[brunog]

The math model that I developed from de Larrard's book embodies a very similar principle to that idea with the difference being that de Larrard spent a decade turning logic like that into math.

Cameron,

Would it be possible to makeout an Excel spreadsheet of the math model and share it, so we can experiment with it?

I know I would be very interested in doing further experimentations in packing densities.

Best regards,

Bruno

[brunog]

Sorry, double post

[ckelloug]

Hi Jonpry,

Good job on the Finite Element Analysis. A lot of the things I have read suggest that doing a finite element analysis is quite helpful to use E/G to its best effect. Your 20GPa modulus estimate is somewhere between the modulus of the materials that folks have sent me to test (about 14GPa) and performance of commercial quartz based materials at about 31GPa. I currently estimate that my own aggregate design will achieve between 27GPa and 34GPa.

Hi Bruno,

I am sorry to report that the model code is not very portable. It's currently implemented as about 1000 lines of C code using linux interprocess communications. This includes the model itself and the custom simulated annealing optimizer I developed to solve it. The model needs a detailed picture of the aggregate size distributions to be particularly useful which is why I have used the well characterized agsco materials in my reference design.

If you can collect up the data on the aggregate you would like to use, I am happy to make a few model runs for you but since I don't use Excel, I can't convert it to a spreadsheet.

Regards all,
Cameron

[johnohara]

I have a new piece to post and can't upload images. Even images that I previously posted are being rejected.

Anyone else having this problem?

~John

[johnohara]

37cm x 5cm x 6cm.



Top and sides are smooth.





Density is getting better. Hand tamped, no vibration. You can see where Zeeospheres would make this stronger.



~John

[johnohara]

1. Newest piece based on Silica Sand.

2. Close-up showing density.

Cast 12 hours ago. Just demolded. Still curing.

[ckelloug]

Jonpry,

That BCl3 amine complex looks interesting. As long as there aren't any problems with hazardous waste, it would definitely be one way to do it. I am encouraged by the fact that the datasheet suggested its use in laboratory countertops. I had not seen something like this. The literature I've read usually talks about BF3 which is somewhat nasty.

Adding a solvent and removing it later with vacuum is certainly technically feasible but I would think on an industrial scale that I wouldn't want to deal with disposing of the solvent, the contaminated vacuum oil etc. To me, reactive dilutants seem like a better choice.

Looking over at the Hanson Company, the Trex tooling resin also looked kind of interesting.

I think you've added to the choices we have on what to try.

Johnohara,

Congratulations on your samples. It looks like you're getting something done.

Regards all,

Cameron

[jonpry]

Cameron,

I didn't notice that the hansonco curative was BCl3 instead of BF3. That does seem much safer. The MSDS doesn't have a lot of info, but it does seem to be fairly nontoxic. They suggest unused chemical be incinerated though. That doesn't really bother me as we have free HazMat disposal here in seattle. I also found a bunch of interesting curatives from Air Products. Like this one, http://www.airproducts.com/epoxyaddi...AMR/H2337U.pdf blood sample possibly required. It has an activation temperature of only 158F, and then cures in like 2 minutes. This is important for me, because I plan to use plastic molds, and the cheap low surface energy plastics just can't stand up to much heat.

While I've never dealt with Air Products in particular, the air companies seem to be pretty easy to deal with in my experience. They are used to the welder that comes in during the day to get work supplies, coming in on the weekends to get gas for their home projects.

Trex tooling resin does look interesting. There seems to be an awful lot of options for doing latent cure epoxy. It may even be the industry norm. Nothing in the ITW video really precludes a bake. In fact everyone mentions a bake to get good cure properties. I don't know why we all assumed there was any level of curing before the bake.

[ckelloug]

Hi Jonopry,

I took a look at the air products Ancamine 2337S that you linked to. It is interesting but I wonder what kind of modulus you can get with it. It's probably not a problem but some strange hardeners don't make a very stiff epoxy.

In regards to the cheap low surface energy plastics, I've made my first sample mold with UHMW. Unfortunately, it distorted in my curing oven. I think the temp was about 200F. This allowed the epoxy to leak out onto the oven floor making for a somewhat annoying mess. Furthermore, this mold warped so much in the process that it couldn't be used again. A quick web search found UHMW has a coefficient of thermal expansion of 9.1*10^-5 inches/inch degree F which will make a high temperature cure give a distorted part.

I've come grudgingly to the conclusion that steel will be required for precision parts. Plastics and wood might be sufficient for some parts depending on the precision required but if you need to cast a part that is flat and on size to a few ten-thousandths of an inch when demolded, steel appears to be the only option.

Baking is a good thing but some hardeners attain sufficient properties without it albeit slowly. My understanding is that IPDA for example will be completely hard after about a month at room temperature. A few hours at high temperature preferably ramped will also cure it completely and to a higher modulus and heat distortion temperature. There's a balance between curing temperature and acceptable distortion and I don't know what it is at this point.

Regards all,
Cameron

[altaic]

Hey guys,
Going by the epoxy I have (Pro-Set 117LV/229), it's supposed to room temperature cure for 15 hours before an 8-16 hour elevated temperature post cure. The nice thing about that is you can carefully demold the part after curing at room temp and then post cure without the mold.

I think Scott Bader (sink manufacturer) does it this way, going from their youtube video (although, I don't know if they cure at room temperature for longer for greater stiffness before demolding, or if the epoxy manufacturer's instructions work sufficiently well):
[ame="http://www.youtube.com/watch?v=vrgbCrhkJTI"]YouTube- scott bader epoxy granite[/ame]

Actually, I'm kind of curious how curing for longer at room temperature before post-cure affects strength and modulus. For instance, if I cured something for 24 hours at room temp before post curing and 110°F for 16 hours, would I get the same or better strength and modulus as the recommended 15 hour room temp cure and 16 hour post cure at 110°F? Something to ask Pro-Set, for sure, but maybe one of you epoxy wizards can shed some light on curing schedules.

Will

[jonpry]

Cameron,

I am curious about the cured properties as well. There seems to be very little data on that. I was also unable to find out what kind of amine it is exactly to try and draw comparisons with other curatives. My limited understanding of epoxy chemistry suggests that it is possible to mix Ancamine 2337S with other things, such as cross linking agents, which effect the modulus to a large extent. It may also be possible to get an epoxy that cures enough with 2337S where you can remove it from the mold and cure it with higher temperature, possibly by activating another higher temperature curing agent.

Polyethylene is probably a no go for molds. Polypropylene can handle much higher temperatures. Even 158 is pushing it for polyethylene. UHMW while having some interesting mechanical properties, is largely useless imho. It exhibits creep on a time scale of seconds under little or no load, and machining it is similar to spinning it into fibers. Most of the literature seems to suggest that UHMW could have exceedingly high flexural strength. But the pieces I have seen make pvc/acrylic look like engineering plastic.

I am considering making a plastic/wood or plastic/metal composite for the mold. There are still limits on the temperature, but strength at temperature will be far less of an issue. I can't really machine metals to the same tolerances I can plastic, so if there is a way to use plastic, I want to find it.

[altaic]

Hey Cameron,
Something just occurred to me while I was thinking about vibration compaction and spin-casting-- If you don't mind, what are your parameters for your simulated annealing solver? We may be able to use them as a guideline for our compaction methods.

Will

[RomanLini]

Hey guys,
Going by the epoxy I have (Pro-Set 117LV/229), it's supposed to room temperature cure for 15 hours before an 8-16 hour elevated temperature post cure.
...
Actually, I'm kind of curious how curing for longer at room temperature before post-cure affects strength and modulus. For instance, if I cured something for 24 hours at room temp before post curing and 110°F for 16 hours, would I get the same or better strength and modulus as the recommended 15 hour room temp cure and 16 hour post cure at 110°F?
...

I try not to cure anything at "room temp", in my opinion that's quite unprofessional. Once the thermoset has finished we do an initial cure for some days (generally up to 5 days) at 45'C then do the elevated temperature post cure as per the epoxy datasheet for about 12 hours.

From my experience the longer the initial cure, the better the overall result.

[ckelloug]

Will,

I'd worry very much about curing the material outside the mold. I have an epoxy blob on several oven racks due to this expedient. . . I started by curing Hexion 813 with IPDA and got a clear hard block. I put this in the curing oven on a rack for a few hours and when I got back the "hard" blob had partially oozed through the oven rack in a manner not unlike a steel wire cutting ice. The part was completely ruined. This effect will be much smaller when aggregate is added but this experience very much affected my way of looking at curing outside of the mold.

I'm far from an epoxy expert but this is what I have learned: Post curing epoxy involves heating it above the glass transition and allowing the material to further react in turn raising the glass transition temperature. The material is no longer rigid when heated to the current glass transition temperature so there will be serious problems maintaining tolerances in post cure without a mold.

It's hard to generalize about the right temperatures and cure times. The correct approach is to take the proposed hardener and epoxy and run them in a differential scanning calorimeter (DSC). Lacking a DSC, I try to get cure cycles from academic literature. The epoxy vendors would probably do this for us if there was a sign of it helping to sell more material but I'm not ready to approach them yet.

My research and experience with IPDA show that the highest stiffness and glass transition come from curing and post curing at specific temperatures from a DSC experiment.

I'll look at the simulated annealing solver and see if there is anything useful I can explain about it. I just hacked together something that worked. The real issue was developing and debugging the model from de Larrard's text. There might be an even better model somewhere but I have not found it.

Johnopry,

I agree that UHMW is not really helpful for molds.

Romanlini,

It has been my experience so far that room temperature curing epoxies that are slow enough to be useful are quite rare if they exist. There is a tradeoff between what you use and the cure cycle required. I got the impression from poster roach in the UK that they use IPDA and complete room temperature cure at the company they work for. This is a much simpler problem for small parts than large ones.

Regards all,

Cameron

[altaic]

Hi Cameron,
The Pro-Set epoxy data sheet lists both a 2 week room temperature cure and a 15 hour room temp cure plus 110°F 16 hour post cure, with glass transition temperature at 121°F and 122°F, respectively... Although I guess it makes sense that the listed glass transition temperature is lower earlier on in the cure. Was I mistaken about Scott Bader's post curing without a mold (I thought they said it near the end of the video)?

As far as 100% room temperature cure goes, I agree that the strength and stiffness is lower, but OTOH the distortion due to heating the mold is not present. For precision surfaces, that's a really attractive point to E/G.

Will

[ckelloug]

Will,

I won't disagree with what they said. Your mileage may vary depending on the hardener and the epoxy used and the tolerance required of the part. All I can say is that I've seen a really bad counterexample in pure epoxy.

My result with the prototype E/G mixture didn't move visibly under similar conditions but it was a tiny piece cast in a cup not suitable for quantitative analysis. I don't trust post curing out of the mold at this point for a precision part but of course, not all parts need to be precision parts.

Regards all,

Cameron

[RomanLini]

...

As far as 100% room temperature cure goes, I agree that the strength and stiffness is lower, but OTOH the distortion due to heating the mold is not present. For precision surfaces, that's a really attractive point to E/G.
...

I know that was directed at Cameron, but I'd like to make a point there. "Room temperature" varies a lot, from day to night and depending on weather.

Using a low temperature just above room temp (we use 40'C or 45'C) provides a consistent temp with no mold expansion/contraction, which is the opposite of what you stated in that the result is less mold distortion caused by heat and cool cycles while the epoxy is at different hardnesses during the cure cycle. A constant temp keeps the mold at a constant dimension, and being able to use consistent temperatures is a big part of getting good repeatable cures and good cured properties.

Also curing for 5 days at 40'C or 45'C gives a much more cured part than a week at "room temp" which is probably a lot lower average temp. So you get a harder and more stable product before you do the critical post cure.

It's fairly trivial to build low temperature incubators that will keep a very accurate temperature. I recently designed an open-source LCD temperature controller which you can build very cheaply, and use any simple cabinet for the incubator. It is on my hobby electonics web page here see the SH1-Temp and SH1_Temp2 projects.

I wouldn't even consider working seriously with epoxies without having equipment to control temperatures. Many of the big firms use temperature controlled rooms, so their "room temp" is not like an average person's "room temp".

[jonpry]

...

I'm far from an epoxy expert but this is what I have learned: Post curing epoxy involves heating it above the glass transition and allowing the material to further react in turn raising the glass transition temperature.
...

Not that I doubt the results of your experiment, but I'm pretty sure the above statement is false. I'm calling the temperature required to change the Tg, Tcg. It seem that one can derive Tcg < Tg as a property of all thermosetting polymers. I've constructed a logical proof of this. Strangely enough it involves no chemistry. Proof as follows:

Assume Tcg > Tg. Now make a bunch of epoxy pellets with a known Tg, say 160C. Since it is possible to injection mold a part at Tg. We setup an injection molder at 160C and produce a molded epoxy part. Since Tcg>Tg, our new plastic part will also have a Tg of 160C. So it would be possible to grind it up and make it a new part. Therefore epoxy is a thermoplastic. QED


I think as long as there is unreacted curative. Raising the temperature to anything above the original cure temperature will result in raising the Tg, but how long it takes is anybodies guess. A temperature ramp would be very important, and very difficult to develop.

[lgalla]

"Therefore epoxy is a thermoplastic. QED"
Jonpry
I did not know epoxy was a thermoplastic.news to me.
Larry

[jonpry]

Igalla, I can't tell if your joking :-). But it was a reductio ad absurdum.