[lgalla]

I agree with Mcgyver or Mcgyver agrees with me.
The modulus of elastic bands or elasticity of E/G is not up to the task on its own.Show me a 4' gantry beam with a 100lbd 10hp router and I will adjust my thinking.
A steel structure damped with E/G and utilization of surface plate technology is the way to go.In simple terms,E/G cannott support its own weight in a gantry or total E/G machine.
Walter has discouvered and made good recomendations on gloves.Epoxy is such a good glue or binder as it penetrates almost any material.
If you want or desire 10% epoxy ratio,which is bread dough consisanty,save your dough and don't buy a vacuum pump.It will "suck"or not suck air out of such a heavy mix.
Due to inflation,this is my $2.00 worth
Larry

[lgalla]

Quote from Mcgyver"I do admire the efforts made but continue to be a doubter on EG as machine material unto itself - if you want the rigidity, better learn to weld boys "
Squaring steel structures with epoxy has been documentated by members such as Madvac,which eliminates welding and stresses.Bolting and epoxy is a good system to eliminate welding and stressing the structure.E/G is used for weldments.How about use it for epoxyments?

[walter]

So many good contributions on this thread. All the comments and ideas are greatly appreciated!

[lgalla]

I have read some of the members testing are trying to break their samples at 24 hours.Various epoxy's can be 2,3to7 days to acheive full cure and max strength at room temperature.It is very boring waiting for cure.Ask your supplier the full cure time.
Slow long pot life is also boring in sample tests but necessary as when doing large batches the pot life will shorten.
The small amount of epoxy between the aggregates may also extend the full cure time.
Patience is necessary.
Larry

[PlasticWorker]

Hi
Not to beat a dead horse, just one thought about your random carbon fiber layer. While carbon fiber has approximately 4.2 x the stiffness of e-glass ( probably less if you compare s-glass) your figure says that the continuous fiber has 5 x the strength of random. So it would seem that a continuous layer of glass is slightly better than random carbon. Since glass is probably one third the cost of carbon, you could use 3x the glass for the same money and have a 3x stronger glass layer. Carbon is great if you really need high strength and light weight at the same time, but in this instance maybe it isn't the answer.
OR
Us Composites ( http://www.shopmaninc.com ) has carbon fiber Uni-Tape 11 0z 13" wide for $ 10.50 / yd. That is a tape where the carbon runs the length with very few glass bindings running across the width to hold it together. Sounds like that would be affordable and stiffer.

Will

[Eson]

I wonder how you guys 'wash' your aggregates.
If you use alcohol it could be the reason why the aggregates crack. I know a guy who was milling in his home built cnc. The holding tool is made out of plexi and when he used ethanol as cooling fluid the plexi cracked at the screw joints.
It could be that the ethanol if you use it to clean, could sneak in the little pores and crack the whole quartz stone.

[ckelloug]

Will,

Your point about continuous glass vs. random carbon is an excellent one!!! I had been generally avoiding suggesting continuous fiber on the grounds that it might be harder to do and none of those who had initiated the thread seemed like they were going that direction. I know Walter had done a few tests with glass cloth in the mix but I'm not sure exactly where in the piece it was or what the effect turned out to be although I know he was having all sorts of bonding problems at the time.

At any rate, putting several layers of continuous glass especially on the bottom of beams and maybe several more on the top for good measure would greatly simplify making a strong enough material to make beams and other items.

Out of sheer academic joy, I'd like to make an E/G piece that didn't need reinforcement by any other material at all since it seems possible and has been done but just so everybody knows, I doubt that the pure E/G with no reinforcement idea is the most practical way of building a machine. Mcgyver was right when he said that learning to weld was a good idea.

The 1/5 figure for strength of continuous vs. chopped fibers came from pg 520 of <u>Materials Science and Engineering: An Introduction</u> by William D. Callister Jr. Copyright 1995 J Wiley and Sons. ISBN 0-471-58128-3.

You're definitely not beating a dead horse, you obviously have knowledge in this field which is a great contribution here whereas I am a generalist with a large pile of books. Carbon may very well not be the "right" answer here but it looked like it might be interesting to try and fun.

Keep those ideas coming, the more the merrier!!!

Eson,

I would have difficulty believing that Alcohol will shatter either the quartz or the epoxy. I do have a couple of references though that water can cause or amplify cracks in glass/quartzlike materials. A chemical engineer, harryn on this thread I believe, posted earlier that washing the aggregate in alcohol and drying in an oven might help remove the tiny layer of water that often clings to aggregate thus improving bonding and getting rid of water vapor that might encourage fracture.

If you have any other thoughts though, please post them as it's a rarity that we actually have the data at hand on a given question like this.

Larry,

You are spot on with your observation about waiting for the epoxy to fully harden. Most of the epoxies used for commercial aerospace composites and laminations are hardened at high temperature like 100-200C for up to eight hours depending on the epoxy used. This both speeds up the cure and also increases crosslinking and thus stength and high temperature strength especially.

Your comment here was both timely and helpful.

Regards to all,

Cameron

[greybeard]

Hi Eson.
I have had a little experience of this, and I believe that the mechanism is one of chemical attack on the acrylic rather than a physical one.
Under stress, many plastic materials are attacked by solvents that in normal circumstances have no effect.
I noticed this many years ago when embedding acrylic fibres as light guides through polyerster casting resins.
Straight fibres are unaffected by the various additives in the resin(mainly styrene monomer) but if I pre-bent a fibre with heat, then embedded it, it promptly dissolved in the area of the bend. I did several experiments with different resin/fibre combinations and the effect was always present if there was any stress in the fibre, locked in by previous treatment, or during the casting process.

John

[Geof]

Greybeard has it correct; it is a stress cracking phenomenon. Acrylic, Perspex, Plexiglas, whatever the name is very susceptible to this, particularly extruded acrylic.

[greybeard]

Geof, I wish you had been around when I was spending hours trying to find out why the light went into one end of the fibre but didn't come out the other !
I'd never come across the phenomenon, and it was a pretty frustrating period.
I did manage to cut 3mm leds down to 1mm cubes to embed in the same project, but the failure of the fibres was the last straw.
The project ? A fan with embedded chips and optics to produce a "firework display" across the surface five seconds after you opened it.
Happy days...!

John

[Mazaholic]

Christ...Now you guys got me to doing some research.
I was thinking about vacuum and it was mentioned that the consistancy of the epoxy might be too thick to vacuum.
I think thats probably right..only solution i could think of is mixing the aggregate and epoxy in a vacuum enviroment,then i ran across this.

Aggregates
High-hardness mineral aggregates including quartz, basalt, and granite are typically used, though recycled glass is another option. Granite, for example, leaves a jagged edge when it is broken up so it better grips the resin. But the jagged edges can also hinder flow into mold features. In contrast, high-strength, high-purity (99.5% SiO2) quartz aggregate has a more round shape that improves flow and compaction. Vibratory compaction during the molding process tightly packs the aggregate together, which boosts part strength. The quartz has a Mohs hardness of 8 (diamond = 10) and makes up about 92% of a part by weight.

Resins
Resin systems include epoxy, polyester, vinyl ester, methacrylate and furan, the most common of which is epoxy. Epoxy has excellent chemical resistance and long-term stability, as well as good mechanical properties. The addition of wetting agents improves resin/aggregate adhesion, while antifoaming agents reduce trapped air. Volumetric shrinkage of epoxy is minimal and additives can nearly eliminate it.

Has anyone tried wetting and defoaming agents or am i on the wrong track?
This thread is so long i'm not sure if i'm behind or adding ideas.

[ckelloug]

Mazaholic,

I personally think you are on exactly the right track with the bonding and deairing agents. I don't think anybody has tried them yet however. I've been advocating for dow z6040 bonding agent and BYK A525 deairing agent for a while now. The datasheets should be linked in that index post a few posts back.

Anbody got any comments on whether they've managed to obtain samples and try the deairing agents?

--Cameron

[Zumba]

Why bother with a couple strips of carbon fiber or fiberglass? Use rebar or steel pipe. Stiffness will be exponentially higher.

[Adam M.]

Finally, made it through. Unfortunately, I am a woodworker, not a chemist. I do have a degree in biology for what it is worth. I have read every post and have enjoyed the friendly banter and even the more tense moments. Anyways...

I have a couple of observations. I should note, I have never used epoxies so my thoughts and ideas may by way out in left field.

1) One problem that keeps rearring its ugly head it insufficiet wetting of larger aggregates. Using Igalla's tip to heat the resin to lower its cps, I thought about putting the larger aggregate in a tumbler and coating them with the thinned epoxy. Put the tumbler under vacuum and it becaomes doubley effective. On a larger pour it may be difficult to warm the entire mixture, so this could be a comprimise.

2) People are talking about thousanths of inches. Some of these same people are talking about using rebar. Since rebar has been prestretched, it is in a constant state of compression. Could that not distort the E/Q (or whatever we are calling it now)? If some kind of artifical support must compliment the E/Q, I really like Cameron's post-cure tensioning. What I like the most is it is adjustable.

3) I am unclear, is it preferable to keep the E/Q under vacuum for the entire duration of cure, maybe in an enclave (everyone has one, right? ), or just long enough to degass? It seems I have heard it both ways.

4) I saw on the Discovery channel, a show where they made carbon fiber masts for sail boats. There were no internal support structures, unless they were proprietary and they didn't show them, and these are able to resist the force of wind in a presumably large sail. What about wrapping an entire gantry with carbon fiber and vacuuming the whole structure?

5) Since we are not ready to pour a gantry 8' long (on day I hope, just not yet), I assume most are focusing their attention on smaller, >36", router. On a smaller router, what really are we talking about in terms of forces? Everyone seems to be scratching their heads trying to get a 36" E/Q beam to support 100lbs of weight. Wouldn't a spindle typically used on a smaller router be half that or less? I could be wrong, and it wouldn't be the first time.

6) It was mentioned on here somewhere wa back there that when you double the height, you increase the elastic modulus by 2^3. Does this apply to E/Q? If so, would a triangle shaped Xsection be stonger then a square Xsection with the same area? If so, you could trim alot of weight by focusing on building a higher gantry beam, triangular shaped, as opposed to a square/rectangle.

7) I am suprised I havnen't heard more about pouring in layers, degassing each layer as you go.

That's it I think. I need time to digest this and do some of my own research on epoxies and their uses. I think if I had a better ideaof what some other uses of expoies are, maybe i will get some ideas. I do think, or rather know, this is plausible. There are definelty ingreients missing, but I don't think they will come without testable experiments. Someone correct me if I am wrong, but wasn't the combustion engine invented about a century before Henry Ford was even born? For that matter the steam engine was invented centries before the first railroad tracks were ever laid. It took people working upon other people's successes and failures to get us to where we are today. And people coming onto this thread saying this will not work while offering no constuctive critisism is a complete waste of everone's time.

Great thread. I am excited about the possibilities and would like to try some of y own exeriments, but need to some moer homework.

I'll buy a beer to who ever can count the most spelling errors. If you could see the position I am in typing, you would understand. We are moving and have no furniture.

[ckelloug]

Zumba,

I think you're right that metal reinforcements are an easier route than pure composite and I've said that to anybody who's specifically asked for advice. Since there seem to be some folks on this thread that would like to build a machine without having to cut any metal, I figured I'd like to find out what it takes to make structural parts with no metallic reinforcement. Once the material engineering is done, I don't believe making the E/G parts that strong will be any harder than making weak ones although the E/G parts might need to be bigger.

I suppose I'm climbing that mountain because it's there rather than because it needs to be climbed but I still believe it's reasonably possible to make an entire machine without either glass, carbon, or metal except for things like dovetails and wear surfaces. It's good brain exercise. I'm hoping I remain as clever as greybeard is when I get another 30 or 40 years under my belt.

I prefer to think of myself as eccentric rather than crazy. . .

--Cameron

[Mazaholic]

ckelloug,
Thanks for the reply,i just wanted to be sure i was up to speed.
This stuff has really sparked my interest.I can think of many uses besides making a machine structure.
I'm going to start a few experments.

[greybeard]

Hi Adam, and welcome to the thread. I can see all your posts, so I hope you've sorted out whatever the problem was.
Hope the move goes smoothly. Last did it 5 years ago, and we still don't want ever to do it again.
Mind you it was the first time in 40 years.

I, too, am more concerned with wood and a much lighter weight router. I confess that I've just bought a bottom of the range one, 1020 watts for £25, and on that basis it is virtually a consumable item. It only weighs about 10lbs, so the cross-sectional sizes mentioned are way over what I may need.
I, too, considered the idea of pre-coating the aggregate as a way of making sure that all surfaces are covered, so I'm with you on that one.

Regards,

John

[Adam M.]

I just can't seem to stay away. My mind is racing 100 mph on this. There is alot to absorb in a just a few days of reading. I really do need to get to work. I have bills to pay.

John,

Don't get me wrong. I am one of the big dreamer guys. I want huge router. (I process alot of panels.) My main concern is actually the software, but that is a different thread altogether. I plan on using E/Q for the base and table. If we can get this figured out, I would gladly us it for the gantry, but we have a ways to go before we get there I think.

Cameron,

The reason Silane appears to be a superior coupling agent is because it contains amines. These react with the epoxy resin through condensation polymerization, similar to how they make Kevlar. According to the data sheet for Silane you posted, these bond to the resin molecules turning them organoreactive. This is how I interpret it: resin forms polymer chains, Silane bonds to the appropriate binding sites on the resins and, in turn, bind to each other, producing a very strong polymer. Carbon Black, however, seems to interfere with the polymerization process. I am not sure how, there isn’t a lot available on it that I can find. But if it does interfere, either weakening the epoxy bonding process or causing the polymer chains to be shorter, this would conclude why Walter’s sample bent so easily. This is also consistent with what little I have been able to find about Silane and Carbon Black. Carbon Black is added to rubber to increase tensile strength; i.e. makes it stretch. Silane is added to lower rolling loss on tires. (This is courteousy of Wilkepedia) I assume that to mean that it makes the tires more resistant to compression, hence the stay rounder and roll more efficiently. Hence, we are talking about two totally different animals. Carbon Black makes the E/Q more pliable, Silane, harder and more brittle. Perhaps the ideal E/Q has some combination of both.

My assumptions could be way off. Without knowing the exact formulas for Carbon Black and the epoxies in question, I don’t know how much further one could take this. I believe Igalla said there was something like 13 hundred different forms, or maybe it was just 13 different forms. I think it will have to be trial and error. In the paper you posted, I seem to remember it saying they used 0.2% Silane to dramatically increase strength. I think it also said that much more would cause an oily pooling of some sort. I would have to go back and reread it.

In my Wilkepdia research on epoxies, I came across this. Might be worth something to someone:
Epoxies will not stick to mold-release compounds recommended for use with epoxy, and polyethylene sheeting, like disposable paint tarps and sandwich bags. Epoxy does not stick to the shiny side of packaging tape or paraffin wax

[ckelloug]

Adam,

Here's a long response to your post questions.

Rebar and reinforcement:

Rebar may be expanded but that means that it was plastically deformed in order to make the modulus higher although it does mean that the ultimate energy absorption to failure goes down. Incorporating rebar into the E/G is a fine way of reinforcing it: that how bridges and buildings are built. Just inserting rebar that was not being tensioned by some external force when the concrete set will not add force to the system. Post tensioning on the other hand has you create a channel through the material through which tension can be applied by metal rods and screws. This allows you to transfer some of the deflection to the internal metal which would otherwise have been in your beam. It also allows you to ensure that materials with low tensile strength like E/G are maintained in compression where they are strongest. If done correctly, rebar would likely make a beam cheaper to produce and perhaps easier to produce too as a smaller cross section could be used. Post tensioning is nice as I have said because it allows you to get an E/G beam completely true. If you route your machine surface or spoil board to the beam and the beam is warped then the spoil board is potato chip shaped. Incidentally, I was not the one who brought post-tensioning to the thread, incidentally, poster the4thseal did it back at post 230! I've just explained and popularized it.



Wetting:

One the topic of wetting, I believe that it's actually the small aggregates that cause the most problem. I have enough theory research to contend that large aggregates are more of a problem with getting high strength than a solution anyway due to the fact that large aggregates have a low fracture toughness and thus break very easily. I think concensus is moving away from the larger aggregates for this reason so they should not be an issue. The CNCZ1 formula that I posted with all small aggregate used the aggregate distribution from an epoxy manufacturer's recipe for Acrylic cement. Vacuum, heat and stirring will undoubtedly solve the wetting problem in some combination. Nobody has tried vacuum yet however so I'll be getting a pump as soon as things quiet down a bit more here.

Elastic Modulus:

Elastic modulus is an intrinsic property of the material itself, not the size of the piece. The deflection under load however is proportional to the height and width of the piece. For a beam sitting on supports, loaded in the middle, the equation is Delta=FL^3/48*E*I where Delta is the deflection that the center moves down under load, F is the applied force, L is the length of the Beam, E is the elastic Modulus and I is the moment of Inertia. I for a rectangular beam is bh^3/12 where b is the width and h is the height. It's easiest to look I up in an engineering handbook but it can be computed with some simple integration for bonus points.

Load:

The 100 lb number was given to me by Larry as the amount of weight you would ever want to put out in the middle of the beam and I have continued to use this value as it's ensures a conservative design. So far, the sample properties have been lackluster and it thus hasn't mattered cause nothing could be built to the calculation strength anyway. A beam length of 36 inches seems to be the largest practical machine one can make at home out of E/G since the 12x4x36 beam solid tends to use about 1CF of material which is about all you can mix without bigger equipment. The last calculation I remember with the 100 lb beam load at 36 inches was at a safety factor of about 6. It turns out to keep the static deflection down to about .0002 inches, the high cross section is needed and it just happens to make it a lot stronger than a failure criterion would dictate. Frankly, PlasticWorker pointed out the problems with loads in other directions besides Z that nobody here has addressed yet and as a result, I think there's a lot more work to do on the machine design aspect independent of the E/G aspect.

<B>Has anybody weighed these behemoth motors they want to hang off the beam?</B>

Small layers:

I think that it's generally accepted practice among professionals to pour the epoxy in small layers. It has to be done before the epoxy has gone too far towards fully cured however or you end up with mechanical but no chemical bonding between the layers. At least one of the epoxy manufacturers best practices sheets suggests this. Nobody on the thread except Larry has a vacuum pump however.

Time under vacuum:

All of the epoxy manufacturers best practices I could find are in accord that the smallest time under the highest vacuum is best. The low pressure over too much time can cause some components of the mixture to boil off drastically and deleteriously affecting the chemistry. The biggest dispute is how much vacuum is needed. It's contested on the thread, but if you're going to buy a surplus pump, I'd personally recommend looking for one with 29.75 inches and as many CFM as you can afford. It looks like the limit for this is about 29.75 inches at 15 CFM in surplus pumps. It looks like much below 28 inches of vacuum is irrelevant as it is unlikely to be helpful and much more than 29.75, is complete and total overkill.

Silane chemistry:

I hate to say it but I think you misinterpreted the data sheet for the Dow Z6040 and gotten the chemistry a bit wrong. This compound doesn't contain amines, but some of the other formulations do. This can be observed by checking the structural formula printed on the data sheet. This particular silane compund contains an epoxy group and the silane group. The silane side tends to stick to the glasslike material while the epoxy ring tends to bond with the epoxy in the matrix, chemically bonding the aggregate to the matrix. The amounts of silane to use that I posted in my trial formula were directly from the dow data sheet: it looks like a little goes a long way.

Carbon Black:

Carbon black is nanoscale particles of relatively pure carbon. They are so small that they interfere with dislocations in the crystal lattice of the matrix binding the aggregate. Pinning dislocations generally causes a material to be stiffer and may cause it to be a bit more brittle. The chemical effects of the carbon black in the epoxy are currently not available as openly published info. I talked to an apps engineer at Cabot Chemical and he thought we ought to see some benefit from it. Any type of tiny particles do tend to greatly slow epoxy curing so it's currently unclear whether it was a chemical disruption of the process like an adsorption of a chemical component or a purely physical effect associated with slowing the hardening reaction. It's also possible hypothetically that it accentuated a flawed epoxy mixing practice and had nothing to do with the problem in and of itself. At any rate, dispersion hardening is a cutting edge way of improving materials and if carbon black doesn't work, Cabot silyated silica fume or sol-gel silica imbued epoxy additive from Nanoresins should work without bollixing the cure.

As for the wood router guys, I agree you probably don't need to hold tenthousandths of an inch in any direction. I've set my personal target fairly high as my main uses for E/G aren't in routers. At least for material strength, I think that the research to get a high strength is harder than just mixing $hit together but I really think that in the end, it will be just as easy to publish a recipe for very strong E/G and I doubt the materials will cost any more. It's also true that the better the material design, the higher the partial safety factor for poor material manufacture. I definitely like safety factors when I hear Larry (lgalla) talking about 10Hp routers on gigantic beams.

As for the ebb and flow of the thread, it seems to vary as people get more or less busy. I'm back to work form 2 months of sabbatical so I can't spend days researching things like I have in the past.

Regards to All,

Cameron

[Eson]

Epoxy doesnt stick to cured epoxy either so if you are going to mold it in layer you have to pour it just before it cures.

Meny pages back someone reported from that german forum Peters CNC-Ecke. You can see that everyone that use E/G have a steel frame as reinforcement.
You must register to see pictures. I can post more pictures here if you want.
Some very good topics:
http://5128.rapidforum.com/topic=110773130998
http://5128.rapidforum.com/topic=110772920369
http://5128.rapidforum.com/topic=110773689090
http://5128.rapidforum.com/topic=110773130998