[walter]

Well...
I'm getting fascinated by this 8% Holy Grail mix, even though I don't plan on using it for my projects (my sweet spot is in safe and predictable 20-25% area).

I sort of tested it with water- about 10% by volume. Surprisingly, the mix behaves very well. It is very dry, but the particles fit together so well it doesn't seem to matter. Very interesting. I'm not sure if I want to try it with epoxy but that's not the point. The point is dry mixes compact better and enable you to use press compaction. At 20% epoxy by volume there will be no press compaction- you have too much liquid, which doesn't compress. Smaller particles are suspended "midair" and there is nothing you can do. It barely flows, and you can't even press compact it. The upside is, you don't have to worry about results- you dump it into the mold, shake a bit and there's your 1000psi E/G. Air bubbles or not. And it doesn't matter if you use 20% or 23% epoxy, or miscalculate the aggregate, etc. It's foolproof. You get your 1000psi no matter what. Pretty awesome, innit?


Not so with "on the edge" 8% mixes (IMO). You have your 92% of aggregate and instead of drowning it in 20-30% of epoxy- you decide to use only 8% and make a bet that you will be able to squeeze it and fit together just right. If you succeed, you triple the strength. But if you miss just a bit - the mix will fail like an old cupcake. There is no in-between result- you have to do it 100% right, every time. That bothers me.

I sort of worked on it from post #1084 when I did the super dry batch and decided to press compact it. A pleasure to work with these powder dry mixes.. I was able to squeeze them to 1/2 the size (tests and pictures in post #1165 page 98). They were pretty dense but not dense enough and ended up as cupcakes. As I recall, I did not use vibrocompaction- which is the key to the process (the particles need to rotate in place to fit and lock in place). I think I didn't understand what I was doing.


My point is: I'm pulling for you Cameron. I found my peace, I hope you find yours. You really deserve it.


Speaking of BYK A525 deairing agent, the left sample seem to have less air voids! I dug something else up: "BYK®- A 525 will also improve the substrate wetting because it reduces the surface tension." (from BYK technical forum).

Cheers!
_

[alexccmeister]

Are all these mixes going to be used with a reinforcement cage or some sort? So far the mixes are with fine gravels. But no steel reinforcement. I am sure the reinforcement will increase the strength some what.

[greybeard]

Walter - Your figure seemed a little low.

Take 7% resin by weight mixture = 93 gms aggregate + 7 gms resin.

93gms aggregate @ 2.6 gms/cc for solid quartz = 35.77 cc

Assume it has a random packing fraction of 65%
2.6 X 65% bulk density = 1.69 gms/cc........... = 55.03cc

7 gms resin @1.1 gms/cc ........................... = 6.36 cc

% resin by volume = 6.36/(6.36 + 55.03) x 100 = 10.4%

Regards
John

[ckelloug]

Walter,

The sample with the BYK A525 definitely looks like it has fewer voids. I think you're on to something. If you still have any experimentalism left, you might try more A525.

Your 1 ksi mix is definitely a good solution for machine building because it is simple and foolproof. Your mixture has the additional benefit that it will do an excellent job of sticking to the aluminum embedments which I suspect the packing-density-optimal mixture won't do so well at. Something similar might very well end up in parts of whatever I end up building. You beat the real granite which is farther than a few people on this thread ever thought any of us would get!

Your mixture was predicted by the model to require at least 12 percent epoxy. This is within a percent of the optimal mixture of the components you used without resorting to cheating. You should be proud of what you did.

In addition to using very carefully graded aggregates and the same additives you are working with, the two types of cheating that I might be able to do to improve upon it are nanofillers and tens or hundreds of PSI of press compaction. I'm interested in the nanofillers because according to the model, it should be possible to get to 7 or 8 percent using the same type of vibrocompaction you are using. Hydraulic and pneumatic presses don't excite me that much for this. . .

I'm thinking that with sieved aggregates, a 9% mixture will end up with so few voids that it will actually behave like a wet mixture once it starts getting vibrated. I'm hoping the safety margin can be 10% of the epoxy volume rather than the 50% of the epoxy volume in your mix. Here's to hoping. I'm going to have to get started on my experimental program now that I've mostly run out of numbers to crunch. (I still have to do a final aggregate design and send it to Agsco).

I'll be posting here once I figure out whether it is possible to make a near optimum mix without resorting to insane tooling. My initial hypothesis was that it would be just as easy to make optimal as suboptimal E/G and I hope I am right.

The Iranian site that Ger21 the moderator posted gave 4ksi as their flexural strength so I think the 2 ksi goal I started with is easily achievable. (You never broke that sample with weights so it may have even made it!)

Do you know if your blocks are breaking in the aggregate or the epoxy? That will be very interesting as it tells what could be further optimized.

I'm looking forward to seeing your machine posted here. You'll have one of the nicest and only homebrew E/G routers in the world!

As for your comment about peace, I'd have to say that I got quite a bit from your report that 10% water was all that was needed to wet out that aggregate mix. It means the model I've been calculating with isn't broken.

I think I've decided that my E/G project will be a machine that takes nice veneer plywood and churns out kitchen cabinets at a rate of about 10 or 20 per hour.

Regards All,

Cameron

[ckelloug]

John,

I've thought about testing your cylindrical specimens and I notice that you said Tensile Strength. That would be very difficult to test with a cylindrical specimen. I would imagine that you could make a jig much like the ones you've posted that had some curved ends for the little pyramids holding the samples up. This would allow you to get flexural strength although you would have to take some measurements and account for sample geometry.

I've had a lot of difficulty thinking up any sort of testing rig that can simultaneously be made cheaply, accurately calibrated, apply enough force to break our samples, and collect enough data to get flexural modulus and flexural strength.

As for the high speed spinning, the particles shouldn't separate if you use an accurately graded mixture with uniform portions of each component. Where it could potentially go wrong is with a mixture containing gaps with no particles of a given size in the middle of the size range. I can't say that it would separate, only that it could.

I'm not much help on this one. :withstupi

Regards,
Cameron

[walter]

John,

You're right. My numbers are based on Cameron's mix, which has 3 parts of Zeeospheres. So in Cameron's formula 7% epoxy by weight=8.75% by volume. But with pure quartz- it's probably closer to 10%. Thanks for doing the math.


Cameron,

I will try one more with BYK A525 and maybe a few drops of 3M Novec (I hope I can find that body suit- they gave me 8lb sample lol, it will be like opening Pandora's box ). But anyway..

You mention cabinet making CNC machine. It is the coolest thing! Hobby metalworking CNC machines are rarely impressive- not many people are willing to go 5-10 tons to get the proper rigidity. But for a woodworking machine- it's so easy to build that Thermwood killer- with only $10-15k in quality components. Not counting the spindle (of course). I hope to see a building log!

And again, fascinated with your 8% epoxy system- I had to do another "water" test.. I don't have tools for 8%, but I did 10%. Picture shows 400ml of your Formula and 40ml of water.

Aggregate:

1 part #6 Agsco Brown Aluminum Oxide
1 part #4 Agsco Quartz
1 part #2 Agsco Quartz
1 part #2/0 Agsco Quartz
1 part G850 Zeeospheres
2 parts G800 Zeeospheres

At first, it seemed hopeless. But slowly, it started turning into mud and I managed to wet the components. The shaker itself did nothing- aggregate does not vibrate- it rolls in the bucket like a soup (my guess would be "not enough VPM", you may remember our discussions on high freq vibrocompacting). I added press compaction and the aggregate still traveled around the bucket.

But check this out:
_

[greybeard]

Just a bit of info I've recently aquired re zeeospheres ... "G850 is actually a refined grade of G800 so it actually has a narrower PSD itself. Also, I'm not sure that G850 is will be around in the future so it would be a shame to spend to much time on it."

John

[walter]

Do you know if your blocks are breaking in the aggregate or the epoxy? That will be very interesting as it tells what could be further optimized.

Anything from 0.5mm to 5mm seems to be broken. Including every particle of Agsco Brown Aluminum Oxide- which is pretty amazing. You don't want to be around when that thing goes off- it's like trying to break tool steel lol.

This particular sample appears to be 10% stronger than my 536lb-test-beam.

Bad and Nationwide..:cheers:
_

[greybeard]

John,
I've thought about testing your cylindrical specimens and I notice that you said Tensile Strength. That would be very difficult to test with a cylindrical specimen. I would imagine that you could make a jig much like the ones you've posted that had some curved ends for the little pyramids holding the samples up. This would allow you to get flexural strength although you would have to take some measurements and account for sample geometry.

Cameron - I understand the problem, and I'm not hung up on specifics. Really, it's just to get some comparative data on what I achieve that will be useful to others in our pursuit of the strongest frame we could all make.
Just tell me what to test, and I'll figure out some way of doing it.

As for the high speed spinning, the particles shouldn't separate if you use an accurately graded mixture with uniform portions of each component. Where it could potentially go wrong is with a mixture containing gaps with no particles of a given size in the middle of the size range.

Thanks for the encouragement. I must admit the possibilities of particle segregation were fairly low on the agenda.

Regards
John

[ckelloug]

Wow,

Lots of incredibly good information today.

Walter, your picture of both the broken sample and the compacted aggregate are enlightening.

If it's true that the majority of the sample shows that aggregate broke then coupling agents and wetting and epoxy strength are not the limiting factor: it's aggregate!!! In that case, an all graded aluminum oxide or zirconia(probably insane) beam would likely do better. On the other hand, if the cross section had more aggregate in it, the forces on each piece of aggregate would have been less and it would have been stronger. If the aggregate pulled out rather than actually breaking then it is a bonding issue. If I remember right, this mixture was a mixture that was going to need at least 12% if compacted they way I describe. Remember that this 2% difference between 10% and 12% represents 20% of the water or epoxy fraction and you'd never say mix AB epoxy with that level of error.

As for the water compaction test, try a drop of dish soap in the water next time if you're interested to simulate the novec. You've got all the equipment I'm planning on using initially: a vibrator and some aggregate. The trick to it is to make a wooden mold with all six sides such that you can put the material into the mold and then press down on the top side such that you can apply pressure across the mixture. Attach the vibrator to the side or bottom and set a box of aggregate on top of the mold. That's all the pressure de Larrard used in his density studies and this should compact to within a few percent of where it will compact to with all the pressure you can apply with an arbitrarily large hydraulic press.

As for the comment about G850, IIRC I originally specified G200 zeospheres and G800 rather than G850 and G800 because the G200 has a lot of very fine round particles which make up for a deficiency in the lowest fraction. The model results that Walter was commenting on account for the fact that the material is rough quartz instead of round zeospheres: the model allows a different value of the compaction coefficient for each fraction! I used a low ballpark one for quartz and the one that de Larrard uses for round sand for zeeospheres.

John,

Your methodology should work well with excess epoxy since the aggregate will be forced to the outside and any extra epoxy will congregate around the neutral axis where it doesn't effect anything. Those of us not spinning don't have that luxury.

Walter,

Since you have an adequate material, don't let this discussion stop you from building a machine because you think you'll do better otherwise it will never get done Save it for machine number 2. My machine is still hypothetical so I have the luxury of tweaking on the material until the cows come home. I'll have to look at Thermwood and see what I can learn from their page.

--Cameron

P.S. I just ordered an arm and a leg worth of electrical parts to wire the laboratory! They'll be here tomorrow. I'll just need to get the shotgun to keep the copper thieves away

[greybeard]

On the surfactant thingy - a very silly idea came unbidden into my head.
What about WD40 type material. Hydrophobic, highly penetrating(= wetting ?) available and relatively cheap.
Would it work in principle to any useful degree, before I spray it in all directions ?

John

PS just noticed I've got to 1812.
De Dah de Dah de dah de Dah Dah Dah................BOOM

[lgalla]

Cameron,I buy time on a 5X10 Thermwood.If you have any questions such as rail size,ballscrew diameter,servo HP etc I will try to find out.Measuring components on someones machine is a bit tacky.
The linear guides appear to be 20mm,but are mounted on a ground steel piece about 1"X2" and this is mounted to the structure which appears to be 1/4" steel plate.The ballscrews are possibly 2"dia.The toolchange spindle is 10HP.The vacuum is 20HP.Feed rates in 3/4 hardwood,300 to 600ipm.
Thermwood has E-cabinets software free for people in the business.Request a copy ,it is good but only usable on their machines.
Any questions?I will try to answer.
BTW I only pay $85/hr for machine time.One hour on the machine equals 10 hrs saved in manual labor.It may not be cost effective for me to build a machine now but will pursue it anyways.
Larry

[ckelloug]

Larry,

Thanks. I'm still in the "Hmm what a good idea phase" as I don't even have a place to put a huge machine unless it can double as a workbench. That Thermwood sounds like a serious piece of equipment. Much more capable than I had planned but I suppose that the mechanisms are all about the same really.

Walter,

I got the Dow Corning Z6040 (courtesy of Dow Corning) today but I need to find something to ship it in. Would it be more useful for me to send you a bottle or would a few vials with 1 ml each turn out better? I'm expecting the Dow Corning Z6020 tomorrow. I'll probably have to order bottles so it may be the middle of next week before I can get this out to you but since you have to wait, I'll make like a Ginsu Knife set and throw in some Z6020 too and perhaps a couple of special bonus items.

It sounds like the optimal method of applying the silane would be to make a water solution with some vinegar according to the data sheet but this has to be dried at 220 degrees F for the chemistry to behave properly and to get rid of traces of alcohol and water. I wouldn't want to suggest anything that put this stuff through an oven that might contain food at some point (just in case) so I think this method is out. Also, it would require some pH test paper to get the acidity of the water right so it's probably "a bit too PhD" for this go-around to quote Xerxes. At some point, I may be able to produce a kit for treating aggregate that has everything that's needed once I understand the actual need for drying etc but not today.

I'd recommend the alternate application method for you as described in the Z6040 data sheet. It is simple as it just involves the addition of a small amount of silane directly to the resin mixture. They recommend between .5 and 2 percent silane be added.

John,

I haven't forgotten about your slew of questions on de Larrard's model or the data you asked for from the book, I'm just a bit behind right now.

:cheers:

Regards all,

--Cameron

[walter]

Cameron, just a vial or two please- I’ve got too much of that stuff. 200-300lbs total E/G for this machine, that's all. And I'd like to use it with resin please- I'm already way too deep into PhD territory

Still at least 2 weeks away- Harbor Freight didn't ship my motors and the mold is in pieces. Pushing slowly forward.

Slowly but surely.

[greybeard]

Cameron - thanks for the pm. No rush, "Rome wasn't built ..." etc
I, too, have an arm's length of projects on the go, but you always want to do the fun ones first.

[greybeard]

Having just discovered a fundamental flaw in my ideas about close packing of spheres, which for the uninitiated is how I've been looking at the aggregate proportions, I ran a search on "close packing spheres" and eventualy found this link. http://www.ac-noumea.nc/maths/amc/polyhedr/packing_.htm

The poem at the bottom of the page on http://www.pballew.net/soddy.html is some consolation, but I've now decided to lay down in a darkened room with a large G & T.

Cheers to all our readers
John

[lgalla]

Finally found a weight of a commercial engineered stone claiming 93%quartz to 7% epoxy.It is 48.2kg/m2/20mm thick.Can reverse engineering be applied to determine if the 93 to 7 ratio is by weight or volume.There seems to be no standard.My mixes were always by weight.Manufactures are saying parts per whatever.They never state %by weight or volume.Camerons research papers are by volume which in this case of multiple sizes for packing is more accurate as a cuft of solid granite is 165lbs while a cuft of crushed granite may be 100 lbs/cuft or so.
Bruno has stated 16% by volume is 8% by weight.This may be true with one sized aggregate.I wish everyone posting would accept a standard of weight or volume.Weight% is an industry standard but useless in multiple aggregate sizes.A block of granite is 165lbs/cu ft while granite aggregates may be 100lbs/cuft.This throws out the %age by weight theory.Hope the De-Larryard papers are considering bulk density or apparent density when mixing multiple aggregate sizing.Confused?Yes!
Surprized Walters samples are showing aggregate failure.Is the epoxy stronger than the aggregates?
In cabinet making 1/64" accuracy is bang on.The accuracy obtainable withE/G is not necessary in wood working,while the vibration damping is a +.Total E/G machine is a waste of money for woodworking,while tube damping is A+
Larry

[the4thseal]

you guys are moving so fast that i try to make samples and then send some out but you just zoom by me. it is almost to the point that as i am making a batch better info comes around. where i can i get BYK A525 and is it work getting, is so where?this is not a complaint just a happy observation.

[brunog]

.
Bruno has stated 16% by volume is 8% by weight.This may be true with one sized aggregate.Larry

Larry, epoxy has approx half the density of quartz or granite,unfortunately I don't have the exact numbers on hand, 1 lb of epoxy will have twice the volume of 1 lb of granite/quartz, as long as the aggregate mix is compact.

Best regards

Bruno

[brunog]

Apparently all counter top manufacturers use the same equipment made by Breton: www.breton.it

There is a video of the manufacturing and finishing processes, interesting.

Best regards

Bruno