[greybeard]

... and then switches to "poly-granite" later in the brochure, as if the mention of epoxy was an embarrassment.
"Plastic - ugh, nasty, cheap stuff. We use poly-granite, so you've nothing to worry about."

One of my life-long rants I'm afraid.
John

[ckelloug]

greybeard,

Do you have any thoughts or references to suggest on epoxy hardener reaction mechanisms or the lab technique for producing epoxy/hardener adducts?

Zumba,

Good catch on roach's link.

Regards all,

Cameron

[greybeard]

Hi Cameron.
Unfortunately no. My own experience was in thermoplastics. Thermosets were outside my field.

( That was 50 years ago ).

I think I'll go and lie down.
John

[roach]

funny?

[Preston E]

Hello Cameron,

What your asking if fairly easy, but knowing what the end result will be on physical properties is the more chalanging question...

Very basically, one mixes and Amine in excess to the stoichiomitry of the epoxy , typically end capping all of the epoxy reactive sites with at least a diamine. The resulting hopefully still liquids equivalents can be calculated from the beganing precurcers.

Overly simplified, but one would heat the amine slightly, and then under agitation and possibly nitorgen gas...slowly drip feed the epoxy resin, such that the epoxy was totally end caped with amine groups...and you could formulate with residule amine as a secondary reactive group to the aduct.

If you need more details. I'll try checking back in more regularly than once a week like in the past...

My background is Thermoset Polymer Chemsitry, more in polyureas, but we modified many amine adducts for our research and factory production.

Regards,

Preston

[Preston E]

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

Hello Again Cameron,

In order to get to a HDT of 100 Deg C your likely going to need some aromatic diamines or adducts there of...

The Jeffamines are nice products, but the D230 is too fast, and the T403 still leaves you short on HDT...

You might look for something based on DETDA diethlyene toluline diamine, but the down side would but a likely need to heat cure the whole casting afterwords...as I doubt that it would crosslink completely without outside heat.

The up side of aromatic diamines is the long pot life and slow viscosity rise, with the added normal benifit of very little exotherm or shrinkage...

Keep up the great work.

Best Regards,

Preston

[ckelloug]

Preston,

Wow. Thanks for the help. In the adducting what glassware would you reccomend? I was thinking 2000ml Erlynmyer with a 3 hole stopper. One hole with a glass tube connected to a teflon one coming out of a buret with the amine, one for nitrogen purge with a glass tube with a rubber tube to a nitrogen regulator on a tank and one with a glass tube to allow nitrogen to flow out. The erlynmyer would then be set on a magstir hotplate.

One reference I found was using a four necked flask but that seems like a lot of trouble. I took general chemistry in school but not O-chem and O-chem lab so it's harder for me to know the best execution of the best laid plan.

Examining vendor datasheets and reading the posts of roach our contributor at a British E/G company, there seems to be a strong preference for the cycloamine isophorone diamine in curing E/G. The easiest source of IPDA appears to be Degussa Vestamin IPD. Extrapolating from roach, they seem to use it straight so the question becomes will it behave miserably in adduct form?

Reichhold 37-601 is an IPDA adduct but it concerns me about the level of benzyl alcohol. It's my understanding that benzyl alcohol does not crosslink into the structure and weakens the chemical bonds therein. I assume they are using it as a solvent but perhaps also as a cure retarder by affecting reaction kinetics by lowering the probability that active species collide?

Thanks also for the comment about the DETA being too fast. The Huntsman folks suggest Jeffamine D 230 for epoxy concrete in the paper on their hardeners. They also hint at synergism with isophorone diamine.

The most interesting thing that I have realized is that the Reichold 37-606 that we have been using is amino-ethyl piperazine from the catalytic family of hardeners and thus it is not nearly as ratio sensitive or reactive in the specified dosage on the data sheet as most other amines. For the final properties desired, it is among the worst hardeners we could choose from what I can tell.

I don't know about the rest of the readership here but I can deal with a heat cure as I have a 48x36x36 curing oven at my disposal.

For the ovenless, what about using a tertiary amine or titanate catalyst in addition to the aromatic or cycloaliphatic amine? Some of the multifunctional titanates can both lower viscosity of the resin aggregate mixture by orders of magnitude and catalytically homeopolymerize epoxide groups.

To whet the appetite of those following the thread, I had an immensely productive day in the library on Sunday and learned a lot of things from old books that will likely influence the final E/G design. Most confirmed things I said a long time ago as best guesses but some didn't. There were also some models for predicting modulus and strength of the composite.

Regards all,

Cameron

[ckelloug]

I forgot to mention that DETDA looks like a great hardener. It's available from abermarle as ethacure 100 and it also is low toxicity. See www.abermarle.com

--Cameron

[Preston E]

Cameron,

The adducting container is not critical as long as it is not steel...ss or pyrex...3 or 4 hole is fine...

But then we need to discuss the EPA and the rules regarding manufacturing a NEW COMPOUND!!!! IE an new adduct...

Not a happy place or cost...at least in PUBLIC!!!!

IPDA with T403 would be my choise...but fast...

But an aromatic amine would give you the most bang for you buck...so to speak...

And with that you would need HEAT...

In the old days MDA would have have been the best choise..(methaline dianaline) not a nice chemical to work with...a solid at room temperature...it would freeze out at 70 de F...in a formula and then need to be heated to above 160 F to reconstitute...

When one starts to try to fix the prime functions in the reactions with other side reactions...things start going down hill quickly>>>

I spent years studying Kenrich additives...some 80 or so and finally had a couple that worked for very specific work very well...but we are talking about thousands and thousands of hours...for a Great effect...

And yes, stay away from anything with benzyl alcohol that is used and a diluante... and does not add to any strength..

And the cycloaliphatic amines are far too fast for your system...

If you look at Bis F epoxy resins, I believe you will up the HDT a bit with what ever cure your looking at in the end...

Let me sleep on this and reply tomorrow...

Best Regards,

Preston

[Preston E]

I forgot to mention that DETDA looks like a great hardener. It's available from abermarle as ethacure 100 and it also is low toxicity. See www.abermarle.com

--Cameron

Yes My point exactly...

Kindly look further into it's uses as and epoxy curative...

Best regards,

Preston

[jhudler]

Using 37-127/37-606 the pot life is 35-40 Min which is extended somewhat by the high particle density.
So when you're discussing 'fast' pot life.... what is fast?
Personally I think 37-606 is too long however, I don't want excessive exotherm either.

Jack

[Preston E]

Hi Jack, The typical way that pot life is measured is for a 100 gm mass at 70 deg F...start mixing up 5 gallons plus, even filled at anything above 70 deg F and that 35-40 minute pot will shrink dramatically...

The longer the pot life, the lower the exotherm in general.

The higher the exotherm, the higher the HDT in general...for systems that do not require an outside heat source for the cure cycle.

The likely reason for the nonyl phenol in the adduct in the IPDA adduct produce is for viscosity reduction and increase in pot life.

It's all a balancing act when developing polymers for special uses...it's rare for the solution to be easy, as the easy ones are already avaliable and easily found...

I've spent years with several formulations only to have to compromise something in the end, in order to get something to market...

If toxisity were not an issue, MDA would be your obvious choise more than likely in a blend with DETDA and Jeffamine T403...and is still used today, though carefully.

I'll be following and thinking about these problems...

Regards,

Preston

Using 37-127/37-606 the pot life is 35-40 Min which is extended somewhat by the high particle density.
So when you're discussing 'fast' pot life.... what is fast?
Personally I think 37-606 is too long however, I don't want excessive exotherm either.

Jack

[ckelloug]

Jack,

It's interesting that you saw the pot life as 30-40 minutes. I either got the mix ratio wrong, my lab was too cold, uscomposites hardener isn't 37-606 like they say, or something but the time I got for the hardening was approximately never. The neat resin was still tacky overnight and didn't set up hard until oven cured.

Preston,

Thanks again for the DETDA suggestion. My comment about it was after reading all I could find on the web on your suggestion. I don't want to get involved in trying to manufacture adducts commercially. That's Reichhold or somebody else's job. I'd only make adducts to try to prove any theories we develop. There already appears to be a CAS# for IPDA/DGEBA adducts so this might make it easier.

Both DETDA and the Reichhold 37-601 IPDA adduct have relatively low toxicities. If we can live with the benzyl alcohol in the 37-601 (probably just a thinner): 37-601 or DETDA/37-601 look to me to be along the lines of something we can accomplish.

I've spoken to Salvador Monte at Kenrich and I have samples of the four titanates he recommends as coupling agents for our system (I'd give part numbers but the stuff is back at my lab). One of the books I found on Sunday had a long article by Monte on the titantes. I suppose I could talk to him about using titanates as hardener/coupling agents as well as talk to dupont about tyzor. (For the record I also have the two main silanes the dow apps engineer recommended.)

So, for an engineer who works mainly on software, how bad is my chemistry

I'm working on drawing the graphs of the modulus and strength models (for a thread post) that I found described in chapter 6 of <I>Structural Adhesives: Developments in Resins and Primers, Ed. A.J. Kinloch, Elsevier Applied Science Publishers 1986.</I> This book chapter also lists about 25 references to academic papers where the info on particle filled composites came from. There is also some fascinating discussion on the use of siloxanes and titanates in the book suggesting that the use of coupling agents at the fill levels we are trying to achieve is essential.

Regards all,

Cameron

[Preston E]

Cameron, Salvador Monte is a good fellow to know...

Did he recommend Lica 38J ??? If not kindly request that to help with lowering the viscosity of the filled material and helping to keep the particles in suspension...

I worked with this for years after finding these benificial effects...this was not known at the time and may will be the first time it's been made public...but since I owned the company and made the discovery, use it to your advantage...

As to the DETDA, your most welcome...as to adducting any of these or other amines, that in general will lower your HDT by making the amine larger and more flexable, the opposite of what you want I'd believe...

Another family of aromatic diamines lies in the TDA family...Toluenediamines and diravites there of...

Also MBOEA also known as DE-DDM (Methylene bis-o-Ethylaniline) avalialbe from the Hanson Company...I believe Ray Scott is still involved there, and you may say that I recommended you contact him...discuss with Ray what your trying to do, he may recommend something very useful.

Regards,

Preston

Regards,

Preston

[ckelloug]

Preston,

I do believe Lica38J is one of the additives I have. I remember it was the odd one because of the 4 he sent, it was the only one that was flammable (due to Isopropyl alcohol content). Do you mean that you were one of the founders of Kenrich?!

Thanks for all the help. Someone like yourself with actual knowledge on this subject will be a great boon to the community effort on getting a usable E/G formula. With help like yours, we will not only succeed, we will likely make a few commercial vendors mad!

Thanks again Preston. I suppose I need to go back to the Hex dump of Scanning Electron Microscope data files I was supposed to be reverse engineering.

Regards all,

Cameron

[Preston E]

Cameron, No not a founder of Kenrich...they worked closely with me when I was working on developing a sprayable highly filled 100% solids polyurea coating ...spray would build to an unlimited thickness as the dry time to touch was less than 5 seconds. Spraying 1/2 inch of material was very easy at a rate of 2 GPM...sold hundreds of thousands of gallons.

We were using Zeospheres and A50 clay activated with the Lica 38J kept the Zeospheres from settling out and hard packing in the bottom of the 55 gallon closed head drums.

It was actually a combination of the Lica 38J and if memory serves corectly KR44U..the U standing for urethane grade...I'd look at both the KR44 and KR44u in combination with the Lica 38J to reduce viscosity, increase packing density, decrease hard packing, and getting some surface bonding...

Adding A 1100 silane will also like aid in bonding to the surface of the fillers. Be sure that the additives are not reactive with the amine side and put them in the amine side...as you want them reactive when the epoxy components are mixed and curing...

You'd be supprised just how many people would put the A-1100 in the epoxy resin side and then state that it just did not work...no wonder, it's an amine termintate silane and reacted totally with the epoxy resin to such an extent that there were no reactive sites left to bond to anything...when mixing with the amine side.

I'll help out where I can...

Regards,

Preston

[Preston E]

Hi Cameron, I must say that I have followed this tread for months on months...and am very glad that I did....

Best Regards,

Preston

[ckelloug]

As promised, I had some nice results from the library. The book <I>Structural Adhesives: Developments in Resins and Primers, Ed. A.J. Kinloch, Elsevier, New York</I> describes in Chapter 6 by R.J. Young, models for filled resin composites.

The lower bound for the modulus of the composite is described by R.J. Young as being governed by the following equation (composites normally do not exceed the lower bound by far):

Ec=(1+vf/(m/(m-1)-vf^(1/3))*E0

Where Ec is the modulus of the composite, E0 is the modulus of the epoxy matrix, vf is the volume fraction of the reinforcing particles, m is Ep/E0 where Ep is the modulus of the particles.

This equation is somewhat complex but it's effects are illustrated by the attached graph.

The magenta line represents composites with the properties of the 37-127/37-606 mixture we have been studying as measured on my flexural test machine. The x axis of the graph is the volume fraction of particles (vf in the above equation). (Quartz is assumed as the aggregate).

The first thing that that can be gleaned from this graph is the reason that DIY epoxy granite has not traditionally worked well: The mixture is extremely sensitive to the particle packing density. <font color=red>It is so sensitive that the 14% difference between a 74% solids and 88% solids is a factor of two in the rigidity of the material!!!!!!!! </font>

In short it's almost impossible to design an effective aggregate mixture without some work. When the industrial manufacturers say carefully graded aggregate, it's an understatement as near as I can tell. The good news is that we have a model that can get us into the ballpark for the proper mixture. The bad news is that the approximations of Beta used in the models so far are not accurate enough to produce a mixture with a high enough density to be in the ballpark of Accures.

The book noted several interesting effects that also temper our design. Pg 169 shows that as particle sizes get larger, modulus decreases! Yes, that's right, bigger particles make the composite floppier.

Pg 174 discussed a topic near and rear to my heart: Strength also goes down as particle sizes go up. The relationship is said to be of the form:

sigma_f=A+B*D^(-1/2)

This means that the strength is equal to a constant plus a factor that diminishes with respect to 1 over the square root of the particle diameter. A couple thousand posts ago or so I correctly surmised the reduction in strength with large particles effect from first principles so I feel vindicated now :banana: This effect has to do with larger particles serving as defects in the matrix (crack nucleation sites).

With the Reichold 37-127/37-606 combination, we would need to achieve 90% particle density to match Accures whereas 92% is probably the highest particle density practically achievable.

Switching to some other hardener like IPDA (Vestamin IPD from Degussa or Epikure 3300 from Hexion) brings the needed particle percentage down to about 88% according to the model which gives a bit more margin for error.

The article by Young also suggests the beneficial effects of silanes and there is a chapter on coupling agents which suggests the titanates as does Preston a few posts back. This is another post on another day.

In short, we're on the right track but some more work is needed to go from where we are no to world class.

Regards all,

Cameron

[greybeard]

... an old man's fancy lightly turns......

I've just put in a request for the Tony Kinloch book you mentioned, so thanks for that link, it looks very apposite.

The spinning lathe is now waiting for the first test mixes to go into the molds, so your latest posting Cameron came at a time when I'm now again puzzling over particle sizes.

Way back, we discussed a formula based on de L's work for a recipe of equal volume parts of a progression of aggregates, using sieves to delineate the size range within each part, and a simple mathematical relationship between them.
I think I expressed my concern that the within that range the distribution could be slanted any which way, so that two mixes, with the same total volume of each part might contain completely different particle populations.

To cater for that, I tried, unsuccessfully, to produce an air powered separator that would give me both an understanding of what the particle size distribution in dry sand might be, and also a way of fractionating the sizes that I wanted for my own spin casting.

Moving back to the idea of using manufacturers' graded materials, I've looked again at the published figures for zeeospheres, g200 and 800, and now I properly understand the implication of percentile figures being equal to or less than, the size quoted, I can see that my thought of the simple recipe is seriously undermined by the wide range of particles present in "off-the-shelf" graded materials, they being nothing like the limited ranges that de L proposed.



So what now ?
Well, I shall start with single sand/epoxy, as stiff as I can pour into my test molds(they've got to go in via the open end in my set up), followed by a mix of two different sand sizes, based on domestically available sieves, then progress from there.
I just have to hope that the release system works, and I soon have something to send you.
Regards
John

[Steven.ji]

In China, more and more manufactuers apply fluorin plastic (Turcite-B from Shamban is famous in the world) attached on machine base. The fluorin plastic combine advantages of slide guide and hydrostaitc guide, moreover the cost is much lower than linear guide.
My question, is it possible to attach fluorin plastic on machine base made of polymer concrete?