What's new
What's new

Considering Single Workpiece Flow

Atomkinder

Titanium
Joined
May 8, 2012
Location
Mid-Iowa, USA
Also known as one piece flow.

Considering experimenting with this a little with a part that we always do in batches of 12 (sometimes more).

Currently the fixturing allows for two parts to be made at once, but we flip the parts to do Op2 on the same fixture. The process is relatively simple, the initial operation drills through-holes and counterbores, and the workpiece is bolted to the fixture with three 1/2-13 SHCS'. After the first operation (about 50 minutes for two peices) they are removed, the fixture cleaned appropriately, and the parts are flipped onto the second side where they are located with bushings that fit into counterbores on the fixture, clamped, new counterbores milled in the parts, then bolted to the fixture again where Op2 continues, for about half an hour.

So far reprogramming has yielded more in-machine deburring and a cycle reduction for two parts of half an hour (yes, 30 minutes). However I'm curious about the benefits of turning this into a one piece flow operation where Op1 is done on the first station, then Op2 is done on the second simultaneously, reducing the total number of tool changes (multiple tools get used on both sides of the part, but not all of them - tool count currently is 23 tools) per part and getting one part per cycle complete (well almost, there's a short 4th-axis drilling/tapping routine that could be integrated as well).

So I would like to know others' thoughts on one piece flow and whether it holds more benefits or drawbacks in the long term. We're just finishing a run of 15 parts though, so part of my consideration of this style of production is integrating a macro that lets the operator set the number of parts to be made so that separate programs are not needed (single vs double part) and the setup can be more automated in this regard (running Op1 only for the first part and Op2 only for the last, for instance).
 
We do the one piece thing here a lot. In our situation though it's 2 parts at once. 2 op 1, 2, 3, 4's with one cycle start. Parts flow from one vise on the left (raw material) to finished parts on the right. This job was run on an old slow mill so I think we saved 2 minutes in extra tool changes alone.
 
We do the one piece thing here a lot. In our situation though it's 2 parts at once. 2 op 1, 2, 3, 4's with one cycle start. Parts flow from one vise on the left (raw material) to finished parts on the right. This job was run on an old slow mill so I think we saved 2 minutes in extra tool changes alone.

Interesting. Currently the program is run on a Haas VF-6SS, and that's not going to get any faster. Most of the program is now pretty well optimized to result in good quality and short run time without much left to go, so I'm curious if running it through the one piece flow might make it more efficient or not. I think I might just have to try a program set up that way and do the math regarding the cycle time.
 
I approach all of my complex project like SMT described. Its really nice to get one part completed each cycle as well as to be able to make changes to the process before its too late. I don't do it just to save cycle time, that's just a bonus.
 
I think always, not so much for saving tool change, which can be substantial on quick parts but won't make much of a difference for these particular parts you are talking about, but so that 1.You can check all the finished dimensions on the part, and know all the ops are good. 2. The less the machine stops the better. 3. No piles of half done parts. 4. How many times do you start running a job, and then they need parts right away? If it's all in one program then you're getting done parts to ship.

And usually anything that repeats that uses 2 or more vises I'll make a fixture with all the ops on a plate or block with one pickup hole to cutdown on setup time,.
 
In my small shop I try to make it as easy as possible for the operator to make the parts. Ive had nothing but trouble with trying to have 3 different setups running in the same machine at the same time. Load one of them wrong and break all the tools and back to square 1.

Just my opinion...
 
However I'm curious about the benefits of turning this into a one piece flow operation where Op1 is done on the first station, then Op2 is done on the second simultaneously

It can decrease handling time and storage requirements. OP1 parts go straight to the OP2 fixture without exiting the machine. No need to store OP1 parts on a cart. OP2 parts go straight to cleaning, inspection, and packaging.

On the flip side, dstyr's concerns are valid. You also have more unique tools in the machine at any given time, so the chances of a worn tool pausing the cycle are higher.
 
I think always, not so much for saving tool change, which can be substantial on quick parts but won't make much of a difference for these particular parts you are talking about, but so that 1.You can check all the finished dimensions on the part, and know all the ops are good. 2. The less the machine stops the better. 3. No piles of half done parts. 4. How many times do you start running a job, and then they need parts right away? If it's all in one program then you're getting done parts to ship.

And usually anything that repeats that uses 2 or more vises I'll make a fixture with all the ops on a plate or block with one pickup hole to cutdown on setup time,.

This is probably true. We did just run one part, total time at 45 minutes and a few seconds. By making two at a time we are saving a whopping three, maybe four minutes.

In my small shop I try to make it as easy as possible for the operator to make the parts. Ive had nothing but trouble with trying to have 3 different setups running in the same machine at the same time. Load one of them wrong and break all the tools and back to square 1.

Just my opinion...

I can definitely see this. On the other hand, this SHOULD be about the simplest version of such. Op1 is a 12x12x2 block of 6061. Op2 has the outside profile finished, part would be flipped over into rear station, and located by features machined on the first side and cannot be located without them. Op3 is drill/tap on the 4th axis. Hard to screw that progression up IMO, but then again, better idiots, etc.

It can decrease handling time and storage requirements. OP1 parts go straight to the OP2 fixture without exiting the machine. No need to store OP1 parts on a cart. OP2 parts go straight to cleaning, inspection, and packaging.

On the flip side, dstyr's concerns are valid. You also have more unique tools in the machine at any given time, so the chances of a worn tool pausing the cycle are higher.

In our current situation the only operation that is waiting is Op3 to go on the 4th axis. Integrating this currently can't be done because there are two parts done per cycle now, but the rotary fixture only holds one. With a single part flow it could be, and spit out one complete part per cycle. There aren't really parts sitting around now though in the sense that they are waiting for a significant operation. The parts are 98% done per cycle. I may have just argued myself into doing it though with being able to complete via the 4th. :scratchchin:
 
One other reason I have the guys set jobs up like this is (and hopefully this hasn't already been mentioned and I missed it. . .) it increases the length of time the machine runs. I'm not talking total cycle time, but combined cycle time. Take our 4 op parts that are a few minutes per op.

Run them all at once and my guy is free to go do something else productive for about 23 minutes. 5 more minutes to unload/load vises and he walks away after hitting cycle start.

It's not exactly free labor, but it's close enough for me :)
 
I try and do all of my parts that way. Most of my parts end up being OP1 in a Vise, OP2 in a second vise and then 4th axis work on a fixture. Generally I write two sets of programs. The first program is just the parts in order of operation (ie all of OP1, all of OP2 etc). I use that for proofing out the program. Then I re-arrange all of the tool paths to minimize travel distance and tool changes. I keep the original program for starting the job and finishing the job so I don't need to run through a bunch of air cutting.

This system drastically reduces my WIP and makes me very flexible on how many I need to run. Also I can QC a finished part every time I open the door... and tumble and ship if the customer all of a sudden has an emergency.

I used to run
 
I used to build machines for the greenhouse industry. Customers would ask how long the machine would take to do a task before it could be moved into another greenhouse. They were often surprised when I told them that you don't want it to work too quickly. If it does the operator won't have time to do other things and will stand there doing nothing waiting for the machine to finish!
 
One other reason I have the guys set jobs up like this is (and hopefully this hasn't already been mentioned and I missed it. . .) it increases the length of time the machine runs. I'm not talking total cycle time, but combined cycle time. Take our 4 op parts that are a few minutes per op.

Run them all at once and my guy is free to go do something else productive for about 23 minutes. 5 more minutes to unload/load vises and he walks away after hitting cycle start.

It's not exactly free labor, but it's close enough for me :)

I try and do this as well when allowable, but usually we are making a handful of parts that we make on a regular basis. We aren't a job shop, we only make parts for in-house use. So generally I have been focused on extending tool life, shortening cycle times, and reducing setup time, as we do more setups than anything else. The argument that we should be keeping some inventory to reduce the time dedicated to setting up the same job twice in as many weeks never seems to sink in, and I've been making it for over two years now. But that's a different thread. :crazy:

I try and do all of my parts that way. Most of my parts end up being OP1 in a Vise, OP2 in a second vise and then 4th axis work on a fixture. Generally I write two sets of programs. The first program is just the parts in order of operation (ie all of OP1, all of OP2 etc). I use that for proofing out the program. Then I re-arrange all of the tool paths to minimize travel distance and tool changes. I keep the original program for starting the job and finishing the job so I don't need to run through a bunch of air cutting.

This system drastically reduces my WIP and makes me very flexible on how many I need to run. Also I can QC a finished part every time I open the door... and tumble and ship if the customer all of a sudden has an emergency.

I used to run

Used to run what????? Inquiring minds would love to see the end of that sentence! :D

I used to build machines for the greenhouse industry. Customers would ask how long the machine would take to do a task before it could be moved into another greenhouse. They were often surprised when I told them that you don't want it to work too quickly. If it does the operator won't have time to do other things and will stand there doing nothing waiting for the machine to finish!

I can't disagree with the sentiment, but at times only one machine is running anyway (we have two VMC's, a turning center, a TL-1, and a "manual" area, each section of those get an operator/machinist at once) so cycle time is more important there than anything else unless we are completely slammed, which hasn't happened since the end of 2014. But then cycle time counts too because we don't have any way to track setup times (another argument I've been placing on deaf ears).
 
I have several situations where I do this, and always choose to get a finished part off every cycle. Perhaps if deburring was required between steps I would think differntly
Reduce handling time, or non spindle time

If the machine runs too fast you put more parts on the machine. Never slow down. Speed isn't everything, it is the only thing. Unless you don't like making money

50 minute, 30 minute cycles, man I guess I have been running aluminum too long, but sounds like there might be some optimization yet to come there
 
We do a few jobs this way, but only on the larger quantity stuff like 100+pcs.
Unless it's a repeat job that we get over and over then we will make dedicated fixtures and combine the programs. But being a job shop, the time spent in setup and programming that style of work for 20 pcs isn't usually worth it.
 
We do a few jobs this way, but only on the larger quantity stuff like 100+pcs.
Unless it's a repeat job that we get over and over then we will make dedicated fixtures and combine the programs. But being a job shop, the time spent in setup and programming that style of work for 20 pcs isn't usually worth it.

I would agree with this. For something like 20 pc we may very well just do all op1 then all op2.
 
I've pretty much always done one piece flow, because prototyping.. prove out the process before risking all time/material. It transitioned into production because on small batch production because of a bunch of unrelated benefits; being able to inspect hand off a finished part to QC for inspection, reduced handling/storage, reduced scrap, etc..

It's virtually never an improvement to overall process time. Sharing tool changes is more efficient with two pcs of the same op than two pieces of different ops, etc. It's all ancillary benefit, and whether that outweighs the loss to efficiency is fairly process specific.
 








 
Back
Top