What's new
What's new

Custom broaches for laser cut knockout die? Weird custom tooling alert

Alberic

Cast Iron
Joined
Jan 15, 2011
Location
SF Bay
HI guys,

OK, here's another weird one.
We make hand saws. Fancy hand saws. The spines look like trusses. They're laser cut in 6061-T6, .190 thick, so there's lots of little slugs to knock out of the holes. They're single-tabbed, so they stay in the blank during cutting. If we cut them clear, one or two of them will only partially drop, then tilt and stick up proud of the sheet. Then the laser will home onto those spikes like we were aiming for them, and knock its lens out of alignment. (at best) So they're tabbed, and going to stay that way.

The problem is that we're blowing a lot of handwork time (A) knocking the tabs out, and (B) cleaning up the little pip that's left over from the tab on the inside of all the windows. (There are 15-18 windows per saw, and we punch out many hundred of these things every month. So *lots* of windows to clean up.)

So my problems are two: knockout and pip cleanup.
The obvious answer for knocking out the slugs is just to make a sort of pannini-press sort of die, with a bunch of steel dowel pins to punch the slugs out through a custom cut steel baseplate that matches the shapes of the windows, plus 0.020".
Other than the PITA factor of making it, not such a big deal, except perhaps figuring out which press to run it in. I have a bunch of kick presses, so that may be the way. Use ball bearing guide shafts, and let the jig align on itself, with the press just supplying power.

The kinky part comes when I had a thought to combine the knockout die with some sort of a cleanup function for all the pips.
The pips are always in the same place. So my idea was to do something like a broach to file or shave off the pip as the die was closing to knock out the slugs. The thing I'm picturing is like a gear or pulley keyway broach but shorter. All it really needs to do is shave off that little teeny pip, 0.200" tall. So it doesn't need to be much. In a perfect world, it'd be round, with a flat spot for the broach face, and a stock item for somebody, doing something else, so i could just order a bunch of them off the shelf somewhere. I'd like a pointed tip so I could aim them into a drill jig bushing for support on the bottom side.

Have any of you out there seen small short broaches like this? Who uses them? For what?
I have a die filer, but I really don't want to do them that way if I can avoid it. Waaaaaay too much repetitive work for me to ask any of my guys to go nuts doing it that way if I can help it.
Anybody got a better idea?

As a final wrinkle, we do spend a fair bit of time cleaning up the inside of the windows with dremels, to smooth out the cut marks. (Because the customers require it.) So I'm also pondering what it'd take to just make up a die that had broaches that'd fully broach all the windows by a few thou, just to clean up the cut walls. I'm assuming for something like that, I'd be wanting to talk to a broach house?

Anybody ever done anything like this? Any ideas or suggestions?
Thanks,
Brian
 
Depending on inside radii of the punched holes, maybe get a small CNC router table and fixture it with some location pads and quick clamps, then profile the holes with an 1/8" (or whatever) endmill. With a good router (or small CNC mill) you'd clean the insides including the tabs and be set.

No custom broaches, and flexibility for other purposes.
 
Some times the quickest way isn't the best way. Aiming for the finish your customer is wanting, I'd go with a CNC mill to save just about all your hand work.
 
Is it this saw?

CNC mill would probably would be best...

I wonder is a sinker EDM would knock these out with minimum finishing?

An idea don't cut the holes at 90* to the face of the part. The tabs stay put until you flip the part over...


455301-4d.jpg
 
Yeah, that's us, Knew Concepts. I try not to turn things into a PR campaign when I'm asking for help. Seems...wrong somehow.
That's not exactly the current design, the windows are simpler, and the corners all radiused, but yeah, that's one of the ones I need to deal with.
I'd go with the mini-bot and vac table route, if we had another 6 inches in the whole shop to swing a dead cat. But we don't.
The room for the little die-filer that I've already got would be an issue. (It's currently in the back of my (covered) truck, if that gives you any clue.) I may well end up doing it that way, but I'd still need to blow out the slugs first, although at that stage, a simple knockout jig becomes reasonably easy.

Ah, the joys of being able to say "It *is* my circus, and god help me, they *are* my monkeys..."
Before I give up on the broach idea, have any of you seen small (?cheap?) broaches in use anywhere, for anything?

Thanks,
Brian
 
My first thought (former tool and die maker) was a stamping die.- Probably more expensive than warranted unless you're selling thousands of them every month. An adequate sized press could be easily outsourced.

Second thought- Water jet and vibratory media de-burr. Yes it needs to be done outside of your shop but in exchange for the costs you only stock ready frames and minimize hand work prep. The laser could make money on other jobs. Frame blanks could be stacked 3 or more in the water jet to lower cut time. Minimal (if any) hand work, no chips to sweep up.

Third thought- Farm out to small CNC mill. Full cut-out profile but adds cost and possible (probable) warped frame.

Fourth thought- Broaching as you suggested? PITA building fixture for locating broach(s)that line up with cut-out nib. The mismatch will be one more area to need smoothing out by hand IMO.
 
From what i can work out, the material is sitting proud of the laser bed. Ive only seen smaller powered lasers,(120W) which had a honeycomb bed. Maybe larger ones dont?
Can you use a sacrificial plate or something under the sheet to stop he slugs falling and or catching?
that way you could completely cut them and eliminate the tab.
 
I use jeweler’s saws frequently in doing my retirement personal projects, and have almost ordered one of the your frames on more than one occasion. I restrain myself because of my age (89) and continue to use my nice antique. From your description, I precisely visualized the frame concept that you posted in a subsequent reply in the thread. A die to punch the entire frame out would be the ultimate, but considering the number of variations you offer, the cost would be enormous. I do not know the cycle times for laser and water jet cutting, but the water jet cut items I have seen were good. Could you set up a jig or plate for each variation that would support the cut-outs to permit the laser to trim each tab fully? I am visualizing a plate with a pattern cut out to clear the laser cut. The support parts shaped like the cut outs could have holes connected to a manifold with vacuum to hold the tabs.
 
Another thought. A powered custom shear that would locate off the laser cut on both sides of the tab and cut away the tab with a single stroke. Could be foot powered, up top lever powered, or activated by a cylinder using air pressure. This woul provide a single action in removing each tab. The cutter would have to be guided from below to provide access by placing the frame over the open top of the shear.
 
Certainly with all the laser's out there, the tip up problem has been solved by programming or some other solution.

Let's hear from of the regular sheetmetal members input on tip up problems.
 
Make an x cut in the slugs before final profile, or at least one cut dividing the slug into two pieces. That way they cannot tip but will fall through. A bit more "beam on" time and a bit more gas. The dividing line may or may not want to exactly touch the cutout line. The line may also need to look like a sharp "J" to keep the pierce away from the profile. I know my software can do it but would need to try some scrap first.
 
HI guys,

Thanks for all the replies.
A little more information: the laser is a giant +3Kw industrial cutter. We're running full sheets of aluminum 10x4 feet at a pop. The bed is a little weird: instead of a honeycomb, it's strips of sawteeth set up on edge. So you only get point contact under the sheet, and only the width of the metal sheet in terms of width that will intersect the beam. And that's mostly down below the focus point, so *mostly* it doesn't get too chewed up. But they regularly make their own replacement sawtooth strips. Which is what lets the slugs drop out and cock. We've tried all sorts of oddball tricks to get them to stop doing that, with very little success. (The laser is owned by a laser house down the road from us, we've been working with them for years. Believe me, *they* want it simple and easy too, as it's their guys who're doing the deburr at the moment, and they are heartily sick of looking at them.)

I've looked into getting punch dies made, except that the saw in the picture above is one of our 'skinny' line. Only .125" thick. The 'heavy duty' line is .190" thick, and has much smaller windows. (and more of them) Which drove the punch house nuts. The rough ballpark quote said about $5K per die, with an expected life of about 5K strikes. So roughly a dollar per strike in terms of die wear. Yikes!

We already do large scale vibratory deburring to break the edges. We've got a pair of large tumblers that load roughly half a ton of media each, but unfortunately while they can break the edges just fine, they can't get the insides of the windows totally smooth, or really deal with the pips. (I've tried.)
We just picked up a used Timesaver wide-belt sander to automate the finishing on the flat faces of the saws. Now that it's finally fixed and online, that should help some. (Bought it used from a dealer who said "it was working when they unplugged it....." Famous last words. So 8K of parts, and 3 months of wrenching later, it's actually working. I got it for a song, so I'm not entirely tweaked, but dear god, the *coolant pump* was cooked. Those things *never* die. How the hell do you kill a coolant pump?? (the main contact roller was grooved, two separate motor control boards were blown, the air knife blower was packed full of not-quite-concrete, etc, etc.....)

The more I think about it, the more it looks like we may end up with a small gantry CNC router with a vacuum table. That seems like the most versatile option. Now I just have to figure out how to run it while it's bolted upside down to the ceiling.

I'm still open for ideas on ways to do it fast, cheap, and easy. Ideally without making any of my guys nuts from doing it over, and over, and over....

Thanks,
Brian
 
The more I think about it, the more it looks like we may end up with a small gantry CNC router with a vacuum table. That seems like the most versatile option. Now I just have to figure out how to run it while it's bolted upside down to the ceiling.

I'm still open for ideas on ways to do it fast, cheap, and easy. Ideally without making any of my guys nuts from doing it over, and over, and over....

Thanks,
Brian

Have test cuts made by whoever you buy the router from. You need to ensure it's got good enough stiffness and path control that using a small endmill will give the finish you want. If it has a toolchanger you can also edge break with a 90 degree tool, in fact you can do both sides with an undercut version of the cutter. Your fixturing will have to allow a small pocket for the cutting head to fit into.

You could even get custom cutters made up that incorporate the two edge breaks with the inside profiler, sort of like some router bits. But this would be best for finishing, not cutting from sheet. And of course you'd need different sizes for different sheet thicknesses.

Once you have the machine, your next task is fixturing. You'll want something easy to clean (not often thought about), able to quickly swap parts out, and hold them well without a lot of clamp actuation time. The clamps should be low profile and not deform the part, or interfere with tool paths.

With the right machine you can do your own cutting from sheets, perhaps that's your goal?
 
Hundreds a month?

Skip the laser, skip the handwork and mill them start to finish. Vacuum fixture seams like a good place to start.

My slowest mill would make a couple hundred of those in a week. I bet my little HMC could do a hundred a day no problem.
 
HI guys,

Thanks for all the replies.
A little more information: the laser is a giant +3Kw industrial cutter. We're running full sheets of aluminum 10x4 feet at a pop. The bed is a little weird: instead of a honeycomb, it's strips of sawteeth set up on edge. So you only get point contact under the sheet, and only the width of the metal sheet in terms of width that will intersect the beam. And that's mostly down below the focus point, so *mostly* it doesn't get too chewed up. But they regularly make their own replacement sawtooth strips. Which is what lets the slugs drop out and cock. We've tried all sorts of oddball tricks to get them to stop doing that, with very little success. (The laser is owned by a laser house down the road from us, we've been working with them for years. Believe me, *they* want it simple and easy too, as it's their guys who're doing the deburr at the moment, and they are heartily sick of looking at them.)

I've looked into getting punch dies made, except that the saw in the picture above is one of our 'skinny' line. Only .125" thick. The 'heavy duty' line is .190" thick, and has much smaller windows. (and more of them) Which drove the punch house nuts. The rough ballpark quote said about $5K per die, with an expected life of about 5K strikes. So roughly a dollar per strike in terms of die wear. Yikes!

We already do large scale vibratory deburring to break the edges. We've got a pair of large tumblers that load roughly half a ton of media each, but unfortunately while they can break the edges just fine, they can't get the insides of the windows totally smooth, or really deal with the pips. (I've tried.)
We just picked up a used Timesaver wide-belt sander to automate the finishing on the flat faces of the saws. Now that it's finally fixed and online, that should help some. (Bought it used from a dealer who said "it was working when they unplugged it....." Famous last words. So 8K of parts, and 3 months of wrenching later, it's actually working. I got it for a song, so I'm not entirely tweaked, but dear god, the *coolant pump* was cooked. Those things *never* die. How the hell do you kill a coolant pump?? (the main contact roller was grooved, two separate motor control boards were blown, the air knife blower was packed full of not-quite-concrete, etc, etc.....)

The more I think about it, the more it looks like we may end up with a small gantry CNC router with a vacuum table. That seems like the most versatile option. Now I just have to figure out how to run it while it's bolted upside down to the ceiling.

I'm still open for ideas on ways to do it fast, cheap, and easy. Ideally without making any of my guys nuts from doing it over, and over, and over....

Thanks,
Brian

Find another laser shop. Most lasers use saw teeth points as support. They may complain about deburr but if they are getting paid for it the management hears no complaints. They can do it but will loose the labor side of cleaning up the openings. Those parts shown have very small drops that would fall through unless fused to a saw tooth point. If so fused they would stay put. Seek another laser shop. I cut 48 x 120 all the time. No big deal.
 
If I had too many tab remnants to trim by hand, I'd think about taking a 2nd kick press and making a sort of nibbler punch/die for it. Fixture could locate on the untabbed perimeter of a window, and just trim the tab remnant out.
 
We've cut a ton of parts with lasers over the years. Most were not critical and "normal" support under the sheet was
all we needed to get accurate parts. We've had the odd run with a lot of small cutouts similar to what the OP is dealing
with and had the same issues with the cutouts hanging up.

We solved the problem by making a plate for the table that had a series of studs sticking up to support the finished
parts in enough places that they would not move around but let the cutouts fall fall way. Not that hard to do because
we just laid a grid of holes over the nesting pattern for the parts and used the laser to to cut the holes. We also used
the laser to cut out good sized holes in the plate so the cutouts could fall through. Never had any issues after that...
 
I'm not qualified to comment on lasers, it's one of the few machines I've not run. Just FYI (OP) the "quote" you were given about a die having a life of 5K hits sounds like something is wrong. A stamping die should be just getting warmed up at 5k hits and batch runs between sharpening should be measured in multiple 10's of thousands. That's for difficult/exotic materials like spring steel or beryllium-copper (for example). With the aluminum flavors the production runs between sharpenings can run into 100's of thousands (or more) depending upon conditions. Now multiply this by the number of sharpenings you can obtain with a die, a function of die design itself, before the die sections and/or punch/perforators need to be replaced. This determines what the "life" of the die is. A die that only gets 30k hits (conservative numbers) between sharpenings and has .187 "land" (before taper) would get 18-19 sharpenings (at .010 removed for sharpening). That's over 1/2 million hits with conservative ballpark numbers. The "quote" for 5k for die life or even for number of hits between sharpenings sounds ridiculously low, is using unhardened material for cutting, or there's something else I don't understand. This is NOT to persuade you to seek a stamping die, they're expensive for low volume batch runs. My post is meant to clarify/correct what I feel is bad information given to the OP. Whoever gave the OP those numbers may be good at some other subject but die design isn't one of them. JMO YMMV. Ganbatte.
 








 
Back
Top