[Zumba]

Cameron,

Excellent idea with the fake peanut butter. However, the sad thing is, some people these days are so gullible that they will think that the epoxy is hardened, dried peanut butter. They won't understand the gag. After you try explaining it to them, they'll have a fit and run amok. Idiots....

[greybeard]

Cameron - re you post #1624, page 55, and various observation/thoughts of others along the way, it now seems to me that the distribution of particle size in the final recipe is going to be a key issue to maximize the qualities of the E/G.
Given that de-airing/compaction, along with good adhesion to the aggregate is achievable by suitable choice of technique and chemical additives respectively, the ultimate strength achievable by diy workshop technicians will depend on their ability to match the required size distribution.

If the final recipe is, in part, based on a bag of "mixed ballast" from "Honest Joe's Builder's Supply Co.", even with "Honest Joe's National Supply Chain", there's no guarantee that the contents will have anything like the same size distribution. I'm assuming that Joe doesn't ship aggregate around the country, and uses local sources.

Where this is leading me is to make the point that knowing what's in your own mix is going to be very important if the final result is to be as good as hoped for.

There would seem to be only two ways to achieve this.
Either you have to make up your own mix from individual materials, or be able to analyze what you've bought in, and adjust it where necessary.
To be able to analyze it means you've got to be able to separate the mix into its parts, and then make good any deficiencies.
So to gain the advantage of the low cost of a typical aggregate mix you have to add in some cost/work to know what you've got.

Needless to say, I'm working on it, but I'd be happier if others thought the same and to hear anyone else's ideas.
I've been looking at a patent showing a fairly simple(!) method using an air stream to separate the sub-1mm particles.

It may seem to be another complication in achieving what we're striving for, but at the moment I can't see any way around it.

John

[ckelloug]

John,

On this side of the pond, there's a company called agsco which will either sell raw aggregate or sieve it down to a specified grading at what appears to be reasonable prices. If you've seen Walter's boxes of all sorts of strange aggregate and the fact that he has a reasonable idea of the size range when he posts, it's because he happens to live near agsco and dropped in and picked up numerous grades of quartz. It will be difficult without careful sieving to get a high solids density mixture. Speaking practically for a moment, it will probably be possible to get a fairly dense mixture with only 3 or four reasonably available components but the minimum strength goes up exponentially with the last few percent of aggregate in the composition so the strongest possible mixture will need help from the pros as near as I can tell.

Sieves are about 20 £ a pop and 6 or 8 are probably needed along with a couple hundred £ for a sieve shaker for lab scale production. I fully intend to calculate what's needed from de Larrard's formulas and buy aggregate graded to that spec. I'll be looking for aggregate just as soon as I stop getting great results from the de Larrard computations like 275% packing densities and other nonsensical figures. While converting ridiculously complicated formulas to software is one of the things I do for a living, this is giving me an awful lot of trouble. De Larrard may have published what he did but he sure didn't go out of his way to make the whole mess easy to follow.

The pros in the aggregate world have laser diffraction aggregate characterizers and other insanely expensive gear that I doubt any of us could afford to be in the same room with, much less own. Even the aggregate research institute at the University of Texas sends samples out to a specialist laboratory at the (US) National Institute of Standards and Technology for some of the critical calibration measurements on aggregate sizing.

Because a lifetime supply of reliable precision graded aggregate costs on the order of a couple hundred pounds on this side of the pond, I'm concentrating on figuring out exactly what's needed and I'll leave producing it to the pros.

In short, I note your problem and have no solution. I am bypassing the problem by leaving aggregate to the professionals. I don't know what the proper industrial directory to use is on your side of the pond but I found agsco looking through the Thomas Register of American Manufacturers.

Best Regards,
Cameron

[walter]

John,

I'd be happy to send you a sample pack of my aggregates, free of charge, so you can test it against ordinary sand. I've got some experience and aggregate isn't as crucial as you think. Vacuum, additives- yes. But without them, you can epoxy old pair of jeans and get the results. My opinion of course but please test it before you spend the money.

I will offer the same sample pack to Cameron, let's see what we can come up with

[WayneHill]

I have been following this thread off and on for a while. Has a sawdust mix been mentioned?

http://en.wikipedia.org/wiki/Pykrete

Also saw a rubber mold maker video where the mix was poured from a height of 30 inches or so to stop the air bubbles from forming. They called it the "bombs away" method. They used this method instead of a vacuum.

[ckelloug]

John,

I'd be happy to send you a sample pack of my aggregates, free of charge, so you can test it against ordinary sand. I've got some experience and aggregate isn't as crucial as you think. Vacuum, additives- yes. But without them, you can epoxy old pair of jeans and get the results. My opinion of course but please test it before you spend the money.

I will offer the same sample pack to Cameron, let's see what we can come up with

Walter,

Thanks for the aggregate offer. I may take you up on it when I think I understand the problem better. I believe you are seeing the effect in the rule of mixtures graph I've reattached here where changes in aggregate percentage make a very small stiffness difference between about 60% and 85% aggregate fill by volume. If you look at the slope on the bottom curve however, you'll notice that between 85% and 92% (about as good as is feasible), the slope takes off.

http://www.cnczone.com/forums/attach...9&d=1177824890

I contend from arguments based on the rule of mixtures theory that theres probably a factor of two or more difference in the stiffness of the material between 85% and 92%. Since it's almost impossible to get an actual mixture with these fill rates without some hardcore calculations, I personally believe that you just don't have the data to see the effect I am predicting yet. Anybody with two bags of different sized sand can get a 70% by volume aggregate mixture. Anybody with large sand and zeospheres can probably make the low 80's.

Until I get my aggregate simulator to work, calculate out a trial mix and then make it and hopefully have it work, it's my opinion that 92% is a lot stronger than 85% vs. your data that 70% vs 80% fill rate doesn't make a difference. I wouldn't suggest anybody trust me on this one till I have some hard data unless you want to push the envelope and see what you can figure out on your own.

In short, Walter, I think you're sorta right but I have faith I'll prove you wrong in the next few months.

--Cameron

[greybeard]

Walter - this is a quick thank you for the offer, which I will take up.
What I should very much like to concentrate on is putting together some simple kit that will enable me, and others, to compare my own work with yours. I had initially thought that it might be possible to test my own material(when it appears) here, and have you test yours in a similar fashion. We could then compare the properties of any future recipe, all done "at a distance".
I can see that the first stage should be that I am able to compare the actual material that you are using with mine, and design any testing kit that I come up with using both. PM will follow.

By the way, I must stress that in my last post I was contemplating(as I think Cameron is doing) on the "best possible" recipe when I talked about the importance I saw of the particle size distribution.

I've managed to contact 3M UK, and in turn Lawrence Industries, so I'm hopeful of obtaining a evaluation sample of Zeeospheres.

Cameron - You ought to have guessed by now that I have an aversion to spending any more money than my ingenuity will allow. I've managed to come up with a "shoe-string" design for obtaining a size distribution of a sub-mm sample from my aggregate, so I aim to put it together Saturday onwards. (Friday is a bit of a social day off with friends, :cheers: music with grandchildren in the evening, and hopefully a clear head by Saturday afternoon.) When completed I'll post the design with photos.

The spin casting rig is built, but that too will have to wait for the weekend to put into operation.
I've test run it empty, and noted that the vibration of the off-center bearings will introduce a vibration which will either benefit the compaction process, or throw it through the wall.

The latter I doubt, having done a fair bit off wood turning, generally starting with very poorly balanced blocks !

Regards
John

[ckelloug]

John,

I understand you on the trying to do this work inexpensively and it brings out the genius in you. I currently hold the thread record for least material and the most books bought yet and also the most free industrial samples acquired.

For me, once I get proper aggregate design through simulation I just intend to buy the aggregate from the guys who can make what I specified. For the research I would like to do, a quick analysis makes me thing it would be cheaper to buy aggregate from people with the right tools and access to odd aggregates than to dedicate the time and resources to making my own. Most of the materials I want to test aren't off the shelf here where I live anyway.

For instance, I want to get aggregate in 5 specific materials sieved identically so that I can correlate fracture toughness with ultimate flexural strength. I am quite concerned with aggregates exactly meeting the grading curves once I determine them and as a result, I really want accurate aggregate this first go around so I can attribute problems to other experimental factors.

When I get my stupid simulator running, I should be able to tell you the packing density of random mixes of say pool sand from Joe's Hardware Emporium LTD. and zeospheres etc. I did notice today that deLarrard's older simulator is available from the french government department of roads and bridges for download. It runs on Windows and provides results that may be good enough for our purposes. There are at least three drawbacks in it for me however: It doesn't correlate with reality quite as well as one based on his newer work, it won't run under linux in my windows emulator and everything in it is all in French which is my next best language after english and computer languages.

DeLarrard's old simualtor can be found by googling for rene-lcpc. This will take you to the page for the Laboratoire Central du Ponts et Chaussees download page.

http://www.lcpc.fr/fr/presentation/o...esult/rene.php

There is also a demo version of their new expensive simulator which I have not tried to run yet. It's current but has an ouchy price tag of 1500 euros.

http://www.lcpc.fr/fr/presentation/o...etonlabpro.php

Best of luck in all of your research, John.

[walter]

You've obviously put a lot of thought into this, and you know what you're talking about.

But I've done 8% mixes and there is simply not enough epoxy to hold it together. I said it can't be done without vacuum and bond strengthening chemicals and I'll stick with that.

Which brings back the E/Q countertop guys. They do 5-7% mixes.. With heat activated hardeners and $20mil worth of machinery. They squezee the mix to 1/4 of it's size, vibrate and apply heat at the same time. I'm sure the results are stunning.

I've done these mixes and they're pleasure to work with. But they need tons of pressure just to hold em together. Can you get that with vacuum?

[ger21]

But they need tons of pressure just to hold em together. Can you get that with vacuum?

14psi = ~1 ton/sq ft.

If you can seal it with no leaks, and hold your mold together under that kind of pressure.

[ckelloug]

Walter,

You're right that I put a lot of thought into this but you've also got data which says I may very well be wrong. You would be wise to take what I say with a grain of quartz-- or salt although I'm sure it's not good for the mix

On the other hand, I'd like to think that hours of reading books and writing software allow me to see a direction to go that pure experiment would conclude was a waste of time. In my professional life I usually come up with solutions that work due to their extraordinary attention to details in strategies that would otherwise fail.

The reason I think I may still have something on a very high solids mix is explained by the attached chart. I've taken the usage of the word dominant from de Larrard who uses it to mean that a grain of this size or size grouping is controlling the mixture's overall density. This grain size class generally also controls mixture flow (rheology).

What my silly chart shows is that for a 10 component mixture, it can reach 92% density in two ways: it's either there because the pile of aggregate has a 92% packing density or it's there because there isn't enough epoxy to hold it together.

Of the 92% mixtures that are there, a mixture that is dominated by each grain size will at least in theory have different flow characteristics.

Ger makes an interesting point about vacuum being 14 psi. De Larrard compacts all of the dry mixtures he's done experiments on under pressure and vibration. What's interesting in this context is he only uses 1.5 psi of pressure during vibration and that's applied with a 40 lb weight! It more than doubles the amount of compaction compared to just vibrating the mixture.

With the mixture that I talked about a few posts back where there is a uniform distribution of log (size) ranges, it can be vibrated continuously for the pot life of the epoxy since it won't segregate. Using a mold that has a top and weight on it sounds like a good cheap experiment on any mix of your choosing. In fact, tethering the vibrator to the top plate of the mold would likely provide the weight that would be good for the sample bars you make.

The weight is also consistent with the work of B.W. Staynes from the 1975 paper who suggests that setting under pressure made an appreciable difference. Near as I can tell, all mixtures are a bit thixotropic: vibration causes them to thin and gravity will compact them some, but extrenal pressure, even a little bit from a weight, greatly accentuates the effect.

Bob Warfield may be right that I'm crazy on this but I'm going to keep at it until I figure out which. If I'm wrong, I suppose I have to make a world tour buying beers for the people who thought I wasn't crazy. . .

Back to the simulator debugathon.

--Cameron

[lgalla]

Walter here is a copy from engineered stone:engineered Stone (ES) is comprised of ground up granite dust mixed with coloring and epoxy to hold it together. These companies will tell you that ES is 93% stone, 7% epoxy. What they won’t tell you is that these percentages are by weight, not volume. As granite is much heavier by volume than epoxy, the actual stone to epoxy ratio is closer to 75%-25%."
Don't know why you worry so much of your mixes which I beleive are volume and right in the ballpark.
http://www.jgreer.com/Pebble%20paving%20page.htm
Interesting link for pebbles in epoxy for flooring.Mix is 6gal epoxy to 600lbs stone.Epoxy is almost 10lbs/gal.making this a 10% ratio by weight.
Have a look at the photo,no way this is 7% epoxy.
Larry

[lgalla]

Too late,too many beers,]make that 4[.Please excuse non technical observations.
One ton of vacuum sounds like a lot,but yes it is and no it is not.
Normally the whole mold is in the vacuum bag equalizing the pressure all round.This is not a strain on the mold.Since the epoxy mix is fluid at the time,the pressure is also fluid.You can push you finger on the bag and move or dent the epoxy mix as it is still fluid.To get the ton of pressure a caul is necessary which would be a stiff material placed on top of the E/G mix.Now you will have 1 ton on the mix.
Larry

[Zumba]

Larry, I'm pretty sure Walter's mixes are by weight.

My tests (which were more for testing the procedure rather than the composition) were similar in that 10% epoxy by weight results in a very dry mix.

[ckelloug]

All of the theoretical predictions I have posted are by <B>volume</B>. While volume isn't very reliable for proportioning mixes in the lab, The material properties are directly governed by the volume proportion of the mix, not the weight proportion except inasmuch that the weight proportion is a way of estimating the volume proportion.

I've almost deceiphered the tricky part of the de Larrard formulas. There are a whole bunch of qualifiers on the domain and range of the equations that he dances around in the text but doesn't state directly. They aren't obvious until the equation blows up when you try simulating it. The whole mess makes about as much sense on the surface as the airport signs in Charles De Galle airport: Arrows pointing to nowhere. . .

Until I can reliably produce simulation results for the given cases, I won't feel confident in trying to predict the properties of anything and can't. I'll give it a rest for a few days on the notions of why we get dry mixes until I have useful data but here's a summary of my theories

If the low epoxy mixes that you guys have are in fact mixes where the aggregate's intrinsic container fill by volume plus the epoxy volume equal 1 then you guys are right that low solid mixes aren't going to work.

From what I've seen of the grading curves however, I have a suspicion that what has been produced so far are mixes where the intrinsic container fill volume of the aggregate plus the epoxy is less than 1 resulting in voids that may or may not be obvious and thus the near zero strength mixes.

If you vacuum out the air in the mold, you've got 15 psi of hydrostatic pressure coming from all directions. This is a lot more than the 0 psi of hydrostatic pressure you'd get from just setting the epoxy by itself. De Larrard's technique only used <B>1.5psi (not 15psi but actually 1.5psi)</B> applied by a weight on top of the container to compact his dry mixtures along with vibration. We should have no trouble duplicating this aspect of his testing.

WayneHill,

Your Pyckrete post reminds me of an art museum exhibit I saw of a barbell made of vaseline sitting on a thermoelectic module plate to keep it from melting. Sawdust has too low a modulus for this application but it is interesting. If you manage to build a model aircraft carrier made of ice anytime soon be sure to post here even if it is OT Thanks for the post.

--Cameron

[walter]

Looks like I'm going to need a bigger fire extinguisher...

_

[Zumba]

Sweet jeebus, Walter. It's been nice knowin' ya...

[Zumba]

Hey Cameron,

I agree that vibrating the dry mix and applying slight vacuum should be easy money. 1.5psi is attainable by a shop vac.

I'm still baffled though as to how void removal should be done in a wet mix. I've been thinking about all this de-airing business. In my tests with high epoxy ratios, I get air bubbles. But as I get closer to the recommended ratio of 10% epoxy by weight, the mix gets dry/sticky to a point where voids would not be perceived as air bubbles, but rather as gaps in the aggregate. Think "rice krispy treats".

What is required to get the gaps out of something like that? Vacuum won't work. Even if you suck the air out, the structure is too sticky for something to come in a fill the void. I'm not sure how vibration would penetrate when you're building a beam that's 12" x 6".

I'm leaning towards a combination of hydraulic pressing + a wand type vibrator that actually goes into the mix. I've got some hydraulic press equipment as well as plenty of steel. Maybe I'll fabricate a cylinder type press using DOM steel tube and a 12L14 piston.

Any ideas on how I should design the thing?

[lgalla]

Sweet jeebus, Walter. It's been nice knowin' ya...

I am sure Walter will still be around as I guess only a few drops are used in a mix.Air release additives are not magic,they assist in air removal while under vacuum or vibration.
Larry

[walter]

Have a look at the pictures-I believe these are 7-10% mixes.

Light grey is Zeeospheres and 40micron silica. Black has a lot of pool filter sand and a little bit of carbon black. I didn't use my shaker, just hammer and piece of wood. I was able to squeeze it to 1/2 the original size- I thought I hammered them pretty effectively. The strength test showed that samples were like old cupcakes- only 25% of typical strength. But like I said, they're great to work with- no sticky mess..

So..You'll probably need a "real" press, I don't think vacuum can do it.
Two quotes from EQ countertop guys:

"Resins, binder, and pigments make up 7% of the product along with 93% purified quartz. Manufacturing quartz begins with the raw white quartz or clear quartz cut out of the mountains. First, the quartz is purified, ground up, and mixed with large quartz particulates—some dime size, some smaller and made into a slab. Through a process of vibration and vacuuming out all air inside the slab, (called vibrocompaction) a 4 and ½ inch slab of engineered quarts can be vibrocompacted down to a ¾ inch slab. After the slab is heated to 180 degrees centigrade, it is polished to a high shine by drums with small diamond bits that act like sandpaper—a process, by the way, that is the same for polishing a granite slab."

"Quartz surfacing is made from raw quartz crystals. After being mined, the quartz is separated according to size and type, inspected for quality and stored. At the time of production, the quartz and a very small amount of resin and color pigment are mixed until thoroughly combined and then poured into a large tray with a paper liner. Once the mold is filled, it is moved to the vibrocompaction area, where massive presses squeeze the material to roughly half its original thickness. At the same time, a vacuum is applied and the mold vibrated to evacuate the air. Curing is further accelerated by moving the mold with its compressed load into an oven, where it is baked at 100 degrees Celsius for almost half an hour. Then it enters a special chamber where is it air-cooled to room temperature. When it reaches ambient temperature, the product is fully cured and ready for calibration (thicknessing) and polishing."


I'm not doubting Cameron one bit. I wasn't really paying attention to his graph, so I'm only starting to understand the whole 92% thing. The idea does seem pretty clever. I'm just saying it may not be as easy as it sounds.

The point is this: We are in midst of a revolution and there is no telling where it will go. I already have cool mixes.. even at 17%, they're mighty strong and I'm dying to get them on my machine. Now imagine if he actually gets that black swan mix.. Holy cow. This place will be upside down.
_