[romihs]

Hi Roach,

I am hoping to avoid post grinding the concrete. I don't have a clue where to have it done here....

create two lands where your rails go then it will ne easier to grind sandi
r you dowelling your rails or setting them upif dowell the in you can put a ground plate into your mould which is ground with dow holes in
hope it helps

I am not familiar with the term dowelling, if this refers to bolting the the rails down, then yes, I plan to do so. I am thinking of casting threaded inserts into the concrete.
Initially, I was thinking of casting an insert, the length of the rail, into the concrete, but I have concern with this as the epoxy I intend to us needs a post cure at 80C for ~12 hours. I am worried that this will permanently distort the base due to the different expansion coefficients. Using small inserts should hopefully eliminate this problem.

Thanks for the help.
Do you have any pictures of parts you have built, or their molds?

Regards

Sandi

[roach]

dowells to pin the rails in the right place so no set up
how big r your rails or are you not machining much off
bases let me see http://www.sterlingint.com/nanoform250.php
Nanoform 250
http://www.sterlingint.com/optoform50.php
Optoform 50

[roach]

no curing raises the glass transition temperature to increase the stiffness of the casting but if you want damping you dont need to just let the casting cure cool slowy
if any of you really want to damp your infills try cleaned lead shot and resin
u need to make sure your fabs are clean inside otherwise the resin wont stick
just using sand will raise your resin %
if u imagine each grain needs to be coated it needs more to coat each grain than a 14mm stone 7% is really needed
hitting an infill that isnt clean will get a better result than an empty one but if it hasnt bonded then u will get resonance fronm the skin
ideally you need to shot blast your tubesand stress releive thm if your realy serious we make test beds for engines which stops resonance

[roach]

by the way i use metric so using imperial i struggle lol

[romihs]

Igalla,

That is going to one huge table. :-)

I think that if you built the legs with those dimensions, they should carry the weight of the table with ease.

Building the table will be a tough one... Are you thinking of using EQ for it as well?
I have some ideas about building it with EQ... I would suggest that it not be solid, but rather constructed so that you have a table top between 4" and 6" thick supported by a box structure of 1' to 1,5' square, about 8" high. Since the bottom surfaces of the box structures are in tension, I would consider casting some steel reinforcing close to the bottom of the box walls. The walls would be about 4" thick.

The process would be complex though (the mold will take a lot of time to assemble), you would probably have to cast the table in place on top of the legs.
The beauty of this method of construction is that the top surface of the EQ can be leveled of quite well before it sets, and then to finish the top surface off, you would pour a thin layer of TT epoxy. The end result is a table built solidly, which probably will not sag much at all and an almost perfectly fat surface.

Well, that is my 10c worth.... considering that I haven't built anything yet...

Best regards

Sandi

[sigma relief]

Igalla

I understand the reasoning for making the legs out of a resin based material, but I have to think that for the $1800 you estimated, you could do better with alternate materials. Your table will have so much mass that any vibrations will be soaked up by it alone. That said, the weight of your table will also require special attention when it comes to creating a leg system so that it is sturdy in response to large impulse forces (ie you bump into it with something big) A set of steel legs could certainly be built for less than $1800 without welding, but it would require lots of pre-planning, cutting, and bolting. Strong structural steel I beams are cheap, but will take longer to fab than a plywood mold for the legs.

Another option could be concrete, or even cinderblocks. Cinderblocks have the perk of being cheap and user friendly. A few hundred bucks will buy enough cinderblocks to form a perimeter wall and internal support walls. I just haven't figured out how to cast the table in place, as I assume you don't have a forklift at your beck and call to lift the finished table onto the legs. You will need a leveling system regardless of what legs you use (something that heavy is bound to creep, even if it is just the flooring below it that does), if you build this into the table, your legs become a lot easier to manage. I am in a position where mobility of the machine is a consideration, I will move before I die/loose ability to work in the shop so I would like to take my gear with me. If you are making your machine at your home, it may well become part of it and be difficult or impossible to move. Using a disposable cinderblock wall around the table reduces the mass of the part you want to move and would simplify rigging as you would only be moving a big flat item rather than a heavier one with legs. While I am sure you have thought of transport/disposal of the table, I am curious to hear your thoughts on the issue.

Sposl

What size lathe are you looking at making? I doubt any of us will have access to a place that can/will surface grind the finished machine. Most local granite counter top shops buy prefinished slabs and cut profiles and edge them. Countertop and headstone shops may be our best bet, but even those will be a long shot. While I am in no position to begin fabricating anything, I keep thinking of using the $100 granite 2 foot by 2 foot B grade surface plates to cast in flat features. I have even thought of casting my own plates off of the "real" plate to save $$$ and get the size I need. With a little ingenuity, you could make a 2 foot by 2 foot replica with a "rough" edge and perhaps with a few bolts and other protrusions, then slide it over and cast another 18 inches or so on the end to make it larger. Like copying anything, you loose accuracy with each pass, but I can't think of a better way to get a long mold +- a few thousandths of an inch over a long length. If you are using linear rails, they will only rest on the high spots and will even themselves out, you just need to be concerned about the total height deviation across the length, not the local surface finish.

Casting our own ways directly from the mold would be the holy grail because it eliminates costly linear hardware, but if you have access to equipment that can make an accurate mold, you can probalby afford the linear gear anyway. Besides, the real perk of casting directly from the tool is long term cost savings. I doubt any of us will make more than 2 (I can guarantee there will be some duds due to improper control of the mixture)

I know machine design is putting the cart before the horse at this stage in the game, but I think the general principles will take a while to work themselves out. The MDF router guys have taken years to develop their lessons learned, now most designs appear to be just diffent takes on the skate bearing rail idea with more or less internal stiffening. Without a proper mixture formula, any build is almost doomed to failure, but unlike the MDF/aluminum extrusion CNC machines, any failures in our machine design will trash a costly casting and much of the mold.

I apologize that I am just talking about this stuff and not building anything, I really don't want to be one of those guys that pops up, fires off a few post and then vanishes. Due to my lack of programming experience and shop space, I am limited to what I do for a living, mechanical design, problem solving, and modeling. I love CAD, but have spent 10 hours a day for the last 2 months in Pro-E, once that calms down, I am hoping to layout some basic molds (that is what we have to design, the finished part is really the easy part to achieve) for an X3 sized mill. I will bother Roach about nominal dimensions and feature sizes he usually sees later when I have more energy to write a detailed list of questions.

John K

[ckelloug]

sigma_relief John,

Don't worry that you haven't built anything. An actual mechanical engineer is an awesome thing to have around here. While I've been researching and modeling the E/G material for a little under a year, I haven't built even a test part yet though I'm getting closer. I've taken the opinion here of why build something that is suboptimal if optimal will cost the same. . .

Regards all,

Cameron

[jhudler]

I've taken the opinion here of why build something that is suboptimal if optimal will cost the same. . .

Ok you stole this quote from somewhere! If you didn't; well said!

[lgalla]

Sandi,John,
Thanks for the comments.Glad to see we are thinking the same way.The original idea was a table top of I beams and E/G and steel base.The top would probably weigh 4,000lbs.I thought about eliminating the E/G in the top and using the E/G in the base.I like the cinder block idea as it is so cheap,I could put the money into the top.I want the top and base separate fore moving.I have a forklift.
Most builders start with the base and table.As a build may take months,I am starting with the Z axis first,then the gantry, then the table so shop space is not taken up for a long time.The Z&Y will use surface plate technology.
Thanks Guys for the comments.
Larry
The drawings are just the concept.There would be more steel cross members.

[jhudler]

Larry, Is that real concrete (portand cement based)?

What precision are you trying to maintain?

Jack

[lgalla]

Sorry Jack,
that is an old drawing.The concrete should read E/G.I am not after extreme accuracy,just chatter free cuts.My table saw is 2500lbs 8hp and so massive there is no vibration.Machines in my shop of lesser weight just don't"cut it"This is why I want a huge mass for the router.
I have parts cut on a large commercial machine and have to sand all the edges.Hopefully I can do better.At a woodworking show the Komo sales men were promoting the fact their machines were stiff and massive resulting in smooth cuts without any secondary sanding.
Larry

[ad_bfl]

Hi Larry

Stiff does make a huge difference in cut finish. I am not processing sheet goods, but am working with highly figured woods.

My first design was not stiff enough and the resulting finish was way to rough, required a lot of post processing to get the finish I need.

My current redesign is going much further to achieve high stiffness. Some testing I have done really pointed out how much that affects surface and edge finish.

A few swipes with a hand plane/scraper and it's good to go.

My current plan is to mount my spindles in a EG mounting "block" at the end of my cantilevered axis (not a gantry router), built out of normalized welded A36 tubing.

It will be interesting to see how much vibration isolation the EG mounted spindle provides, especially as I crank up the feeds and speeds.

Al

Sorry Jack,
that is an old drawing.The concrete should read E/G.I am not after extreme accuracy,just chatter free cuts.My table saw is 2500lbs 8hp and so massive there is no vibration.Machines in my shop of lesser weight just don't"cut it"This is why I want a huge mass for the router.
I have parts cut on a large commercial machine and have to sand all the edges.Hopefully I can do better.At a woodworking show the Komo sales men were promoting the fact their machines were stiff and massive resulting in smooth cuts without any secondary sanding.
Larry

[LN-JET]

Hello all, lurker here.
I have never built a machine, but I do both welding and epoxy composite work. Lately I have been trying to think up something in the X3 size, just welded from 6-10mm sheet steel. Now I think maybe using 2+mm and lots of internal webbing and thread inserts, and then fill everything with E/G. It would be like leaving the mold in place.
Roach writes about cleaning the infills. How would one go about to get maximum bite, and be sure that epoxy and steel stays together for the next 50 years?


regards

Lars

[romihs]

John K,

I am thinking of a lathe (CNCed) about 1 meter long with ~ 160mm swing over the ways (~300 mm maximum workpiece diameter).
I'm thinking of a tool changer on the cross slide, but that I can make once I have a basic lathe running.

I'm busy 'planning' ahead. Right now, I do not have the facilities to build anything. We are in the process of building a house, and I will set up a workshop in the basement. I estimate (hope) that by then a good formula will be reached on this forum.
I plan to make tests (formula, insert design, mold techniques, etc..) of my own first, before making my first serious parts.

Regards to all

Sandi

[brunog]

Larry,
This might sound ridiculous but have you ever thought of making a slanted bed machine?
Seriously for cutting, routing and making holes pockets on large flat panels, I think it's the way to go, however you would definitely need vacuum to hold the pieces in place, but still structurally it just would be simpler in the sense that the would need less support to keep the bed straight you could brace the machine on a building wall. Think of the space that can be saved by not having a 6 foot by 12 foot machine in the middles of the place.

Best regards

Bruno

[lgalla]

Hi Brunog,
Vertical is great for saws like the Home Depot panel saw or you cannot afford the space.A router centralized in the shop actually saves space.A table saw could be eliminated with a router and considering in feed and out feed a saw needs25'X10'.The router can eliminate virtually every machine in a wood shop such as saw,shaper,bandsaw,pin router,drill press,hand routers etc.Usually the router is in the middle as it becomes the central machining point.Sheet goods are loaded on one side,parts come off the end and offcuts are removed from the other side.
When you consider the router can replace so many machines,they are not really that costly.Vertical is great if you need the space and I prefer to have my legs or feet on the ground.
Thanks
Larry

[jhudler]

Roach writes about cleaning the infills. How would one go about to get maximum bite, and be sure that epoxy and steel stays together for the next 50 years?

Lars, A good media blasting or good degreasing would suffice. Media blasting exposes more surface area, thus makes a better bond. Combined with the extremely low shrinkage of epoxy and a great design, gives an almost unlimited lifetime. (geez that sounds like marketing copy)

[jhudler]

Larry, I've had in the back of my mind to recast my table saw top in EG and mount a stationary router or shaper in it as well.

Don't know about replacing all the tools, I'd love to see how a CNC router can handle making a 1 inch radus by 1 1/2 inch thick cut in cherry without blowing up. But given that a CNC can takeoff any amount of material it's programmed to do, I suppose I shouldn't have any worries.

Had a heck of a time building the tops of a sleigh style crib (see picture. the end caps were not ready yet), ended up cutting close with bandsaw and finsihing on drum sander. Kinda sucks when you blowup 8 BF of S/B Cherry, only took me twice! (nuts)

[ger21]

A router centralized in the shop actually saves space.A table saw could be eliminated with a router and considering in feed and out feed a saw needs25'X10'.The router can eliminate virtually every machine in a wood shop such as saw,shaper,bandsaw,pin router,drill press,hand routers etc.

If you're using these tools in a business, trying to replace all those machines with your router will probably put you out of business. Especially a table saw and shaper. Most operations done on a shaper would probably take from 5 to 20 times longer with a router. And you can walk to a table saw and cut your part faster than you could even sit down to program the g-code, let alone figure out how to hold the part down. My unisaw with 50" fence and 4x8 outfeed table will always have a spot in the middle of my garage.

[ckelloug]

Hi all and especially lerman,

I'm attaching the modeling paper I previously posted with the correction to the Ki formula. I've also added a section on measuring beta values.

I'm also attaching a result from my solver finding the minimum S distribution for the case used to draw the graph on page 69 of <U>Concrete Mixture Proportioning: a Scientific Approach</U> along with the raw data values.

The format of agtheo.txt is Beta,size. This is the solver input. The initial condition was 10% of each.


result.txt is the solver result and some data to help see what happened. The format of result.txt is:

<preformat>id,Ki,diameter,volume percent, cumulative volume percent</preformat>

Regards all,

Cameron