[jsage]

Greetings Folks,

I have put in my time reviewing different projects, resources on this fine site. Decided if I am going to go through the process I might as well share my success and or failure with anyone interested.

I just finished an interesting pdf posted on the site. A deterministic process for evaluating design option towards a specific outcome. It's an MIT Doctoral thesis that seemed to fall under philosophy. I guess at MIT the technical component is assumed. Of value I think it provided a good structure/methodology for developing a CNC machine. Kudos to the poster and the writer. Coverage of beams, solids, tubing, finite element analysis, pugh charts, damping were interesting.

Thanks to Joe and Mr. Steele for sharing their projects.

One thing I notice that there is an expectation by some to have everything figured out in the beginning. I find that problematic since in many design cycles I have seen The process is iterative and a change in one feature changes the others.

Anyhow enough intro.

The goal:

Universal Platform- the capability to work with a variety materials and project sizes within reason. Expandability within reason, additional axes, fabrication methods.

Honest 8' x 4 x 2?

base welded steel tubing (on hand) concrete dampened, torsional box upper frame with basin, polymeric concrete work surface.

ball screw, linear slides (fabricated where possible) determine most affordable compromise

capabilities: standard woodworking, rough lumber milling, green concrete, aluminum, mild steel, FRP Sectional mold prototyping.

Target Initial investment for base platform $2,000

[Madclicker]

You lost me at $2000. That's an unrealistic number for the machine you described.

[jsage]

I can see your point. To some degree it is based on materials on hand. A better way might be saying my incremental investment. I'd say I have 90% of the structural components which leaves ball screws, and drive systems. I'm looking to fabricate the initial linear slides in aluminum.

It may be unrealistic but it is a starting point. I probably have a couple hundred feet of steel pipe (schedule 40 type) mig welder 20 inch band saw aluminum capable. 2 1/2 hp router. Pentium desktop system (using laptops).

So the goal is use materials at hand with carefully chosen bargains.

The X will be driven by a single screw below the working area in its own torsion box.

[jsage]

The major goal here is rigidity.

Nixed: The four pipe construction is intended to approximate the mechanical characteristics of a 6 inch square tube.

Nixed: The four slide design on the long axis is designed to give three point support and allow stability for a long z axis.

By committee ; ) it was decided that the pipe method was two challenging to maintain true and square dimension, beside being a lot of work. The value of the lower slides and ballscrew placement was similarly thrown out for the conventional dual screw method.

First images include revision to 4" frame 2" gantry. 4 inch gantry material looks really odd.

Obviously with the different design I need to consider ballscrew placement and if I want to shield the rail, ballscrew. Since these haven't been pinned down I focused on a first pass design. On my own review it needed some additional span support.

Did the concrete and rebar reinforcement experiment last night, giving it some time to cure to measure improvement using the pretzel pipe.

Bought a chopsaw, Taskforce ain't bad for the money, smooth movement, good clamping, ok angle setting, I can make it bog. The cuts are far better and feel safer than using cutoff wheel on grinder.

Miss the pipe barge a little but not too much.

[dertsap]

wow ,you have some serious cutting and welding ahead of you

[jsage]

Yep,

Was thinking the same . I'm going to try some 1/10th scale for feasibility to optimize and minimize the complexity figure out how to incorporate missing structural detail. Jigging, and a tool I recollect for mating round tubing might make it tolerable. It seems it is always hard to get away from the chicken and egg problem.

[jsage]

Today I'll do a little experiment to what the deflection rate is for the 1" steel pipe I have available and see what magnitude of stiffening etc I get from the design, addition of internal concrete dampening. Should be interesting. Still looking for post on pdf mentioned. Will put under resources.

[jsage]

Resources used in this project

Rapid Machine Design: Key topics: Deterministic Design process, Finite Element Analysis (FEA), dampening http://www.mech.utah.edu/~bamberg/re...e%20Design.pdf (200 pages worth reading)

http://oneoceankayaks.com/madvac/madvac_index.htm High Precision 4X8 Metal capable
https://www.onsrud.com/home Specialty End Mills
http://www.cnczone.com/forums/showthread.php?t=8813 Ball Screws

http://www.rockfordballscrew.com/bss.htm Ball Screw Selection

[jsage]

:withstupi Steel tubing:

Did some very basic analysis on the characteristics of the steel tubing. Outside diameter approx 1 3/8" wall thickness 1/8". Approx 1.5 pounds per foot.

Method: spanned a 6ft section over two uprights with two commercial buckets set 16 inches apart with a home made plumb. Measurements done with cheap mechanical ruler. Measured filling with water and removing water, consistent in both directions. Some increments interpolated since the process was a bit tedious.

Did a quick measure with 1/2 the span and the deflection decreased by more than 1/2 (approximately)

Next step add concrete then rebar. Then test actual design. Will probably do harmonic test and dynamic test if possible.

Conclusions were that an unmodified pipe with large span did experience significant flex under higher static loads. At approximate 100 pounds the 6 ft length flexed .186 inches. So I'm more comfortable that I am not overbuilding unecessarily.

Had a small epiphany today on how to measure deflection in an easier fashion. As Dr. Evil would say "Laser". I have a "mini laser" from an air rifle. set perpendicular to pipe and center. Parallel might be more interesting. Mark initial level then just add water and mark the movement. Might add a glued on flat marking piece to simplify measurements.

Never found that damn "laser" but I did find a local steel supplier who has what I'm looking for at the right price. The steel composition and characteristics need to be figured out but this what I found a the Metal Mart.

Damn have my quote for roofing but not the steel prices. It's primed and inside at least it looks cold rolled.

Anyhow 4" x 4" by 24' 11 gauge ran $76

The base works out to $234 with tax
rebar and concrete will be extra

$40 for rebar
$40 for concrete (horizontal structures)

So I expect when it's said and done approximately $500

[jsage]

Below is a first pass spreadsheet which should allow you to compare desired features versus expected version capabilities
2nd pass

[faststeve]

Use one large steel tube instead of that small tube welded truss. Don't you realise steel bends where you weld it !

That thing would be bent and twisted like a pretzel.

[mxtras]

I, personally, can not imagine welding that thing up and having it resemble anything other than a pretzel - especially with many of the tubes having welding performed only on one side down the length. The structure will resemble a bannana unless you have serious fabrication experience. Just my impression.

Scott

[jsage]

Appreciate the feedback. I am aware of the deformation. Honestly I haven't done much work with steel in a long time. Aluminum is much easier to cut, bit harder to weld. Haven't found a reasonable tig, but for this it would just take longer.

I know someone who used square stock and the cost based on his comments was probably around 2k. In the end it showed some signs of lacking rigidity. I'm challenged on a number of fronts but am evaluating some methods which will overcome some of the sows ears and chicken and the egg issues. Yet to be proven but there is reason the steel will be less likely to deform, less likely to need to be destressed. I'll have to assemble a portion to see how much cross member support will be needed.

The pipe is actually drawn as 1". The pipe I have laying around is 1 and 3/8ths 1/8th thickness. Looked like 1 inch to me. It's free and I have some larger material although I believe I partially buried it with front end loader.

In the PDF under resources I recollect the concrete formulation they used expanded. The rebar should have some insulation from the welding and provide additional rigidity. (added excerpt). for pound dollar for dollar round stock was better than square in terms of handling torsional forces and overall. However, it acknowledges as I will square is more practical for mounting additional hardware and of course fabrication.

I can rent a plasma cutter for $60 a day and purchase a chop saw. I'll finish up with the sand blasting at $160.

Been finishing an Acre of pond and considering my options. Making the workspace a better place, my girlfriend calls it a garage, is another project. Been a homeowner for year, had to quit the dayjob to take a vacation and to get some work done around the house.





:cheers:

[JerryFlyGuy]

Well, I might as well jump in here

I'm just finishing up my mill, it's overall dim's are 20 x 10 x ~5 feet.

Reading the first bit of your thread several things come to mind.
1) If your serious about producing a high end router, the money you've set out is about 10% of what it will really take to do the job. Your linear rails are going to cost you ~ $3000, that is unless you plan to buy little pc's off ebay, in that case.. it'll be a real struggle to get any precision.

I realize you have lots of mat'l already in hand, but I question if its the correct mat'l. Someone mentioned larger tubing, I would give some SERIOUS thought to that point.

2) It would appear that you've taken a simple fixed table design and extrapolated it for size. I'd re-look at that design, having a center drive for that large of a gantry is just not going to work, it will be one racking headache after another. Dual drives from either end of the gantry is the ONLY way to make that work decently.

3) Check out the 'Polymer Concrete' thread, there is some very interesting dicussion on how to build larger structures which are flat and will give planer rail mount surfaces. If/when I build my next machine there will be some changes on how I approach my build/design, based on this info.

4) The less welding you do, the better. If you came make use of large structural sections instead of little pc's [ like 1" dia pipe] do so, cutting and fitting that 1" pipe is going to drive you crazy and as was mentioned.. you'll end up w/ a nice large curvy pc of art, not a router table. If you've ever bellmouthed and fit pipe, realize that doing it for a router table is going to be that much worse. [just laying out the cuts is a huge PAIN]

5) Buy the tools you need, renting the plasma is fine, but get decent machinist tools for layout and what not. Precision level/type tools are a must.
[again budget several thousand dollars]

I was where your currently at about 1.5 yrs ago and have learned much since then [mainly the hard way ]. I was a little more conservative on my cost estimates and came out w/ ~$10k. Later I changed up to linear rails and really started looking over how well my idea would work and realized that alot more had to be spent to get a decent machine. So far I'm WELL over three times my initial estimate. Can it be done cheaper, sure but not a whole lot. If you want 'high end' then there is a cost that goes along w/ it.

Keep at it, and realize that there is ALOT of help for offer on the forum [ and no I'm not talking about my presence here ] there's lots of great people on here!

Ok, I'll get off my soap box now
Jerry

[jsage]

LOL,

I've already looked at some bigger materials. After bellmouthing 1 piece of pipe with very primitive means I was reconsidering a number of things. Free in this case means free to do a lot of work to make it work.

On the single ballscrew, wouldn't the double lower rail help to keep things in alignment. The racking concern was why I put them there. Not sure about the interaction of span and screw diameter more sag more problems.

On the bigger materials I probably went too far in another direction and saw some thinner gauge large diameter pipe which could get most of the strength from the rebar and concrete but decided the guage was too small.

Yes, from reading a lot of posts I got the feeling that improvising to save money many times ended up creating headaches. So yes I did try to discern what worked or what would work. The higher end was not meant to be a put off only that I would go with higher end linear systems.

P.S. the post in the linear section wasn't what I originally wrote, seemed a reasonable substitution. Had a crazy idea that on reflection had some serious merits. Last time I had a more reasonable idea on a web site I saw a major manufacturer come out with it a year and half later. Probably not feasible but if it was it would be worth checking into.

When I bought my two phase welder I bought a book, the most common one. Tacking and filling in small sections and would guess transverse sequences would be how I would address that issue.

[skippy]

I would have thought that being in the US it would be very easy to buy second hand steel beams very cheaply (U channel or H section, can't remember the English names anymore, RSJ?). If weight isn't a problem then I think that would be a better route. I have made quite a few race car chassis using notched tubing and Tig and if you are set on continuing with the tubing you will gain considerable strength by adding triangulation. you only need thin wall small diameter tubing for the triangulation, 1/2" or 3/4" is enough.

[JerryFlyGuy]

Good to hear, I'd stay away from round pipe completely if possible. But then, thats just me.

The bottom rails are just a waste of time and money, because they are located in the same plane [along the y axis] and are not staggered down the length of the x axis [at least not by a noticable amount] they are just wasted use of rails. Judging [scaling] by the size of your rails, they're good for several thousand pounds, the two up top on the x and the two on the y are more than enough. The rails rating will FAR exceed your strutural rating [FAR meaning the rating is over the horizon in comparision]

If you have to drive the gantry from the center, the only way to make it work w/ and reasonable expectation for success, you'll have to space the trucks/carriages on the x axis by at least 1/4 of the length of the Y axis, this will still [probably] not be enough but will help. If you want to illustrate this, take a ruler and place it on your desk/table. Its pretty easy to move it w/ your finger in the middle pushing across [ 90degrees to the length of the ruler], now put a tape measure [or something of comparible size/weight] in the way of the ruler, try and push it, when it hits the tape measure, all it does is twist outta the way, watch as its twists, see how little the center of the ruler 'moves' compared to the outside edge, this is the exact same way the gantry will react when you start using a tool in its spindle. The tool force is going to twist the gantry outta line, the center rails aren't going to help one teeny bit. Do the same experiment but grab the ends of the ruler and only allow them to move the same amount, hits the tape and doesn't twist outta the way.[there I just saved ya buying 33.3% more rails than you need ]


Driving the gantry from both ends is really a non-event. The control of the motors is taken care of by the motion control software, the only part that is a negative is the extra motor, wiring and ballscrew or whatever your using for a drive mechanism. However, on the upside, you'll still have hair after you get it working..[assuming you started this project w/ some ] If you take the extra rails off, you can more than compensate for the added cost of a motor and drive.

Your right about improvising to save bucks, its all about what your time is really worth, AND how badly you want your machine to work correctly. If you build it wrong today and have to fix it tomorrow anyway, it's gonna cost alot more to fix.

This reminds me of the 10x rule.
If your supplier makes a part it wrong & costs him $1 fix. If the part gets to you and you find its wrong [on delivery] and have to send it back it costs $10. If you get it welded into the structure and ready for delivery and then find its wrong, it's $100. If your customer gets it and finds it its $1000 [all from a $1 part] Some will argue its the 3x rule, whatever ... the point still stands.

I missed the post in the Linear section but will go take a look @ it.

Sometimes people have great idea's and lots of people don't think it will work, this is true. However I'd sure hate to hear someone spending lots of money on something, only to find it doesn't work. I've done this myself, but then.. I thought I was smarter than the other guy.. until he proved it wouldn't work.. all he had to do was let me try

Welders are so much fun eh? I've taught myself to weld.. I'm not no expert/papered welder, but may stuff doesn't fall apart when hit w/ a hammer [anymore..] I trust its a mig?

Also, realize that building this machine is going to frustrate the heck outta ya, [as well people offering their advice ] but if you keep at it, you'll be amazed at what you can do, even w/ little to know experiance building stuff of this nature!

Jerry

[jsage]

Skippy,

On the reclaimed metal, good idea I'm going to check it out My uncle works or worked for pretty big salvage company here in town. Used to get some copper and other stuff. I'm sure they do steel. Just have to be straight and relatively good shape. If that doesn't work I think there some abandoned rail road track about a mile from here (%

The cheapest steel runs around $2.50 per pound per the metal supply here locally. Taking that route I would need to figure out how to maximize the investment.

Jerry,

Thanks. need to find a ruler and absorb what you said . I'll send you the concept, then you can laugh, and I'll move on to something more productive.

[JerryFlyGuy]

Sage, I'd think you could get Steel cheaper than that. I bought all my Steel for less than $1/# here in Canada, most of mine was HSS which is going to be the most expensive. I've got lots here that I'd sell for $2/lb

Btw, don't take my posts as crusty or poking fun, I'm only posting to help. The ruler illustration is pretty self explanitory but it breaks things down to the basic entities

I'd love ta see your latest concepts!

I can honestly say that I wouldn't be where I am [on my project] today w/out the help of the people here on the forum. [ I'd probably have given up on the whole idea and be sulkin ]

My applogies if I came across as talkin' down or anything..to me there is nothing more gratifing than helping others succeed. I can honestly say there is nothing more I'd like to see than someone like yourself taking the 'CNC bull ' by the 'proverbial horns' and winning in the end, on the first try as well!

Jerry

[jsage]

Skippy,

Not giving up the small tubing in general but in this case going big is a lot more expeditious. I wanted to add some triangulation or transverse? stiffeners but felt I had already painted myself in the corner.

Still would like to find a good TIG but for now I'll sharpen my MIG skills. I dropped about $600 on a lincoln. Told it wouldn't work well for aluminum, took some time and more than few burn throughs and now have functional weld though not as pretty or precise as TIG.

Building a monocoque race type frame would be more than fun or something baja type.

Well as I told Jerry it was good to be talked down from the pipe. I can start fresh with a more fundamental approach that doesn't revolve around what's laying around. Get the work area in order.

Thanks.