[ckelloug]

Hi again.

I've been in the lab today testing Hexion EPON 813 resin with EPIKURE 3046 hardener. The low viscosity is amazing. I think that this pair is a much better match to the application than Reichhold 37-127 resin and 37-606 hardener.

I made my sample mold out of UHMW Polyethylene but there was a 0.5mm gap at the bottom with a depth of about 15mm and a length of about 50mm. The Reichhold resin from my original resin tests mostly stayed in the mold.

The resin viscosity of this new Hexion stuff was so low that about a third of the resin ran out this gap before the epoxy gelled after a few hours. Room temperature was 30 degrees C. Every drop would have run out if the epoxy hadn't gelled.

I had poured the mixture and gone home. I came back a few hours later to find a small pool of epoxy on the floor that had run out of the mold, wet the paper I protect my lab bench with and ended up on the floor. Seeing this and the fact that the gelled epoxy had sealed the mold, I made a small batch and refilled the empty section of the mold.

It sure is easy to get air bubbles in the stuff. I suppose I'll have to get a sample of the BYK A525 that was mentioned a long time ago.

As an interesting aside, it must be pointed out that the Amine number of a hardener cannot be used to derive the Amine Hydrogen Equivalent Weight.

Regards all,

Cameron

[brunog]

Cameron,
what do you mean by fractions?? is it number of different rock sizes??

best regards

Bruno

[ckelloug]

Hi Bruno,

Yup rock sizes. I've been thinking out loud in the terminology of the model I got from de Larrard's book which is a bit opaque at times. Sorry for the confusion.

So Bruno,

Hows the engineering degree?

Regards all,
Cameron.

[greybeard]

Cameron - you had me very worried for a while yesterday. Thought I had a senior moment, or was posting in my sleep
I think your post #3241 must refer to sigma John, not me.

Happy to be known as gbJohn from now

[ckelloug]

gbJohn,

Unfortunately, my study in Post 3241 was based on your data. I think that your P7 mix has too many fractions (rock size groups) that are too close to one another. Your original mixture had a far greater size range and would be apt to get a much higher density according to the model.

So how'd that dinner party go?

I poured two batches of plain epoxy today and milled a seal gland into my mold so that I wouldn't end up with all the epoxy on the floor and one in the mold. I had to make a special gland cutter to mill the gland so I chucked up (in the lathe) a piece of hardened grade 5 bolt I had sitting on the floor, turned to size and then hand ground it into a D-Bit. The viscosity of the Hexion EP 813/ EK 3046 is like light cooking oil. Molds pretty much have to be watertight or they end up empty

During Batch one I started second guessing myself and bungled the mix ratio. I did an experiment pouring this into a slick on a big piece of cardboard to test surface plate theories. The mix ratio is fairly forgiving in this system in that you will get a solid but it might not have the desired properties when poorly or incorrectly measured.

My first batch Friday had drastically too much hardener due to a math error. It hardened but it is more ductile and much less stiff than more correct batches. It behaves like extra strong taffy.

Batch 2 made it into the mold although I suspect the measurements might be off by a few percent. The sealing gland combined with a piece of teflon valve packing running all of the way around in the gland is holding. I also sprayed the mold with a bit of teflon impregenated wax lubricant I spotted at the hardware store getting the valve packing. I'd get real mold release but there was some superficial adhesion of the last sample to the mold which caused cracking due to thermal stresses in the curing oven and I was hoping this wax stuff would mitigate it since I couldn't order real mold release and get anything done this weekend.

Yesterday's batch with the correct mix ratio ran out of the mold partly until the epoxy gelled. I added more epoxy but the mold glued itself to the bench. I finally was able to free it by whacking the mold (with the sample inside) very hard with a big wrench but this caused a crack which ruined the sample so thoroughly as to get no testable pieces.

It's amazing how fallible a process it is to mix and pour glue. . .

--Cameron

[brunog]

Cameron,
Thanks for the info.

My intensive CAD-CAM training at techschool (they call it CEGEP around here) is going great, I still have a month to go in class, then internship for a month and I will be back to work hopefully in September. I've gone through projects in Catia V5, Autocad, Mechanical Desktop, Inventor, Mastercam and Vericut along with courses in metrology, materials, geometrical and dimensionnal tolerancing, conventional and CNC turning and milling etc ...in the last 14 months and I am enjoying every minute of it.

I have been looking at the thread about every second day it has greatly evolved in the last year or so.

Does anyone know what Walter has been up to lately?


Best regards

Bruno

[greybeard]

Hi Cameron.
The dinner party was a blast, though it ended earlier that usual as the host started to fall asleep. Everyone was very kind, putting it down to my treatment, rather than the alcohol level.
Fairly obvious side effect on the memory though, as you can witness.

On that topic, can you remind me of the derivation of the "beta" value.

Regards
John

[ckelloug]

Beta is the theoretical packing density. It is computed from a formula to correct for container size after taking a measurement of the density of a tube full of material.

It hasn't been discussed much but the derivation for Beta can be found back at <A HREF="http://www.cnczone.com/forums/showpost.php?p=413378&postcount=2871"> Post 2871</A>

Beta can be measured by trivial apparatus. I had considered last week sending you a beta apparatus by making an extra when I got around to making one but all of my machine time this weekend went to fixing my mold and making a couple bridgeport parts.

[greybeard]

As I suspected, this is where the old man is getting confused.
Which does the figure 0.6 refer to ?
Is it to the individual fractions, before mixing all of them together ?
John

[ckelloug]

The Beta=.6 was used for all of the fractions. It's probably a bit low but since I don't have measured data, I pulled a not-crazy number out of the air. . . Beta values pertain to each fraction individually. The idea is that in an infinitely sized container, a fraction would fill (Beta*100) percent of that container at the highest achievable compaction.

The Compressible Packing Model tells us the actual packing density of a mixture of materials with different sizes and Beta values.

--Cameron

[greybeard]

So could I suggest that the actual packing density of the total mixture for some individual recipe, as predicted by the model, might be referred to as Beta-subscript-Zeta, or Bz ?
Might help me, if no one else
John

[ckelloug]

Phi is the actual packing density.

[greybeard]

phi it is then.
John

[greybeard]

Cameron, how would the following mix, P6, shape up ?
2500/1200
1200/600
600/300
300/125
125/75
75/0 microns.
It's the biggest spread I can manage just using my sieves, and limiting myself to 2.5mm as the maximum size in a 12mm cast section.
Could you let me have an optimised quantities list for this ?
Many thanks
John

[ckelloug]

gbJohn,

Here is the simulator output. It looks like this mixture will manage 75% density assuming Beta=0.6 for all the fractions for lack of a measured number. I assumed that each of your fractions was the average of the upper and lower sieves. For example 2500-->1200 micron sieves yields 3700/2=1850 microns average for that fraction.

Columns in the simulator output from left to right are:

ID, K_i, avg size in microns, percentage of solid volume, cumulative percentage of solid volume

5 1.500619 0.000037 0.134719 0.134719
4 1.501823 0.000100 0.115172 0.249891
3 1.497447 0.000212 0.119779 0.369670
2 1.497451 0.000450 0.140424 0.510094
1 1.501823 0.000900 0.186249 0.696343
0 1.498191 0.001850 0.303657 1.000000
Phi=0.755396

So reading the plot above, 30% of the biggest, 18% of the next biggest, 14% , 11%, 11%, 13% going from ID 0, the largest to ID 5, the smallest.

Your mixture looks like an excellent high density mixture from the model's standpoint.

On the bright side, I finally got the hexion epoxy out of the mold without destroying it. It's hands down the stickiest stuff in the universe if you ask me. I'm gonna have to go for real mold release and and although the new seal I added to the sample mold works, I will have to replace the seal each time to keep the epoxy in the mold.

Regards all,

Cameron

[greybeard]

Thanks for the figures, Cameron, I'll try that next, and if my new "rod" mold works ok, that will be my next test samples.

By the way, thanks for considering making a beta tester for me. That would provide another useful piece in the tool chest.
On that value of beta = 0.6, how much difference would it make to phi if beta = 0.7 ?
Regards
John

[ckelloug]

John,

Here's the sim output for Beta=0.7


5 1.501549 0.000037 0.109677 0.109677
4 1.501549 0.000100 0.102547 0.212224
3 1.497410 0.000212 0.114460 0.326684
2 1.497410 0.000450 0.142793 0.469477
1 1.499584 0.000900 0.199573 0.669050
0 1.501549 0.001850 0.330950 1.000000
Phi=0.831250

Ultimate density goes up and the distribution changes slightly

[greybeard]

Interesting. The increase in phi is more than I might have expected.
I've just done a rough check on the betas of the new mix, P6, by weighing a measuring cylinder with the sands tamped down as best I could. They all came between 0.58 and 0.61 so your approximation seems more than adequate.

First rod is spinning now, but my clever idea of using wax melted into each end of the tube didn't quite do the business. Fortunately only a small leak, contained within the paper shield I put up for just that eventuality
John

[ckelloug]

John,

Cool on the beta approximation. Measuring it certainly isn't rocket science. The method described in the book is just a fancy grad cylinder with a weight on top and some vibration.

I understand your pain with the epoxy leaks. I finally milled a sealing gland in my mold and now I need to lay in a supply of teflon packing since getting a reasonable amount of the stuff is tricky. The local home center only had 3 ft and I need about that per sample set.

Best of Luck.

--Cameron

[greybeard]

I've just measured the workroom temperature as 25C, up from my last mixing when it was about 18C. Noticable difference in pourability. V. useful when trying to get the mix down a small hole
John