@Bastelmike

Your inputs, even if valid, are mostly negative, Mike.  Do you have suggestions on how to get around the problems you have identified?  Is there a better way of accomplishing what we are trying to build?

For example, if we use a polymer concrete, suitable reinforced, would that provide the stable base that we need?  Regarding the multimachine, would it be better to create separate machines with a drive unit that can be moved between them?  How would you go about designing these machines to meet our needs?

- Mark

Oh, BTW here is a book I just had open in another browser window, which looks pertinent:
http://www.scribd.com/doc/37301862/Principles-of-Precision-Machine-Design

I have also been looking for a top-down view of existing manufacturing systems which I think would help a lot with the flexfab facility design.

@gregor

>  Well, it may be negative, but first of all if it's true it's true.

My point was to encourage solutions to problems, not just pointing problems out.  In my experience, people who are smart enough or experienced enough to know when something won't work are often quite capable of contributing to solve the very problems they rais.e

>  Knowing it won't work without knowing the solution is a step
forwards..... 

Naturally, it's  matter of perspective and definition, but I don't agree.  Moving backwards away from things that won't work or cause serious problems is part of the overall process, but I wouldn't call it moving forward.  The problem is that people sometimes place a value judgment on moving away from the goal.  This has been shown mathematically to be necessary in some circumstance (see Simulated Annealing, form example).

>  and secondly you don't want to bite the newb.

Oh, tsk.  I didn't bite him all that hard.  :)

- Mark

@mjn
-
How would you go about designing these machines to meet your needs ?
-

For lathe work, I would go for a design like conventional lathes. A rigid non moving horizontal spindle and the tool performs movements in X and Z axis.
Not sure about the need for a lead screw, because it complicates design significantly. Having no lead screw is a drawback, except the machine is CNC-controlled and there is a synchronisation between spindle and X-axis available.
Many shafts have large threads, 2", 3" and more, for example to fix ball bearings. If you can't cut those large threads, you have a real problem, because there are no affordable cutting tools available for, let's say a 3" thread on a shaft.

For milling, I am not sure between two possible designs.
For sure, I wouldn't build a knee mill. The design is too complicated and its stiffness is bad.

One design is similar to the original multimachine setup, which in fact is more like a horizontal boring mill than other milling machines. Horizontal boring mills are phantastic and universal machines, but for enjoying all the advantages of that design you need a turning table on the machine too.
A drawback to that design is that milling is more difficult because you miss a vertical spindle; except you have an angle head for your boring mill.

Another favorite design would be a bed-type mill (not sure if this name is used in USA).
I am talking about a mill with a table, that performs the feed in X direction, while Y and Z feeds are performed by column/spindle. This setup makes usually for stiff designs, well suited to CNC operation.

Which of those designs is preferable, is a big question.
I think it depends on the kind of workpiece that is most frequent. For square box-type like workpieces, the boring mill layout would be my favorite.
If you have long workpieces with smaller width and height, the bed-type mill would be the type I prefer.

Mike

I'm not trying to rack anyone's jewels over this, I'm just trying to inject some realism, & practicality.

What are the goals for the machine tool?  Everyone may have different answers to this question, and as such, their ideal machine tool would probably look very different from anothers' because of their respective different needs & goals.

I'm trying to frame this in terms of what would allow a labor or time savings in manufacturing things like the CEB Press, & the LifeTrack, or similar, and none of those goals require any level of precision that is not easily attainable with a generic MultMachine design.

It's certainly accurate & precise enough to allow one to start making electric motors for things like the torch table, and it is even precise enough to be able to manufacture hydraulic motors and pumps, when OSE progresses far enough to be able to make their own parts for those too.  By the time OSE grows a need to have more precision than what is easily attainable with the multimachine, everything else kinda falls into the category of an academic discussion - which isn't all bad in and of itself, as long as it didn't hold up actual production or product development, all for the sake of a better mousetrap for that mouse that hasn't yet evolved...

@Allen15

> I thought this was for OSE, not for our dream shops, though?

Personally, I think it does support OSE - in conjunction with my own wants and needs.  Gathering metal working tools allows me to work on GVCS parts and even whole tools.  If I have a decent mill, lathe, drill I can still build a PowerCube.  I don't HAVE to use an OSE Multimachine.

- Mark

It is unlikely that OSE or anyone else will obviate or obsolete a good machinist (ever), any more then teaching some machine to play the flute will be able to put Hubert Laws out of business. They are artists at work.

The best that is likely possible, is that the minimum quality level with little or no training can go up with the right tools.  That being said, the people who are expecting to be able to use this equipment will become better designers if they can at least get the basics of machining down on the respective equipment that they are needing to use, so that they can become more familiar with the limitations of their own equipment, and that has to start manually, not CNC, or there will probably be many more and more expensive mistakes made in the learning process.

Yes, the process has to be something that CAN be automated, to a degree, but combining the possibilities of this platform with even average machining experience can increase the payoffs considerably, and faster too.

gregor:
But Mike,
remember to keep this in context of what they want to make with the thing, basically making all the gvcs tools and things like them. they won't be turning anything that weighs tons, there is nothing like that in the GVCS and probably never will be.
Yes, I have seen what is planned in the GVCS. I found there things like 200hp-tractors, trucks, bulldozers and 50MW wind turbines.
These all have parts you want make on a table lathe ??!
I'm dealing with OSE since a few days, but for me there are big discrepancies between what they plan to do with GVCS and the OSE specifications; and how they plan to do it !
One OSE specification demands OSE has to be of equal or better performance than commercial available tools !!!
Does anyone think, that can be achieved with cheap simple tools, made from scrap parts.
Another example:
There are basic drawings of the OSE shop on this site. Do you really think, that is a shop for building trucks and bulldozers ?
I made a simple basic layout for a shop where you can build trucks. You have the space and most of the necessary tools. Its huge, look for yourself. Building size is 164 x 58 ft.
Its no factory for mass-production of trucks, I would guess its capacity at 1 truck per day. Do you still believe, with the OSE shop and a multi-machine you can manufacture trucks ?
Mike

I've been around the corner with OSE on this topic.  What is that saying?  "A little knowledge is dangerous"?....:)

I operate professional quality CNC machines for a living and there is no way that a untrained person could even run one never mind program it.  Even a specially made, easy to operate Multi-machine will be next to impossible to program without a fair bit of training and experience.  The cheesier the machine, the more skill you will need to get it to do anything useful.

It takes 5 to 10 years to create a decent machinist that can make commercial quality mechanical parts.  My purpose here is to put things into perspective.  To recreate 10 years of human experience into some sort of program is well beyond OSE's abilities at the moment.

I agree totally with Allen15, manual first.  The computer that controls a CNC machine does not know how to machine.  YOU must know how to machine first.  It's just good at going fast and repeating your commands over and over.

The Dawg

@Gregor

LOL...  I must have missed any episodes about prototyping a nickel-iron battery with concrete...  Scary, but maybe entertaining?

I do have some ideas about water treatment that I'd love to get your feedback on, as a sanity check, if you're willing, though, in the OSRLiving.org group, currently under Aquaponics.

 

gregor

How does your truck factory work, and why is it not possible to make it smaller?

Basically I have tried to put all equipment in it that you need to build trucks. Built it mostly on your own, not buy most of the components and just assembling them. Two important areas are still missing, but they are better performed in an auxiliary building: paint jobs and heat treatment.

I designed that factory such that the material flow works from the right side (raw material) to the left until a completed truck leaves the hangar.

It has space for assembling up to 6 trucks simultaneously - thus I guess, with enough personnel you can complete a truck each day.

Why is it not smaller? Well, gregor, I now made a smaller (83ft long) version for you!

Look at it and you will see how it performs compared to the large version.

This is what I did:

- removed storage for raw materials

- removed 1 bandsaw

- reduced area for welding operations

- removed 1 small and 1 medium lathe

- removed knee type mill

- removed 1 drill press

- removed 3 workbenchs and space for small assembly

- removed most space for final assembly, now just one truck can be assembled

- additional shelfs and storage space removed

- free space between all equipment reduced

Disadvantage of this reduced concept is, that you can't work with many people in that shop. Too tight spaces between machines, not enough space to store larger unfinished parts. Only 1 assembly area. You save half the shop size, but production rate will drop to 1 or 2 trucks per month as you can only work with a few people.

I don't think you can remove much additional equipment; or you will have to buy most of the components. But thats not OSE, is it ?

Mike

Just seen on the wiki, see GVCS Rollout plan

Astonishing, the capabilities of this machine !

CNC integration with tool change capability, works as #1 mill, #2 drill, #3 lathe, #4 surface grinder, #5 centerless grinder, #6 nut&bolt machine, #7 bandsaw, #8 cold cut saw. In the same document, but other place, it also serves as #9 ball-bearing grinder.

Thats the solution, one machine replaces nine

I regret very much, that I personally don't  believe in such a miracle machine. But maybe someone in this forum can convince me that this 9-in-1-machine is more than a useless piece of metal, going soon to the scrapyard.

Mike

Ashes to ashes, dust to dust, & scraps to the scrap yard...  From whence it came, so shall it go forth, but not until it hath toiled for mankind yet again :)

While it is true that this machine will be a challenge to automate with CNC, that doesn't automatically relegate it to the scrap heaps (where most of it came from in the first place).  One has to have tools to make better tools, and this is a pretty good start to making better tools, if nothing else.

Do you perhaps have a better alternative to suggest?  Instead of us trying to convince you that the multi-machine is more than useless, how about you convincing us that an alternative is more feasible?  (Like providing the alternatives or pointing the way to them?)...

We'd all love to hear better options, but without them, right now, this is the best thing going that can get anywhere close to doing what we'd like to do with it.

(Remember, feasibility also involves replication in low-tech environments too...)

Your turn :)

Well Just a word of warning Mike: don't think of this as "we". I stopped saying we when it became clear that is not how it works at OSE.  It's Marcin doing what he wants and you can fetch coffee if you want.  Don't expect anything you say to change anything on the project.  I only regard the forums as an interesting discussion, and it might be useful to others in the future.

Anyway, about the machine, having looked at that precision machine tool design book, and having watched a bunch of instructional videos from MIT recently on machining....  IIRC there is currently no stated explicit goals regarding many parameters like the size of the workpeice it can handle which means trying to talk seriously about it is sort of like trying to shoot a fly.  But if you are willing to make compromises on some things like size and speed, it does look like you should be able to get the 1 thou precision they say they want with good surface finish, probably more, and the roundness OK too.  The thing is stiffness and vibration damping (at all suitable frequencies) are stiffness and damping.  The rotor speeds are a productivity issue, it appears to be practical to use the same RPM range for the spindles of the lathe and mill if you are willing to compromise on speed.

For bearings basically you would be just reusing the motor and nothing else really so there may be little point, but if you can accomodate the possibility of the functionality without causing problems elsewhere maybe it's worth it so you could get by in a pinch or use it for bootstrapping.

For the actual machine tool it might be a good idea to look at hydrostatic bearings.  I saw an interesting doc on an ultra precision lathe that used hydrodynamic bearing, was made of ceramic, and required relatively low precision itsself to make.  They provide an inherent degree of smoothing much more than roller bearings so you get a larger step between the precision of the motion, and the stiffness, you get from the bearing, and the precision needed to make the bearing, and can nearly eliminate backlash when used on the leadscrew. 
 
Although others point out that manual would be a better start, actually the idea of leapfrogging to CNC might not be so bad IMHO.  The thing is the design of the machine can be changed to give better stiffness with less material because the operator does not need to be able to reach in there as much, and secondly it opens up the possibility of a scanner to compensate for inaccurate fixturing of the peice.  Instead of spending all that time to get the workpeice aligned with the machine tool to say, mill perpendicular faces (or use special holders) you can scan in the peice and simply have the machine do all the alignment work - it knows which plane one face is in, and where all the other features are within the XYZ space in the working volume of the machine.  

Thirdly, as I mentioned in prior ramblings, it allows the possibility of scanning, machining, then  repeating.  I came across a doc a while ago that used a profilometer (highly precise scanning device) along with a very poor machine tool and the authors got excellent precision out of the setup.  If you think about it doing something like pointing a rifle at a distant target with the terribly imprecise actuators that are human muscles is pretty darn amazing, and that can be used here too.  You want a good scanner anyway so that may be a good way to go (and should be added to the functionality list).  CNC also eliminates the transmissions for power/manual feed switching, and also also for thread cutting, as well as some other bits and peices.   Also, you may be able to compensate for very low frequency distortion of the machine issues automatically.
Also, BTW I have seen advertisements for 5 axis milling machines that can supposedly make most parts that a lathe can make and are being advertised as such.  There appears to be a sort of continuum actually, extending from a regular lathe, to lathes with live bits, to 5 axis mills.
 

Hi everyone. I am really glad to be sharing with you. I heard about Open Soucre Ecology on CBC Radio.  I am a canadian from northern ontario. My occupation is machinist. Most of the machinists up here are not certified. Our machinist trade is a open trade in ontario. We have a certification program. But most employers not require it. I did my machine shop basic training at Canadian forces Fleet School Engineering Division in Halifax Nova Scotia in 1981.  I also took machine shop for two years in high school. Ten years ago I took a one year CNC  course at a trades college.  The very first bit of advice I was given was "If you want to become a good marine engineering mechanic you need to first become an expert at removing as many variables that would cause failures and do it in a dynamic fashion ".  I think he ment when the steam driven lub oil pump needs oil you better do it quickly. The first lathe I go paid to operate on was aboard the Hmcs Saguenay. We had a newly minted machinist specialist just back from fleet school. And he was deterimed to teach machining to any one he out ranked. Talk about removing variables that would cause failure. At sea we would make valve spindles and face valve bonnets. The spindles were hard to make because the half nut would disengauge when we hit a waves. We over came this problem by making a master thread spindle and attaching it to the end of the blank to be cut. Attaching a piece of wire to the carriage and used it to follow the rotating grooves in the master and produced the acme thread by manually moving the x and y. After a few months of this we discovered if we just blued the blank and scrached pick line we didn't need to drill and tap the end of the blank to attach the master. So the very first piece of imformation I learned from the very first professional engineer I ever met gave was me the most important  to my  career as a machinist. " Remove as many variables that could cause failure and you will consistinty be successful"  

What I hope to bring to this and hopfully other forums.Is a bit of experience maybe some humour canadianized.

When I did the questionare that allowed me to join this forum I was ask " what I have to offer this project?"

The answer is : I have been a metal butcher for 26 years. I

I started my trades training when I was 17. As a marine engineering mechanic for the canadian Navy at CFFS halifax. I served on the older DDH class Destroyers. They ran on y-100 steam boilerers(2). As a Sailor I got to visit  local machine shops and fabricators in differend parts of the world. I was always amazed that in some parts of the  less modernized world people could find ways of achiving big things with very little. I am a bit of machine tool history buff.

My Occupation. I am a CNC Hybred lathe operator I run and maintain a Boringer DUS 560, I do some leadhand work like set-ups and training in the machine shop. We manufacture minerial sampling equipment. We are a metal fabricating shop so the bulk of our work comes from our fab (welding) shop.

My Creditals: Certified General Machinist (inter-provincal red seal)

 

My Experience: I have been exposed to alot of older machine tools and alot of different shops I have visited and worked in quite a few. I have a small garge based tool and die shop. econo mill, surface grinder, a small manual lathe, heat treating and casting oven. I have built a few machine tools. My lastest build was a metal cutting hack saw from plans I found in Popular Science Feb 1964. In an article by Walter E Burton. The cool part about this build is that the rocker component is two used Piston to Crankshaft Con_Rods from a V-8. I think it might be a good fit for GVCS. I up loaded the article in PDF and the BOM. But I am slowly learning WIKI.

I will pick up a stripped v-8 engine block. The first questions I want to answer is exactly how strong is engine block and how easy is it to machine on. I know when it is fully assemblied it is very strong, but what if you remove the heads and the dome and the oil pan and the crank shaft, I wil search for these answers with a 5 ton bottle jack with a pressure gauge and a hand dril and a couple of dial indicators. Collect engine dimentional data and see if I can destroy a v-8 engine block with a 5 ton jack and if I can destroy it , how much force is needed. Should be a fun test. Might even get some usable data.

I hear my bed calling

Bye