[roach]

ill try a pic of the matrix it looks like stone
i can tell u when we make a casting wrong and try to get the inserts back out you cant smash them out you have to put the in the oven

[ad_bfl]

What I am not so sure about, is if the silica will start to fuse together during the anneal.

That would definitely be a problem

[ckelloug]

sigma John,

Strength and modulus models are sometimes confusing. I myself was confused initially when I posted that graph (8 months ago). I cited it again a few posts ago noting that it was wrong but pointing to the fact that it correctly applied to modulus and not strength (No longer confused thinking this could also model strength). I really need to redo that graph with good labels and numbers. In that incarnation, it was being used to show potential modulus vs. aggregate percentage.

Hashin Shtrikman theory produces a set of tighter bounds for the modulus but I haven't read this seminal paper yet. De Larrard cited this paper and gave the equations in one of his chapters and that is how I found it.

I am no longer confused: The above two models are definitely germane.

rowbare,

I've seen you lurking here before; welcome to posting and thanks for the videos on casting. Most of my problems were based on equipment failure (too small a balance) and stupidity: being excited and not transferring the correct amount of material. A swift kick to the posterior is about the only way to fix gross errors although I need to get a new balance and a mixer so that I can mix thoroughly.

roach,

I test samples 19cm long x 1cm thick and 1.3cm wide. I agree that the crack should go through the granite and not just the resin. I'm looking forward to your pictures. If you can send me a test block, I can diamond saw it to size and run the tests.

The deflection numbers we have are about 1/5 of reasonable. I'm fairly convinced we have a hardener that isn't likely suitable for this purpose. Hexion and Reichhold apps engineers both owe me an email right now. What we have now would probably work OK for something like a router where holding 1000um i.e. 1 mm is good enough. Precision metal working equipment would require massive design to get by with a material with this low a modulus.

ad_bfl,

The heating of aggregate is an interesting idea. We already know that the materials should be dry and that at least moderatly high temperature drying is required (I'm thinking vacuum might be better however). Invisible residual water on the surface of the aggregate makes a big difference but this effect can be minimized by the use of silanes and or titantes to displace the water from these sites and form covalent bonds there. As a bonus, slinaes and titanates have a pendant molecule that has an affinity for epoxy and chemically bonds the aggergate and the epoxy. We know that water greatly influences the cracking of glassy materials including epoxy.

lgalla,

This isn't about producing a truly brittle composite, it's just about producing one that's less compliant than the one we have now which is a bit too rubbery for precision applications. Thanks for the comments on hardener, I'll put them to use soon. The epoxy we and others use for E/G is chocked full of diglicidyl ether and is rather tough but not incredibly stiff. We won't end up with something that is brittle like glass when we get it right.

Gizmot,

Good luck on your machine. A decent E/G formulation should be reasonably satisfactory for most of these properties. Cost isn't quite mutually exclusive with all the other properties as maximizing aggregate density which minimizes creep and maximizes modulus also minimizes cost.

greybeard John,

One must keep in mind that epoxy and granite have fracture toughness values that are similar. As a result, the energy to propagate a crack through either is quite similar. The problems like you are describing with packing seem to me that they will come into play when you use aggregate like aluminum oxide which has a much higher fracture toughness than the epoxy. If there is an easy path around the high toughness grains, the fracture is more likely to go that way. Fracture affects the ultimate strength of the material but for tool applications, we must design our systems so we are nowhere close to that stress level.


All the above being said, I guess I need to finish reading Lazon on damping models and Hashin and Shtrickman's paper on modulus.

Regards all,

Cameron

[jhudler]

What I am not so sure about, is if the silica will start to fuse together during the anneal.

Silica will start to melt (depending on quality) from 600C to 1100C; which hardly matters as the epoxy has long since burned off.

There wouldn't really be an annealing process to EG, beyond the post cure dictated by the epoxy formulation.

[ckelloug]

This is a repost of a link that jhudler (Jack) found many weeks ago.

http://www.cnczone.com/forums/attach...9&d=1204501651

If any of the European contingent on this discussion needs some E/G related entertainment, contact your local Hexion office at www.hexion.com and see if you can get samples. It's not available in the U.S. right now although I have the Hexion apps engineers wrangling for a U.S. substitute of for making the formula available here. Hexion will sell small quantities of stuff Stateside so if you EU folks can buy small quantities of this stuff, it is definitely what you want.

Kudos to Jack for posting it.

--Cameron

[sigma relief]

hmm, this is twice I've posted a reply and it never went thru, Ive got to figure out why.

Anyway, I agree that DOE on its own is useless, if you don't have general bounds for your experiment, you are only doing scientific guessing. This is where the traditional control/single variable experiments can best be put to use.

I have been putting a lot of thought into how to fit EG into the traditional DOE format requiring complete variable independence. To keep the math happy, we need to keep the overall amount of aggregate constant, but need some mobility in terms of ratios. The best way I can figure it is to establish pairs of aggregate and adjust the relative percentage, but keep the combined total the same. The limits of the two aggregate sizes in each pair do not need to be symmetric, 15% of A and 3% B can be the upper limit (normalized to 1) and 11% A to 7% of B could be the lower limit (normalized to 0). In order to ensure we don't have too much large or small overall granite, pairs of aggregate should not be sequential (pair large and small, medium and extra small, or something to that effect). This allows us to fine tune the relative amounts of aggregate, but requires even number of aggregates. Another problem this creates is controlling the resin ratio and I see two options. Option one is use a percentage variable that is calculated for each batch based on computer calculations of packing density. This it a bit tenuous as it is not true variable independence and could lead to problems at the end. The other option is an absolute percentage of epoxy scaled to weight of aggregate. This will probably be more stable at calculation time, but the changes in large to small aggregate will require a large range of epoxy test values. Because EG performance drops off rapidly once we get too dry, the low epoxy data will be garbage. This will require a minimum of 3 value testing, with better resolution coming at 4 or 5 values. The last "good" epoxy rich data point will serve as damage control for the low performance dry test point by doing linear interpolation over only the last bit of data, preserving the good data from the wet samples. Unfortunately, if our theory of reducing epoxy is correct, it is this data on the cusp of going dry that we are most interested in.

Well, unless someone gets lucky and guesses the magic numbers, or the computer simulation is verified, we may have a lot of testing to do. Actually, we have a catch 22, without the testing, we don't know we have the optimal solution, but without the input data, we don't know if we are looking at the correct variables.

[harryn]

Silica will start to melt (depending on quality) from 600C to 1100C; which hardly matters as the epoxy has long since burned off.

There wouldn't really be an annealing process to EG, beyond the post cure dictated by the epoxy formulation.

Hi - I wasn't very clear. I meant to anneal the fumed silica (or aggr. mix), then cool down to room temp - prior to mixing with epoxy, not after it was EG.

[ckelloug]

sigma John,

Sorry not to key on the DOE (not Dept. of Energy) trial thing. I've been busy trying to solve a riduculously annoying but simple software problem in the paying portion of my life for the last week or so.

In general, I would consider the aggregate problem to be solved. Given any mixture of aggregate, I can say within less than 3% what the packing density Phi will be after taking a Beta measurement on each component. (I'm using estimated beta values right now and thus my results are only relative to one another right now). The French Researcher, de Larrard, whose models I have implemented spent about 10 years validating them with data and has published many papers and the book I am working from. I have shown that I can duplicate the numbers published in his book.

Given that the aggregate packing density Phi can be calculated to within 3% for the model, My engineering judgment tells me the DOE experiment design doesn't need to account for individual aggregate types but should instead specify Phi values. Phi goes from 0 percent (all epoxy) by volume to about 92% (8% epoxy) by volume.

It is known from the rule of mixtures model that modulus follows a uniformly increasing nonlinear curve with percentage of aggregate. It is also known that eventually there is a point where the modulus becomes zero with increasing aggregate content due to insufficient epoxy to bind the material.
The minimum amount required to fill all of the empty space in the solid is 1-Phi. Any less than that and the mixture doesn't contain enough epoxy to hold itself together. More epoxy than 1-Phi that causes the particles to be fully encapsulated by epoxy in model terms.

So, I would say the first unknown is what the optimal amount of epoxy for maximum modulus is. Of course this assumes that all mixtures with the same Phi value are equally good for modulus. This assumption is true according to the rule of mixtures but there could be effects related to the quality and type of aggregate that cause fractures which affect the result.


Here is the summary of factors,likely effects on modulus, and range is as below:

Packing Density of Particles: positive linear: 0----92% by volume aggregate
Epoxy volume percentage in composite: negative linear: (100% - Phi)----50% by volume aggregate
Titanate Concentration: concave down parabolic: 0----1% by weight aggregate
Silane Concentration: concave down parabolic: 0----5% by weight aggregate
Cobalt Acetyl Acetonate Concentration: positive linear: 0----1% by weight epoxy
Epoxy to Hardener Mix Ratio: concave Down Parabolic: 2.1-----2.5 w/w for 37-127/37-606
Well Dispersed Nanomaterial Concentration: positive linear: 0----10% by weight epoxy

Once the gross problems like bad choice of hardener etc have been worked out, a DOE trial of the above factors would be a good way of getting the optimal material.

Regards all,

Cameron

[greybeard]

Cameron - The only "modulus" I'm familiar with is Hooke's, stress/strain = constant(?) in the guise of Young's modulus, er .... extention proportional to the weight on the end of a wire etc.

If this not the one you are refering to (chair) please give me a definition.


John

[ckelloug]

John,

Engineers like me (who are sloppy about terminology), often refer to modulus.

Modulus most correctly called Young's Modulus is k from Hooke's law in the simple tensile case. Mother nature hates us however so non-metals behave differently in tension and compression thus we have tensile (Young's) modulus, and compressive modulus, the same idea calculated in compression instead of tension. Non-metals behave differently when a beam is flexed then they do when either in pure tension or compression so thus we have flexural modulus.

Flexural modulus is measured using a 3 point bending test and is the parameter that affects how a beam made from the material will deflect. This test requires the least powerful testing machine and is the one I am doing via procedure ASTM D-790.

Regards all,
Cameron

[greybeard]

That's good news - we're speaking the same language then

John

[roach]

to me your way too hot for your glass transition temperature
try lower and longer to anneal it
said enuff lol
want more i need more cider lol

[sigma relief]

Cameron,

I have a lot of faith in your packing simulations and agree that it will probably be the single most important tool to creating optimal EG. You have proven the ability to optimixe an arbitrary mixture for packing density, but I'm not sure that is the whole story. In theory all granite is the same regardles of size, but because it has grain and the surface texture changes, I don't know what the modulus trends look like. In general smaller items are considered to be "stronger" because they are less likely to have critical flaws, and the flaws that do exist are contained to individual pieces.

Because we are staying below the failure threshold, this principle does not directly apply, but I'm not sure if something similar may be at work when it comes to the epoxy/aggregrate load transfer paths.

The minimum epoxy thickness has been mentioned a few thousand posts ago, but as we try to squeeze a variety of small aggregrate into the mix, the increased surface area and epoxy required are not accounted for in the model. The result of this is the small particles are effectively larger than simulated and do not behave quite as predicted. We can add a fudge factor in the size inputs of small aggregrate to better predict the results (fudge the input sizes, then go back and re-calculate the true density based on the true particle sizes) The epoxy layer will also disturb the beta values for small particles, they will probably behave as more spherical than dry, but I can't say that for sure.

My reason for dredging up ancient history is not to question it, but to make sure we determine if the small aggregrate benefits outweigh the penalty of increased epoxy. My personal opinion is small aggregrate will be a part of the ideal formula, but in smaller percentages than predicted by the packing model. I really don't know if DoE will even work for adjusting aggregrate, it is just an idea. It is the other addatives that will benefit most from the test. If we ran the addative test on a "standard" aggregrate mix including epoxy percentage, the optimal result should be usable in aggregrate testing. Two DoE tests with 5 variables will go much quicker than one 10 variable test If we assume the two data sets are independant, we can mix and match results. If the aggregrate results end up being significantly different than the initial addative tests, a few short tests to checks to compare against origional data may be in order.

[ckelloug]

sigma John,

I think your last comment is spot on. I'm trying to figure out how to either model this problem or design an experiment right now.

The quick summary is that 1-Phi is the least amount of epoxy that could work. This is the inflection point between the theoretical volume of the composite not going up as epoxy is added and the theoretical volume starting to go up as epoxy is added.

The volume percentage of epoxy goes up as we add more epoxy after the inflection point and the amount that it does up is proportional to the specific surface area (surface area per unit mass) of the aggregate. Hashin Shtrikman can give us bounds for the modulus so if we couple the Compressible Packing Model to find Phi, Hashin Shtrickman to Bound Modulus, and Gamsky/Gupta's epoxy shell model, we ought to be able to predict how this affects our composite if we have numbers for specific surface area.

Unfortunately, we have a bigger problem. After considerable thought and a look through the MSDS for 37-606, it's 30% nonyl Phenol with a dash of AminoEthylPiperazine. This is designed for paint applications where incredible flexibility and toughness are paramount.

We need strength and high modulus. It appears that several non-formulated hardeners would do substantially better either singly or in combination:
Diethylene Triamine
Isophorone Diamine
Jeffamine D230 (diamine Polyetheramine with mw 230)
Jeffamine D410 (triamine Polyetheramine with mw 410)

If there are any armchair epoxy chemists out there, comments are appreciated.

Regards all,
Cameron

[roach]

isophorone diamine
:wave:

[ckelloug]

Well,

After a day in the library, it looks like the epoxy is going to be fun too. First off, all of the hardeners similar to the one we're using are extremely toxic. Our current 37-606 would be even more toxic if it weren't for all of the nonyl phenol in it and nonyl phenol isn't all that nice either. Guys like Larry know this but pure amine hardeners decompose on exposure to water vapor and in some cases CO2 and have to be stored with nitrogen in the headspace of the container.

I don't know how everbody else feels but if our E/G is going to be suitable for home use, it might be worth looking into some of the less toxic hardener families than primary amines. What do folks think: are we comfortable with seriously corrosive hardeners with deadly vapors?

Isophorone Diamine looks like a nice hardener for E/G but it is still quite toxic and hazardous.

Secondly, for maximum damping, it looks like we want to control Tg so it is between about 60C and 100C. Too low or too high a Tg appears to substantially reduce damping. Too high is a much worse damper than too low but an operating range just below Tg seems best from the vibration damping book by B.J. Lazon as cited in Slocum's book.

The literature also shows that best Tg and modulus are not necessarily stoichiometric for reactive hardeners. E.G. for the hardener DETA (Di Ethylene Tri Amine), best results are obtained for slightly less(about 8 pph) than the 10 pph that the chemistry would suggest. This demonstrates the need for trials. It is also a sign that we might be better off with one of the catalytic hardener families which do not require exact mix ratios. The pure amines require very accurate and sometimes odd mix ratios.

On the bright side, some of the raw amine hardeners have viscosities in the 10 cps range and could substantially reduce the viscosity when used properly. It looks like some of the catalytic hardeners also have similar advantages.

I'm also getting the impression that we would be better served by adding quite a bit more reactive dilutant to our 37-127.

I don't know if this post is much better than epoxy-babble but if there are any lurking epoxy chemists, input is good. Input on whether we should pursue the best system or one that's safe enough outside the lab is another question.

If anybody has comments on either toxicity or anything else, please feel free to comment. This thread often seems stuffy these days without Larry's jokes and Walter's insights and pictures.

Regards all,
Cameron

[jhudler]

37-606 may not be ideal however, what's been produced so far doesn't suck.

As long as you take the appropriate protective measures, I'm not put off by the toxic nature of epoxy or its adducts. We deal with far worse stuff around the house on a daily basis.
Isophorone Diamine is about on the scale of Gasoline in flammability, but more toxic by skin contact, which is OK as we wear gloves mixing epoxy.
Silane on the other hand is Pyrophoric! That's just scary stuff in undiluted form.
Though I plan on using it.
But heck, who here hasn't played with Potassium permanganate, or tossed a chunk of sodium in the lake! Anyone.... Anyone...

[jhudler]

Oh.. I forgot... define deadly vapors; which chemical?
Is it worse than the chlorine and hydrochloric acid that I keep (in close proximity) for my pool?

[ckelloug]

For the record,

Silane is a gas and bad news as far as handling it. Siloxane is the more correct name for the chemical family we'd be using as a bonding agent and it's really not too bad.

Pure diethylene triamine and other hardeners are less nice than we would hope for.

Regards all,
Cameron

[lgalla]

OK,OK,I will try not to make any jokes or puns.
Jack had a spelling mistake in a previous post.

"As long as you take the appropriate protective measures, I'm not put off by the toxic nature of epoxy or its adducts. We deal with far worse stuff around the house on a daily basis."
Adducts,should be spelt"addicts"If you don't know what you are getting into,take appropriate protective measures such as a condom,er gloves.
In general terms,if the epoxy ships non hazmat,it is not corrosive and one of the safer formula.Generally non hazmat epoxy has low vapor pressure and toxic mainly on skin contact.
Nonyl is also added to make easier ratios possible.Cameron,I remember quite a while ago,the epoxy rep for 307-606 said you could add more nonyl.This would possibly be more brittle or more rigid???
If I can't pronounce the name of the hardener,I assume it is too toxic.
Microspheres are necessary for their ball bearing action.
Hardeners needing nitrogen blankets are not for home use.
Let's stick with[no pun intended]307-606.Wish Reichhold payed me commission.
Larry