[walter]

There's also a lot of very very fine aggregate in the commercial formulations for countertops that means that the surface is mostly aggregate, not epoxy and thus pretty hard.

I was talking about this (unless this is a counter top sample?)

[walter]

Finally, I need to figure out how to get some deairing agent and some bonding agent pronto since these are our problems and they are what the apps engineers recommend.

I can visit Cary Co tomorrow and do some begging :stickpoke

[walter]

Thanks Groger,

That is a good info and I appreciate you joining the discussion!

[qroger]

Thanks Groger,

That is a good info and I appreciate you joining the discussion!

Now you have done it, you encouraged me..Probably the best thing that concrete does is that it is strong in compression. The trade is that it is weak in tension. The tension part is related to the stickiness of the binder, (cement) and the compression part is related to the strength of the aggregate and voids or lack there-of. Not knowing anything about the polymer concrete, or epoxy concrete, or I forget what it is.. The bond between the binder and the aggregate is probably way better than portland cement. So this material has way better strength in tension, and probably just as good in compression. What I am worried about is where the damping properties come from. It might be that it is just that cast iron is crystalline and concrete is not. If so, it could be that fiber reinforced concrete from your local readi-mix plant would work. Just as added info, in concrete they work like crazy to get little bubbles in, (air entrained) to lower the viscosity and it helps to absorb freeze-thaw expansion.

[qroger]

One last post, then it's off to dinner for me. The aggregate plants have been working on the size recipe for ever. They call it the gradation, and it is the specification that they are judged on. I get on a little weaker ground here, but for instance they have a product called "57 stone" that must meet state specifications for concrete. Where they diverge from the counter top guys is that it doesn't have to look pretty.

point 2. If Epoxy is two component stuff, couldn''t you dump the aggregate in with the main part, mix it around, maybe even put it under a vacuum, get it all wet, then add the hardener, mix it up some more, and maybe put it under pressure to squeeze the bubbles back into the mix?

[ckelloug]

Groger,

I believe the main reason that epoxy concrete is a good damper is because it's hard to get the particles to vibrate in unison. Even if you get one going, it's moving out of phase with the next one as the epoxy phase compresses such that the small motions quickly die out. This is caused by the fact that the epoxy is springy but it isn't perfectly springy. It absorbs some motion on each cycle and quickly reduces vibration to 0.

I imagine that portland cement concrete is also a reasonable damper but I believe that the cement fraction has a higher compressive modulus compared to the aggregate and lacks the springy but not perfectly springy characteristic of epoxy.

You are right that E/G polymer concrete does have a much higher tensile strength than regular concrete and is only a bit better in compression. This allows you to make beams out of it unreinforced which don't transmit vibration as much as reinforced beams.

If I interpret it correctly, consensus on the thread is that epoxy concrete is more stable in terms of changing dimensions after being cast which is important on machines that ideally would like to hold .0001 accuracy. It's also more compatible with ferrous metals and better at sticking to all of the things that are cast into a machine. Epoxy concrete also tends to shrink less. I'm sure somebody else will have other reasons.

Thanks for joining us and posting.

[ckelloug]

Groger,

In short, I personally don't know what whether you're suggesting in terms of mixing the aggregate with the resin and then adding the hardener will work or not. The hardener and the resin have to be mixed rather uniformly as does the epoxy fraction and the aggregate fraction. How thoroughly is a point of question which can only be answered by experiment. How-to's of mixing theory is something we need to study more on this thread.

You're right that the concrete guys have been working on aggregate designs for a long time. The basic principle is to get an aggregate fill ratio as high as possible while leaving just enough room for epoxy which apparently needs to form a 30-36 micron layer around each macro particle according to the paper by Gupta. This is a bit different than portland cement concrete where the number is different but I don't have it.

We need good surface finish on our parts and to expose mostly aggregate at the surface rather than the binder/matrix materials as does portland cement concrete when it is trowel finished. We also need to be able to accurately mold much smaller details than those possessed by the average bridge which means that the aggregates we use have to be smaller. Also epoxy concrete does not tend to have good working properties once cast such that it needs to be perfect on the surface when it comes out of the mixer.

It is my impression (which might or might not be confirmed by someone like ahlbebuck the concrete engineer on the thread) that structural concrete is generally engineered so that mortal concrete contractors can mix up a batch and not have the bridge fall. As such, things are designed conservatively to make consistent batches under field conditions rather than the very best batches under lab conditions. Finally, I don't think ordinary concrete is trying to hold tolerances in the ten thousandths of an inch for either design deflections or mold tolerances.

The concrete industry tends to use what has worked and that I believe is traditionally Fuller's Rule, a parabolic curve showing percentages of aggregate sizes.

The NISS paper that was posted by greybeard last week tends to show that statistically speaking, the random distribution of particles in 3D isn't parabolic though it is not a bad second order approximation: http://www.cnczone.com/forums/showpo...postcount=1185

Using the statisticians' paper I've been working on a 9th order fit of the parameters which I believe ought to exhibit better packing ratios in theory. We're also exploring the use of nanoparticle aggregates for dispersion hardening which are like pozzolans in ordinary concrete.

[lgalla]

Anyone having trouble posting?
Anymore than 5 lines and I lockup.

[qroger]

Ckelloug,
I wonder if you aren't asking too much of this material. The exposed aggregate can be provided by grinding it back, like a terrazzo floor. The casting detail can be helped by going to a smaller aggregate size, and a more rigid mold. If it was possible to have it swell just a little as it hardened that would be good. (The swelling is one of the concrete casters tools.) I wonder if the contribution of the vacuum environment isn't a momentary decrease in viscosity. Some things can be gotten by casting inserts into the base material. I guess we are looking for something that will do everything cast iron will do, only better, cheaper, easier.
roger

[harryn]

Hi - Just came in from working in the yard / sprinklers. It is always interesting to look down at the ground. I have a pathway that is made from a stone mix called "quarter x dust" which is roughly its size in inches. It is a highly angular stone mix, which self compacts into a tough walking bed almost by itself, with only nominal tamping.

It seems to me (brave, dumb words, of course) that if you used this as your agr, and just bonded it together with your epoxy, it would be quite tough and stable.

I was able to break up a small section of the walkway and rinse the stone mix with some rubbing alcohol. I let it drain over some sand to keep from changing the mixture. It will take some time to dry out, but that would be my pre-treatment prior to epoxy mixing.

I tend to agree with the poster who thinks of this material as more of a stiffener / damper material than a precision ground surface. I tend to want a system where the bed is stationary and the head is moving, so the mass of the beams needs to be managed.

My current thinking is to
- Start with a modest wall precision square tube for the exterior / wear surface
- Insert a separate square tube inside of it with perhaps 1 inch of gap between the two
- Mold the EG into the gap, leaving a hollow core
- Fill the hollow core with a light weight structural foam

Of course, I am very new at this hole machine building area, so that may be just a crazy idea.

[ckelloug]

Groger,

You obviously know a lot about concrete. You wouldn't happen to be another one of those there concrete engineers lurking on the thread would you? I do ask a lot from this material but I believe it will deliver. I don't think I'm inventing impossible criteria, I'm just collecting up the criteria used in similar materials and writing how I see them applying to machine design.

We know that several of the companies that cast resin parts for machines regularly cite accuracies in the ten thousands for cast parts. Grinding is a no-no because us home shop types don't have precision grinders capable of holding tolerance to grind back aggregate. We know from Dewayne Harlow on another thread who is making mill tables for mini-mills that it is possible to cast good surfaces against a surface plate. We know from countertop materials like dupont Zodiaq that tiny aggregate near the surface embedded in epoxy make for a hard abrasion resistant surface as was shown by tests in Belgium in the 1970's. Vaccum is necessary however because air in the mixture will likely be around the aggregate grains causing you to have aggregate that's not well connected to the matrix. This is borne out by several papers in the London Symposium book I just got.

harryn,

You're spot on on your comment about the quarter by dust as aggregate. The reports I've been reading lately suggest that the aggregate is not as critical as I once believed. A perfect aggregate distribution is not as important as getting the epoxy to stick to it. Good aggregate sizing is an optimization but if we're going to bother engineering a mix, we might as well start with near-optimum aggregate.

You're also on with your comment about using a metal exterior and interior as far as it being the easy way. Hollow beams are also essential to spanning a significant distance with minimal deflection. In some studies I did of brunog's design, beams of size around 4x12x36 have less static deflection when the sides are only about an inch and a half thick than when they are solid.

Some of the loonier folks on the thread like me believe that good engineering will allow a nice machine to be built without having to cut any metal supports but this requires a reliable design for the E/G mixture which I think we are almost upon. Not everybody agrees with me and the others who believe a nearly non-metallic machine is possible.

We do have to remember that we on this thread are currently relearning 40 year old technology! Huntsman still makes the resins used by Staynes in his 1975 paper.

[walter]

Both Agsco and Cary Co are closed today which means testing will be on hold for another week.

I'm rapidly starting to lose interest in this whole thing.



Also, did some search on BYK-A525 de-airing agent. Found a patent for "Machinable, high strength epoxy tooling compositions" that use iron powder. They use BYK agent and there are some other bits of info. These are cast-to-size forming dies for sheet metal stamping - but check out their tensile strength!

http://www.patentmonkey.com/PM/PatentID/5280053.aspx


"..A machinable, high strength epoxy tool which is the polymeric reaction product of a mixture containing 5 to 12 weight percent of a bisphenol A epoxy, 3 to 8 weight percent of at least one polyoxypropylene amine catalyst, 60 to 85 weight percent of a mixture of no less than three different sized particulate fillers, the major portion of which is iron powder, 5 to 15 weight percent of short (<250 .mu.m) glass fibers, and 0.02 to 1 weight percent of a surface active agent."


"..The BYK additives were a series of wetting and dispersing agents made by the BYK Chemie Company. It appears that these additives enable better bonding between resin, glass and fillers. We found that certain of them produced higher physical properties in our epoxy tool compositions."


"..Our epoxy tool compositions contain, by weight, about 8 to 20 percent epoxy resin-curing agent system that consists essentially of 5 to 12 percent of a diglycidyl ether of bisphenol A and 3 to 8 percent of an aliphatic di- or tri-amine curing agent."


"..Our epoxy compositions are highly filled--using 60 to 85 percent of three distinct size groupings of particulate fillers, mostly iron powder, 5 to 15 percent short glass fibers up to about 250 .mu.m in length, and a small amount, suitable up to one percent and preferably 0.02 to 0.2 percent, of a suitable "surface active" agent that enhances the contributions of all the ingredients to form a surprisingly strong stamping tool."


"..We use a particulate filler mixture comprising interstitially mixable particles of at least three different average sizes or size ranges. Iron powder of at least two different average sizes (e.g., 300 .mu.m and 30 .mu.m) constitutes 80 percent or more of the total particulate filler by weight. Any suitable other filler material (of complementary third particle size, e.g. about 3 .mu.m) such as calcium carbonate is employed to accomplish the desired filler loading. Our particulate filler mixture constitutes about 60 to 85 percent by weight of the total moldable mixture. The total mixture serves, in part, as a heat sink to absorb and conduct away the heat of reaction of the resin-forming constituents. The fillers are not reactive with each other or the epoxy resin-curing agent system. The iron powder assures that the cured tool is machinable."


"..In addition to the particulate filler, we use short glass fibers (average length less than 250 micrometers). Surprisingly, such fibers are found to not unduly increase the viscosity of the castable mixture. The fibers do not increase the strength of the tool. The fibers make up 5 to 15 percent by weight of the moldable composition."


"..We have discovered unexpectedly that it is beneficial to use a combination of short glass fibers and certain additives that we call surface active agents in conjunction with iron fillers to produce machinable epoxy tooling formulations that exhibit excellent tensile strength. By surface active agents, we mean those materials that, together with the glass fibers, iron powders and/or other constituents, are found to markedly increase the tensile strength of the cured stamping tool. We set a target tensile strength at 58 to 60 MPa as a requirement for our stamping tools."
_

[martinw]

Grinding is a no-no because us home shop types don't have precision grinders capable of holding tolerance to grind back aggregate.

Dear Cameron,

I lot of people skim the X axis beds of their MDF X,Y,Z axis routers to get the bed true. The concrete countertop people grind the surface with diamond abrasive disks mounted in a hand held angle grinder. This is done to expose the aggregates for aethetic reasons.

I would have thought it would be possible to mount a grinder in the router tool holder and skim the E/G in-situ.

Could be wrong...

Best wishes

Martin

[ckelloug]

Martin,

Skimming will get the machine as true as the beam is. It might be realistically possible to diamond grind the table in-situ but I don't have much faith that I would be successful at it. If you are interested in metalworking grade precision on your table, I don't think hand grinding it will get there. I've been looking at a couple of studies on abrasion resistance in the London Proceedings book and they talk about embedding microparticles of carborundum in the epoxy. I don't disagree with you that the material can be ground, I'm just not sure that it can be done well without industrial facilities.

[martinw]

Martin,

Skimming will get the machine as true as the beam is. It might be realistically possible to diamond grind the table in-situ but I don't have much faith that I would be successful at it. If you are interested in metalworking grade precision on your table, I don't think hand grinding it will get there. I've been looking at a couple of studies on abrasion resistance in the London Proceedings book and they talk about embedding microparticles of carborundum in the epoxy. I don't disagree with you that the material can be ground, I'm just not sure that it can be done well without industrial facilities.

Dear Cameron,

I don't think hand grinding would be an option and I don't know if machine grinding would work. Those diamond disks don't come cheap, and if they are being asked to take on some carborundum, you could be looking at the wrong end of mega-bucks.

Best wishes


Martin

[brunog]

Their formula is 10-12 mm aggregate and sand used with huntsman gy-250 epoxy and HY830 and HY850 hardeners.

Huntsman gives a formula for Epoxy concrete on page 33 of the linked pdf: http://www.huntsman.com/performance_...theramines.pdf

60% 3/8 aggregate, 30% coarse sand, 10% flu ash as the percentage of aggregates out of total aggregates. Their aggregate to epoxy, hardener and accelerator ratio is 11:1 by weight.

I also found out that EEW means epoxide equivalent weight after calling

Cameron,
Aggregate mixture from Huntsman is very similar to Fuller's parabola. I cannot give you the exact ratios because don't have the size of flu ash.

I still have a lot to catch up to do. I have started an intensive technical course in mechanical engineering that's gonna require most of my time for the next 16 months. However I go to the site whenever i have the opportunity.

Best regards

Bruno

[qroger]

I know this is awful lazy of me, but I came into this thread at about page 104, and I wonder if someone could point me to a spot that describes what it is that we are building, and what we want to optimise. Please don't make me go back to the beginning.
Roger

[walter]

check out posts 942, 946, 963, 965, 966, 967, 969

The Whole Shebang on E/G: post #541 (page 46)
E/G Tips & Tricks: post #364 (page 31)
Useful links: post #241 (page21)
_

[ckelloug]

After all of the work on this thread, here is the first formula:

<B>Aggregate: Round Grained Silica Sand</B>

<table border=1>
<TR><td>100%</td><td>Passes 1/4 inch sieve</td></tr><tr><td>12%</td><td>Retained #7 Sieve</td></tr><td>35%</td><td>Retained #14 Sieve</td></tr><tr><td>20%</td><td>Retained #18 Sieve</td></tr><tr><td>15%</td><td>Retained #30 Sieve</td></tr><tr><td>12%</td><td>Retained #230 Sieve</td></tr><tr><td>12%</td><td>Passes #230 Sieve</td></tr>
</table>
This aggregate recipe is a translation to sieve numbers of the Reichhold formula for polyester polymer concrete. Based on the sample formulas in Gamski, it ought to work quite well.

<B>Epoxy U.S. Composites #635 (Reichhold Epotuf 37-127 with 37-606 amine hardener) </B>

Mix epoxy thoroughly before addition of regulating charge or aggregate. Use of an epoxy dispenser with static mixer is preferred as it will provide the most uniform results. If an epoxy dispensing system is unavailable, mix thoroughly with electric paint or drywall mud mixer trying to avoid entraining air in the mixture. Be sure to add de-airing agent and if used, directly adding coupling agent while mixing the epoxy before the addition of the aggregate.

After mixing the epoxy, add the regulating charge and then the aggregate and de-air the mixture at 5mm Hg of vacuum (.19 inches Hg). A paint pot with a vacuum pump has been suggested as a possible method but a venturi system is likely to provide insufficient vacuum.


The exact amount of epoxy required is currently unknown but Gamski puts it between 10 and 14% of aggregate without carbon black. Gupta places it at 16% but Gupta's formula had much less fine aggregate. Gamski's formula used with Gupta's 33 micron binder thickness equates to 35% of aggregate weight. Huntsman puts the value at 9% by weight but their formula uses much less fine aggregate and no carbon black. The epoxy required goes up with the amount of fine aggregate and carbon black and an overabundance of epoxy seems to be less detrimental than too little.

In absence of better data it seems like 20% of aggregate weight is a reasonable starting amount.

<B>De-airing Agent BYK-A525 from BYK Chemie</B>
Add de-airing agent equivalent to 0.2% of epoxy weight when epoxy is mixed.

<B>Coupling Agent Dow Corning Z6040 </B>

Adjust the pH of a large container of water with vinegar to approximately 4 using ph test paper available from http://scientificsonline.com/product.asp?pn=3021313.

Add 0.5% Z6040 to the volume of water which is about 0.5-.6 oz by volume per gallon and stir for 15 minutes. Treat the aggregate to be used by soaking it in the treated water and stirring for 5 minutes. Dry the treated aggregate under heat lamps until it appears bone dry and does not smell like methyl alcohol.

Alternately, add Z6040 equal to 1% of the weight to the epoxy directly to the mixed epoxy.

<B>Regulating Charge/ Dispersion Hardener Cabot Monarch 120 Carbon Black</B>
Add carbon black in an amount equal to 1.5 times the weight of the epoxy. This recommendation is from a russian paper cited by Gamski however it is not clear if it is by weight or volume but it seems like way too much.

<B>Molding</B>

Pour the vacuum de-aired resin into the mold taking care to entrap as few air bubbles as possible. After the epoxy has been added, cure the mold under heat lamps to avoid creep at higher temperatures.

<B>Notes</B>

I have not tried this formulation. It is my best attempt at a good simple formulation based on my research on the matter. The precise carbon black and epoxy fractions will need to be adjusted by experimentation. Use this formula at your own risk.

[walter]

has a weird weight to volume ratio- 1:14 by weight resulted in 1:4 by volume.

Maybe these results will help you fine tune the Formula.



I grabbed a bag of pea gravel and some aquarium crap and started mixing.

Rock 1 ___________17.5%
Rock 2 ___________17.5%
Sand 1 ___________17.5%
Sand 2 ___________8.75%
Zeeospheres _____13.125%
Fine quartz _______12.25%
Carbon Black ______0.875%
Epoxy ___________12.5%


This resulted in 35% of rocks, 52.5% of sand-fines-carbon black and 12.5% of epoxy.

Epoxy-hardener first, carbon black next, then rocks, then the rest.

Larry's mixing tips were right on the money but the mix was still unusable, it looked like asphalt to me. Way too dry.

I might actually use it on my driveway lol..
_