave him the quick shop tour and, you know, which it's great, but it's like, again, a guy like that seen plenty of EF2s, like, you know, even horizontal. And then he, we were out of the corner and he saw the Willman and most people don't know the Willman.
00:12:44
John S
He's like, Oh wow. He's like, we, you, wow. We know those guys. Like that's awesome. I think he's expected to see it, which is kind it's kind of fun.
00:12:51
johngrimsmo
Yeah, yeah, exactly.
00:12:52
John S
Yeah.
00:12:52
johngrimsmo
Good, good view.
00:12:55
John S
So you're, where are you at now? Or maybe ah is it spoil the video? Sorry.
00:13:00
johngrimsmo
um
00:13:00
John S
I'm like,
00:13:00
johngrimsmo
ah The clip is, there's a couple toolpath, visual toolpath things that I've been tweaking. I think I've made the tweaks, I'm ready to make another clip. um But as I tackle each of these projects,
00:13:09
John S
Okay.
00:13:13
johngrimsmo
I get to a point where it's like, I'm good. like I'm ready to you know essentially start producing. Tolerance-wise, everything's good. I can fine-tune surfaces and things like that over time.
00:13:20
John S
Yeah.
00:13:23
johngrimsmo
um So the clip at this point is good. I haven't made any sellable clips yet, but I'm ready to do that. um However, the Willeman is currently in pieces getting all of the airlines replaced.
00:13:36
johngrimsmo
um which thankfully, Angelo and Jeff just took on. They're like, yeah, we can do that.
00:13:42
John S
Great.
00:13:42
johngrimsmo
And so I come in every day and it's just, there's more airlines um on the ground and in the garbage can and new ones. And they said doing it one by one is absolutely the answer. It's tedious, messy, oily, chippy, dirty, but it's not hard.
00:13:58
johngrimsmo
It's just one by one lets you really snake it and go, you know, entry to exit and make sure they match.
00:13:58
John S
Yeah.
00:14:06
John S
like Right.
00:14:06
johngrimsmo
and and And they're labeling as they go.
00:14:07
John S
Good.
00:14:08
johngrimsmo
And so far it's going really well. And I'm not touching it, which is awesome.
00:14:12
John S
It's great. Even better.
00:14:16
johngrimsmo
So I can't make clips until that's done. I have some clamps I need to make on the Wilhelmin when that is done. so that But that should be done probably tomorrow, which is good.
00:14:25
John S
Good. Oh, good. That's not bad.
00:14:26
johngrimsmo
Yeah.
00:14:27
John S
That's a huge win.
00:14:29
johngrimsmo
Yeah, exactly.
00:14:29
John S
Sweet.
00:14:30
johngrimsmo
And then it should be like good for another 20 years, at least airline wise. And that like with any old car or whatever, you're like, it's good until something else breaks, but it's fine.
00:14:35
John S
Yeah.
00:14:38
John S
Right.
00:14:41
johngrimsmo
And then same with the hard milling. um I tested it um on one blade and I was like, sweet, it works. And then I did other stuff. So I've hard milled once and it worked great.
00:14:53
John S
It was okay.
00:14:54
johngrimsmo
But i I have more to do.
00:14:54
John S
Yeah.
00:14:56
johngrimsmo
I have lots more to do. and just haven't done it That hasn't been the biggest thing right now.
00:14:57
John S
Yeah.
00:15:01
johngrimsmo
So I've been making more fixtures. I changed ah changed around the way I'm holding some of the parts and strategizing them. um I was making the initial blade, the soft blade like before heat treat on the speedio.
00:15:17
John S
Yes.
00:15:17
johngrimsmo
which which goes insanely fast with the speedio tool changer. um But the plan there was to make it in speedio, pallet change it away, schedule it again later to have the current pick it up and then engrave the serial number because the serial number is a five axis tilted op.
00:15:30
John S
Mm-hmm.
00:15:34
johngrimsmo
That's a four axis static tilted op.
00:15:35
John S
Yeah, yeah. Yeah, sure.
00:15:39
johngrimsmo
But on the current is where I have all the logic to do the automatic serial number count up engraving. ah steal choosers with rights rw alert, rights damasteel, rights whatever.
00:15:51
johngrimsmo
So I worked on that over the weekend and then I realized the speed savings that I'm saving on the on the speedio is not quite offset by the time it takes to palette change it away and for us to like manually reschedule it.
00:16:06
johngrimsmo
um So I'd rather just make the blade in the kern
00:16:07
John S
Oh, yeah, right.
00:16:11
johngrimsmo
because we make all of our other blades in the current, and and all the tools are in there, and tool life is managed very well, and surfaces, and tolerances, and all that.
00:16:14
John S
Yeah.
00:16:17
johngrimsmo
So um that's what I've been working on this week, because I made ah reprogrammed all that for the current. So now the blade is getting made in the current. It's significantly slower than the speedio, surprisingly.
00:16:28
John S
Really?
00:16:29
johngrimsmo
Yeah, the tool changes are just are just that much slower.
00:16:29
John S
but
00:16:31
johngrimsmo
It's not 0.7 seconds. It's several seconds.
00:16:34
John S
Three sec, yeah, right.
00:16:35
johngrimsmo
Yeah.
00:16:35
John S
And there's that many tool changes.
00:16:37
johngrimsmo
Yeah. I mean, there's 10, 20 and something, but I, I kind of want to like, no, this is as a six minute part.
00:16:38
John S
Huh.
00:16:41
John S
This isn't the like 57 minute.
00:16:47
John S
Okay, there was at one point you were pulling something off the current because it's like an hour long, like the chop grinding or whatever.
00:16:47
johngrimsmo
Um, but like, yeah, blade grinding. Yep.
00:16:53
John S
But you're still doing that on this video.
00:16:54
johngrimsmo
and And that's perfect in this video.
00:16:55
John S
Okay. Yeah, right, right, right.
00:16:57
johngrimsmo
Yep. it's It's not quite as accurate of a form as the as this the current.
00:17:05
John S
Yeah.
00:17:05
johngrimsmo
like you can You can see a tiny bit of faceting, but it's good enough. It's fine. um So then the other night, I copy and pasted the RASK serial number engraving macro code, which is like many, many hundreds or thousands of lines of code that says, okay, the counted serial number is you know three, four, five, six.
00:17:09
John S
Okay.
00:17:14
John S
Okay.
00:17:28
johngrimsmo
And then it breaks it down into three is the first digit, four is the second digit, five is the third digit, six is the fourth digit. And then it jumps down into the code to the file where it engraves a four, spaced over four times.
00:17:41
John S
okay yeah
00:17:43
johngrimsmo
And I've been doing that for years and it's pretty complex code. So I copy the rask and change it over for the field. New fixture, new rotations, new angles, things like that. um and spent a couple hours doing that.
00:17:56
johngrimsmo
And then I went to test it and it went great. I had one error at the very end where I forgot to do a semicolon and I just did brackets instead and the curve like that. That's fine.
00:18:06
johngrimsmo
um And then I go look at the serial number and it should have engraved 0000 because the serial number macro was empty.
00:18:11
John S
Okay, yeah, yeah.
00:18:13
johngrimsmo
And it engraved 3480. And I'm like, hold on.
Fusion CAM Variables for Efficiency
00:18:19
johngrimsmo
ah
00:18:19
John S
I've made more of these than I realized.
00:18:20
johngrimsmo
but Where are you getting those numbers from? And then I looked in the macro list and the rasp blade is 3, 4, 8, 6. So the first three digits were the same, the fourth digit was was correct, is the the zero.
00:18:32
John S
Yeah, yeah,
00:18:33
johngrimsmo
And so I looked through my macro again and it turns out I forgot to change the macro field for the first digit, second and third digits. And then I haven't tested a second one, but I'm confident.
00:18:45
johngrimsmo
I got it nailed now. And you kind of look through, you test them tune and tune, right? You find your mistake and you're like, oh, idiot. Okay. Well, it should work now.
00:18:53
John S
Yeah, that's a very, I think that's a very normal human approach of, yeah.
00:18:54
johngrimsmo
So it's good.
00:18:57
johngrimsmo
It's kind of fun and silly to 3000. Wow.
00:19:03
johngrimsmo
um But in this process, I designed instead of a single blade fixture, I designed a two blade fixture that patterns them 180.
00:19:11
John S
Yeah. Yeah,
00:19:12
johngrimsmo
So I'm, even though it's in the somewhat slower speed current, I can optimize two at once instead of, you know, you're consolidating totally, which makes a big difference.
00:19:19
John S
yeah amortizing the tool changes, sure.
00:19:24
John S
That's literally most of my life this past week has been, but I'm just going to call it patterning hell and fusion where undoing component pattern on those valve covers to sort of avoid having to program the same part twice because um I'm going to do a little video on this because it's worth everything I've learned on tricks to deburr parts in Fusion, but suffice it to say that these valve covers have, you know, it's like 80 different things to click across multiple different deburring operations and
00:19:55
johngrimsmo
yeah Yeah.
00:19:57
John S
You're not going to catch it all because of how small the chamfer is. I guess, now that I'm talking about it loud, I could program the chamfer width as a variable. I could increase the variable from like 5,000 to 30,000. Then I could simulate to see where I'm missing it. It's actually not a bad idea now that I talk about it. I'm going to make that as a note.
00:20:17
johngrimsmo
I don't know if I follow.
00:20:18
John S
Okay, hold on one second, make it as a note. Program, program, huge chamfers just for simulation. So you know in Fusion now, you can link CAM values as variables.
00:20:32
John S
So like, let's say you have 17 different whole operations. Instead of doing whole bottom, you could set ah you can set each one to be negative 0.83 inches and it's all linked back to one variable.
00:20:44
John S
CAM variables.
00:20:45
johngrimsmo
I don't know if I've seen this yet.
00:20:47
John S
Oh yeah.
00:20:48
johngrimsmo
I know you've been asking about it for a decade now.
00:20:50
John S
So yeah, um i to be totally honest, I did ask for it for a long time. I haven't really used it. So there's like a 1% chance I'm lying right now. I don't think so. Pretty sure this works. In fact, I will test it when we hang up because what what I could do is I want 5,000 chamfers. I could program them as a variable and increase them to 50,000 chamfers and then that would let me see in simulation more easily see where I missed a freaking selection of a .22 inch section that didn't get chamfered and then go with red, yeah, exactly, or something.
00:21:13
johngrimsmo
Oh, okay.
00:21:20
johngrimsmo
yes because you'd see you'd see a million holes that are big chamfers and you'd see one with no chamfer or yeah okay interesting that makes sense yeah
00:21:29
John S
um But the component pattern works, but component pattern is fussy. Like if you do copy, paste new it will in the CAD side to duplicate a part.
00:21:42
John S
Copy, paste new ah doesn't work for a component pattern. Component pattern will not honor that relationship. Whereas if you just duplicate or copy, paste, for those that are fusing that beginners, paste new like means a change to one.
00:21:56
johngrimsmo
Breaks.
00:21:57
John S
Yeah, yeah.
00:21:58
johngrimsmo
Yeah.
00:21:58
John S
So, but that's just annoying because sometimes you're like, why isn't my component pattern working? And then you got to redo the joints, you reduplicate it, but yeah. It is now i'm now, I am not happy with it. I'll say that.
00:22:08
johngrimsmo
Good.
00:22:10
John S
Yeah.
00:22:12
johngrimsmo
Yeah, cam variables would be interesting.
00:22:17
John S
yeah wish i Yeah, I should not do this in the middle of our talk, but let's let's make sure that's a real thing.
00:22:22
johngrimsmo
Yeah, maybe it's not done.
00:22:22
John S
but um the Oddly enough, you know I've definitely asked often for viewer input or viewer help. Of all the things, the deburring stuff was like, drew a huge response from folks. So um I think last week or two weeks ago, I talked about using the deburr toolpath infusion to um automatically deburr. And i I said something that was not true, and i but I didn't mean to say it the way I said it, which is that you can't control the height of the deburr
00:22:53
John S
ah The tool on the deeper and that's not true. You can there's a slider that's like zero to 99 It's a little bit weird because it's meant more along the angle like if you were swerving Something but it would still work But my point which was corroborated by folks that wrote reach out was that it's just still fussy Like sometimes it won't reach
00:23:02
johngrimsmo
Okay.
Deburr Toolpath in Fusion
00:23:12
John S
certain areas.
00:23:12
John S
There's some chamfers that are super tight. So I need to really specifically have a almost at the very tip or almost at the very edge of the chamfer tool to get it to work to fit in the hole or the feature or to maximize the amount of chamfers along a short slot by, you know, almost chamfering where the taper meets the square.
00:23:31
johngrimsmo
okay
00:23:31
John S
So it's just fussy and it didn't look great. I didn't like it. And that's kind of why I bailed on it.
00:23:36
johngrimsmo
And are you deburring with, you're using the deeper toolpath with a chamfering tool?
00:23:36
John S
Yeah.
00:23:41
John S
Which you effectively have to on a three axis. I mean, you could use a ball, but
00:23:45
johngrimsmo
Yeah, I've only used it with a ball.
00:23:48
John S
Yeah, it works for the champion.
00:23:49
johngrimsmo
Cool.
00:23:49
John S
Yeah. I mean, a ball that you're not servicing would technically create a half pipe, but.
00:23:55
johngrimsmo
Yeah, depending on how big the chamber is, like a little five thou edge break kind of thing that you can't even tell really.
00:23:57
John S
yeah Tiny, yes. Right. Right. OK.
00:24:06
John S
Yeah, so let's see here. I'm in fusion axial sucks, so let's hear. Let's go to. um
00:24:17
John S
If I do this, let's hear, like, stock to leave negative.
00:24:22
johngrimsmo
doing this live, people.
00:24:22
John S
just
00:24:25
John S
Expression, reference parameters. Oh, man. Well, I but i don't want to waste time. let me look Let me report back on this. I'll throw something up on Instagram, too, more quickly on our next podcast, in case anybody's super anxious, because obviously this is a huge feature.
00:24:43
John S
um Again, this is an exact example. like I have right now on this part, um I have open, I have six or eight different chamfering operations for a part, and you may want them all to be 10,000 chamfers, but there's eight different operations. If I want to change it from 10 to 13,000, there's no reason that can't be a can variable to keep those linked.
00:25:02
johngrimsmo
I mean, you can do the compare and edit, select them all, compare and edit. I do that all the time.
00:25:06
John S
True. Yeah, fair point. ah Probably another fusion feature that's frankly underutilized. the yeah
00:25:12
johngrimsmo
Yeah, probably. Yeah.
00:25:15
John S
Um, this, so the NS tool rep was in, he's really a micron tool rep. They happened to also be the exclusive NS tool. I was asking him about tool life, like what's expected for tool life, time in the cut, you know, any of that sort.
00:25:27
John S
And he basically had no idea.
00:25:29
johngrimsmo
Yeah.
00:25:29
John S
Um, so if you learn anything or I mean, I thought about texting, I got like Adam, but I just see anybody else reach out first before I bother him. Um,
00:25:39
johngrimsmo
I think it sounds like it's so application specific, um, both the exact steel, yeah, I know, but the exact steel, but also like how you're cutting it.
00:25:42
John S
Okay. I hate that answer though.
00:25:47
johngrimsmo
If like in, in the Maldino magazine catalogs, they basically say, do you want a two thou step over two thou step down, whatever.
00:25:54
John S
True.
00:25:58
johngrimsmo
And it just, these parameters and, and with that in that steel should give you a fairly consistent life, but every application is different. Um, if you're only floor finishing.
00:26:07
John S
I'm still trying to figure out
00:26:09
johngrimsmo
Yeah.
00:26:10
John S
It's like a 60 second cycle time for this part. And let's assume I'm doing the correct, hey it's only 2,000 engagement. um yeah Should I expect 50 parts, which would be about 50 minutes of cutting, or is it way less, or is it 500?
00:26:22
johngrimsmo
me yeah Exactly.
00:26:25
John S
You hear about some of these guys doing these die applications, which I think is what Adam has said in his videos, in certain applications with certain tooling, and obviously super rigid machine, consistent thermal, no run out holders, and they're getting really, really long tool life, is my understanding.
00:26:39
johngrimsmo
But yep, I think you should be able to get one hour plus plus is my is my educated opinion.
00:26:45
John S
Okay, let's get it out.
00:26:47
johngrimsmo
And I'm looking at the same thing, like how many blades can I get before I gotta replace a tool? And what variable am I looking at to determine the where? Am I inspecting the tool every 10 parts?
00:26:57
John S
Yes.
00:26:57
johngrimsmo
Or am I looking at the parts? Or am I gauging tolerances? Like we we're using pin gauges for some of the features. um So for the first 100 blades, we're just gonna track everything, we have to.
00:27:08
johngrimsmo
um We're going to inspect the tool every, I don't know, 10 blades or something. But ideally the same person is doing that inspection of the tool because it's a visual thing. And you're like, how do you how do you know what's what?
00:27:22
johngrimsmo
What's dead and what's not?
00:27:23
John S
You literally took the words out of my mouth.
00:27:25
johngrimsmo
oh And I get it.
00:27:25
John S
I'm like, how do I even know?
00:27:26
johngrimsmo
I'm in the same place. um
00:27:28
John S
I assume service finish will start to go, but even that can be a frog boiling in water, you know?
00:27:31
johngrimsmo
I don't know any hard. I think hard will stay shiny all the time. It might get a little smeary or a little not as shiny if it gets really dull.
00:27:39
John S
Okay.
00:27:40
johngrimsmo
But I mean, what we hard mill um the Norseman blades with a 3.8 ball mill, not hard milling grade.
00:27:47
John S
Yep.
00:27:48
johngrimsmo
It's just like an off-the-shelf lakeshore carbide 3.8 ball.
00:27:52
John S
Yeah.
00:27:52
johngrimsmo
um Not the right tool for the job, but that's what we're using. It works works okay. And that tool gets absolutely trashed, like chips everywhere. Like the edge is just, it looks like a ah serrated edge.
00:28:05
John S
Yeah. alert Hilarious.
00:28:05
johngrimsmo
um And it still leaves a shiny finish.
00:28:09
johngrimsmo
So I don't know if shininess is the answer there. um But depending on how you use the tool and how you fif see it, you might see streaks in your finish, like a microchip in the tool might transmit a chip in your finish.
00:28:22
johngrimsmo
um And for you, I would just inspection like the way you're gauging the parts and see how that holds over time.
00:28:22
John S
Okay. Yeah.
00:28:31
John S
So we bought a Haas 3 16th end mill that they say is up to 55 Rockwell, which is about two points under where we're at. So it's for sure not.
00:28:41
johngrimsmo
I'm not used to hearing a Haas end mill.
00:28:44
John S
Yeah, I know, right? um But it's it was, I mean, look, it's super cheap here the next day, like, but whatever, great. And they are working great, but what I've noticed is, like your three-eighth inch lakeshore, like, this tool looks like it's lived a rough life in the, like, six parts it's cut. And I didn't see anything apparent other than, again, the flutes look terrible, and then I quickly realized the way you were gauging it, you're getting an extreme amount of taper.
00:29:13
johngrimsmo
ye
00:29:14
John S
which number one, I'm gonna switch that out, which actually kind of stinks. We'll get another and NS tool and it's just like, it's gonna be three weeks distributor in Japan. And so Mitsubishi had one here in two days and it's like, hey, same price.
00:29:22
johngrimsmo
Yep.
00:29:24
John S
Like, so we're gonna try the Mitsubishi or Maldino on that one.
00:29:27
johngrimsmo
ye
00:29:27
John S
um I still might use those Haas's to rough out um that slot. I don't really, honestly, I might go, good.
00:29:37
johngrimsmo
What? Three sixteenths you said? ah
00:29:40
John S
A four millimeter, yeah.
00:29:42
johngrimsmo
And it's a slight, like how how deep, how wide.
00:29:45
John S
It's an eight millimeter, it's the clocking pin slot on our puck chuck.
00:29:49
johngrimsmo
Okay.
00:29:49
John S
So it's eight millimeter wide, um about eight millimeter long, open-ended slot that's ah four millimeters deep, I think, about 0.2 inches.
00:29:56
johngrimsmo
Okay.
00:29:57
John S
So we don't, do anything with that feature prior to heat treat. So that whole feature is cut his hard mill.
00:30:03
johngrimsmo
Yep.
00:30:05
John S
So I'll probably do a beater and mill to rough it out from Haas. I'll probably use another Haas duplicate tool to do a semi finish and then the Mitsubishi. Oh, no, shaking your head.
00:30:15
johngrimsmo
No, no, I'm in the same thing.
00:30:16
John S
Well, talk to me.
00:30:17
johngrimsmo
Um, cause on the field blades, I, I tried to like ah a soft machining, the slots of the slot was there. And then I used just a beater end mill to clean it up and the beater end will dies after one.
00:30:30
johngrimsmo
part
00:30:32
John S
Interesting.
00:30:33
johngrimsmo
and And hard milling tools are just the bee's knees.
00:30:34
John S
They just work.
00:30:36
johngrimsmo
So I ended up with a four flute. I think it's a four flute, a two millimeter high feed end mill from Maldino. And I'm high feed milling this slot.
00:30:44
John S
Okay.
00:30:45
johngrimsmo
And it takes, I think it takes a minute or two. Cause I'm going down three thou at a time through an eighth inch piece of material.
00:30:51
John S
Yeah.
00:30:52
johngrimsmo
um but So what? um You could get a four millimeter tool and probably fly it a lot faster. The tool is a hundred dollars, but probably going to last a hundred parts.
00:30:59
John S
Yeah.
00:31:03
John S
Okay. So you're saying just don't dick around with the non.
00:31:03
johngrimsmo
Where's your, that's where I'm at anyway.
00:31:07
John S
Yeah.
00:31:08
johngrimsmo
Yeah.
00:31:08
John S
Fair enough.
00:31:09
johngrimsmo
Cause there is some, there's magic to these hard milling tools, the grade, the coding, the grind, the geometry, like they're made for it.
00:31:09
John S
Okay. I'm fine with that.
00:31:18
John S
I still wanna know, this host the three 16th host tool is $11 though. And I'm able to remove, this slot is about the size of your pinky nail.
00:31:24
johngrimsmo
I hear you.
00:31:28
John S
So in about, but literally about as deep as your pinky and the size of the nail.
00:31:28
johngrimsmo
Yeah.
00:31:32
johngrimsmo
Okay. Not big.
00:31:33
John S
So if I can remove 98% of that with an $11 end mill and just bring in the hard-built tool to finish it up, still feels like a win. But look, I hear you, like,
00:31:42
johngrimsmo
But there's also, there's a real thing to consistent stock to leave for your finishing tool.
00:31:47
John S
Yes. Tool pressure.
00:31:48
johngrimsmo
If you wanna hold one 10th on this feature, then you're you're your tool is gonna leave a wall taper. And now your expensive finish tool is gonna come in and cut harder and harder and harder and harder.
00:31:59
John S
Yeah.
00:32:02
johngrimsmo
And that's that's hard to to consistently rely on, you know?
00:32:04
John S
No, it's a very, a very apt point.
00:32:08
johngrimsmo
Yeah.
00:32:08
John S
I hear you.
00:32:09
johngrimsmo
Because like reading all this documentation about making molds, they're like, stock to leave is is critical. They do like roughing with hard milling tools, semi-finish with hard milling tools, semi-semi-finish with hard milling tools, and then you come in with a PCD, whatever, MCD, and you just finish it kind of thing.
00:32:22
John S
Yeah.
00:32:27
John S
Yeah.
00:32:30
John S
Good boy. And some of this will be fixed by finally getting the four millimeter Mitsubishi or Maldino in. I was, and I was, I didn't have the, the only other tool I had was that tiny two millimeter, which does a different feature on the part.
00:32:36
johngrimsmo
Yeah.
00:32:43
John S
So that'd be good.
00:32:45
johngrimsmo
Unless there's a way to use, uh, on the same minimal in two different ways. I've done that before, like use the floor or the corner radius for one feature and then the sides for another feature.
00:32:50
John S
Mm-hmm.
00:32:54
johngrimsmo
And you're like, sweet, double it up.
00:32:57
John S
We just started doing that on, I don't know if you guys have hat tops in your production, like when you flip parts.
00:33:03
johngrimsmo
Uh, sometimes, yeah, where you just got to rough it out, like.
00:33:07
John S
What I started doing, I don't know why I didn't think of this earlier, was, you know, we have beater end mills, like roughers that just, they do their job.
00:33:13
John S
We still don't usually end up using the full shank of those roughers.
00:33:14
johngrimsmo
Yep.
00:33:17
johngrimsmo
who
00:33:18
John S
In mills, like if it's a one inch, three-eighth inch diameter, one inch flute length rougher, a lot of times we're only using the first three quarters of an inch. I just programmed my hat tops to cut with the very top of the flute where the flute start transitioned into solid carbide, because that's otherwise a virgin at cutting edge.
00:33:28
johngrimsmo
Yes. Yep.
00:33:35
johngrimsmo
Yeah, I do that. I forget what for what op. But you realize you're not using this tool, and you're replacing it regularly anyway, this this portion of the tool.
00:33:43
John S
Yeah.
00:33:44
johngrimsmo
like That's perfect for this job.
00:33:46
John S
When you always beat the heck out of the the tip of the tool, like you're always using something on down there.
00:33:51
johngrimsmo
Yep, yep. That's good. I've got a hard milling the Norseman blades. There's this radius lock face. and I'm using a one eighth inch ball in three different five axis orientations to use different parts of the flute.
00:34:08
John S
Oh, sure. Sure.
00:34:11
johngrimsmo
So like, it's a feature that gets finished with the radius of the ball mill. um But I'm like, let's rough it out with the side flutes because they never get used.
00:34:19
John S
Yes.
Tool Path Optimization Strategies
00:34:20
johngrimsmo
So I five axis tilt it and use the side flutes to go and then I tilt it a different way and then semi finish it and then I get kind of that middle zone of the radius where it's like, that's my finishing zone.
00:34:31
John S
That's really cool.
00:34:32
johngrimsmo
Works great.
00:34:33
John S
The adjusting it the ah bob toolpath guy was showing off a thing he did where they wrote a script that will auto calculate the right
00:34:45
John S
chamfer tool I coincidentally this is all chamfers on the right chamfer tool offset to get to machine all of the features to avoid and they have this like sample part that's the worst case where you've got floor features you have to avoid with the tip of the tool you've got sidewall features that you have to avoid with the shank and it just calculates the best solution for everything um and then
00:34:57
johngrimsmo
Okay.
00:35:04
John S
When we're having this conversation, I was like, well, now what you want to do is number one, drive the surface footage of the tool off the effective cutting diameter instead of the OD of the shank.
00:35:11
johngrimsmo
Sure.
00:35:11
John S
And number two, start thinking long-term about some variability to where every time you repost the part, can you vary it by one or two thou to start spreading out the wear of your chamfer tool?
00:35:21
johngrimsmo
That'd be cool. Or, yeah, or program different features at a different side depth so that in one cycle, you're using more of the tool.
00:35:22
John S
A little bit complicated to actually implement, but yes.
00:35:35
John S
yes
00:35:35
johngrimsmo
That would be next level.
00:35:43
John S
Yes.
00:35:46
johngrimsmo
So speaking of programming stuff, I forget if I told you last, I think I did. The Raspberry Pi that's running all the scripting for our current.
00:35:53
John S
Yeah, the script that purges.
00:35:54
johngrimsmo
um Exactly. It hasn't been working for a week now. And the deeper I dig into it, the more I think I'm breaking the pie, like the the all the scripting in the backend.
00:36:06
John S
Okay.
00:36:08
johngrimsmo
I keep trying things and changing things. And I'm at the point now where I'm installing a fresh version of Raspberry Pi, copying in my old files, kind of starting fresh again. But I think I can get it.
00:36:19
johngrimsmo
And then I'm going to get our buddy Phil to help us help help us put the finishing touches on it, because I'm like, my Google Fu is only so so good. And then his is probably the better.
00:36:29
johngrimsmo
I'm happy to pay him for his time together. you know, bring this home for me.
00:36:32
John S
Your whole script broker, just the script you wrote that wipes the file every month.
00:36:37
johngrimsmo
Not that one. Oh, I'm trying to build that into the regular thing. Like the one I wrote was a Google Apps Scripts thingy, which kind of works
00:36:45
John S
Okay. The one that tracks toollight predictive tool life, right?
00:36:49
johngrimsmo
Yes, the one I think I talked about last week was purely an app script to delete unused rows and and make the file smaller.
00:36:57
John S
Yeah, yeah, right.
00:36:58
johngrimsmo
And while that does work, it takes quite a long time, probably hours to clean the whole script of tools, but whatever. um But the same thing could be done on the Py in the Python script automatically, very quickly, using MySQL and other big words that I know very little of.
00:37:16
John S
okay yeah yeah
00:37:16
johngrimsmo
um
00:37:18
johngrimsmo
and And yeah, like I can Google, and I can figure things out, and I can use chat GPT, which is amazing for this kind of stuff. um But i'm I'm kind of stuck. or I don't have another 10, 20 hours to put into this, so I'm going to pay Phil to do it for me.
00:37:27
John S
yeah
00:37:31
John S
I give Phil a ton of credit. Like we all got to know Phil when he was the MJK doing it phenomenal five axis parts.
00:37:33
johngrimsmo
Absolutely.
00:37:37
John S
And like he is really seems to really yeah but very sharp and like his programming skills, post editing skills, like pretty dialed.
00:37:39
johngrimsmo
Is this Mark I?
00:37:46
johngrimsmo
Yep. Yeah. He's now basically a self-taught software engineer. And so I'm, I'm sending him work.
00:37:50
John S
Yeah.
00:37:54
johngrimsmo
And I'm also going to lean on him and DSI to finalize the, um, my post for the current. So I can do it all through fusion now.
00:38:00
John S
Good. Oh, right.
00:38:02
johngrimsmo
Yep.
00:38:02
John S
Bail ditch on complete.
00:38:03
johngrimsmo
Yeah. Which expires in 30 days. So I'm like, Oh, okay. I gotta, it's going to be a busy Christmas season.
00:38:09
John S
I assume the interface is no longer like I used it five years ago when it was.
00:38:13
johngrimsmo
complete, it hasn't changed.
00:38:14
John S
Yeah.
00:38:15
johngrimsmo
I mean, maybe there's updates, but no, the intro screen is now it says Autodesk complete, but everything else hasn't changed.
00:38:15
John S
Oh, I thought honor desk kind of put their flare on it.
00:38:23
johngrimsmo
And maybe I haven't updated it, um, cause it just works and I don't care, but, but yeah.
00:38:23
John S
Hilarious. Yeah, right. It's funny.
00:38:32
johngrimsmo
So yeah.
00:38:32
John S
um Okay, speaking of shop like stuff, I think I'd ask you quickly offline, but I kind of wanted to hear your more public take on, do you do you use chatter?
00:38:44
John S
Should I use chatter? Why should anybody use chatter? What is chatter for?
Chatter Software for Machine Monitoring
00:38:49
John S
Well, I think it's a machine monitoring, but I CJ's doing maybe slightly different stuff with it, et cetera, et cetera.
00:38:51
johngrimsmo
o
00:38:56
johngrimsmo
I think Dennis was using it too, or is.
00:38:58
John S
Izzy, okay.
00:38:59
johngrimsmo
He hasn't talked about it in a long time. anyway so as as
00:39:01
John S
Well, I think Rob too, same thing, I don't know, yeah.
00:39:03
johngrimsmo
Yeah, it's it's something that you know some people use, don't talk about enough. um As I understand it, and I have been using it on and off for a couple of years now, um our buddy Pete kind of invented this software, this website, that interfaces with each of your machines, whether it's Haas, Dusan, Fanuc, Morey, Brother, Kern.
00:39:23
johngrimsmo
um and and tracks information. It's just collecting cycle times and spindles and and utilization rates and all that stuff. And he's created a sweet little dashboard to monitor what's going on. um And depending on how you want to use it, there's like your current dashboard that's like, oh, I've got three spindles running right now.
00:39:41
johngrimsmo
Sweet. And my current shop utilization is such and such. um But what I ended up using the most for was the historical data tracking, where I can look back the past week, tell me how much the current the the speedio ran and when it ran.
00:39:50
John S
Okay.
00:39:56
johngrimsmo
Oh, I'm only averaging four hours a day? What the heck? I can do better than that.
00:39:59
John S
Yeah, right.
00:40:00
johngrimsmo
Um, and it, it's like a slap in the face, the reality of that. Um, and then you look at like the tornos and you're like, sweet 18 hours of like a day or, or we can do better than that.
00:40:09
John S
Yeah, right.
00:40:12
johngrimsmo
Um, and I never got to working for the Kern and that was kind of bugging me and I never got to working for the Willaman and that was bugging me too. I was working back and forth with Pete. This was a year ago. Uh, he's probably got it all figured out by now, but.
00:40:22
John S
Yeah.
00:40:25
johngrimsmo
When it comes to all the FTP backend stuff where it gets complicated, I kind of get lost.
00:40:25
John S
Is it?
00:40:28
John S
Yeah.
00:40:32
John S
Cause I know I followed it a little bit more closely when he just came out with it and it was more like, Hey, you bought hardware. And now my understanding is he's gotten it to where a lot of times it's just software. You don't need to actually buy a hardware box for each machine.
00:40:42
johngrimsmo
Yeah. He used to send you a Raspberry Pi um with his script on it.
00:40:47
John S
Okay.
00:40:49
johngrimsmo
You plug it into your network, and it would then, I think the Pi is what did the calculations, and now he's probably using some online server to do those calculations instead, I think. um Because yeah, he created this desk desktop connector program, which I think my speedio is connected to.
00:41:06
johngrimsmo
So my speedio is not running through the Pi, but my Maury, Nakamura, and Tornos are running through the Pi.
00:41:11
John S
okay
00:41:12
johngrimsmo
And, uh, like I said, I haven't, it's one of those things that either you as the owner says, guys, we're doing this and everybody like gets on board. Um, whereas I was just playing with it myself and I never really got everybody else on boards and nobody else used it. So it's very underutilized. Um, and it's the kind of thing where you really want to dedicate a computer and a monitor and just have it.
00:41:34
johngrimsmo
open for people to see.
00:41:35
John S
Yeah, sure.
00:41:37
johngrimsmo
um Because when it's on the 14th tab of your Chrome browser on your computer and you forget to look at it, it's it's there, yes.
00:41:38
John S
Sure.
00:41:43
John S
Yeah. Yeah.
00:41:45
johngrimsmo
ah But even still tracking that information and being able to look back once or whenever um at the history and go,
Machine Utilization Reflections
00:41:55
johngrimsmo
how's the current been the past few weeks?
00:41:56
johngrimsmo
Or how's the, but you know, you as the owner can look at the whole shop and see what your utilization rate is. And that's freaking cool.
00:42:04
John S
yeah I fix mixed feelings.
00:42:06
johngrimsmo
I kind of have this love-hate with it, you know?
00:42:09
John S
Yeah. Part of me is like, ah. I measure our success in different ways than than that, but I also very much understand well we're not realizing, oh my gosh, we're really not using the VF6B machine or or whatever.
00:42:21
johngrimsmo
Yeah.
00:42:22
John S
um And certainly it doesn't take a rock scientist to start thinking about where this where can go to help with like terms of like, you know, overall like job scheduling or um truly trend tracking and point of tolerancing and yeah.
00:42:29
johngrimsmo
Hmm.
00:42:35
johngrimsmo
That's the big thing. Like it's, it's kind of easy and wrong to look at it and go. Like if I showed Eric the results, he'd be like, what, we're not running. We're only running three hours a day.
00:42:43
John S
Yeah.
00:42:45
johngrimsmo
Like a, we don't need to run more cause that's enough parts to like.
00:42:48
John S
Right.
00:42:49
johngrimsmo
You know, it's it's hard to keep a machine running for 24 hours a day, especially one that's not automated. Like, it's not impossible.
00:42:55
John S
Yeah.
00:42:56
johngrimsmo
um But there are massive benefits from it. Job scheduling, like you said, like, well, that machine is totally underutilized. And could we put this job on there that's maxing this machine? And as you get more and more machines, it starts to make a lot of sense.
00:43:11
John S
Yeah, and my understanding with the dashboard is that, let's say you have, you know, a Kern, a Wilman, a Torner slave, the knock, like, not only can you log in from anywhere or maybe there's an app and like see it, but it can even like give you alarm history information or current state information.
00:43:23
johngrimsmo
It can, it can text you when there is an alarm, which I never wanted or needed, but I know some guys love it.
00:43:26
John S
Yeah, yeah.
00:43:30
johngrimsmo
Like CJ, who sets up his Wilhelmin and goes away for 24 hours.
00:43:31
John S
Yeah. Right.
00:43:34
johngrimsmo
Um, and back when he was working at Autodesk, he's like, his night job was keep the Wilhelmin running and he's programming at Autodesk and it's, and he gets a notification and goes, oh, tool broken.
00:43:38
John S
Yeah. Right.
00:43:46
johngrimsmo
Good to know. I'll take my lunch break and go fix it.
00:43:47
John S
Yeah. Yeah. The amount of work we went through to get that to that functionality on our Kuma, you know, i standardizing that in a sort of machine tool agnostic platform would would be phenomenal.
00:43:57
johngrimsmo
Mm-hmm Yeah Yeah, so I think it has a lot of power a lot of potential.
00:44:03
John S
Interesting.
00:44:05
johngrimsmo
Um, I I wish we used it better more I'm still trying to figure out how like specifically what?
00:44:05
John S
Okay.
00:44:13
johngrimsmo
Is the
Underutilized Tools Impact on Production
00:44:14
johngrimsmo
biggest win for us? Like do I i actually literally have a Raspberry Pi set up with a 17 inch monitor? That was like meant for this to just display it all the time, but I never did and it
00:44:22
John S
Yeah, yeah, yeah. Never did, yeah.
00:44:26
johngrimsmo
It wasn't needed. And the the guys know if the machines are running or not. It's their job.
00:44:29
John S
Right.
00:44:31
johngrimsmo
So what's this for? Is this for me to just spy on everybody? Kind of. That's nice. But, you know.
00:44:38
John S
Oh, yeah, no. Well, like I don't... um Yeah, I totally hear you. um And a part of me doesn't really want to be bothered. But on flip side, if I'm coming home from dinner somewhere and I now know that there's a silly error and I can just pop in the shop literally on my way home and fix something, I guess, I don't know.
00:44:49
johngrimsmo
Yep. Yep. yeah
00:44:56
John S
It would be more interesting in the long term to start looking at the trends. And um I should ask CJ, because it's like, okay, can you start proing pulling in the probing data to start understanding deviations that way?
00:45:06
johngrimsmo
I think CJ does that a little bit. and What's interesting is, if like imagine you have a part as as we do. When we're grinding a blade, say it takes 30 minutes, and currently it's one by one cycle, we're not palette changing it.
00:45:21
johngrimsmo
So every 30 minutes, somebody's got to be there, put a new blade on, hit go.
00:45:21
John S
Yeah.
00:45:25
John S
Yeah.
00:45:25
johngrimsmo
And that lag time of like, oh, I'm having lunch, I'm in the bathroom or I'm going to the store or something, that 30 minute cycle turns into a two hour cycle because that hour and a half, somebody wasn't there to put the next blade on, you know what I mean?
00:45:37
John S
Sure. Sure.
00:45:38
johngrimsmo
So there's this theoretical maximum of how many parts you can make, but to see it graphically on a chart, Like, this is what this video did today. And it's like, oh, you can clearly tell, ran apart, two minute gap, ran apart, six minute gap, sure.
00:45:52
johngrimsmo
Ran apart, you know, two hour gap, launched, problem, machine broke, coolant swap, like who knows. um And I think you can, there's a feature where you can log, like if the operator has access to chatter, they can be like, I'm changing coolant.
00:46:07
johngrimsmo
That explains this, this problem.
00:46:07
John S
Oh yeah, sure. sure Yeah.
00:46:11
johngrimsmo
So how deep do you want to go? You know?
00:46:13
John S
yeah What I will never forget the phrase that I heard from you. I don't remember if it, you came up with it or, or it was a book, but it's kind of like what gets measured gets managed.
Measuring Work for Productivity Gains
00:46:25
johngrimsmo
Mm-hmm. Yeah, I got that from Tim Ferriss kind of thing Oh,
00:46:25
John S
I'm not. Okay. I'm not there yet, but I'm starting to understand and appreciate what that could look like.
00:46:32
johngrimsmo
yeah, I'm teaching my kids that like it's it's a big one um I want to finish up this pie stuff
00:46:37
John S
Yeah. What do you have to do today?
00:46:46
John S
Oh yeah, in.
00:46:46
johngrimsmo
I'm at the point, yeah, the Raspberry Pi stuff.
00:46:49
johngrimsmo
Like, I have a big email to send to Phil, but it only works if he can log into my Raspberry Pi.
00:46:57
johngrimsmo
Yeah, but I have, I'm putting a new Pi together, basically.
00:47:00
John S
Oh, okay.
00:47:01
johngrimsmo
Because that was the problem with the old Pis.
00:47:02
John S
he he He actually might still be able to do it.
00:47:05
johngrimsmo
Yeah. um even with Part of the reason the old one broke is because it's an older version and I couldn't set up a remote login for somebody outside of my network.
00:47:17
johngrimsmo
So that's what made me go nuclear to just start over pretty much.
00:47:17
John S
Oh. Got it, yeah.
00:47:21
johngrimsmo
So that's where I'm at. That won't take long. And then after that, it is. What is the other thing I gotta do?
00:47:31
johngrimsmo
Check a couple features on those soft blades that I made, and then I've got to redesign a top clamp to be able to hold a hardened blade onto a fixture. but with as small of a clamp as possible because I have some tool holder clearance issues on side features.
00:47:43
John S
Hmm.
00:47:47
johngrimsmo
And it's I thought I had a good clamp, but now it's way too big for what I'm now trying to do. So I have to like imagine if you had an M6 buttoned cap screw and you just use the screw to tighten that apart, so you're hooking on with the lip.
00:47:52
John S
Okay.
00:48:00
John S
Yeah.
00:48:02
John S
Yeah.
00:48:02
johngrimsmo
But I feel like just the screw could kick it sideways a little bit.
00:48:05
John S
Yup.
00:48:06
johngrimsmo
um So you want to create some sort of balanced top feature, but it has to be as low profile as possible, like not much higher than the screw head itself.
00:48:17
John S
um so okay think big here your old school bridge port where you've got the set of the stair the set of stairs and the big strap clamp and the big post coming through it and then in your big toe clamp well the toe clamp
00:48:25
johngrimsmo
yeah
00:48:32
John S
can have a raised up lip that grabs the part and the rest of the clamp can be, I don't know how much Z height you have to work with.
00:48:36
johngrimsmo
you
00:48:39
John S
And then that way your screw head, which itself could be in a counterbore, can be below the work plane. So you're kind of reaching up, like you put your hand like this and grabbing the part.
00:48:47
johngrimsmo
Yeah, yeah, yeah.
00:48:51
John S
You with me?
00:48:51
johngrimsmo
Yeah, you're absolutely right. And I realized that this morning that that is a a possibility. I was actually sitting there doodling in my notebook. You can see my little clamp sketches there, different options.
00:49:01
John S
Yes. Yeah. Bingo. Yeah. Sure. Yeah.
00:49:04
johngrimsmo
um And exactly what you said was one of them. And I'm just holding an eighth inch blade, so I have some room. but
00:49:10
John S
Oh, not much.
00:49:11
johngrimsmo
the The other theory is do I want the screw head over the part, not down beside the part, so I'm relying on the clamp itself to do all the not flex pretty much. I don't know, I've got options.
00:49:22
John S
You. Yeah. You just want to make sure it's good.
00:49:23
johngrimsmo
it It just needs to not move for hard milling. um For, you know, not super aggressive hard milling.
00:49:28
John S
Yes.
00:49:31
johngrimsmo
These are the two millimeter tools kind of thing. It just...
00:49:34
John S
Oh, I'm still treating hard milling as work holding security is critical because yeah there's no tool pressure, but I also need it to be stable for sure.
00:49:40
johngrimsmo
I agree.
00:49:43
johngrimsmo
Yep. Yeah. You're trying to hit a 10th of tolerance.
00:49:44
John S
we actually
00:49:45
johngrimsmo
You can't let it move ever. Like.
00:49:48
John S
we put ah We put the magnet, Chuck, I bought that China magnet um like it just with a the, what do you call it?
00:49:53
johngrimsmo
Okay. Yeah.
00:49:56
John S
The dumb magnets where you just use a lever bar to turn on and off, or not turn on and off to turn to function make the magnet work or not work.
00:50:02
johngrimsmo
yeah
00:50:02
John S
There's no electronics. and precision machine to boss in it turned a part that fits in that holds the center of the part so that constrains XY and then the magnet constrains Z. 3D print ah jig kind of gives us approximation on the clocking which is all we need.
00:50:19
John S
um Works great. Works great.
00:50:21
johngrimsmo
And is this for milling or grinding?
00:50:23
John S
Hard milling.
00:50:24
johngrimsmo
No way.
00:50:25
John S
Yeah. Yeah, it works real good. I'm guessing your blades aren't magnetic though.
00:50:29
johngrimsmo
There, yeah.
00:50:29
John S
Or can you hold them with a magnet?
00:50:32
johngrimsmo
You could. Don't all the chips stick to it?
00:50:34
John S
don't don't Don't turn your nose up at this.
00:50:36
johngrimsmo
I'm just that thinking, I don't know.
00:50:38
John S
So great question. I think the chips are so small plus the air blast relative to the polarity of the magnet um itself and the body that we we are not seeing any material amount.
00:50:53
johngrimsmo
Interesting.
00:50:53
John S
But yes, every time we take a part off, we're obviously cleaning the white, stoning or wiping the magnet for sure.
00:50:59
johngrimsmo
forever
00:51:00
John S
But that's still way better than clamp screws, which, you know, yucky on pressure and distortion and um yeah yeah.
00:51:06
johngrimsmo
Yeah, that's what I'm working with right now.
00:51:09
John S
Right, right.
00:51:10
johngrimsmo
I mean, holding it flat is shouldn't distort, but you have to have eat enough clamping around the part so that you're not just clamping on the ends and having the middle bow up.
00:51:18
John S
Right.
00:51:20
johngrimsmo
yeah So that's what I'm working on.
00:51:21
John S
Sweet.
00:51:22
johngrimsmo
What are you working on?
00:51:24
John S
Um, fish, the pattern stuff is basically done. I'm just kind of just double checking stuff. And like, you know how, like I said, there's eight parts of these about covers, like on the seventh one, six are already done on the seventh one.
00:51:35
John S
I discover a new way. I like doing something. And so I, you go open all the other six ones and they work your back way through it to change it.
00:51:38
johngrimsmo
ah Yeah.
00:51:41
John S
And you just, you just pour yourself a cup of coffee, smile and pound through it.
00:51:44
johngrimsmo
Yes. Do it. Yeah.
00:51:45
John S
You know, we're in a good, we're in a good place on it though.
00:51:48
johngrimsmo
Good.
00:51:48
John S
Um, and then it's super annoying, but I guess maybe a PSA for all of the fellow Americans out there, the very just disgusting governmental overreach that is block of the U S beneficial owner information request thing. It's like got a wall. They got sucked through in 2021 for good reason. There are a bunch of like, apparently SUS like real estate.
00:52:11
John S
shell-holding companies where nobody could figure out, like the government could figure out who actually owned entities with subinterest with subinterest and it's like probably tax fraud and all that stuff but like basically now the US government is making every United States LLC fill out this new information list and information request which is a whole bunch of of baloney.
00:52:16
johngrimsmo
okay yeah
00:52:31
John S
You know nothing good happens when governments start making lists and you know you actually you you don't have to do it if you're like a huge business like um We don't come close to qualifying, but if you have 50, I don't know if there's 50, 50 employees or 50 million, it's like, okay, you're not some secret shell company that's doing sus stuff, but um so we have to, it's like, ugh, so I'm just trying to, I'm basically, it's the start of the end of the year paperwork is really what I'm saying.
00:52:57
johngrimsmo
Yeah, yeah.
00:52:58
John S
Yeah, yeah sounds good.
00:52:59
johngrimsmo
All right, have fun with that.
00:53:01
John S
I'll see you next week, bye.
00:53:01
johngrimsmo
Bye.