[ckelloug]

Hello all.

I've been lurking/reading for a few weeks now. I have a hard time believing those who read this thread in a couple of days! (Must not have any children or a job!)

Anyhow, Thank you all for all this great information that are posting and the work you have done to this point! Now I would like to help where I can without stepping on any toes.

From what I've gathered to this point:
1. the thread index is great. Thanks ckelloug
2. The simulations that ckelloug can run on aggregate sizing is an incredible tool, but an "optimal" aggregate formula has been derived. (?)
3. roach has given some really great ideas for custom epoxy formulations, while bloefeld has given us some off the shelf alternatives.
4. You guys are almost there in terms of getting the puzzle pieces to fit together nicely.


So is the best way to help by producing samples to test, theoretical input or working together to make something happen?

I think I understand aggregate. There is no optimal formula that doesn't take into account what is available to use, and some measurements on it I haven't even taken on what I do have. The rule of thumb is that for a density in the high 80% range, 4 orders of magnitude size difference are needed between the biggest and smallest particles. The other observation from the model is that aggregate designs that use equal intrinsic volumes of all the components should tend to work fairly well although they have a greater tendency to segregate under vibration. I'm what you might call an aggregator I can run the model if you have a component list you want to try.

I posted a trial aggregate formulation using several aggregates from Agsco in Chicago but I'd have to search the thread to find it. One question is whether an accurately made version of this formula has bad flow characteristics and can't be molded.

That's a good Question: what to do next.

I'd say that the trouble is in the epoxy right now. Roach suggested Isophorone Diamine as a hardener which seems to work great with the Hexion low viscosity resin I've been working with but is quite brittle at the initial cure. I don't know how long it would take to get a reasonable cure at room temperature but I know that overnight won't do it without an hour at 125C using Hexion 813 resin. I haven't tried the Vestamin IPDA with Reichhold 37-127. I also haven't made a D790 test of this Hexion 813/IPDA formulation due to the first batch deforming during the oven cure to the point I couldn't cut samples. I've solved this in the second batch by using 4 hours at room temp, 4 hours at 70C and 1 hour at 125C but I need to make another batch before I can test it.

bloefeld suggests heated DER331 medium viscosity resin and Huntsman curatives with vacuum infusion which I haven't been able to try yet. I'm worried about having to heat the resin before initial cure as it seems like it might be hard to hold machine tool like tolerances due to the extra shrinkage it might cause. I also need to fix my vacuum pump and small vacuum chamber.

So Chris, if you're used to working with epoxy, dealing with epoxy vendors, and you have a balance, you might try getting some samples and investigating the properties of bloefeld's Dow Epoxy Resin 331 / with Huntsman curatives like T403 and D400.

Another thing that needs to be done is more work with both epoxy and aggregates together. I haven't made any samples with aggregate yet as the work of others convinced me that we started with the wrong epoxy hardener. I have tested 5 samples of each of 4 different formulas with different aggregate sent by jhudler and greybeard. Three of these used the Reichhold 37-127 / 37-606 epoxy which isn't strong or stiff enough while graybeard used some kind of epoxy he got in England. All had flexural strengths around 2ksi and moduli around 200ksi.

Once the epoxy choice settles down, there are tests with epoxy and amine active silane and titanate bonding agents. There also need to be some tests of the performance of hard surfaces made with the Inhance Titanium Carbide particles. bloefeld also suggested Wollastanite spicules as an aggregate which would be very good. Unfortunately, long needle like particles will break the model I have been using to predict density until I add that theory to the model from de Larrard's book.

I provide free ASTM D790 flexural tests to this group for all samples sent to me provided I can saw them to 7 1/2 x 3/8 x 1/2 so if you want to make samples of known components, I'll be glad to test them and post the results.

We all usually work together here so if you would like to contribute: theory, samples, and collaboration are all things that help. I personally also would gladly hand over the job of indexing the thread to somebody else as I haven't managed to work on it lately. (I have written some software to link referenced posts in the index provided that the index entries are kept in my special text format).

For what it's worth....I'm in the Madison Wisconsin area.
I have access to some pretty good sized CNC's as well as a few walk in ovens. I also have 2 vac pumps for composite type stuff as well as an array of bagging materials.

I'm not saying this to come off as bragging, but I would like to help, and I feel that if I use E/G for something and don't contribute, I feel like a thief after everything that has already been done.

-Chris

Edit: While I am very eager and capable of helping, I suffer the same problem as many of you out there; a job to pay the bills, a 2 year old and a 6 month old. As you can see, I have been very efficient with some aspects of my life, but I am finding that sleep is a great thing.

Glad to have you aboard Chris and hope you get some sleep. I find that I lie awake wondering about epoxy molecules

Regards all,

Cameron

[crr]

Thanks Sandi and Cameron!

This has to be one of the friendliest threads I've ever read.
Thanks for the direction. I am very familiar with working with epoxies, and am familiar with all of the places you have listed (Huntsman, Shopmaninc.....)

I am more than happy to do some batches for testing. I'll look into the last suggestion of Dow Epoxy Resin 331 / with Huntsman curatives like T403 and D400. Do you think it would be advantageous to also try some with an aggregate that has been vibrated, and infused per bloefeld's idea?

I'm a huge fan of vacuum!

-Chris

PS: Thanks for doing the testing! That is a very generous offer, and I almost feel guilty utilizing it!

[bloefeld]

Hi,

Welcome aboard Chris. Walk-in ovens, are going to be a great addition to the process, nothing beats a good post-cure heat.

From CK's last post it seems that there are concerns with brittleness in some of the formulations. I suggest that reading the Huntsman document I posted should point toward some good solutions. The shrinkage factor as a result of heating the resin is nil.

There are many options available to add toughness to the resin pack. I suggest adding 10 to 20 percent Inhance HD 1800 particles to the resin. They will both toughen the mixture, improve adhesion to the aggregate, and increase the bulk properties of the resin. Obviously reactive rubberisers are available and require some simple calculations to adjust the stoichiometry. I prefer to keep my resin systems as neat as possible and alter their characteristics with particles.

Because you have an oven available, I would suggest that the matrix be cast and then cured at around 200F for about 2 hours. This will reduce the likelihood of cracks forming in the resin. Then follow one of the post-cure schedules laid out by Huntsman.

The main idea behind my preference for systems like Huntsman curatives and accelerators is that they have been very well investigated and are zero or very low VOC systems.

Cheers,

Tony

[ckelloug]

Here's what I've learned about IPDA today:

Abstract:
IPDA reaction mechanisms have a strong temperature dependence. The elevated cure temperatures required along with the temperature dependence (not to mention the toxicity and corrosive nature) mean that IPDA won't make a good hardener for the average reader here even though it can perform very well under laboratory conditions.

Diatribe:
I have been frustrated in my testing of isophorone diamine as a hardener as it seems to behave a bit differently every time. I finally went to the library and pulled a research paper: Journal of Applied Polymer Sciences 106:4169-4173, 2007. Scary that this paper has been published since we started working on this here on CNCZone.

There are apparently two reaction mechanisms in the cure of epoxy with IPDA and which one is active depends on the temperature. It's also true that if the temperature is too high or low that the epoxy will not cure as completely as it is going to.

The paper presented the results of a differential scanning calorimetry study using IPDA and DER 332. The net result was that the ideal curing temperature for maximum cross linkage of 95% is 85C. 90C or 80C are about 4% less. Furthermore, extrapolating from the paper's data, I expect that room temperature isn't going to be good for much more than about a 20% cure unless copious quantities of accelerators are used.

If we have better luck with the Jeffamines, I still think we may be better off with the low viscosity epoxies we have been using.

Aside:
As an aside for blofeld, one of the reasons that elevated temperature cure isn't ideal for this application is that it will heat the fillers and cause thermal expansion. This expansion will cause the epoxy to shrink from mold surfaces when the part cools.

One of the principles described in Precision Machine Design by Alexander Slocum is that of a replicant: An epoxy replicant is able to be cast to dimensions conforming to the mold and being accurate to ten thousandths of an inch. Not all of the castings folks want to make need the replicant property but if you wanted to cast a machine way that was truly flat using either a surface plate as the mold face, weird thermally induced stress and shrinkage from a high temperature initial cure could be bad. A post cure at high temperatures is probably okay but curing the entire thing at high temperature isn't necessarily good.

Regards all,

Cameron

[bloefeld]

Heating the epoxy will reduce its viscosity. It will flow better and wet-out the aggregate better. Using Huntsman Jeffamine curatives gives a great many options as to cure rates and peak exotherms. Adding 1 or 10 parts per hundred by weight of Accelerator 399 changes the cure rate a great deal. But it isn't a necessary additive if a slow cure can be tolerated.

I would formulate as follows; all ratio's by weight.

100 Bis-phenol A epoxy epoxide weight 185 to 190. Dow DER 331 is one of many possible examples.
32 parts Huntsman Jeffamine D-230
1 part Huntsman Accelerator 399
Add 5% by volume Inhance HD 1800 after resin pack is mixed.

Prepare the aggregate and apply vacuum and vibration to consolidate.

Heat the epoxy to reduce its viscosity, it should pour like motor oil.

Pump the epoxy into the mold while drawing with vacuum.

Allow epoxy to cure for 2 to 4 days and demold.

Then apply a post cure heat. Put the part in the oven and increase the temperature by about 10 degrees C per hour and hold at 125 C for four hours. This will reduce internal stresses, complete the cure and cross-link of the epoxy and increase the glass transition temperature of the matrix.

Expect shrinkage of .0005" per inch.

Without the fillers this epoxy is nearly as strong as the published tensile, compressive and flex modulus of the published properties of many of the commercial casting companies.

The exotherm and the post cure will have I predict little or no effect on the after post cure dimensions of the casting.

Cheers,

Bloefeld

[crr]

Hi Everyone!

I am going to pursue Tony's recommendation from his last post.

Currently I have on order:
The huntsman D-230 I don't mind about the slow cure, so I didn't order the accelerator
Inhance HD 1800
D.E.R 331 from a local supplier


I called the Cary company, and they said they cannot get G800 zeeospheres anymore. They are supposed to call me back with an alternate product, so I will let everyone know how that turns out.

Is the best place to get the aggregate still Agsco or US Silica?

My plan is to do the vibratory compaction under vacuum, then infuse, as I already have the tools to do so.

I was wondering if some samples with 1/4" chopped glass fibers would be interesting to see? I remember someone before mentioning it, but I'm not sure if it was actually done.

I can get some chopped carbon, but I already have the glass on hand. Is there anything I'm missing, or something else that might be more beneficial?

-Chris

[lgalla]

Chris be careful.
Sorry I have not posted in a long time and may not be up to date.Basically glass fiber carbon fiber and the like as cloth or rovings under vacuum infusion plus injection is cool as the fiber/resin ratio is about 50/50 by weight.Trying to get 90/10 epoxy aggregate is not possible with a small % age of glass fiber no matter how much vacuum or injection pressure you have.The space between the aggregates is minimal and the glass fiber will complicate the procedure.The fibers will "wash" at the injection point plugging the resin path and building high pressure in the mold.If you are infusing with a high pressure pump,good luck.The mold must be very strong to resist the pressure buildup.Forget the glass fiber for now.
Larry

[crr]

Larry-

Thanks for the input. The carbon/glass that I would be using is not in cloth form, but rather, 1/4" loose chopped fibers. For the test samples that I am going to do, I am going to make a split mold, with the resin injection points at either end and the center on the bottom. I am then going to put the whole thing in a vac bag and compress and vibrate simultaneously. I'm going to try and get as many of the sizes as close as possible to save the effort of cutting too many edges on the wetsaw. The first trial that I am going to do, I am going to see if I can get the resin in purely under vac after the resin has been heated.

On a side note, I have all of the supplies en route, so I look to be contributing more fruitful information in probably 2 weeks time.

-Chris

[Calthorp]

I think the loose fibers would be good as it will act as large aggregate.
Going to cast a sample today if I can get my vaccum pump to go.

[Low-Alloy]

Bloefeld wrote:

"Pump the epoxy into the mold while drawing with vacuum."

Instead of using a gear pump to pump the heated epoxy, would a paint pot at around 50psi do the job? It would certainly be easier rig and clean up would be a lot simpler as well..

I was also thinking that heating the aggregate to around 140f pre-infusion shouldn't be too hard; and if the casting was massive enough it would stay warm long enough to pump the epoxy and keep it's viscosity thin during the entire process.

Possible problems I see with the procedure is that it may cause loss of accuracy as the different aggregates cool their coefficients of expansions may stress the casting out of shape. Also the heated aggregate may cause cause too fast of a cure of the epoxy.

However, I like the idea of heating the epoxy to lower its viscosity if it allows for a superior casting, but the aggregate would have to be heated as well, because I doubt that it would stay warm after infusing though more than just a few pounds at room temperature.

What do you all think?

[probinson]

My thoughts on......

Using a paint pot:

You can use air pressure to drive the epoxy out of a sealed container (I use a pressure cooker with the safety plug filled myself) but you will be using MUCH lower pressure than if you pumped the epoxy directly, approximately 100 PSI instead of a few thousand PSI.

Thermally induced stress:

One of the most attractive properties of the aggregate is low thermal expansion so you don't need to be worried about that. If the aggregate expanded significantly with such a low temperature differential it would be worthless as a machine base to begin with.

[crr]

This is what I was going to use for the test articles. This way I can keep it cheap until I'm ready to do something big.

If you don't want to follow the link, they are basically refillable caulking tubes. I figured I could put the mix in, de-gas, and heat it up in a easily transportable and workable manner. Considering how thick caulk is, heated epoxy should be a piece of cake.....or pie.


http://www.westmarine.com/webapp/wcs...4&classNum=165

[probinson]

Sounds perfect a a first trial. Should be more than sufficient to give insight into what it will take to make thing work.

[Low-Alloy]

My thoughts on......

Using a paint pot:

You can use air pressure to drive the epoxy out of a sealed container (I use a pressure cooker with the safety plug filled myself) but you will be using MUCH lower pressure than if you pumped the epoxy directly, approximately 100 PSI instead of a few thousand PSI.

I was hoping that heated, low viscosity epoxy would get around that high of pressure. Even at 50psi it would take a pretty stout mold to hold shape. If it takes say 100psi to pump through the aggregate the mold walls would be reacting the load pretty early on during the infusion I would think. It would'nt be a matter only seeing the load when the epoxy/aggregate is 100%.

What do you say?

[probinson]

Actually, I really don't see how you can get complete air removal in a large mold trying to force the epoxy in to compacted aggregate. My thought is that you need to premix with excess epoxy then vibrate and compress the mixture forcing out excess.

That's not to discourage your attempts, just my untested opinion. I am sure we will all be tickled to death if it works.

[lgalla]

Resin transfer molding and infusion are similar.Usually in RTM a heafty mold is necessary as resin is injected at 1500psi or more.The reinforcements of choice are continuos mat and cloth or rovings.Compared to E/G there is tons of space for the resin to flow.The pressure on the mold is low as long as the resin has somewhere to go.When a closed mold is full it must withstand the full injection pressure.The problem with a good aggregate mix is there is little space for the resin which will cause pressure build up right away.Addition of fiber will cause more packing and local build up of pressure.
Infusion is just vacuum pressure but would be too slow.
RTM and Infusion are great methods for molding parts with fiber,but not to wetout well packed aggregates.
The big boys use vacuum mixing and vibration tables and this is the proper path to follow.
Larry

[ckelloug]

Howdy all,

I wanted to say that I too will try the pressure/vacuum approach out of interest as Tony definitely has some experience in the area that I lack. If Walter were still tracking this however, he'd have a fit because molding under pressure and vacuum in heavily prepared molds will turn out to be the very antithesis of easy: even if the results are great.

Pressure and vacuum may very well be the way to go for volume manufacture of machine parts and I'm also in favor of vacuum degassing. I'm not so sure however that a pressurized resin vacuum infusion process will work well for guys trying to build routers with improvised tooling.

For all the work done on this thread, we still have a ways to go. I just sawed some samples of Hexion 813/IPDA with different cure cycles. Some of the samples appear quite rigid but due to other tests I needed to do on this set of castings, I will only get to take a small amount of ASTM D790 data on them until I make some more castings.

Regards all,

Cameron

[qroger]

I hope I am not repeating. In the current issue of Civil Engineering, there are two places where "waterless concrete" is announced for possible use in poured in place structures on the moon. Peter C. Chen working at NASA proposes to spin cast a reflecting telescope dish using epoxy, carbon nanotubes and lunar dust.

Houssam S. Toutanji, and Richard N. Grugel, report that NASA has also considered using molten sulfur reinforced with glass fibers to cement lunar aggregate together, creating a waterless concrete suitable for construction purposes. The thrust of the article was the effect of the glass fiber on the strength of the concrete.

[snookered1ca]

I just came across an EG machine base on the CNC Cookbook site ( http://www.cnccookbook.com/ Blog 10/12/08) that referenced it to the German CNCZone site. The machine was described as a German Epoxy Granite Milling Machine and is made with R L & G epoxy resin. This looks like the finest example of a DIY EG casting. Is R L & G epoxy available here in North America ? We are looking to try this for the first time and we need to learn as much as possible about this process.

Bruce

[romihs]

Hello Bruce,

Yes, the Germans do some impressive work.

I also experimented a little with the EG mixture closer to the beginning of this thread, and I used the Epoxy the Germans use: L1100 from R&G Epoxies.

Now, I don't know if the R L & G epoxy that you have mentioned is the same as the one I have posted, but I think it is.

I also don't know if this range of epoxies is supplied in North America...

Regards

Sandi