[brunog]

Calculations of final volume are difficult as 1 cu/ft granite is 166 pounds and 1 cu/ft sand or gravel is much less because of air space.
Basically a mix would be 100%,10% resin,90% filler.Is this correct?10lbs resin to 90lbs aggregates.
Larry

Larry,
Your calculation in weight ratios is correct.

Here's where the challenge is:

Getting a proper sand/aggregate/gravel... mixture to get as close as possible to about 157lbs/cu ft.
Epoxy has a density of half of granite (approx 83lbs/cu ft). in order to make 1 cubic foot of E/G, the best possible result would be with 141 lbs of granite and 16 lbs of epoxy. I've rouded up the numbers that's why the epoxy weight might look a bit high.

[martinw]

10% resin,90% filler.Is this correct?10lbs resin to 90lbs aggregates.
Larry

Dear Larry,

I completely agree that whatever mix you use, you really must measure the ingredients by weight rather than volume. There would be far too much uncertainty using volumes.

I came up with the 17% figure for the volume of epoxy in the mix by assuming a 1:10 weight ratio rather than the 1:9 weight ratio you mention. I just thought I'd mention it in case someone checks my dodgy maths.

Best wishes

Martin

[lgalla]

To get that 157lbs/cuft it may require much higher ratios of sand and or finer aggregates.Possibily two grades of sand and aggregates not bigger than 3/8".Sand blast sand comes in different grades or sizes.I mentioned a long time ago a box of spheres occupies 48%of the box requiring 52% resin.We can lower the resin demand by filling in the spaces with smaller filler.Vibratory compactation helps to push the particles to-gether and micro-pack the mix.
Larry

[walter]

ok, time for another review:

EPOXY GRANITE - Tips & Tricks



1. Epoxy/Granite vs Polymer Concrete

It's imperative to keep clear in mind the difference between "polyester" resins (familiar as glass reinforced plastic materials), and "epoxy resins" more familiar as potting compounds. The first has large shrinkage in polymerization and contains lots of viscosity lowering solvents adding to the problem, and the later with next to zero shrinkage.

2. EPOXY/GRANITE.

- E/G has better properties for damping than solid granite.

- Epoxy's are generally considered zero shrinkage if they are "NEET" eg no solvents or reactive dilutents. The aggregate fillers will result in zero shrinkage. E/G is more thermally stable than solid granite or cast iron.

- Why epoxy over portland as a bonding agent? It is more expensive. Simply put epoxy is a superior "glue" which does not shrink and is fully cured and stable in one week. Portland can take years to settle. Water+steel=IO2=rust

- Disadvantage of regular portland cement is that it uses the water in the mix as the curing agent so voids are left within the cured mass. In addition portland concrete shrinks during curing and may continue to shrink for a long time such as many months.

- Companies are filling weldments or machines made of steel tube to eliminate resonance. Since epoxy has 0 VOC or solvents, when filled it has ZERO shrinkage. If you poured raw epoxy on the floor it will self level to .005 or better. What a way to get an accurate table surface to measure and build your machine on.

- Epoxy is only nontoxic after full cure, one week. People tend to think it is not toxic as it has no flammable solvents. Epoxy can pass through your skin. Wear gloves and mask if sanding.

- There is still a lot of issues to be dealt with. For eg air entrapment, vibratory compaction, heat generation, viscosity , vacuum degassing.

- You need to use a mix of aggregate sizes based on the part being built. The bigger the better, up to a fraction of the feature wall thickness. The German book recommended around 1/5 wall thickness for the largest aggregate, the Zanite site say 1/2. For a two inch thick feature, you are looking at rocks in the 1/2" to 1" range, and below.

- The idea is to use different particle sizes to fill in the spaces between the larger sizes.

- On the German site, http://5128.rapidforum.com/topic=110...8&search=beton there are a few test pieces guys have done, using various ratios of the different aggregate sizes and resin ratios. The guys are testing resin ratios from about 8% by weight, which is what the commercial outfits are using, up to about 12%.
The vibrating table seems to have a very large effect on the final result. They are also having some issue with mold releasing agents. I believe the highest quality cast to tolerance molds also use vacuum degassing.

- Epoxy is an extreme glue which seems to penetrate anything. That mold has to be waxed 4or5 times following the manufactures directions to the T. If you could imagine violently shaking a bunch of sand and rocks on a mold surface whats going to happen? It is called the sandpaper syndrome. the abrasion is going to take the mold release off and probably scratch the mold surface. Solution: Thoroughly wax and apply a gellcoat first.

- Quartz sand is a good starting point for a fine filler as some of the machine castors are using quartz or granite. Interestingly high quality granite is 60%quartz and a very hard stable mineral.

- Whatever size you choose for the biggest "pebble", make the next size 1/5 the size/diameter.
If you consider the pebbles as perfect spheres, the smaller one will just fit in the gap in a tetrahedral packing.

- The aggregates are a commodity product, you can get locally at building supply or landscaping places, cheap.

- Aggregate mix could be play sand [appears to be clean and fine& smooth], builders sand or sharp sand and pea gravel. 3 grades mixed should be enough.

- Builders sand is soaking wet. Not suitable for epoxy. The bags of play sand seem promising as they are very dry and clean. The Home Depot suggested sandblast sand as it is graded and dry. Bagged gravel it is not suitable as it is very dirty. Anything you use must be washed, to remove dust and clay. Crushed stone is not a good choice as it is jagged. Particles tend to "snag" on each other, making the mix thicker than necessary. Simply put smooth aggregates will roll over one another like ball bearings creating a smoother mix.

- If the aggregate is too large, 1" or larger it results in poor strength as the aggregate has different stiffness and properties compared to the binder (epoxy).

- If large aggregates are used, when you vibrate for air release and compaction the matrix becomes very fluid at certain frequency and the large aggregate will drop to the bottom.

- Raw epoxy is solvent free and ships non hazmat. Cheap reactive dilutents are used to thin the epoxy and thin the price. They also thin the properties. Ask your supplier if their epoxy contains Nonyl Phenol.
If your supplier cannot ship non hazmat the epoxy has dangerous additives.
Raw, 'neet' epoxy has low vapor pressure (smell is not bad)
Reactive dilutents have very strong odor, might i add unbearable. Again if it ships non hazmat it is probably OK epoxy.To confirm request MSDS from the supplier.

- You need thin epoxy to maximize filler loading. This would be called Laminating epoxy. Thinnest? Probably 600cps which would be the consistency of #1 Canadian maple syrup. Epoxy glue is 10,000cps or better. Similar to honey.

- Why is epoxy such a great glue? Just before gell it thins out penetrating almost anything. I have in the past taped off mold edges only to discover the epoxy penetrated the tape. This is also why the German guys saw pooling.

- All sand is not created equal. The better stuff is considered better because for one it isn't leacherous - chemicals don't leach out of it. This 'better sand' makes its way into markets like water filtration sand or aquarium sand/gravel - two examples that require the sand to be clean as well as non-leacherous. These sands are usually granite based vs limestone as the limestone will leach, however its still rare, a lot of granite leaches as well. For the lucky owners of such deposits, the price per ton is 10x as high and water filtration sand for example is trucked distances as its is not as common. Bottom line is water filtration sand should be clean and won't leach.

- Avoid crushed stone, look in landscaping centers or architectural suppliers of stone. These markets want aesthetics and the crushed look won't do - like surface treatment of precast panels for example or ground covers. 'River rock' for example which comes from kame deposits and is nicely rounded. Construction uses don't care about leaching or ascetics so thats' the wrong supply chain to look in imo. I believe some pits even 'polish' crushed aggregates to remove the sharp edges/irregular shape.

3. DO NOT OVER-COMPLICATE THE PROCESS....

- Do not over-complicate the process. Many manufacturers optimize in areas such as strict aggregate sizing, specific aggregate mineral formulations, very high granite to epoxy ratios, complicated shake and vacuum arrangements, high precision molds, and many others. Many of these optimizations are only practical when the costs and efforts associated with such optimizations can be amortized over the the product's lifespan.

- As a hobbyist or someone without a need to create many machines, you are free to use materials and processes that may not be practical to a manufacturer. For example, you can simply add additional internal reinforcement to counter effects of voids created by ungraded aggregate and less complicated compaction and deaerating techniques. Also, because you are adding reinforcement you can use a higher percentage of epoxy and smaller aggregate to ease mixing and pouring and still have equivalent rigidity. Since you are not running a production line, you don't have to be concerned about the small additional cost of extra epoxy.

- Here are some important notes on granite, epoxy and molds to help ensure a good casting (some of these have already been mentioned):

Granite
* Size: 3/8 to sand
* Rounded edges ease mixing and compaction
* Clean and dry aggregate well
"Decomposed granite" from landscaping places usually has a good mix of sizes and relatively smooth edges. If you want to get kind of fancy, you could probably separate the granite by size using appropriately sized wire screens, and then combine in an optimized ratio. This really isn't necessary, but I know some of us are overachievers...

Epoxy
* Low viscosity
* Slow cure
* Minimal amount of additives

Molds
* Masonite or similar with a smooth surface and adequate thickness for noncritical mold sections
* Aluminum or steal on areas to be precisely replicated
* Take the time to produce your precision mold surfaces to a higher degree of accuracy than you want your final product
* Don't place inserts closer than 3/4" from an edge
* Use a wax based release that is brushed on. Apply a coat, let dry, buff with a soft rag, then repeat two to four more times
* A gel coat may also be used.

- Here are a few epoxy mix tips. Measure A&B in separate containers, pour into the mix bucket and scrape out every last drop.Use some kind of propeller mixer on a drill. Pause at intervals to scrape the sides of the mix bucket to the center as the epoxy near the walls tends not to mix in. I must add to use relatively low RPM as you will whip in tons of air. I have only experience with micron spherical fillers ie 200 micron and under. We would mix the largest spheres first then the fines.

- If you are having trouble with mold release, never use silicone release agents. Silicone release agents will contaminate your shop forever and you will never be able to achieve secondary bonding such as painting or an epoxy pour. High end car paint shops will not let you in the door if you used Armoral on your tires.

- Adding a small % of 1/32 milled glass fiber- this would greatly reduce the viscosity and result in lower aggregate/epoxy ratios. The benefit is micro fibers would reinforce the bond between the aggregates. Idon't know if this is true today but epoxies 20 years ago would not stick to cured epoxy. Repairs were impossible. I think this problem has been resolved as suppliers say secondary bonding is not a problem. Our solution in the "old days" was to add a small % milled glass. When sanded the resin would sand away leaving micro fibers on the surface promoting secondary bonding on cured epoxy. If you are casting in many pours you have to pour within the WINDOW of your epoxy usually 24 hrs. In plain English layers should be within 24hrs or secondary bonding may be a problem. Calculate what Epoxy you need for the job. You do not to run short and wait a week for more epoxy. Make sure you have enough for the job.

- For the milling and lathe guys filling hollows will enhance the machines performance. I have a lot of molding and epoxy experience and would not consider a totally composite structure. I don't care if anyone follows directions or not but if you fill a hollow with any ratio sand/ epoxy you will acheive better machine performance. I know you are aware of this better damping extends tool life. The grinding people claim 30% INCREASED tool life with E/G damping.

- The largest castings use a combination of a shaking table and shaking motors bolted to the outside of the form.

- For optimal bond, any steel inserts you bond into the part should be sand blasted. The practical method for prototypes and us homebuilders is wooden forms with steel sections moulded in which are subsequently milled or ground to tolerance. The big guys are going more and more to cast to tolerance, but also do grinding of the mineral to tolerance. If you have a surface plate and an eye for detail, grinding and scraping alignment rails could be a garage job.

- Tolerances to +- 0,5mm should be possible with wooden molds. Edge distance for inserts should be >3D. If you intend to reuse the form, it should taper about 5°. Round all internal edges.

- You do not need to go to the complication of epoxy and fancy aggregate ratios.The thesis on rapid machine design, which was designing a big tool grinding machine, that Bob gave a link to goes into damping in great depth. They found that simple concrete inside tubular structures was very good and concrete with an additive to prevent shrinkage on curing inside rubber bags in tubular structures was excellent. It is not surprising that the bag technique worked best because they were effectively making a constrained layer damping system which is the most efficient possible. I have not found any detailed analysis but I am sure the reason the epoxy/aggregate mix damps so well is because the entire structure is in effect a constrained layer damper; it is comprised of to materials with radically different Young's moduli so vibration cannot effectively pass through in any direction.

- Mixing epoxy generates tons of air entrapment.A 10% resin content is so stiff the vacuum will do a poor job. Why is air entrapment a huge problem? When the epoxy exotherm hits 150 C what happens? The entrapped air expands leaving huge voids, not a good thing. If one will accept say 20% resin content we can pour the material and have less air problems. At this ratio, true the epoxy will "puddle"on the surface. Just push in rocks
and add sand to suck up the excess resin. This would probably bring us near the 10% level.

4. REMOVING ENTRAPPED AIR WITH A BLOW TORCH...

- Removing entrapped air with a blow torch. You pour the epoxy in the middle and let it flow over the top and down the sides, then Lick the epoxy with a blowtorch possibly moving 100 to 200ipm or faster. You will be amazed seeing thousands of bubbles popping at the surface. This is because the entrapped air is expanding and being forced out. Although I have done it many times it does not feel right having a flame licking epoxy. A paint stripping gun also works. Just never hold either in one spot.Another plus it thins the epoxy and heats for faster cure. What use to us? I would recommend doing small 1/2" to 1" pours with sand at high epoxy ratios. The heating and air release will result in puddles of epoxy. Now you could sprinkle in the large aggregate and tamp it down with the end of a 2"X4". Any residual wet epoxy can be sucked up by sprinkling in sand.

- The viscosity of epoxy significantly reduces with heat. If I have too stiff a mix I may put the mix pot in a tub of hot water and stir. Now it will flow and I make the decision to add more filler or pour. Remember any added heat will reduce pot life. How do I know when I may exceed pot life?Put your gloved hand on the side of the mix container. If its too hot to hold, ya gotta get it in the mold pronto!!

- Gast pumps are easy to find surplus and 26Hg is probably the max vacuum you would want to subject a resin to before vaporizing some of the chemicals. When you subject a mix to negative pressure the entrapped air expands or boils out the air. Suitable vessels are vacuum bell jars[very expensive] or what I came up with, I use a 10gal Binks pressure pot as the vacuum chamber. I mix 1 gallon of material in a plastic pail and put it in the "vacuum chamber" The mix will foam or expand 3X the volume as the air is released. It is only safe to degass 1 or 1.5 gallons at a time or it will overflow. Negative pressure is very powerful that is the reason for a pressure pot. Vacuum can be stored like an air compressor tank four reserve. An air compressor tank will collapse under vacuum. This is to illustrate why you must have a vacuum bell jar or paint pot as a degassing vessel. I did not really want to post about this as I figured it was beyond the hobbyist. If you get a Gast pump cheap it would also be useful as a vacuum clamp or hold down on a small router. My pump is a bosch capable of 28.8Hg. This gives a pressure[for clamping] of nearly 15PSI this translates to something like 2000lbs on a square foot. Very heavy clamping. On a larger note a 4'X8' vacuum table needs at least a 15Hp pump because of CFM required. A small gast pump would be a handy item in a small shop.

- Other members have asked about pulling a vacuum and injecting the resin. It doesn't work. Rather than directly referring to our E/G, I will refer to RIM and you can figure it out yourself and why it has no use to our process.
Resin injection molding a process whereby a two sided mold or 2 part mold is injected with resin. Fiber reinforcements are laid into the mold and it is closed. A tube is fitted to the injection port from the resin pumps. To ensure void free parts it is recommended to pull a vacuum to remove the air first before injection.A thought occurs: why bother injecting,l et the vacuum suck in the resin. Another one of my great ideas bites the dust. It would take like 5 weeks for the resin to"suck in". The resin pumps inject at 1000psi or better. I hope you get the relationship. We are not going to suck epoxy through sand for wetting.

- Epoxy exotherm can vary between 100 to 200 degrees. Generally slow epoxy is low exotherm fast is high exotherm. High volume of fillers decreases the temperature. I would first try 1" thick and if the cure time was very long, try 2" on the next pour. 4" all in one pour would need low exotherm epoxy. Since we are hobbyists we do not need rapid turn over times and can take our time with slower cures and thinner pours. Air release would be easier in thinner layers.

- The larger the mass the harder it is for the heat to escape. E/G the thicker the mass the more rapid the cure and heat generation. The pot or mix container will harden rapidly because of mass or volume. Thick castings heavily filled should not have excessive exotherm while unfilled raw epoxy could "smoke" in the pot if left to long...

5. www.shopmaninc.com - "Low viscosity epoxy like 30 bucks a gallon"



__________________________________________________ _____________

Big thanks to Larry and the rest of the contributors. Much appreciated!
__________________________________________________ _____________

[lgalla]

Walter thanks for the sumary:cheers: Must have taken hours to complete.Thanks to william for the first person to experiment.I do epoxy sand mold back-ups but have not done tests like William such as bashing it with a hammer and drilling.Glad to see E/G seems to be very strong to abuse.Being noted as long winded I will make my shortest post.E/G is good stuff.
Larry

[lerman]

...
Whatever size you choose for the biggest "pebble", make the next size 1/5 the size/diameter. If you consider the pebbles as perfect spheres, the smaller one will just fit in the gap in a tetrahedral packing.
...

You didn't tell us how much of the smaller size to use. My math tells me that it is a very small amount. If there is one gap for each pebble in the larger size, then there will be one grain of the smaller size for each pebble. Assuming that the pebbles and the smaller grains have the same density, the weight will go as the cube of the diameter. So for each pound of the pebbles, you will need (1/5)*(1/5)*(1/5) pounds of the smaller size. That's 1/125 of a pound, which isn't very much.

Assuming we use a third size, how big should that be? I believe it is *larger* than 1/5 of the smaller size. Also, how much of it should be used? I would guess that my calculation above would still be correct.

So much for the math. In fact, though, because you cannot hope to get perfect tetrahedral packing, the numbers are probably way off.

---------
Does anyone know what the "right" numbers are?

Ken

[martinw]

.....because you cannot hope to get perfect tetrahedral packing, the numbers are probably way off.

Dear Ken,

As I read it, Walter and others suggest that the stones should be as spherical as possible in order to encourage close packing. IMHO, even if you cannot get them to be spheres, avoiding stones with sharp edges may be a start.

How about this.....http://cgi.ebay.com/NEW-6LB-DUAL-DRU...QQcmdZViewItem

Just a thought.

Best wishes

Martin

[jsage]

Walter,

Great summary. Just what I needed. Thanks to the other contributors.

[brunog]

Assuming we use a third size, how big should that be? I believe it is *larger* than 1/5 of the smaller size. Also, how much of it should be used? I would guess that my calculation above would still be correct.

So much for the math. In fact, though, because you cannot hope to get perfect tetrahedral packing, the numbers are probably way off.

---------
Does anyone know what the "right" numbers are?

Ken

Ken,
Assuming that you have 3 sizes 1, 5 and 25 size 1 being the smallest and size 25 being the largest. If you want to mix 10 lbs the proportions should be the following:

1.21lbs of size 1
2.71lbs of size 5
6.08lbs of size 25

If all size rocks are quartz and/or granite, the density being the same, you can use the same proportions in volume


Bruno


P.S. Walter, your summary is real neat!!!

[Zumba]

Does anyone here know how well epoxy bonds to a flat steel surface, assuming no contaminants and a slightly roughed up bonding surface? I've never tested a bond to destruction so I have no idea.

I think that a combination of steel and E/G would be far superior to E/G by itself. Steel can be ground to tenths flatness and offers a much more durable surface than 80/20 aluminum, or bare E/G (which at the surface, is basically pure epoxy, which is quite soft). Also, with steel, you can drill and tap holes for set screws to properly align the linear rails. Merely mounting the rails on a flat surface with no shoulders or set screws is a difficult thing to do.

Weldments with grinding and stress relieving has been discussed in this thread, and the conclusion was that it's too expensive. However, what if you avoided welding all together? Just bolt together a hot-rolled steel structure and fill it with E/G. If the epoxy bonds well enough, then you should be able to remove the bolts and retain 100% structural integrity. Leave the bolts in for added insurance.

A side benefit of all this is that you can be a hell of a lot more sloppy with your epoxy work without detrimental effects, namely the aesthetic ones.

You can also make your structures lighter, which is important if you're talking about a moving gantry router. You don't want a 2000 lb gantry bridge unless you have giant AC servos and ballscrews. Much better to build a steel rectangular tube with 4 long pieces of A36 plate (not A500 tube because it is cold worked with stress built in and has radiused corners), then do a partial fill with E/G. A carboard tube in the center will create a cavity to lighten the load.

[martinw]

Dear all,

It seems to me that what you are aiming for is a complete load of balls.

Best wishes

Martin

[brunog]

Zumba,
Please refer to post#30, it will give you a good idea on how steel is not only bonded but secured in E/G.

Bruno

[307startup]

Hmmm. I was thinking about this today. What about some ceramic ball bearings as filler. Just the balls. Something like this.

http://www.cicball.com/balls.htm#ceramic

I know they aren't granite but they are quite hard which means that they probably do not "grow" as much as steel or iron. Which for our purposes makes them just about perfect. The reason why I say these balls probably do not grow much is because in a ceramic ball bearing, you wouldn't want the balls to grow faster than the steel cage or track, because it would freeze up. Which tells me that they are either grow much slower, or effectively do not get hot enough under the friction they experience to grow. Just a quick thought.

[Zumba]

Zumba,
Please refer to post#30, it will give you a good idea on how steel is not only bonded but secured in E/G.

Bruno

I'm interested more in a steel shell than studs and inserts.

However, that post does help. I think the best way to do it would be to drill/tap holes in the steel plates, thread in some bolts in REVERSE (so the bolt heads are INSIDE the shell), and then fill with E/G. That should secure the plates quite well. You could even weld the bolts to the plate. Tiny spot welds aren't likely to create stresses in the steel.

[Zumba]

Hmmm. I was thinking about this today. What about some ceramic ball bearings as filler. Just the balls. Something like this.

http://www.cicball.com/balls.htm#ceramic

Too much $$. I think the whole premise of E/G is that even though the epoxy is expensive, the aggregate is practically free, making it an economical alternative to cast iron.

[lgalla]

Zumba,
I agree some kind of steel structure and E/G to enhance properties and use epoxy and bolts to avoid welding stresses.Sorry I can't find the link but I think a thread on the Zone called"my first router built in steel"by Haydn is squared up with epoxy.
I read here or Williams thread not to use playbox sand as it is smooth and to use builders sand which is sharp.This is not so.Smooth aggregates roll over one another and allow more fill.Sharp or crushed aggregates also induce stresses at the points and can lock together impeading flow.Coarse aggregates also make for difficult de-gassing as air can be trapped in the roughness of the surface.Stick with smooth sands and river rock.
I spray fiberglass and heavy fillers 50% by weight.The mix is really not sprayable but the addition of 10%ceramic microspheres reduces the friction and allows easy spraying.Any use to E/G?Probably help air release,higher loadings and flowability.I will just give the specs on the product.
Mohs hardness 8
size of grade 850 10 to 200 microns
These are hollow spheres but are thickwalled and weigh 22lbs/gal.
The crush resistance is 50,000 psi.
Cost is about a buck a pound.
Made by 3-M,named Zeeospheres.
Larry

[svenakela]

Zumba,
I agree some kind of steel structure and E/G to enhance properties and use epoxy and bolts to avoid welding stresses.Sorry I can't find the link but I think a thread on the Zone called"my first router built in steel"by Haydn is squared up with epoxy.
Larry

At least Madvac and me are using epoxy.
http://oneoceankayaks.com/madvac/madvac_index.htm

Regards,
Sven

[martinw]

Does anyone here know how well epoxy bonds to a flat steel surface, assuming no contaminants and a slightly roughed up bonding surface? .

Dear Zumba,

If you choose the right epoxy, you should get a shear strength of about

1500 pounds per square inch (roughly 20N/mm2)

Best wishes

Martin

[martinw]

Dear Zumba,

If you choose the right epoxy, you should get a shear strength of about

1500 pounds per square inch (roughly 20N/mm2)

Best wishes

Martin

Oh dear, that is completely wrong. The lap shear figure of 20 Newtons per square millimetre is right, but I did a wonky conversion to PSI.

It should have been 2900 PSI.

Sorry!

Martin

[martinw]

To get that 157lbs/cuft it may require much higher ratios of sand and or finer aggregates.Possibily two grades of sand and aggregates not bigger than 3/8".Sand blast sand comes in different grades or sizes.I mentioned a long time ago a box of spheres occupies 48%of the box requiring 52% resin.We can lower the resin demand by filling in the spaces with smaller filler.Vibratory compactation helps to push the particles to-gether and micro-pack the mix.
Larry

Dear Larry,

This is almost certainly getting a bit too theroretical, and I apologise, but I came across this abstract about random packing of different sized spheres..

http://www.blackwell-synergy.com/doi...1961.tb13716.x

I think it is implying that you can get a 95% solid fill. It states...

A quaternary packing with a density 95.1% of theoretical density was formed from spheres with diameter ratios 1:7:38:316

I may have misunderstood.

Obviously, this takes no account of the practical problem of actually managing to wet the aggregates with resin.

Best wishes

Martin