I’ve been doing some large runs for customers (+300 xmas ornaments in one case) and I’m wondering about optimization of my work.
What I have been doing is, screwing down squared “L” bracket and then placing the stock in there. Then I cut about 10 (v-bit and then 1/8" down cut with tabs), then replace the stock and cut more.
When I’ve cut all the parts I need, I then flip the stock over and cut the remaining side.
Is there a good tutorial for doing multiple operations in one pass? I’d like to set up two “L” brackets, place stock in the first, then flip it over into the second position and load fresh stock back into the first (rinse/repeat).
What software are you using? I use Fusion 360. In this situation, I would design a fixture out of MDF with 2 locations for material and do the first op in one and the second in the other. I would both ops in a single file with a single zero point. It is pretty easy to do that in the CAM environment in Fusion, I just am not sure about other software.
Rather than using “L” brackets alone where there’s still some positional variability depending on how the stock moves as material is removed, I’d recommend planning for your 2nd op to use locating pins that are friction-fit into holes that get cut as part of the first op. You can put these pins in sacrificial “sprues” that you cut away from at the conclusion of the op down to a very slim rib that you remove afterwards.
This eases clamping precision requirement, as you know that your work piece will be in the right place for the 2nd op due to the pins and you can just quickly apply the clamps to hold the material down (rather than carefully ensuring positional accuracy).
If you want I can try to find an example where I’ve done this in the past. I did a slingshot design for my son a while back in Fusion360 that I did this way to ensure the front and back sides aligned properly.
What you are describing doesn’t save time. Yes, using 2 L brackets can double your throughput relative to the number of manual interventions changing tools and clamping. But having “L bracket 1” do “Op 1” and “L bracket 2” do “Op 2” doesn’t save any time versus having “L bracket 1 and 2” both do “Op 1” then “L bracket 1 and 2” both do “Op 2”. If I were doing this I would have both brackets do whatever you can do with tool 1 with all the sheets you will do, then switch to tool 2, then do the flip side tool 1, and tool 2. That is unless you cannot afford to handle all the final output in one batch (need extensive time to hand finish, paint, etc).
If both op1 and op2 use the same tools it will save a bunch of time in setup. After every set of ops, you could then be doing any hand finish work on the finished op2 part that came off. At the end you are all done.
Method A, what you are asking, has a lot of tool changes, less clamping changes, but more steps overall. The benefit is that parts come off out staggered, which may or may not matter. Method B, what I suggested, has less tool changes, but more clamping changes. Note the overall number of bit and clamp changes are the same. But I would much rather change the clamping than the bit. If you have a powered screwdriver (even a T-handled allen wrench), it should be a breeze to do and undo the clamps. I assume you have a BitSetter, but I don’t suppose you have a power drawbar or automatic tool changer? I cannot imagine anyone would rather do more tool changes versus part changes, unless as stated you cannot afford to have all the parts finish in one batch.
I should have been way more specific, the stock is cut to specific lengths (something that can accommodate N number of ornaments) as I’m buying 10’ lengths. The lower left corner is in the crook of the “L” and the upper right is clamped. Even if I’m off by an inch, I care only about the the material from the edge of the last CNC operation to the left of the stock (this is unchanged).
I do use locating pins on some types of operations (tiling), but you’re right, this would be a much better way to do locating for non-controlled stock dimensions!
I’d love to see anything you’ve done - while I’ve been working with cnc machines for years, I’m always learning new/better ways to do things!
Bingo! Cut one side, flip into position 2. Next run does a new operation on position1 and the finishing cut on position2. Then on the next reload, the second position stock gets taken off the machine, position 1 gets moved to position 2, position 1 is reloaded and while the machine works on these two things, I’m off cutting things free from tabs, sanding, etc. By the time I was done, the machine could be reloaded again.
The current process is similar, but I have to flip the stock and tell it to run again. And at that point, I’m done with any by-hand finishing operations. So it’s an efficiency loss.
On industrial scale machines, you’ll see a bunch of different clamps set up inside and it’ll make multiple passes across multiple positions. It’s not a new concept, I just don’t know how folks do this (I’m looking at you @Vince.Fab
Me. With a BitSetter, tool changes are trivial when compared to trying to make sure individual tool paths match up when you have moved the material in between switching tools. This is even more difficult when you add double sided machining into the mix.
I was just going to run all the v bit cuts, then run all the 1/8" cuts. No need to swap bits between flipping operations. I have enough room on the bed of this machine to set up multiple positions, I just don’t know how.
The reason I use fixtures instead of vises when doing this is because it is easier to setup in Fusion. Carbide Motion does not handle multiple work offsets but I bet you could do this with another gcode sender. To do it with multiple vises and Carbide Motion, you would need to very accurately measure the distance between your first and second vise. You would then model them both in Fusion and layout your parts in each one. I find that making a dedicated fixture with built in workholding is easier. That way you don’t have to worry about locating multiple vises accurately.
I am pretty sure you could do the multiple work offsets approach with CNCJS and the NC Programs method you posted the video about. I just have never used that method.
Yeah, I haven’t used motion in forever at this point, I’ve been using cncjs.
I’ve been contemplating going back (and likely will for the first few cuts when I get the HDM), and this is another consideration I should have at this point (multiple fixturing).
Honestly, some of the struggle I’m having is finding the right language to use to describe to google what I’m attempting to do. And honestly, I can do either multiple fixtures or vises (depending on what I’m trying to make). Generally, when I’m cncing something, it’s because I need to make multiples, so figuring this out sooner rather than later will help me with my efficiency.
If you have any videos of your setups/etc, I’d love to see how smarter folk do this
I always use vises or superglue and blue tape for my first attempts at anything. Once I have everything dialed in, I switch to a dedicated fixture if I need to make multiples. See this for one of them I have posted about:
You know the point of the L bracket is to repeatably locate the stock, right? As long as you have two points of contact along one leg of the L and one point of contact on the other leg of the L, you are fixed. In any case, at a minimum you still have two clamping operations per sheet. If for some reason the clamping is so prone to error, this whole process will not go well regardless.
More moving of the material is just more opportunity to incorrectly torque your clamps, have chips get caught between your stock and the stop, have the stop move because of repeated clamping, etc. Everyone makes mistakes. All I am pointing out is that your way of working is not the only way and why I like my way. If you like how you do things, great. Keep doing it. I do things my way because of my experience. I chase pretty tight tolerances in aluminum and have found that I have less problems if I move the stock less.
I use various WCS on my other machines via Fusion CAM. Handy, saves time, flexible. I program the offsets via the controller for various fixtures, no limit that I’ve observed yet. I haven’t found my Shapeoko that consistent, so have not used it in such a manner. But GRBL gives you a few WCS to work with, shouldn’t matter what sender you use.
