PDA

View Full Version : Reintegrating machining in our company



beast
03-20-2005, 06:12 PM
Hi Folks,

I'm not sure if I've posted this in the right forum as their doesn't seem to have a general business forum. At any rate...

We manufacture industrial equipment and have a number of machined parts. I've been tasked to develop the business case of reintegrating mechanical design and machining in house. The reasons include:

- lower inventory with spontaneous build to order
- faster delivery times to the customer
- mass customization options in the future

Outsourcing is generally a batch operation which is inconsistent with our business .

The goal is to develop a machining and painting cell that has the following characteristics:

- flexible cell with 3-axis CNC machine whereby we can switch jobs easily
- ultimately work towards batch-size-of-one so reducing setup is important
- automated as much as possible, no paper processes or anything like that
- capacity is less of an issue. We plan to manufacture 3000 pieces annually (30 min. per piece) but the pieces can vary.

Three questions:

1) What type of equipment would members recommend to get this cell up and running. I was thinking of the following to get a reasonably priced, flexible cell (but again my knowledge here is limited)

- 3-axis CNC machine like a Haas Mini-Mill
- Horizontal Band Saw
- Pro Engineer CAM software
- Pro Engineer CAD software
- End Mills and options
- Air compressor and misc
- Basic metrology equipment

2) How long would it take to commission a cell like this? Is there a how-to manual to do this OR are we better off to hire the machinist first to help us out?

3) What's required in terms of tooling up the facility (HVAC, electrical, insulation?)

Thanks in advance,
beast

trubleshtr
03-20-2005, 06:54 PM
What kind of products will you produce? Do you need a large CNC or small? Will the parts made be made of hard metals, therefore requiring a ridgid machine?
Also don't forget coolants, they can get costly if you are running semi-synthetics, which could be necessary depending again on the metals you plan on cutting. We also added mist collection systems from Nederman so our plant was'nt so "cloudy".
Other companies for cnc you may want to consider are, Hitachi(mori-seki),Okuma, MAkino, they all make "off the shelf" felxable cnc's. Only problem with them is the
t-shooting when they breakdown as most of the software is propiatary, and they won't share things with you like PLC ladders and such, which, puts you at their mercy during breakdown scenarios. You will have to add your tombstone to the cnc if you go with any of the afore-mentioned.
You'll need to consider the power into the shop, transformers and such. Where are you in Canada? I know people in GTA ontario that are good for that.
If you are going to have paints, you may want to consider a fire proof cabinet for storage to keep the insurance/fire dept. happy.
Oils for lubricators and hydraulic systems.. you should have on hand too, I'm just rambling things out that come to mind as I am typing so I apologize if this seems long winded or obvious.
What about a robot load/unload system that can place raw parts into the cnc and then the finished machine part onto a conveyor and deliver it to the paint booth? where another robot does the painting?

Nonoriginal
03-20-2005, 07:33 PM
Ok, I'll have a crack at this... I work in a small CNC shop and I have been dreaming of starting my own shop since I walked into this trade. The only thing holding me back are my skills(time in the trade) and $$$$$$$. At my current shop we manufacture our own products for the off-road racing industry. Typically between 40 to 100 parts for each item or assembly. We also run jobs / batches of parts for the aerospace, and advanced optics industries. As well as small onesy and twosy jobs a.k.a. Prototyping.



- lower inventory with spontaneous build to order
- faster delivery times to the customer
- mass customization options in the future

Lower inventory, yes... but you now have a much higher overhead.

Faster to customer, yes

Mass customization, yes... but you have to have the skill set in house to implement these customizations. More on this later.


Outsourcing is generally a batch operation which is inconsistent with our business .

The goal is to develop a machining and painting cell that has the following characteristics:

1 flexible cell with 3-axis CNC machine whereby we can switch jobs easily
2 ultimately work towards batch-size-of-one so reducing setup is important
3 automated as much as possible, no paper processes or anything like that
4 capacity is less of an issue. We plan to manufacture 3000 pieces annually (30 min. per piece) but the pieces can vary.

Items 1 and 2 conflict or items 2 and 4 conflict...I'm not sure what you mean with "batch-size-of-one."

The best way to manufacture products is to run them in batches, otherwise you will spend to much time setting up. If its only one, you still need to set-up, and run the part. BUT, you'll have a $50K machine and a $40 to $60K a- year machinist standing around with nothing to do. You'll have to do the math to see if it makes sense to your company.


1) What type of equipment would members recommend to get this cell up and running. I was thinking of the following to get a reasonably priced, flexible cell (but again my knowledge here is limited)

- 3-axis CNC machine like a Haas Mini-Mill
- Horizontal Band Saw
- Pro Engineer CAM software
- Pro Engineer CAD software
- End Mills and options
- Air compressor and misc
- Basic metrology equipment

2) How long would it take to commission a cell like this? Is there a how-to manual to do this OR are we better off to hire the machinist first to help us out?

3) What's required in terms of tooling up the facility (HVAC, electrical, insulation?)



Machines: Haas is fine, we have three of them 2 VF-4's and a VF-3. I prefer the Matsuura 800 with box ways, but your talking big $$$$$ $200K. Our Haas's are much easier to run and MUCH less expensive when compared to the Yaznak control on the Matsuura.

You will be extremely limited in your work envelop with a Mini from Hass. Better to go bigger right away for $20K or $30K more IMO.

Stock cutting: A horizontal bandsaw if your stock is bigger than 6" x 6" x XX" irregardless of the material. If smaller and you primarily use Steel....Bandsaw. If smaller and you primarily use Aluminum, I'd go with a cold circular saw. It would probably be beneficial to have both, IMO.

Software: If you already have Pro/E or someone proficient with it... then go for Pro/E Wildfire. If not, I would look into Solidworks and the various CAM add-ons they offer Ie. CAMworks and SolidCAM. We use Solidworks and GibbsCAM at work, but I have seen a Demo of CAMworks. I would pull the trigger for that software in a heartbeat if I was calling the shots. Simply AWESOME..Automatic feature recognition and machining processes sorted and ready to go with a click of the button.

There are many, MANY others and I would try a few to see what you like before spending the gigantic sums of money these people ask for. Your software could cost more than the machine itself. Take your time here!

Tooling and options: $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
Big money, don't skimp on cheap crap unless this is a hobby and I already know its not a hobby. So be prepared to spend $150 to $300 a piece for nice endmills and you'll need a lot of them... as they will break when you don't want them to. A couple of grand on a nice Fly cutter. A couple of grand on some nice vices. A couple of grand on tool holders. You can't have enough of this stuff.

Air: You have to have it...to breathe :rainfro: All kidding aside, its a must have. Haas has many recommendations with them. You'll also need a shop air dryer.

Metrology: Yep, you'll need to inspect the parts. A CMM would be nice, but you could get by with;
Granite plate
Gage blocks and pins: Depending on your tolerance's... we use Grade B the most. Grade AA for the Aerospace stuff. Its nice to have both
Height gage with a dedicated test indicator
Precision Angle blocks and V blocks
Microscope

Other things you need:
Bench grinders
De-burr equipment
Coolant
Way oil
more tooling...always

Other things you MAY need:
Small lathe
Precision grinder (We actually need one, but you might not)
Knee mill
Tumblers


Facility:
Air: Shop air plumbing
HVAC: We have swamp coolers and fans for our shop and they work fine...in AZ... where are you?
Electrical: I think we use dedicated 3 or 5 phase power for the machines. You could get by with a nice VFD.
Shipping and receiving bay
Forklift
Tool Storage
RADIO
waste storage for used coolant and other oils

I know I've left some stuff out...but it will come to me. Let me know if this helps...


BTW - Don't do this without a machinist.

DAB_Design
03-20-2005, 08:33 PM
One thing I noticed that wasn't mentioned, is initial material costs. Probably won't be the biggest purchase, but will definately be a nice chunck of change.

beast
03-20-2005, 08:56 PM
Thanks for the information,


What kind of products will you produce? Do you need a large CNC or small? Will the parts made be made of hard metals, therefore requiring a ridgid machine?

All parts are no larger than 4" x 4" x 4". 90% of the time, it's 6061 aluminum but on occasion, it could be stainless steel.


Also don't forget coolants, they can get costly if you are running semi-synthetics, which could be necessary depending again on the metals you plan on cutting. We also added mist collection systems from Nederman so our plant was'nt so "cloudy".

What are semi-synthetics? Why is the shop cloudy?


You'll need to consider the power into the shop, transformers and such. Where are you in Canada? I know people in GTA ontario that are good for that.

Yes I would like a name if you got one. Is it 3-phase power or something more exotic?


What about a robot load/unload system that can place raw parts into the cnc and then the finished machine part onto a conveyor and deliver it to the paint booth? where another robot does the painting?

Our shop is small as are the parts. So, I don't think automation with robotics is appropriate at this time but who knows what the future will bring.

beast
03-20-2005, 08:58 PM
One thing I noticed that wasn't mentioned, is initial material costs. Probably won't be the biggest purchase, but will definately be a nice chunck of change.

If I remember correctly, aluminum is $1.99 Cdn a pound. I think we're looking at about $15,000 worth of material a year so it doesn't factor as much as the initial capital equipment.

Thanks again,

beast
03-20-2005, 09:13 PM
Thanks for the information,


Lower inventory, yes... but you now have a much higher overhead.

Our initial calculations is that our part costs drop by a factor of 3 and payback on all capital equipment is less than one year.


Items 1 and 2 conflict or items 2 and 4 conflict...I'm not sure what you mean with "batch-size-of-one."

The best way to manufacture products is to run them in batches, otherwise you will spend to much time setting up. If its only one, you still need to set-up, and run the part. BUT, you'll have a $50K machine and a $40 to $60K a- year machinist standing around with nothing to do. You'll have to do the math to see if it makes sense to your company.

Mass production works best if you produce the same item over and over again. We will be producing several hundred products, often with customizations. We need to be able to build products to order which have very small batch sizes. My question to you is this. Is it possible to have a system whereby you feed in the same raw material (same size, same composition) and get a different part on the output with no setup? We ultimately need to get there.


You will be extremely limited in your work envelop with a Mini from Hass. Better to go bigger right away for $20K or $30K more IMO.

Our parts are small. No more than 4" x 4" x 4" max.


Software: If you already have Pro/E or someone proficient with it... then go for Pro/E Wildfire.

We plan to hire a mechanical designer who uses nothing but Pro/E. I'm not sure about Pro Manufacture so if you have any suggestions, I'd appreciate it.


Tooling and options: $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
Big money, don't skimp on cheap crap unless this is a hobby and I already know its not a hobby. So be prepared to spend $150 to $300 a piece for nice endmills and you'll need a lot of them... as they will break when you don't want them to. A couple of grand on a nice Fly cutter. A couple of grand on some nice vices. A couple of grand on tool holders. You can't have enough of this stuff.

What's a Fly cutter? A shop air dryer? and why do I need these items.


Other things you need:
De-burr equipment
Coolant
Way oil
more tooling...always


Can I deburr with the milling machine. ie. chamfer the corners and so on...


Facility:
Air: Shop air plumbing
HVAC: We have swamp coolers and fans for our shop and they work fine...in AZ... where are you?
Electrical: I think we use dedicated 3 or 5 phase power for the machines. You could get by with a nice VFD.
Tool Storage
waste storage for used coolant and other oils

I'm in Canada. Pretty cold up here.
What's a VFD.

I appreciate the feedback.

trubleshtr
03-20-2005, 10:01 PM
All parts are no larger than 4" x 4" x 4". 90% of the time, it's 6061 aluminum but on occasion, it could be stainless steel.


What are semi-synthetics? Why is the shop cloudy?



Yes I would like a name if you got one. Is it 3-phase power or something more exotic?





Maybe you could go with a bridgeport mill that is cnc'd

Coolants that are man made,If you use alot of coolant in your process, the mist can start to collect in the air

Phase will depend on the machine you use, most 200+ voltage machines are 3 phase. 110v and less are normally single phase.

Nonoriginal
03-20-2005, 10:23 PM
Thanks for the information,

Mass production works best if you produce the same item over and over again. We will be producing several hundred products, often with customizations. We need to be able to build products to order which have very small batch sizes. My question to you is this. Is it possible to have a system whereby you feed in the same raw material (same size, same composition) and get a different part on the output with no setup? We ultimately need to get there.

No, not unless you want a bunch of scrap. Set-up for one operation is (not nessisarily in this order);
1. putting the tools together that are needed for the operation.
2. Determine how the work is going to be held
3. Tramming or aligning the vices, Tombstones, fixtures or a combination thereof.
4. loading the program
5. input of the various offsets...ie. tool length, work fixture, and tool radius
6. Run each tool for varification on the part.
7. fix or adjust a.k.a. "dial in" the program to where you want it.

Pretty simple, eh? Well depending on the complexity of the part and the machinists skill this can take anywhere from 30 min to three days. Clear as mud?



Our parts are small. No more than 4" x 4" x 4" max.


Good, go with a cold saw and be done with it. Ma mini mill may be what the Dr. ordered.



We plan to hire a mechanical designer who uses nothing but Pro/E. I'm not sure about Pro Manufacture so if you have any suggestions, I'd appreciate it.


Any of the higher end CAM programs will work with Pro/E...just make sure the person using it...likes it!



What's a Fly cutter? A shop air dryer? and why do I need these items.

A fly cutter is a large diameter indexable (uses Carbide inserts) cutter that basically makes a wide flat cut across a wide semi-flat surface. Its nice to fly cut the material so you have a flat surface to locate on. If your work is 4", I would go with a 2.5" or 3" fly cutter.

Compressed air tanks have condensation issues. You don't want to spray wet/oily air onto your work or your steel machine, and you don't want wet/oily air in your air plumbing... therfore you need to dry the air as it comes out of the tank.



Can I deburr with the milling machine. ie. chamfer the corners and so on...


Yes, it all depends on what type of finish your willing to live with. If you consider tooling marks an issue, like I do, then you need to tumble the parts and de-burr that way. We do both...everyday.



I'm in Canada. Pretty cold up here.
What's a VFD.


Too cold for me!

You MIGHT be able to get away with a VFD if you have single phase power and you need three phase...I'm not sure...check with Haas or whoever you buy the machine from...

A VFD consists of three distinct sections:

1) a "converter" section, which takes single- or three-phase ac and converts it to high voltage dc, either 325 or 650 volts,

2) a "bulk storage" capacitor section, which stores the output of (1), and

3) an "inverter" section which takes charge stored in (2) and converts it ... more precisely, inverts it ... to three-phase ac.

Each section has its own performance requirements.

By properly designing each section, and selecting from a number of standard sections, the following can be provided:

1) 115 volts single-phase to 230 volts three-phase,

2) 230 volts single-phase to 230 volts three-phase,

3) 230 volts three-phase to 230 volts three-phase,

4) 230 volts single-phase to 460 volts three-phase,

5) 460 volts three-phase to 460 volts three-phase.



I appreciate the feedback.

No problem...

HuFlungDung
03-20-2005, 10:26 PM
Beast,
Are your parts in "families", or what kind of relationship do they have to each other? How many parts make one set to complete one order? How many times does the part have to be moved to complete machining?

The number of tools you will use to machine all these parts is an important consideration. Consider the toolchanger capacity. You would at minimum, need a dedicated toolholder for every tool you will use. Offline tool presetter would be important. Automatic loading of tool numbers and tool offsets for each program is also important for what you are trying to achieve.

Consider using fixtures on removeable pallets to quickly change setups. Expensive yes, but if the parts are difficult to set up, the time required to set up for one is a killer.

Pro E is pretty expensive, not? Take a look at OneCNC while you're shopping for software. You want something that most anyone can learn with a minimum of learning curve, in case your machinist quits on you.

beast
03-21-2005, 07:42 AM
No, not unless you want a bunch of scrap. Set-up for one operation is (not nessisarily in this order);
1. putting the tools together that are needed for the operation.
2. Determine how the work is going to be held
3. Tramming or aligning the vices, Tombstones, fixtures or a combination thereof.
4. loading the program
5. input of the various offsets...ie. tool length, work fixture, and tool radius
6. Run each tool for varification on the part.
7. fix or adjust a.k.a. "dial in" the program to where you want it.

I understand. What if:

1) I was able to standardize on the tools used to do the job such that I don't have to change tools
2) Use the same fixture for all designs
3) Have a means to automatically load the program

Would I be closer to no setup?

Thanks again,
Beast

beast
03-21-2005, 07:50 AM
Are your parts in "families", or what kind of relationship do they have to each other? How many parts make one set to complete one order? How many times does the part have to be moved to complete machining?

Yes, the parts have similar characteristics. Some might have a few extra features. Others might have different height offsets. It currently takes three parts to complete the order. We're moving to one in the near future. The part has to move once (top operation and then a bottom operation)



The number of tools you will use to machine all these parts is an important consideration. Consider the toolchanger capacity. You would at minimum, need a dedicated toolholder for every tool you will use. Offline tool presetter would be important. Automatic loading of tool numbers and tool offsets for each program is also important for what you are trying to achieve

What's a tool holder? An offline tool presetter?



Consider using fixtures on removeable pallets to quickly change setups. Expensive yes, but if the parts are difficult to set up, the time required to set up for one is a killer.

I like that idea. Do you have any names or links on this?

Thank you,
beast

Nonoriginal
03-22-2005, 01:58 AM
I understand. What if:

1) I was able to standardize on the tools used to do the job such that I don't have to change tools
2) Use the same fixture for all designs
3) Have a means to automatically load the program

Would I be closer to no setup?

Thanks again,
Beast

A Little. You still have to verify each tool during the first run. Then you need to check the part for accuracy. Make adjustments... check part...and then let it rip until you have enough of whatever it is your making...

Repeat for the bottom operation...let it rip until they are done!

Aside, these machines are not perfect, and no boss I've met enjoys a bunch of expensive 6061 that as been turned into scarp because of lack of attention. (BTW the 6061 is now worth half-pennies on the dollar.) You need to have someone close to the machine at all times to make sure things are going well.

It only takes .0001 seconds to snap a carbide endmill or break a carbide insert.

For example; You have a part that is dialed in sweetly. You walk away and let the machine run. The machine loads an endmill and proceeds to pocket an area...for whatever reason the endmill breaks in half leaving one half in the part and the other in the tool holder. The machine finishes the path and returns home to load the next tool. Lets say its a finishing endmill.

The finishing endmill plunges down and begins pocketing...OH S^%T!...you have a broken carbide endmill in the part...BANG, BANG, BANG they all go down... Best case scenario is you have a bunch of broken tools...Worst case is you need a new transmission and spindle. Your part is ruined regardless. BIG MONEY MISTAKE... Your call.


Also, Keep in mind that a circular saw is only accurate to + or - .010 or so... if your good. More like + or - .100 for the average joe. Band saws are WAY worse. Stock cutting irregularities are one of the biggest problems I've seen. People just assume the machine will take care of it. NOT TRUE!

beast
03-22-2005, 06:55 PM
I appreciate your comments. No setup in my mind still requires paying attention to the machine as you stated.

How long would it take to commission a machining cell given what you know?

Nonoriginal
03-22-2005, 09:47 PM
How much money are you willing to spend?

beast
03-23-2005, 09:18 PM
Equipment costs budgeted at around $60K.

Nonoriginal
03-24-2005, 02:24 AM
I would say its a go! However, you can't neglect to consider the hidden costs Ie. Shipping, unloading, and setting-up. Diesel is expensive right now, so tack a little extra for shipping. We were going to buy a machine from a shop across town (30 Miles) a few months back and the people we contract to unload and position our equipment quoted $4K...30 miles...OUCH!

It takes a while to level a machine...the concrete quality becomes apparent at the wrong time... Plan on a day for this task alone...

There is more...

You need to call a Haas distributor near you to check out the current lead times..Perhaps even a distributor in the Northern states.
http://www.haascnc.com/products/vmc/prcMiniTM.asp

From the time you get the machines...saws, CNC, etc. It shouldn't take more than two weeks to see some chips flying. Hire some help to get rolling, and you may see some chips in as little as three days!

Good luck!

murphy625
03-24-2005, 02:59 AM
Wow.. this thread is getting long.

Ok.
1st.. If you want to automate the process you will have to consider the following:
A) How long each operation takes to complete.
B) How to get the part from one operation to the next with minimum effort.
C) Space constraints. (Never be afraid to go vertical if you have the head room)
D) How much automation do you want.? No industrial process is totally automated. There must be human input at some "point". The less human input "points" you have, the cheaper to run but more expensive to build.

Start by drawing a block diagram flow chart of the process. Use each block in the flowchart as an "index" and reference that to a page of solutions or ideas to complete that part of the process.

I recall you said something about a CNC machine but you wanted to stay away from a Robot.. To me, this is like saying I like rocks, but not stones. Maybe I misunderstood something.

Who do you plan on having build this thing? Welding? Electrical? Saftey? bla bla bla...

Not much talk about the paint process in this thread. What are your finish specs? What kind of coating do you have in mind? What quality of coating do you need?
Are there colors to be considered?

You will also have to calculate energy costs because some processes need heat and natural gas is going up fast $$$.

One other suggestion, perhaps it would help if you posted a photo of what your trying to do... Seeing a finished part and knowing its finish requirements goes along way in understanding the best route to take.

Just some thoughts.
Murphy

DareBee
03-24-2005, 08:51 AM
If you are doing top and bottom work on your parts it may be possible to fixture a 4th axis to do the part all in 1 set up. If it is 100% machined would be an idea to run 2 fixtures in the CNC cutting the top in 1 and the bottom in the other, reducing tool changes and operater attendance time.

I can hardly imagine putting together a "cell" for $60K but without full and complete specs sitting in front of me it is almost impossible to devise a cell plan.
You also dont want to limit your budget to inexpensive equipment if you plan on long-term manufacturing, "cheap" equipment should be considered as disposable because you will be replacing it!
You get what you pay for.
Last cell I designed and installed we had close to $10K just in consulting, procuring, etc - just generally making sure things were planned, purchased, and going to work.