[johnohara]

. . . Nice. Is that a choir?

~John

[stevv05]

Their is no special formula. Common sense. Read the other posts here. A mountain of information. Plan your project and do it then move on to the next phase. Keep it simple and have fun. enjoy

Steve

[ckelloug]

stevv05,

With all due respect, some formulas will have better mechanical properties than others. If all you need to do is fill a base, any E/G formula will work along with concrete or just about anything else. Commercial E/G on the other hand has a stiffness more than twice that of any impromptu E/G material folks on this thread have sent me for testing.

The information I post here is based on 2 years of engineering research I have done on producing a commercial quality E/G material. If you are interested in a structural material for a good sized machine, I think I am fairly close to having the design right. If you aren't interested in a high modulus structural material, use whatever you want.

For just putting something together without caring about the material I can say the following. Perfect epoxy is less important than the aggregate. Use something that is easily available. More aggregate sizes will generally make the stiffness higher. In the absence of serious calculations, equal parts of the different sizes is a start but for a stiffer material you need more of the larger sizes.

For a small desktop sized machine, it won't matter what you use as the forces involved are too small to have the structure move much. Good Luck and Have fun.

Regards,
Cameron

[altaic]

Hi all, I'm back from the dead.

The alluring thing about E/G is that the basic idea is easy to comprehend, yet it gets much more complicated when one attempts to understand how a particle system works in reality, and then it's even more difficult to evolve that understanding to eventually end up with a good product. Beyond coming up with a good formulation based on a well thought out model, Cameron and others here have vastly increased my understanding of material science, especially WRT composites. Kudos!

Actually, that reminds me of a few fluid dynamics videos I saw from the University of Twente few years ago-- Gallery of Fluid Motion. Really cool stuff!

Cameron, you mentioned the piezoelectric effect of the quartz in E/G and it occurred to me that E/G could benefit from conductive additives. I hypothesize that it would allow the quartz to more effectively dissipate the energy from vibration and impact.

Revisiting Inhance Ti particles: Beyond their very high hardness and abrasion resistance (which are both desirable qualities), Titanium Carbide happens to be fairly conductive. And, Ti-9100 is apparently Ti-9000 but pre-treated with a proprietary coupling agent for use with polymers. IIRC, Ti-9000 was ~$35/lb and Ti-9100 was ~$40/lb, with no minimum order size (the Inhance people were really nice and had no problem selling to me as an individual). Anyway, they list the average particle size as 1.2-1.4µ, so I'd guess it could replace some of the 3M G-200 Ceramic Microspheres. Would you kindly try it in your model for me?

Will

[johnohara]

Larry,

To answer your question of what I intend to build with E/G -- I don't know. I do know that I'm moving in the direction discussed in this article:

http://www.wired.com/magazine/2010/0...volution/all/1

I agree that economic recovery will be innovation driven from the ground-up.

~John

[ckelloug]

Hi Will,

Thanks for the kudos. What I've done wouldn't have been possible without everyone who has contributed here.

As for the titanium carbide particles, I am sure they would fill in some of the micron sizes where G200 zeeospheres have a gap. Unfortunately, at the prices you give, I think incorporating them into the matrix like another aggregate probably wouldn't provide much bang for the buck.

I can't model the exact effect they would have on the modulus. In the model, I can only assume they are quartz and that they influence the modulus due to the slight increase in packing density that a few more small particles might provide. I expect the effect would be negligible regardless given the small amount that would end up in the end product.

My take on the inhance Titanium Carbide particles is that they would be better used as a thin surface layer to make a hardened way area on a machine. I haven't seen this done and I don't know whether it would work but I suspect it would with a bit of finesse.

Regards all,
Cameron

[veteq]

Hello people,

Very nice to see the project going and going!!!
Read a lot but not all in this thread, been reading a lot on the German forum to. Very nice to see the EG is engineerd better and better.

Had a discussion on school eith a proffessor and talked a lot about EG, to much to remember..... One point had my attention, does the material deform as time goes by.

Milling machines are such precise machines that a little displacement over time would damage the precise machine.

Could the experts overhere tell me some more about it??

Kind regards,

Roy

[altaic]

Hey Cameron,

I think I remember discussing the Inhance particles with you earlier in this thread. A small quantity may go a long way as far as abrasion resistance goes-- Inhance claims the particles have 58x the abrasion resistance of steel. However, I definitely agree that it's more suitable to a thin surface layer due to the higher cost.

Actually to that end I've been kicking an idea around for awhile... Basically, one would mix the particles desired for the surface layer (in this case Inhance Ti-9100) with epoxy powder which is used for powder coating, with a 1:1 mix ratio. Then powder coat the casting surface of the mold (which already has mold release on it) with a liberal amount of the mix, and bake for a partial cure to allow the epoxy to flow out among the particles (this will keep it from mixing into the bulk material). Finally, fill the coated mold with the bulk material (in this case E/G), and fully cure. I'd appreciate any thoughts you have regarding the chemical compatibility of the epoxies and partial cure of the coating bonding with the fresh bulk epoxy, as I don't know if such things should work just fine or if there may be complications.

Another idea I think I mentioned before was to coat the inside of the mold with mold release, followed by the Inhance particles, such that the particles get stuck to the mold release as the release cures. And then there's always the possibility of instead using a temporary adhesive on the inside of the mold (after applying and curing the mold release), followed by the Inhance particles. I find these two methods less appealing since the coating could only be one micron thick.

I'm kind of curious to try the Ti-9100 mixed in the E/G for a sample, though. If I were to use them instead of the G-200 particles, should I use the same amount by volume? Also, I wanted to verify that your volumetric formula uses specific gravity (also called true density, I think) as opposed to bulk density; please correct me if I'm mistaken.

Will

[altaic]

Hi Roy,

I haven't seen creep specifications for the epoxies I use, so for ballpark figures someone else on here may do a better job of answering. However, I recently read an article entitled "Creep behavior of an epoxy resin cured under hydrostatic pressure" by Jong Keun Lee and K. D. Pae which described increased creep resistance of epoxy cured at high pressure. The creep tests in the article were conducted at 144°C, and creep increases with respect to temperature (up to the glass transition temperature where the epoxy changes state). This leads me to believe that creep would be a non-issue at room temperature, however the type of epoxy and the cure environment & schedule are certain to have a huge impact. I also suspect that the particles in E/G would help resist creep, however again I have no idea how to model or even estimate that.

Will

[veteq]

Hey Will,

Most off the creep we read about is the one that occours when the epoxy goes from liquid to solid, What i`m interested in is: when the epoxy is solid and thus the machine base is ready for use...

Thus the creep continues at a very slow rate???

Pretty important for precision machines

Regards,

Roy

[altaic]

Hi Roy,

We're talking about the same thing

I was just saying that creep is more severe when you raise your already cured/solid epoxy to higher temperatures, especially nearing the glass transition temperature. The article was about using high pressure during the cure to make the cured epoxy less prone to creep, but it also talked about creep in relation to temperature.

Since most precision machines are kept at room temperature, I got the impression that creep wasn't a problem. However, it's true that for precision machines a just little distortion can compromise the machine. I still think that the high packing density (92%) of the particles in Cameron's formula would help resist creep, but as I said I don't know how much. I'd better wait for someone else with more specific knowledge to chime in here so I don't mislead you

Will

[veteq]

Hey Will,

Nice to read we are thinking about the same thing!!
Just back from a long day at the university and had a 2 hour class
in plastic`s etc. Funny thing is that the subject was about the thing we are speaking about. But then about plastic`s in general.

The teacher was talking about 2 things, creep
and (dont know the englisch name...?) imagine:
A force on a block epoxy in glass state, it will deform a bit and as long as the force is on it, the epoxy will give a reactive force (elasticity).

Both cann change in time, what i understood off it.

Funny thing is he said the same as you, the particles will make this effects less over time.

He had the example off a gasket on a car engine, over time it loses its gas closing ability (opposite off creep, through force)), then the engineers thought about it and added some reinforced fibers to the gasket.
Problem solved, time till the effect gets negative is extended in time.

Talked about this in relation to our milling machines after class, he said that the change off deformation is always there. couldn`t say anything about the amount off deforming.

For me the design off a expensive machine with full epoxy is to dangerous, 5K for linear components and a hope for a 15 year use off machine, deforming would be terrible. Going to try in the future with aluminium base as backbone with slots filled with EG.

Maybe somebody cann say a bit more about it, the subject needs to be clear for me before going back to full EG design.

Kind regards,

Roy

[veteq]

Also, when nearing the glass transition temp. the material gets softer, thus less ressience against bending

[johnohara]

veteq,

What kind of downward force or environmental temperature are planning to expose your desktop mill to?

~John

[RotarySMP]

I think there are too many theoretical assumptions in your post Roy.

If that machine will be your first ever use of E/G, then the performance after 15 years or creep at elevated temps are not very relevant criteria.

You will have a learning curve, and the first machine you build might be utter crap (It looks cool in your rendering though).

Just design it so the $5000 worth of parts are only bolted on, so you can salvage them for use on the second, much improved version.

The price of epoxy and fill material you can write off against personal learning. It will be insignificant in the big picture.

At this stage much more relevant are things like what will I make the mold from, how much draft do I need, which filler mix and which epoxy, how do I mount it to what sort of vibration table, how do I machine interface points after it cures etc.

[altaic]

Hey Roy, Nice machine design! I like it.

The problem with aluminum for precision machines is that it's thermal coefficient of linear expansion (TCLE) is high-- something like 23 * 10^-6 /°C. Steel is around 12 * 10^-6/°C and iron is 11 * 10^-6/°C. I've calculated E/G (as per the current formula in the thread) to be 7.0 * 10^-6/°C, and E/Q (using #6 quartz instead of #6 alumina) to be 5.2 * 10^-6/°C.

So if the column of a mill is aluminum, say 20cm x 20cm x 1m, just 1°C will expand it by 0.0046mm x 0.0046mm x 0.023mm. And that's assuming the metal is consistent; if not, that expansion will cause irregular distortion. Furthermore, aluminum distorts as it ages, which is accelerated with fatiguing vibrations.

E/G with aluminum structuring is also problematic. Using the high packing density formula here, the rate of expansion is very different which can cause the E/G to delaminate from the aluminum. Additionally, the stresses of the different expansion rates can cause distortion and create fatigue points in the aluminum. You can adjust the E/G formula to match aluminum's TCLE by increasing the amount of epoxy greatly (40%), but the strength will be greatly reduced.

It would be really nice to know what the rate of creep in E/G actually is, so we can estimate the change of accuracy for precision machines. Actually, I just found US Patent 4497764, which has a table listing creep for both epoxy and epoxy/silica (83% aggregate, 17% epoxy)! Since E/G is 92% aggregate and 8% epoxy, it looks like E/G should have less than 1/10 the amount of creep as pure epoxy. I haven't really made sense of the actual numbers and units yet, though... But, those tests were at 74°C, so I think creep should be much less than those numbers at room temperature.

Will

[johnohara]

Quoting said patent:

3. The method according to claim 2 wherein there is 1 parts by weight of said siliceous aggregate per part by weight of said combined weight.

4. The method according to claim 1 wherein said epoxy composition comprises

(a) from 10 to 20 weight percent of said epoxy resin diluted with about 0.5 part by weight per part by weight of said resin of the diglycidyl ether of cyclohexane dimethanol and at least an effective amount of triethylene tetraamine to cure said epoxy resin at about room temperature; and,

(b) from 80 to 90 part by weight of a siliceous aggregate having a particulate size ranging from 8 to 400 mesh with from 20 to 30 weight percent of siliceous aggregate having a particulate size ranging from 100 to 400 mesh, said weight percent being based on the total weight of said composition.

The important point to remember here is that the aggregate percentages are based upon all the aggregate being "siliceous."

100 to 400 mesh is silica powder. Be careful, wear a mask.

Thanks for the link.

~John

[altaic]

Hey John, I'm not planning to use the formula in the patent, but thanks for the concern.

I was just pointing out the table of creep measurements for various pure epoxy and epoxy/silica. The 83% silica and 17% epoxy figure was from Example 1 as referenced by the table in the patent (IIRC it said Example 1 was composed of 5 parts silica mix to 1 part epoxy by weight). I'd assume that 92% alumina/quartz and 8% epoxy would have less creep by virtue of having less epoxy to creep, but I could be wrong. Just shooting from the hip, here.

Will

[veteq]

I think there are too many theoretical assumptions in your post Roy.

If that machine will be your first ever use of E/G, then the performance after 15 years or creep at elevated temps are not very relevant criteria.

You will have a learning curve, and the first machine you build might be utter crap (It looks cool in your rendering though).

Hey Mark,

I`m a believer that everything cann be calculated in the proces of designing a mill, been designing a machine for the last 5 years and have calculated everything to make it at least a quality machine as the Minitech.

Understand where you coming from, but i really cann say i dont have any doubt the machine will perform motion like i designed it to. Made all dynamic calculations and EEM`s too. For the software i`ve been using Solidworks and Solidcam and take classes for it the last 6 years, dont expect a problem there. The only problem i see is that the EG base could deform over time, sorry, but only calculations will make me change my mind about the creep subject.
All subsystems designs are tought off and made 100+ designs of it.

Dont want to be a smartass student, but could you tell me the facts about my assumptions? You only cann tell it are assumptions if you know the facts, please share and take my creep problem away.

Kind regards,

Roy B

[veteq]

It would be really nice to know what the rate of creep in E/G actually is,
Will

Hey Will,
Like the discussion, so all the idea`s are put together for a better understanding.

Imagine i put the mill in my room where it is the hole year the same temp.
Alu creeps natural because it deforms by age, doesn`t Epoxy has the same proces only faster. There is also a cnczone.nl forum in my country and we are talking about the same subject, there are several service mechanics that have seen in there life problems with proffesionel EG machines that deform over age. Think the designers put a lot more effort in making the right mixture etc. All stand in temp regulated rooms. So there has to be other parameters that make it creep, they also talk about moiture in the aggreget or moisture that enters the machine base in time.
Dont know for sure, but maybe creep isn`t the right name in Englisch?

Kind regards,

roy