[brunog]

Cameron,
great post!

Quizz question: I wish to make a 4" w x 12" h x 36"long beam for a milling gantry, if I add 2 pieces of C1018 FB 1"w x 2"h x 36" long anchored in the E/G beam would it be strong enough not to deflect more than .0001" with beam weight and 100lbs weight, if not what size FB should I use??

Thanks, best regards

Bruno

[ckelloug]

I got information from Epoxytech, they manufacture East system epoxy.

They suggest we use their epoxy resin and hardener 9482 which is actually intended for marble repair and aggregate bonding.

I have attached the specs. Does anyone have any idea what a shore D hardeness represents in compressive strength??

Best regards

Bruno

I don't think they're directly correlatable The better the hardness, the more likely it is to have high compressive strength but the relationship is indirect and no approximations are in my books and I didn't see any on google. Working on your gantry problem. May solver simpler ones first.

[ckelloug]

Bruno,

Assuming I haven't developed incompetence (A big assumption as complicated and wacky as these beams are. . .). . . Here is an answer that isn't to what you asked:

16in.H x 4in.W E/G tube Wall Thickness=1.5 -> 2.25. Delta=.0001

Thicker or thinner walls much outside this range and you either aren't strong enough or extra load from tube weight will cause you to fail the deflection criteria.

I'm off to relearn how to do the computation with multiple materials and work out your fatback/strongback question.

[walter]

Sorry for the delay.

I decided to also test some older/bigger samples and my jig needed to be updated.




The results are in and they are not pretty.

The bad news is that my "no epoxy" samples failed miserably!

There's also a good news: A new contender has emerged...



And when that sample cracked, it stopped my heart for a minute and made me think twice about doing anymore testing



But let's not get ahead of ourselves..It will take more than one post so please bear with me.

Here are some of the pictures:
_

[martinw]

I decided to also test some older/bigger samples and my jig needed to be updated.




:_

Dear Walter,

This is a wild and wonky thought about testing samples.

IMVVHO, if you apply a really concentrated point load to the centre of the span of a beam sample, my guess is that the beam will not fail in bending, but will fail by tension cracking (and/or shear).

Here is an example... worried mothers-in-law often give their daughter's family a handy tool for cutting seatbelts and breaking windshields. The windshield tool is designed to break a brittle material, and the principle of it is not anything to do with bending...

Any thoughts from material testing experts about how to test E/G, E/Q.....?

Best wishes


Martin

[Geof]

.... And when that sample cracked, it stopped my heart for a minute and made me think twice about doing anymore testing ...

You are wearing a face shield and for good measure safety glasses. And possibly a full length apron, leather or heavy canvas.

When you put something under a lot of deformation at a high load you have a lot of stored energy. When the part fails this energy can be converted into small fragments moving fast...in other words; shrapnel.

[ckelloug]

Walter,


You should consider making your samples much much smaller if your mixing technology is up to it. It's safer and likely to yield more data. Geof is right!!!!!!!!! (0.125" thick x 0.5" wide x 2.5" long ) for ASTM D700 flexural test.


The samples that you are working with are big enough to fail at between 1 and 3 thousand pounds of applied force on the clamp with howitzer like shrapnel as your jig is set up. Imagine getting hit with a blast from a waterjet cutter at the 2000psi failure stress on the bottom of the beam.

For data quality purposes , it would be good if you made several smaller test samples of the same size from the same batch and tested all of them reporting the torque value for each rather than just one. With torque values and the thread pitch like Geof and Martin were discussing, somebody on the thread should be able to come up with preliminary values for breaking strength.

For average E/Q ala the nist report, A 12 inch long beam 3x3 will fail at about 3000 pounds force on the clamp. A 12 inch 2x2 will fail at about 1000lbf.

As for martin's comment, we hope this is a flexural test and because this material is weaker in tension, that it will fail in tension. Tension doesn't really matter for the sake of E/Q machines because the design is generally deflection limited by the allowable deflection, not stress limited by material strength. Material strength comes into play in this test however as it is the failure mechanism of this test configuration. On the other hand, martinw, you may have a point that the material could fail in shear. Generally, shear negligible for beams about 10 times as long as square IIRC but it plays a larger role in bigger test specimens like these than smaller ones.

Martin, If I were on your side of the pond, I'd bet a pint of ale that these end up failing in tension but based on your concern, this is my opinion since I don't have a published number anywhere for shear strength.

Is anyone enthusiastic enough to go pull ASTM D700 and find out how to do this test the way the pros do it?

Walter and Martin,

Can you post your epoxy mixing procedures so that DAK3333 can look them over? I suspect that insufficient epoxy coating on the solids is what is causing the lower epoxy samples to fail.

[harryn]

LOL - approaching vice limit, at least for the poor guy that has to crank it.

[walter]

Thanks guys, comments and ideas are always welcomed and appreciated!

* * *

This was my attempt at flexural test, like the one below:




I didn't think my samples were that good so I decided not to bother with normal tests. Just needed some comparative analysis.


I already tested the "high epoxy" idea, it was time to try the "low epoxy" mix.

I am of course aware that Quartz countertop manufacturers are using high dollar machinery, polyester resins, binders, etc but still wanted to try their "5%" approach. I was just curious.

They're using massive presses, vibrators, heaters and the whole process is closer to baking it, and not just mixing, like I thought. The chemicals involved probably play a huge role- they're able to produce solid rock at "5%" mix. My samples were solid but crumbled like a cake when load tested.

There was also a good side: No entrapped air. ZERO.

One picture shows old zeeospheres sample with a lot of entrapped air so compare it for yourself.


As for the numbers, all samples failed at 2.5-7.5 foot pounds on a 20 tpi screw.

Some samples had 4 layers of fiberglass cloth and that didn't help much. It was good for an additional 1 foot pound, that's it.
_

[walter]

`
I should probably add that the black sample showed some major (grain) integrity..

It was mostly large quartz and quite a lot of carbon black so I guess The Nano-Theory works! (credit goes to Larry and Cameron)

Zeeospheres seem to work great in other mixes but fail miserably in this setting. Probably due to low epoxy level.. I recall that the old "high epoxy" and Zeeospheres sample was pretty strong.

(The gray sample is 60/40 zeeospheres/fine quartz)
_

BTW, that last picture in post #1170 shows the old Zeeospheres only sample. That one contained a lot of epoxy. Still, it was a lot stronger than this one.

Could it be the entrapped air? Maybe I should put it back in there..

[martinw]

Dear Walter,

Those samples look pretty good to me.

I wouldn't worry at all about the broken samples in your press.

If you are at the stage at which samples break in tension, you are on the home straight.

I could whack the top of my Powermatic 66 with the sharp edge of a hammer, and the cast-iron would fracture. What would that prove? Well, I suppose it would show that cast iron is pretty bad at taking short sharp knocks.... ( it has, however, been quite useful for a few hundred years).

Stick with it guys...

Best wishes


Martin

[SORCHEROR]

you guys are forgeting one thing,most everyone is using some kind of support inside the mix,like rebar or a steel weldment type sub frame,solid polymerconcretes will break with out it,besides i think your bending past the point of no return anyway

[SORCHEROR]

one more thing,most your mixs look to dry,maybe more appoxy needed

[SORCHEROR]

if your going for less epoxy and compression,have you though of heating it to cure it and possibly help in the bonding with some sort of agent?

[walter]

Thanks Martin, that is a good observation.


Cameron,

0.125" or 0.5" sample breaks in your fingers. Ghetto measurement tools are +-10 lbs - pretty hefty reduction would be needed for that to work.


Sorcheror,

Samples were baked for a couple of hours, post #1084
Bonding agents are still being discussed, post #1097



Thanks for all the comments everyone. I really appreciate your thoughts!



Final thoughts and conclusions will be posted tomorrow.
(And yes, we do have a preview!)




:cheers:

[ckelloug]

Howdy there Sorcheror . Thanks for the comments about mixes, reinforcement and bonding agents. You sound like you know more than your short posts would let on so please stick around and help if you're interested.

We already know that reinforcements are needed for anything longer than 24 inches. We also have rule of mixtures calculations from materials engineering that show that less epoxy should be better up to some point. We're really in need of quantifying what that point is and most importantly why. Can you shed any light?

The testing shown by walter is work on determining the ultimate flexural strength of the material mixtures and thus he's breaking samples to failure.

We aren't crazy. The design loads are 100lbs plus weight of beam and the required deflections are between .001 and .0001 max depending on application. Under these circumstances, we achieve a safety factor of 6 on ultimate flexural strength assuming our material can meet published modulus criteria for run of the mill Epoxy Quartz from NIST.

Finally, thanks to your posts I just had a eureka moment: the epoxy amount will vary according to the surface area of the admixures. This is the skin theory comment from Gupta's paper reference 1 posted by brunog back in post 1092. http://www.cnczone.com/forums/showpo...postcount=1092

I'll be posting more about this as soon as I can make it concrete.

--Cameron

[walter]

Here are the numbers.


- Unsupported length of the sample: 3.70"

- Screw: 20tpi, thread dia 0.62" (0.60" minor) Length: 6.65"

- Less than 1 full turn used in breaking each sample (except for #4)

- Width/Height/Weight:


1 - 2.08" x 1.45" x 1.1001 lb
2 - 1.93" x 1.45" x 1.0847 lb
3 - 2.00" x 1.45" x 1.2566 lb
4 - 2.85" x 1.38" x 1.3316 lb


- Samples 1-3 went at 2.5-7.5 foot pounds

- Sample #4 went at whooping 22.5 foot pounds


Hard to believe but the last sample was "full" of epoxy and had some major air entrapment.

Here's the kicker:

It was the only sample that was repeatedly vibrocompacted and treated with heat gun!


Turned out I had two of those magic samples!

Not believing in my results I decided to go through the heart stopping action again... This time I wore the face shield and my lucky hat..


_

[martinw]

Finally, thanks to your posts I just had a eureka moment: the epoxy amount will vary according to the surface area of the admixures. This is the skin theory comment from Gupta's paper reference 1 posted by brunog back in post 1092. http://www.cnczone.com/forums/showpo...postcount=1092

--Cameron

Dear Cameron,

That article by Gupta that Bruno mentioned is incredibly good. I think it implies (amongst other things) that with small aggregates the huge surface area of the aggregates makes good epoxy wetting very difficult, and this reduces strength. With larger aggregates, wetting becomes less of an issue, but I would guess that it is still vitally important.

It would be really interesting if people noted the order in which they mixed up their batches. My wild guess is that if you add the finest aggregates to the epoxy first, stir it up, then add the second finest, etc etc, you might get the best wetting.

I could be entirely wrong.


Best wishes

Martin

[martinw]

[QUOTE=martinw;294322]That article by Gupta that Bruno mentioned is incredibly good. I think it implies (amongst other things) that with small aggregates the huge surface area of the aggregates makes good epoxy wetting very difficult, and this reduces strength.

I could be entirely wrong.




Martin[QUOTE]


Errrrr, yes, I am entirely wrong!

If I re-read it right, the actual surface areas of the aggregates in both batches are the same.

Sorry,

Best wishes

Martin

[walter]

- 2.95" x 1.38" x 1.5212 lb

- tested with the same setup

- going down hard at 32.5 foot pounds!


Mixing procedures were pretty much the same. Aggregates first, epoxy/hardener second then the whole thing mixed together. Carbon black was mixed with aggregates.


Unbelievable. I've got my winner. This is it.


This is what I'm going to use for my machine.
:cheers:
_