Tips for streamlining setups.

I'm kicking around a few ideas on how to speed up setup times for my mills. My shop is your typical job shop, we might do 4 setups a day between the mills. We've got 6 mills ide like to streamline.

My 5 axis has a full Lang setup, and 60 tools preloaded, I've got it dialed in really well to the point where switching jobs might take me 10 minutes. Ide like to apply similar methodology to my 3 axis jobs.

First idea is to have a large selection of tools preset for each mill, so all my setup guys have to do is load specific tools and go. The problem is, I would need at least 60 tools/holder per machine, perhaps 100k investment. Some reprogramming needed

Another similar idea is to change the method of tool touch off. I'm thinking of having everyone touch off tools to the table (or some standardized point) then offsetting the work coordinate Z accordingly. This would at least keep everyone on the same page, no more of the excuse "I didn't know where z was" This idea isn't dissimilar from using a tool setter, however I don't fully trust my guys with tool setters. No reprogramming needed.

The last idea would be to implement a Lang type setup in all my mills, the problem here is the cost of investment (not an issue if it's foolproof). The other issue, and I see this as a huge issue, is the lack of flexibility, in a job shop you never know what you may have to do. I'm moving vises all the time, taking 4th axis off tables, taking windows and doors off machines, getting locked in to 3 different setup positions may prove fatal for us. Potentially tons of reprogramming.

Please let me know what you think of any of these ideas, and let me know of any other tips you might have. Thanks

I'm not a machine probe user, but it surely sounds like it is an applicable technology here. The trouble with maintaining an archive of tool data is that you never know when it has fallen out of date. The archive requires ruling with an iron hand to get everyone to participate, and then, it will still get screwed up.

You might have to spend some time, and machine some fixtures, to make table mounted sensors quick and easy to install/remove as the situation requires.

BRIAN.T said:

Another similar idea is to change the method of tool touch off. I'm thinking of having everyone touch off tools to the table (or some standardized point) then offsetting the work coordinate Z accordingly. This would at least keep everyone on the same page, no more of the excuse "I didn't know where z was" This idea isn't dissimilar from using a tool setter, however I don't fully trust my guys with tool setters. No reprogramming needed.

Click to expand...

What do you touch off to now? I'm a pretty firm believer in setting your tools relative to something that isn't going to move (like the table), and then using fixture offsets to locate the fixture.

Guys who do things like touch off to the top of the stock and then machine away their datum must have a hell of a time keeping things consistent.

isaac338 said:

....Guys who do things like touch off to the top of the stock and then machine away their datum must have a hell of a time keeping things consistent.

Click to expand...

Back in the early 80s when I went through training for CNC mill programming and operation, that was how it was taught. So, I did it that way for a while, just because that's how a supposed "expert" said it should be done.

isaac338 said:

What do you touch off to now? I'm a pretty firm believer in setting your tools relative to something that isn't going to move (like the table), and then using fixture offsets to locate the fixture.

Guys who do things like touch off to the top of the stock and then machine away their datum must have a hell of a time keeping things consistent.

Click to expand...

I'm really happy to hear that, and honestly it is a garbage system, and the more I think about it the more I realize Ive just been keeping it consistent with how the shop has always done it, even though it's always been imperfect.

HuFlungDung said:

Using G92 can also shift a whole batch of tool offsets globally (per machine) but that is fraught with more danger.​

I'm not a machine probe user, but it surely sounds like it is an applicable technology here. The trouble with maintaining an archive of tool data is that you never know when it has fallen out of date. The archive requires ruling with an iron hand to get everyone to participate, and then, it will still get screwed up.

You might have to spend some time, and machine some fixtures, to make table mounted sensors quick and easy to install/remove as the situation requires.

Click to expand...

I appreciate your help, I definitely don't have time to rule with an iron fist, as much as I would like to sometimes. I program all the machines, as well as all the multiaxis set-ups, and the more complicated 3 axis setups. No time for babysitting.

Vancbiker said:

Back in the early 80s when I went through training for CNC mill programming and operation, that was how it was taught. So, I did it that way for a while, just because that's how a supposed "expert" said it should be done.

Click to expand...

What have you moved to now?

You are a much bigger shop than I am, so take my suggestions with a grain of salt

The first 5 tools in all of my machines are standardized (tool holder, cutter, protrusion). For me that is a 2" shell mill, 1/2" EM, 3/16"EM, 1/8" EM and a 1/4" Spot drill. I base as much of my programming around those tools as possible to minimize how many tools I need to load for every setup. At times it become a trade off between cutting efficiency and set-up time.

Whenever a vise or fixture is mounted on the table, a known datum is written down on a clipboard at the machine. For me, I record the XY coordinates of the back left corner of vises and the YZA location of 4th axis centerline. (This is enforced... date and initials of who ever mounted the work holding) This way, I minimize the number of times I need to probe to get locations for the next part.

All soft jaws and custom fixture have a bored hole to make getting back into them very easy. I also record that location on the clipboard if the fixture is going to be in there longer than 24 hours.

Tool Offsets are all taken 5 inches above the surface of the table. For me that is done with a 4" Edge (brand) gauge on a 123 block. Then the Z offset to the top of the part is the difference from that 5" datum to the Z Zero in the program (usually the top of the part). The operator can then easily use a dial indicator to go from the tool offset block to the top of the part. If I am setting it up myself I usually just use math. Any tool loaded into the machine must be set exactly this way, that way I can glance in there and use any tool in the machine without re-setting anything up.

All of these have to be strictly enforced. But once it becomes habit, you can take tremendous shortcuts on setups. My goal is always to walk up to the machine, load a minimal number of tools and probe nothing. Hit the green button and go to lunch

BRIAN.T said:

What have you moved to now?

Click to expand...

I touch tools off the mounting block for the fixed jaw of my vise. The controller (Mitsu hardware with Mori U/I features) has a touch off routine that has a parameter for touch off block height that comps out the .03" aluminum shim that I touch the tools to. Then the G54 Z is set to where the part datum is is relative to the mounting block for the fixed jaw. I usually set that with a caliper or even scale for first ops. Second ops I'll either set it high and take a skim cut and adjust or I'll measure it with an indicator in the spindle.

Vancbiker said:

I touch tools off the mounting block for the fixed jaw of my vise. The controller (Mitsu hardware with Mori U/I features) has a touch off routine that has a parameter for touch off block height that comps out the .03" aluminum shim that I touch the tools to. Then the G54 Z is set to where the part datum is is relative to the mounting block for the fixed jaw. I usually set that with a caliper or even scale for first ops. Second ops I'll either set it high and take a skim cut and adjust or I'll measure it with an indicator in the spindle.

Click to expand...

Even that for a job shop might be dicey - what happens if you have a setup that requires you to take the vise out of the machine, and you break a tool halfway through?

If you touch off to a 123 block on the table your datum is always there. If you know the height of your vise bed then you can easily set your Z zero to it or a parallel height without measuring anything, it's just basic arithmetic. And if you make your XY zero coincide with the edge of a jaw you never have to probe anything, just adjust your fixture Z offset and run.

isaac338 said:

Even that for a job shop might be dicey - what happens if you have a setup that requires you to take the vise out of the machine, and you break a tool halfway through?

Click to expand...

All our vises were precision ground, so they are all very close.

I still do the touch off on the top of the part. I don't think one way is worse or better than another, they each have their place.

To the Brian the OP. If you are getting 3-4 set ups a day on each Machine, you are already ahead of the game. Standardizing process' is the best way to optimize time. But there is the other edge of the sword, it becomes monotonous and people get bored. IME there will be a guy or two, who are really passionate about optimizing the system and making it better, so use them for that. And there are guys like me, who only want to do the stupid jobs, that everyone else was too smart to submit a quote for. But you can use them for that and be productive too.

There are about 20 Milliions Thousand articles on Lean and 6 Sigma and all that crap, that will generate more revenue, if that's what you're after.

R

litlerob1 said:

All our vises were precision ground, so they are all very close.

I still do the touch off on the top of the part. I don't think one way is worse or better than another, they each have their place.

To the Brian the OP. If you are getting 3-4 set ups a day on each Machine, you are already ahead of the game. Standardizing process' is the best way to optimize time. But there is the other edge of the sword, it becomes monotonous and people get bored. IME there will be a guy or two, who are really passionate about optimizing the system and making it better, so use them for that. And there are guys like me, who only want to do the stupid jobs, that everyone else was too smart to submit a quote for. But you can use them for that and be productive too.

There are about 20 Milliions Thousand articles on Lean and 6 Sigma and all that crap, that will generate more revenue, if that's what you're after.

R

Click to expand...

I should clarify, we do maybe 4 setups total, not per machine. I like the idea of standardization on certain jobs, that may be an easily attainable goal. As for the stupid jobs, it would seem you are my direct competition, we drill so many holes in band sawed stock it's unbelievable. But it keeps my guys busy and paid!

isaac338 said:

Even that for a job shop might be dicey - what happens if you have a setup that requires you to take the vise out of the machine, and you break a tool halfway through? .....

Click to expand...

Yes, my choice of the reference surface would not be applicable to all users. The main point is to use a common, known fixed reference surface.

My situation is not typical of many, if not most, shops. I have the luxury of most everything I make being of my own design. Many of which are one piece only. So I design using vise workholding or fixture plates held in the vise. In the 6+ years I've had the machine, the vise has never been off the table. I have G54-G56 reserved and set at the corners and middle of the fixed hardjaw. I program using those datums as much as possible and use G57 and up for things that don't work out using the fixed hardjaw. This results in rarely having to find and set an X or Y fixture offset.

I have a pretty good collection of toolholders and leave my "core" tools set up with the length value noted on a tag. So I rarely have to touch off a tool, just key in the length so the appropriate offset position.

The end result is that very few new parts take more than 5 or 10 minutes to get going.

BRIAN.T said:

I'm kicking around a few ideas on how to speed up setup times for my mills. My shop is your typical job shop, we might do 4 setups a day between the mills. We've got 6 mills I'd like to streamline.

Click to expand...

Brian,

- If I understand correctly, you are currently setting all tool lengths to the top of the blank, or possibly to a step on a jaw, or something like that? Is that correct?

- How are you setting work coordinates? Edge finder? Haimer? Probe?

- Have you done a quick and dirty analysis on how much time is being spent on the various steps? (tool setting, work coordinate setting, etc.)

- Do all your 3-axis machines use the same tool holders? (e.g. Cat40 across the board, or something like that)


Personally... I touch all tools to the table (well, I touch them to a known-height indicator like the one made by Edge or something like that which is set to zero at an exact height). Then a quick run down the tool table and subtract that indicator height to all the tool offsets so all tools are set to the table. I've found that (for me) this method that makes the most sense on a machine that doesn't have a tool setter onboard. I have a Renishaw probe, and I know its length, which is set as its offset in the tool table. In other words, if I probe the table, the probe outputs Z=0. Since all the tools are set to the table, all that's required is to probe "whatever" is being used as Z=0 on the part, and that positive value gets put into G54/Z (or G55, or whatever being used). Done.


I think G00 Proto makes a great suggestion: Find the tools that are common to most/all your jobs (even if nothing more than a face mill, spot drill and a chamfer tool) and leave them set up in each machine. To really take advantage of that though, it's probably good to touch off on something other than the top of the part blank, otherwise the only time being saved is not having to put those tools into the tool carousel.

A bit more info on how much time is being spent, and maybe how many tools you might be using on a typical job might help some of us to get more specific with ideas.

The reason I asked about the time analysis is... many years ago, I had a repeat job in my shop which I did often. I finally came to my senses one day, and realized there was more time to gain in cutting the time for packaging/shipping than anything I could do on the machining end. I focused on streamlining that part of the process, and that paid off more than anything I could have done in the machine. Of course that might not be the case in your shop, but hey... I thought I'd mention it.

PM

For those unfamiliar, this is the style of Z touch off gage mentioned several times above: Pro Touch Off Gage 04-000 - Edge Technology

BRIAN.T said:

I should clarify, we do maybe 4 setups total, not per machine. I like the idea of standardization on certain jobs, that may be an easily attainable goal. As for the stupid jobs, it would seem you are my direct competition, we drill so many holes in band sawed stock it's unbelievable. But it keeps my guys busy and paid!

Click to expand...

As for the "stupid" jobs, I do the work that is borderline black magic. Not mindless, they are jobs that sane people with a good sense of ROI wouldn't touch with a Shepherds Crook.

But set-up is billable time. Are you competing to win new jobs? Or to improve what you already have going? Quoting new work and making money at it, is what separates us from Monkeys.

R

Disclaimer: I am a one-man shop, one VMC (10 tool), and have never worked in a "real" shop before

My machining business is a secondary source of income, and I only ever work on it between the hours of ~8pm-10:30pm. As such, it is vital that I am able to setup a job quickly, get it in the machine, and run it. Any inefficiencies can cost me the evening which is a huge loss of time/revenue (~$150/hr, $300-400 per evening). My jobs are typically very simple 2.5D with cycle times of ~30 seconds to 3 minutes (per op). I will typically run anywhere from 1-3 different jobs (start to finish) in one night. 80% of my work is done on a fixture plate and the other 20% in the vise. I never keep "permanent" work offsets/datums because my machine is relatively small and not worth the time to track and maintain that datum. Higher volume parts are done with a dedicated fixture.

My process is:
- The majority of my "common" tools are pre-loaded in holders on a storage rack (~40 holders?) with their heights recorded on a hand-written tag which is attached to the holder. The height is the length from spindle nose to tool tip. When I need to use the tool I load it into the machine and input the height into the tool table.
- Any new tools that need to be set up are measured by: Using the "Edge" height gauge to zero out my spindle nose. This is done using any surface.... I then load the tool and touch it down to that Edge height gauge. The Z reading is the tool length. This gets written on the tag and put into the control.
- My part/fixture XYZ origin is found with a Haimer probe. I know the height of my Haimer probe (6.6895" from tip at "zero" to spindle nose), which then gives me the fixture Z origin.
- Run the CAM, modify GCode as needed, and send to the machine.
- This has proven to be quite effective and efficient. My next improvement from here is likely going to be a quick-change pallet system where I can load parts offline for the next OP.

I like my sub-plate. I could (if I had them) have an intermediate plate with Lang setups on them. The sub-plate employs ball-locks (Jergens) that locates and secures the intermediate plates. I also have their nesting vises, also with ball lock interfaces. The vises all locate (I have 3 of them) within tenths, and each solid jaw is aligned within tenths. And since every thing is in a known spot, I can use G10 to set work offsets, no probing. My tool touch-offs are always from the sub-plate surface. Pretty quick changeovers. Works well for repeat jobs, especially.

Tool setting aside, the most wasted time is usually spent looking for crap that someone didn't put away the last time it was used.

I have job boxes for all my repeat parts. Whatever it takes to setup the part and get it running is in the box. That may be soft jaws, drills, taps, reamers, end mills, partially completed parts, material blanks, and any other tools that I use in the production of that part.

Job is finished, it gets broken down and everything goes back into the box. I make sure that if I need a fresh cutter or drill for the next run I order them and put them in the box before it gets put away.

Next job, grab the box for that part and make my setup. Everything I need is right there on the workbench.

guest said:

Tool setting aside, the most wasted time is usually spent looking for crap that someone didn't put away the last time it was used.

I have job boxes for all my repeat parts. .....

Click to expand...

Where do you draw the line at deciding a repeat job justifies establishing a dedicated box of tooling? Weekly, monthly etc.

I've seen other shops use similar methods with mixed success. Mainly due to folks knowing there is a tool they need in a box, getting it, and then failing to replace it when they were done or it is still being used when the box is retrieved to set up the repeat job. Others feel there is too much cost in under-utilized tooling when it sits in a box untouched for a few months waiting on the repeat job to come in again.

I've not yet seen any place have a truly "ideal" (at least in my eyes) system. What works for one shop doesn't at another.