[peteeng]

Hi Thomas - by 3 tests I think they mean a sample set ie 3 of the same test piece. This is for statistical coverage. So ask them do they mean 3 identical samples or 3 different materials.... Statistically material tests typically use 5 samples and for some of the work I do they require 6 samples so the reliability is greater. Follow up the other sources before you jump. Peter

Here's some numbers from a laminate test. I get 6 because it costs the same as 5 at the lab I use then the results are more useful with a particular design code...

[peteeng]

Hi Thomas - Heres a video of a test a person did on EG. If you do something like this and make an aluminium coupon you can ratio the deflection to the Al.. Cheers Peter

[badhabit]

Hi Thomas - by 3 tests I think they mean a sample set ie 3 of the same test piece. This is for statistical coverage. So ask them do they mean 3 identical samples or 3 different materials.... Statistically material tests typically use 5 samples and for some of the work I do they require 6 samples so the reliability is greater. Follow up the other sources before you jump. Peter

Mjea, after re-reading our mail correspondance i actually think that is up to me to do the statistics(if i want them)...then i'll have to send more samples... i've asked for clearification just to be sure...

Anyways, it has proven very difficult to obtain the proper casting moulds for a testsample...a cylinder with 100x200mm or a 150x300mm... but split forms for this is not easy for me to find... actually i cannot find any locally and i've extended my search to china and other countries. But i would like the tests done ASAP so i can't wait 3 weeks for the moulds to arrive from China(or other countries for that matter) :-/ I found ONE locally(In Denmark) but they could not ship until April 1st...and by that time i've already cast all the parts, so no good :-)

By the way: Last week 2.000kg of Silimix282 arrived and today i got the 200L of epoxy. I've almost done with assembly of one of the moulds and hopefully i'll get started on machining the parts for the other mould this weekend... I've attached some pictures from midway in the mould assembly....It took quite a long time to model in fusion but it's turning out great... everything fits and works as expected...

Attachment 459174Attachment 459176

[peteeng]

Hi Thomas - I'm sure the supplier of the Silimix will know the expected stiffness of the mix if you ask them. Peter

[badhabit]

Epoxy time!

[the4thseal]

is that a Phenolic plywood? Just curious.

[badhabit]

is that a Phenolic plywood?

yes :-)

[excitedbox]

Another terrible post by peter. That video has some of the most unreliable testing I have ever seen to the point that taking anything but what not to do from it would be a bad thing. Any air = 0 strength so all his dry samples count as things he didn't test. In the end you have 1 test result for 100% that you might be able to rely on and that is one which industry experience has shown is NOT the optimum case.

A 64% fill rate would assume a SINGLE round object in a box but not any real packing physics. I feel bad for anyone who has followed your advice on here because it is wrong on so many levels. Using multiple particle sizes you should be able to get close to a 80%+ fill rate.

Close packing of equal spheres is about 74% so your maximum will equal: the starting Volume - the space filled by all larger particles + (74% of the space filled by smallest particles). The less round and more smooth the higher your packing fraction can be with enough shaking.

Since for any sphere you can add another smaller sphere or in this case broken particle that is smaller to fill the remaining space the theoretical limit is quite high. Especially when not dealing with spheres as largest particle. When dealing with Shoveled or Vibrated cement / concrete mixes of multi sized particles 80-85% is usually what they shoot for. The rest of the volume is taken up by your epoxy or binder (water).

Flat particles have higher fill rate and longitudinal strength while also giving better shear strength. Round smooth particles have lower shear strength than jagged round particles


A real Recipe/G:
I have seen many tests and industry mixes that are 9-13% epoxy by weight and as much large particles (~1/5 or less) of the thinnest feature as possible with a medium grain of 1/5 or less of the larger. Then fines to fill in the rest. The smaller the fines the better the surface finish.

More small particles means more epoxy because of increased surface area. At the minimum it should be wet enough to settle under it's own weight when vibrated. Best is the consistency of that beach sand that seems moist but seeps some water when you slap it repeatedly. Vibrating it is a must to get out air bubbles and it is best to degas the epoxy with one of those cheap pumps. Heating your aggregate AND epoxy helps make everything flow better.

For round particles the theoretic number for the medium and small particles size is 29% the radius of the larger because when you look at a ball that is the space under the curved part.



https://en.wikipedia.org/wiki/Close-..._equal_spheres

[peteeng]

Hello Excited - If you read the posts you will see I have discussed the same numbers as yourself in various places. They are all theoretical and I'm interested in real outcomes. Packing can be anywhere between 0-80% by volume depending on how you do it. There are limits to how small you can go due to the fact that resin can't fill spaces under a certain size. Concrete can achieve very high packing as water is very small and thin so much smaller particles can be used.

You can use 5% by weight epoxy if you wish just means there will be lots of air. The video maybe poor but at least he tried to do something. I fail to understand why surface area has anything to do with resin ratios if you could explain that to me I'd appreciate it. I think you have expressed the same sentiments in other threads...Peter

[Mikeedd]

Badhabit, it looks like a serious tool, you are making, but why do you use 282 and not 271, as 271 is made for resins?
BTW, I am scared of the amounts EG you will use, as I plan a smaller hobby project, but thumbs up, it looks .. big..
Love to see more of the build,,, maybee I can pop by and disturb you :-)
Just kiddin..

Kind regards, Mike

[badhabit]

Badhabit, it looks like a serious tool, you are making, but why do you use 282 and not 271, as 271 is made for resins?

Good question! Actually the 282 was choosen because i've seen a couple of very successfull projects using this stuff. So a year ago i bought a bag of 282 and have been testing it a bit with good success so far. I did try to grade my own stuff using all kinds of locally sourced quartz in a fuller curved mix, but nothing worked as well as the 282. I alsoed toyed around with aluminium oxid for a while, but it made EG even more expensive. So at some point i just had to take a decision and went with the stuff i saw other people having great success with...

BTW, I am scared of the amounts EG you will use, as I plan a smaller hobby project, but thumbs up, it looks .. big..
Love to see more of the build,,, maybee I can pop by and disturb you :-)
Just kiddin..

Yeah, it kinda scares me aswell, but don't tell anyone;-)

People are allways welcome to pop by! I can't see where you live, but in case it isn't in my neighborhood you can pop by virtually: https://www.cnczone.com/forums/verti...268-forum.html

/Thomas

[Mikeedd]

I know some folks use Dansand, as they even makes the mix you specify, but it is very interesting that Silimix is kind of foolproof, I better talk to them to get some more info..
I also like the idea to use metal as aggregate for machineability as PeteEng, but it is not as strong as using quartz.. and the aluoxide/blasting media is way too expensive. so silimix it will be.
Jeg bor syd for Køge :-)

Kind regards Mike..

[badhabit]

I know some folks use Dansand, as they even makes the mix you specify, but it is very interesting that Silimix is kind of foolproof, I better talk to them to get some more info..
Jeg bor syd for Køge :-)

Uh! A somewhat local guy :-) I'm from Aarhus, ping me if you are in the neighborhood and want to see the progress ;-)

Regarding the Silimix you can buy it from a german website, and shipping are included! If you are considering buying more than a bag or two, make sure to contact them and ask for a quote...

[Mikeedd]

SILIMIX Nr. 271 0.1 - 3.0 mm synthetic resin bounded soils from 8 mm layer thickness
SILIMIX Nr. 282 0 - 4.0 mm liquid-compact mortar
I wonder what bounded soils is, for the 271.(are the aggregate coated, as they sell a resin coated colour mix? or is it soils that will bond to resin?)
I also wonder why 282 gives good results with epoxy, as for making uhpc, you use things that binds the chalk when water is added, plus plasticides, not working with epoxy?
It must be the graduation of the 282, that makes it so well for EG, or?

Disclaimer, I am not affiliated with silimix producers, but the price is good, and if this works for most folks, it must be a nobrainer?
The attached file is a more English version of the lathe controller I plan to buy, but it is outoftread, so sorry, I must disclaim this too :-)

Cheers from Mikeeddit a lot :-)

[Mikeedd]

If elements of the mix is watersoluble, but not epoxysoluble, will it damage the mix, used with epoxy, instead of cement and water? I think that was my question, sorry..

Regards MikeEdd to much..

[badhabit]

Hi,

I just got a package in the mail, the forms for casting standard 100*200 testing-samples:
Attachment 460646

I'm going to cast 5 different samples and send it to a national test institute. This will at least give me the compressionstrength and e-modulus for the Silimix282 + R&G resinL/GL2...i'll make one sample without the postcuring and a sample with postcuring. I want to see what difference(if any) it makes to the e-modulus(my guess is none...or close to). I'll also send a sample with larger aggregate mixed in with the Silimix282. Since the largest size is 4mm in the 282, i'll try to add some ~15mm aggregate. This is to determin how it effects e-modulus and if it would be more cost effective in a larger cast, saving on both the Silimix and a little of epoxy...

/Thomas

[peteeng]

Wow Thomas - That's great., how many sleeps till the results? Peter

[badhabit]

Wow Thomas - That's great., how many sleeps till the results?

Don't hold your breath :-) It will take at least a couple of weeks, i need to make the casts and do postcuring(in my wife' sousvide pan while she is not home!)....and then i'll have to schedule the testing with the institute...

I'll post the results here when i get them...

/Thomas

[peteeng]

Hi Thomas - Boiled, broiled, poached, simmered? Pinch of salt? cheers Peter

[Mikeedd]

It is so nice, you are doing theese tests, BadHabbit, I love to see the outcome, as well as the sousvide you will use to cure the final machine :-)