[webgeek]

So I'm cutting TONS (as in thousands) of small parts, say 1" x 2" out of 1/8" 6061-T6 aluminum plate. I'm using 3-flute high-helix cutters and full coolant. Everything goes great but the way I mount the parts is really time consuming and I want to learn how to speed up production.

Here is my current process:
1) Turn on and home the machine (once per day)
2) Set the work offset to the edge of my fixture jig - this is low accuracy but more than good enough for my process (once per day)
3) Take a 12 x 24 piece of aluminum, a matching piece of 1/4" plywood and a spoil board and clamp them to the fixture jig using c-clamps and/or welding clamps. This is only temporary until the next three steps are done.
4) Touch off the surface of the metal using a Tomach-style dial indicator.
5) Insert a 1/4" drill bit.
6) Run g-code that drills 1/4" holes through both the aluminum and 1/4" plywood.
7) Remove the clamps and pull the spoil board off.
8) Bolt down the aluminum and plywood (which is the new spoil board) to my fixture plate using 1/4-20 bolts.
9) Touch off the surface of the metal using the indicator again.
10) Run actual part program and change tooling as needed.

All of the tooling is using the Tormach system (I love it!) and pre-measured in Mach 3 so tool changes are quick. Additionally, I have all the tool operations quite optimized for feeds and speeds to keep times down. Finally, I have all the tools sorted for efficiency in g-code. Long running operations are at the end so I can leave the machine unattended once the initial holes and the like are drilled. Effectively, a 3 hour operation requires my involvement for only the first 20-30 minutes before it gets to the last tool which runs quite a while.

What I'm looking for is a way to speed this up. Drilling the holes in the beginning takes an absurd amount of time as I need to manually clamp and unclamp every part and swap out all of that. Additionally, all of the bolting and unbolting is a pain too. I'd make a lot more sense to avoid this step entirely but I just don't know how to do it. It seems to me that a vacuum system would be ideal but I'd need to make it myself.

I'm good with making the vacuum pump/tank side of things - I've got a bit of experience with pneumatics so I feel confident there but the vacuum jig stuff is the tricky bit.

So this long rambling post has two specific questions:
1) Do you have any suggestions on improving my process?
2) Do you have any pictures of examples of a custom DIY vacuum jig so I can see the basics of how they work? I've never seen one for metal in person and I imagine mine needs to be quite custom to pull it off.

Thanks in advance for the help!

-Mike

[LeeWay]

If you used smaller pieces of stock and cut fewer parts each time, That might speed things up. At least then you could use a clamping fixture.
I cut three parts out of sheet aluminum in this manner and it's goes pretty quick.
My stock is 4" x 12" but you should be able to shear your stock or cut it on a bandsaw. I just used toggle clamps, but there are many other ways to do that.
I use a piece of phenolic for a spoil board so coolant doesn't effect it. This tool plate is actually on a magnetic chuck. All my tooling plates fit this chuck and are held both mechanically clamping on the side of the chuck as well as engaging the mag. They never move and are always easy to relocate precisely.
Here's a little video. Turn the sound down. The router really isn't that loud.

[nomedia="http://www.youtube.com/watch?v=fswDGVjeZPc"]YouTube - Milling aluminum sheet[/nomedia]

[webgeek]

I watched the video - thanks! Currently I use the bigger material as it's inexpensive and comes pre-cut to size while being right about the max size for my mill. As my parts are small and require drilling, profiling, and channels I'd need more hold downs than the 4 you have there even with small sheets you use I'd still have to clamp from the center.

The phenolic is a superb idea. I'd assumed it was far too expensive but a quick check shows it's quite reasonable and this means I don't need to drill the wood every time. I can use one sheet over and over a bit. The wood I use holds up to coolant for one sheet before the glue starts to break down. It also adds a good .01" of height variability to every part I cut which is a pain in the neck with holding tabs. Ideally I'd get at least 5 to 10 sheets from a single piece of phenolic which would keep the cost about the same as the plywood. Not drilling as much would be a huge time savings if nothing else.

I don't quite understand your magnetic clamping comment. Can you explain more? I'm totally self-taught so my practical experience is nill with any sort of specialized tooling or clamping.

-Mike

[LeeWay]

I haven't wore out any phenolic yet. This is the black paper type, so it is the cheapest as well. My plastics supplier also sells one offs and drops on Ebay. He sells pieces of this there. Cheaper than buying new stuff.

The mag chuck is usually used for grinding stuff, but I also cut parts out of thin sheet steel. I keep my feed rates down and the parts don't need any other fixturing. Only engage the magnet and the part is ready to be milled. Since this chuck is bolted to my table, I just leave it in place and all my vises and tooling plates attach to the chuck.
Just works well for what I do.

Before I used the mill for these parts, I cut 24 out at a time on my router. I too had to drill holes to screw the sheet down to the spoil board on the router. It didn't have coolant, so used WD40. Real pain in the butt with clean up too.
The mill is much easier even though it only cuts 3 at a time and I have a little more scrap. This sheet is cheap enough to make it a better option for me.

One thing you could do is make a bowed jig to hold down. What it is is a tooling plate about the size of your part. Have three hold downs that are kind bowed. Mount them hinged on one end. Install the sheet. Drop down the three bows and have a swiveling bolt oir clamp on the ends. Tighen those down and these bows would put plenty of pressure on the sheet. You would have to redo your drawings to account for the bow locations, but that would eliminate any initial drilling and any bolting later.

[Boltz]

Have you considered subbing these out to a stamping company? 10,000 parts is a couple hours work for a stamping press. You could then make up a couple fixture plates and do the secondaries a hundred or so at a time, loading one plate while the other is running. I'm picturing vacuum hold down using o-rings to seal against the parts. If you wanted to get really clever you could use the vacuum to retract the locating pins. A pump for auto AC service (or if you really wanted to get cheap, a salvaged refrigerator compressor) would provide plenty of flow.

Boosting production is a really sweet challenge to have! Good luck!

-Jim Hart

[BobWarfield]

Make your fixture so you can bolt down the parts without removing the workpiece from the fixture or having to use a spoil board.

For example, consider a fixture plate with locating pins. You need 3, they need to be round, and they need to stick up enough to locate two edges of your workpiece.

I'd put two at top and one on one side. This holds 2 sides against 3 points to avoid what's called "over constraining" the fixture. You want clamping that will lock it down after it is in place against the locator pins.

To use, slide the workpiece along the two top pins until it bottoms against the side pin. Activate the clamps. Toggle clamps are fast and easy. You could use something Mitee Bite style too.

Now drill the tapped holes. In the fixture plate, leave unthreaded clearance so the drill doesn't bung up the threads. Once you drill and tap, you can use bolts to get things clamped better. Put them in with an air ratchet to make it fast.

Now you're all bolted in solid, you never had to remove the workpiece or fiddle with spoilboard. Let 'er rip. You'll need to leave enough waste you can mill down to the fixture plate without running into the locator pins.

Want to go even faster? Make a couple of these fixtures and use them palette style so the machine is working while you are tearing down or setting up a fixture.

There are MANY other ways to skin the fixturing cat, this is just one idea. Carr Lane has some excellent fixturing books available that are full of ideas.

Cheers,

BW

[RCaffin]

Here is my current process:
1) Turn on and home the machine (once per day)
2) Set the work offset to the edge of my fixture jig - this is low accuracy but more than good enough for my process (once per day)

So far so good.

3) Take a 12 x 24 piece of aluminum, a matching piece of 1/4" plywood and a spoil board and clamp them to the fixture jig using c-clamps and/or welding clamps. This is only temporary until the next three steps are done.

This needs to become a high-accuracy single fast step. I assume the edges of the plate are reasonable, and that you have a drill press. Try as follows:

1) At the start of the day (week, month), position a single high-tensile pin on the bed of the mill at (0,0). You can do this by:
a) clamp the high tensile pin in the chuck on the Z axis
b) move the mill to over a T-slot roughly where (0,0) should be
c) set DROs to 0,0
d) lower Z axis until pin is at.in T-slot, and bolt it down there carefully.
You now have the mill origin accurately registered to that pin.

2) Repeat this process to position a second pin at (say) (500,-20) in another T-slot. (Metric - millimetres).

3) Take Al plate and drill (drill press) a single hole in one corner, sized to fit over pin #1 tightly. Make matching plastic underplate in similar manner.

4) Fit plate(s) onto pin #1 and against pin #2. Lock down with over-centre toggle clamps of adequate size around the edges. This will locate and constrain the plate, and IS REPRODUCIBLE for that plate.

5) Optional - clock up the top surface of the Al plate. But once you have prepared the under-plate out of good plastic you should not need to do this every time. Nylon for instance will take the heat and is stable. Yes, I am assuming the Al plate is of constant thickness, although this may not even be necessary.

6) Start milling.

7) At the end of the day, !!!BEFORE you power down!!! position the mill head over 0,0. Most machines do not move during either power down or power up, so this reference should persist for the next day.

If you do lose the origin (power failure etc), reclock the machine to pin #1 using a touch probe.

Cheers

[webgeek]

One thing you could do is make a bowed jig to hold down.

I used something kinda like this prior to my current fixture. It didn't hold down evenly but I suspect that was my problem. Where do you get the "bowed" metal to get better hold down?

Have you considered subbing these out to a stamping company? 10,000 parts is a couple hours work for a stamping press. You could then make up a couple fixture plates and do the secondaries a hundred or so at a time, loading one plate while the other is running. I'm picturing vacuum hold down using o-rings to seal against the parts

While I think I'll need at least 1000 or so of the small parts it will be over the course of a year. As I get paid when they are ordered it's just not feasible for me to pay to get them stamped. I also think many of the parts are too small to hold down pneumatically as best I can tell. I'd love to see how parts this small can be fixtured individually without tons of tiny shoulder bolts (which is what I used to use).

Make your fixture so you can bolt down the parts without removing the workpiece from the fixture or having to use a spoil board.

Wow Bob; tons of great tips there! I never thought of simply starting the threads lower in my fixture plate so I can avoid the spoil board. The pin + toggle clamp trick is great too. It ensures a specific corner is correctly positioned while still not requiring perfect pieces.

Thanks much everyone!

-Mike

[webgeek]

Thanks for all the help everyone. I spent some time this weekend designing up a new generic fixture plate and figuring out how I'd mount it. I've attached two pictures below.

I need the fixture plate to be elevated a good bit off the table so it will be held by my two vises. The vise jaws are perpendicular to the table as opposed to parallel so the vises are mounted longways on the table. My current fixture plate is gripped at the ends by the two vises but this causes the plate to bow up in the middle and the work around for that is quite nasty. Additionally, no matter how hard I've tried, I've never been able to get it more level than about 0.01" over the full length and this is enough off to give me fits with holding tabs.

This new plate uses the second picture + some matching angled soft-jaws on the vises to hold the fixture. The angled block pictured will be attached to the bottom of the fixture plate via the holes in the middle of the plate (the two groups of three with recessing.) The matching jaw in the vise will cause the plate to be pulled down against the ground surfaces of the vise. I THINK this will work well but I'd love feedback. I'd love to just use a square block to clamp and not cut this nasty angle but I don't know if it will hold down properly. Any thoughts there? The vises are not Glacern but look a lot like em - Imagine this vise without the jaw protruding above the ground surface:
http://www.glacern.com/photos/gsv_02.jpg

The six sets of four clustered holes around the outside edge are for the toggle clamps that will go on the outside edge to hold the work down before it's bolted down. Here are the toggle clamps I'd use - each one is rated to 60 lbs:
http://site.toggleclamps.biz/drw/201a.jpg

There are two 1/8" holes on the upper edge and one on the left, they are outside of the grid of holes. These are for the edge locators/dowel pins. Did I position them correctly?

The little square hole in the upper left is a way to touch off using my edge finder. This will let me zero to the edge the dowel pins define. This method has worked for me in the past but I'm open to suggestions.

Finally, normally I'd create a long groove in the fixture so I can ensure the plate is parallel to the table. I'm ignoring that this time as I never need to take the plate off mid operation and I never need to move the pieces between this machine and another. Once the fixture is mounted, it will be used until the piece is done. I'm hoping I can get this accurate enough with the rough edge of the plate so that once my holding blocks and vises are zeroed in, it will be generally consistent. I only need it accurate enough to avoid hitting my bolt hold downs. I think this will work but again I'm looking for any and all feedback.

Thanks much for the help, this is quite new to me so I'm kinda muddling my way though it.

-Mike