[greybeard]

Re the method of using excess epoxy to allow easier mixing/pouring, then removing it as waste.
Whatever method was used to remove the excess, what if rather than waste, it was used to make a second mix, probably not optimum, but used in "neutral stress" positions.
I'm not sure of the correct terminology, but I mean that part of the molding where compression and tension are equal to near zero - down the centre of a rectangular beam for example.
I would assume that in this part of the structure, a localized weaker material would have little effect on the overall structure.
After all, one considers using cast-in blocks of foam as a material and weight saving.

I had also considered that if the excess could easily be removed from the top of the cast mixture, why not add that to the next "optimized" batch to add fluidity, and so on, till it is only the last mix that produces any "waste".

John

[greybeard]

deleted due to duplicate post

[Nivea]

This small sample was a preliminary test to try out vibration under vacuum to see if it is worth further testing. My quartz is just 6mm down to dust straight from the quarry, unsized and therefore as you can see has segregated but for this test it's not important. I didn't bother with measuring ratios I just wanted to make a wet mixture and vibrate and vac. I can tell you that the mixture was still quite lumpy and I did not expect it would pack down to only half quartz!

Compared to the last test which was simply compressed but full voids, this test is rather solid. The backlight shows air bubbles well(green arrows) they are about 0.5mm and there aren't too many. It's quite translucent. Think im gonna cast some LED's in there too!

This test has made me believe that vibration under vacuum is potentially the goods and needs further testing. It's almost foolproof and potlife is not at risk. Next test will be with the 2 stage vac pump (this was the single). I will de-air the resin first and maybe use a lower frequency vibrator. This was a small 20,000 vpm job, too high I think.

[RomanLini]

I apologise Cameron, on re-reading my post it does sound a little judgemental and condemning (re "too much reading") and I was not intending to be negative.

To put it in different way, let's assume you have already found the very perfect mix of ratios and sizes. Now comes the work of making it, the different temperatures, different vibrate/compress/vacuum/centrifuge mixing and/or casting processes, different post cure regimes, etc. I've worked with manufacturing before and there are always so many things that have to be changed or things that behave differently due to circumstances or effects that could never have been predicted.

As an example with epoxy molding of small plastic items it never works as planned even with a lot of experience in the process and mold design. Some parts pull away from the mold where they should not, some stick in places where it impossible, some hot spots or suck-ins that just don't make sense, or mis-timed contractions, and it often requires making new molds or sometimes even a redesign of the shape itself.

As another example you may find that the perfect ratios when applied in practice cause an issue that reduces the modulus, maybe a settling effect or mixing problem and the result is worse than using an imperfect ratio that when put into practice avoids that problem and gives a better finished result.

And post-curing regimes alone can take weeks to months to sort out for maximum strength and hardness as there are a few variables and the process of each cure can be in the weeks, with the full hardness reached months later.

Nivea- Congrats on the samples! With that much epoxy it acts like a lubricant and you will get heaps of settling out (as you found) as you will get a total liquifaction effect on vibrating. Also application of vacuum causes many of the volatiles in the epoxy to expand and become gaseous and when vibrating that will give you a bad settling problem as it increases the non-aggregate volume (epoxy volume) a lot until it goes back into solution, compounding the liquifaction under vibration. Compression and vibration under vacuum may work, but my money is on compression vibration and pressure.

[Kevin_Johnson]

Re the method of using excess epoxy to allow easier mixing/pouring, then removing it as waste.
Whatever method was used to remove the excess, what if rather than waste, it was used to make a second mix, probably not optimum, but used in "neutral stress" positions.
I'm not sure of the correct terminology, but I mean that part of the molding where compression and tension are equal to near zero - down the centre of a rectangular beam for example.
I would assume that in this part of the structure, a localized weaker material would have little effect on the overall structure.
After all, one considers using cast-in blocks of foam as a material and weight saving.

I had also considered that if the excess could easily be removed from the top of the cast mixture, why not add that to the next "optimized" batch to add fluidity, and so on, till it is only the last mix that produces any "waste".

John

Be cautious: a possible side effect of adding material from a previous batch to a newer one is the accelerating of the subsequent reaction. For example, if you are using plaster, adding small amounts into fresher mixes can cause the fresher mixes to set up more quickly. Many technicians keep a container of slurry water for mixing batches that they want to set up dramatically more quickly. This variation in reaction time might also progressively effect the material properties of the finished pieces.

[greybeard]

Good points Kevin, and I would expect to allow for such changes in the way each batch might be formulated, i.e. slight reduction in hardener,.
However, considering the idea might be to divide the whole casting process into say five batches, I don't think that would be a great problem.

Another modification might be to consider keeping the mold at a reduced temperature, then after the last batch has been added, raising the temperature to kick-start the normal exothermic process that is so necessary to get a good cure going.

I'd be interested to hear people's ideas on how the excess might be removed from a mix, given their preferred method of mixing and delivering into their mold.

John

[Nivea]

I'd be interested to hear people's ideas on how the excess might be removed from a mix, given their preferred method of mixing and delivering into their mold.

John

John,

Great question, exactly what I've been wondering too.
For a simple open box type mold where the part is cast upside down, Im thinking of making the walls about 30mm taller than the part needs to be. Cutting a "lid" from 25mm thick melamine that is loose fitting by say, 1mm from each wall, drilling a hole for a barbed hose fitting in the center of the lid to which I will connect a vac pump via a canister to collect the epoxy. Leave the vacuum pump running while it's all vibrating in the mold. As the agg settles epoxy pools at the top and gets sucked out. This lid will also act as a gentel weight to help compress with the aid of suction. A mesh filter may required at the barbed fitting for the agg, not sure yet.
I don't think it matters that there is a gap around the lid. It will slurp and suck air as it dries out but no probs there. Not too difficult to seal the lid with a strip of soft rubber overhanging the walls but this will put the mold at -14.7psi. Too risky for me.

If you simply want to remove the excess epoxy, just the vac pump, canister and a tube will be an easy task for the helper. Nurse, more suction!

As far as delivering into the mold, I'd love to have the pot suspended above the vibrating mold with a poly "chute" fed from the bottom of the pot. I can then direct the mixture in the remote corners of the mold. I have a small pneumatic vibrator I plan on bolting to the pot to keep it all fluid while hosing it in.

Also, after all the testing, calcs and dress rehearsals I'd be surprised if we have too much excess anyway. Should be able to start with only a few % above the optimal mix?

[greybeard]

Also, after all the testing, calcs and dress rehearsals I'd be surprised if we have too much excess anyway. Should be able to start with only a few % above the optimal mix?

Yes, I think there have postings on that.
I always think of the way a critically small increase in water concentration turns wet sand into quicksand. If memory serves, going from 14% to 15% makes the difference.

John

[greybeard]

Cameron - I have been thinking about a comment you made several posts back re calculating the surface area of the aggregate particles with a view to quantifying the minimum volume of epoxy needed to coat them.
With that in mind, and combining that with the often mentioned problem of using commercial mixtures of unknown size ranges, I went back to my own experiment with separating a stream of particles with an air tunnel set up.
At the time, when the results of that were inconclusive, I followed that by looking at an enlarged scan of a particle sample, and wondered if there was an optical program that would do the counting for me.
No idea where the thought came from, but yesterday I found a wiki on "Optical Granulometry".
The brief description led me to chase after an open source program to perform this particular piece of magic which I found at Granulometry | Download Granulometry software for free at SourceForge.net

I wonder if this might be a helpful way forward, and has anything like this been tried by any of our readers ?

John

[greybeard]

Further to the above, I've just found JMicroVision, which might be even more useful at JMicroVision: Image Analysis Toolbox
Same questions, as above.

John

[ckelloug]

Roman,

No offense taken. A lot more experimentation is definitely necessary to get a working design. Because my experiments with very low epoxy levels have been a disaster, I've looked at more of theory of what happens in such a mixture. I'll also admit that I tend to do more theory when I can't, at the time, afford to do the experiments I want to do. Since my day job involves researching mathematical theory on other subjects, theory comes natural to me.

Your suggestion of vibration and pressure or weight on the material is excellent and it is in keeping with the suggestions in de Larrard's book on the most effective packing method. de Larrard recommends 10kPa of pressure via a weight on top of the mixture.

Shae,

That's a great experiment and good job on the pictures. The translucent nature is kind of neat. Go for it with the LED's!

From what I've been reading, this type of experiment represents what must be done to find the thermodynamic equilibrium. Based on my experience with the compressible packing model, it appears that you are missing a lot of the middle sized material and that the aggregate you have has segregated during vibration. Segregated material is not predicted to have properties as uniform or as good as homogeneously mixed material.

In case it helps with future experiments, the compressible packing model suggests to me that it would be almost impossible to get less than 70% quartz by volume in a mixture with as many size components as yours. That may help you keep the epoxy down a bit during experiments.

I am quite interested to try a very similar experiment with an equivalent excess of epoxy as it will definitely be more telling than the ones I've done with a lack of epoxy. I haven't got all the aggregate sizes I want but I suppose there is more value in doing it anyway rather than waiting.

I said earlier that I think a huge excess of epoxy will tend to create material that is at thermodynamic equilibrium.

I postulate that minimal epoxy E/G isn't at equilibrium. My idea is that frictional forces in the material are not conservative and could result in a non-equilibrium condition. I'm thinking that as the aggregate is vibrated, its own weight and/or a weight placed on top will cause it to pack closer and closer together up to a point causing an internal stress that cannot be relieved due to friction. The presence of internal stresses would prove a non-equilibrium condition.

Based on the above logic, I theorize that E/G made in a large excess of epoxy could have a lower modulus than E/G made with a near optimal amount of epoxy at the beginning. Of course I've proved in the lab that it doesn't appear to be possible to make E/G using only the optimal amount of epoxy so figuring out how to handle the excess epoxy becomes the challenge.

I'm thinking E/G made as a "suspension" in excess epoxy will pack only to the equilibrium condition. Even if the excess epoxy is siphoned off before setting, the amount of epoxy between the equilibrium energy condition and the higher energy condition that would be afforded by tighter packing cannot be possible without the application of an external compression force.

I think I'm postulating at about the tea-leaf reading stage at this point. . . but I take the above to mean that external compression needs to be provided to the vibrating mass of particles in the excess epoxy in such a way as to allow the excess epoxy to escape as the mixture is compressed.

I think a perforated metal surface with sufficient weight to provide as much compression as practical applied to the top surface of the material would be ideal and at least 10kPa of pressure on the surface would be nice. This structure would also probably need some sort of cavity to contain the excess epoxy.

I digress but If you could make a hollow perforated piston that exactly sealed against the mold surface with a sealed cavity above the perforated part and and apply vacuum in the cavity, it would be ideal although I freely admit it's probably impossible in a practical sense.

For me, the above set of postulates is a breakthrough because it provides a way to get as close as possible to the optimum epoxy level while providing enough epoxy to adequately coat the particles. My experiments up to this point have been based on adding an amount of epoxy sufficient to fill the voids in the aggregate mixture as measured in my packing density piston apparatuses.

That reduces the process to:

1. Design an aggregate mix with the appropriate packing characteristics.

2. Add an excess of epoxy and mix thoroughly. Add the mixture to the mold and allow it to settle. Evacuate the excess epoxy by some method such as manually vacuuming it out into a catchpot.

3. Place a layer of bleeder fabric on top of the free mold surface and place a flat plate and as much weight as practical the same size as the free mold surface over the bleeder.

4. Vibrate for several minutes with weight and bleeder in place.

5. Oven cure if desired leaving the plate and bleeder and optionally the weight in place.

See somewhere like breatherbleeder.com for a cheap source of nonwoven bleeder fabric. (I just googled this up, I'm sure there are a zillion sources).

Roy,

Your machine is looking marvelous. Perhaps one of the other insane folks here can help you make the aluminum parts. You have excellent solidworks models and I have the software and ability to do the non-linear FEA gratis if you're interested. I'd be happy to do it under a non-disclosure agreement if necessary.

E/G is kind of an open source graduate school for me and it's fascinating.

Regards all,

Cameron

[RomanLini]

...
2. Add an excess of epoxy and mix thoroughly. Add the mixture to the mold and allow it to settle. Evacuate the excess epoxy by some method such as manually vacuuming it out into a catchpot.

Keep in mind you can't just "pour" a lumpy thixotropic mixture and expect no voids. Ideally it should be vacuumed to reduce voids (eliminate gasses out the top) before it gets too deep. So that probably needs vacuum-degassing every time you add a couple of inches of depth.

Vibrating under vacuum improves de-gassing BUT without the compression weight on top the epoxy will aerate and increease in volume and decrease in density and viscosity (see my cooments to Nivea above) meaning it will liquify and settle out. It will be a trade-off of more vacuum steps vs chance of settling out.

3. Place a layer of bleeder fabric on top of the free mold surface and place a flat plate and as much weight as practical the same size as the free mold surface over the bleeder.

4. Vibrate for several minutes with weight and bleeder in place.

I think having it under vacuum during this process is vital. Also it will probably need a top-up as you would expect 5% to 10% settling in height. It is looking like a fast way of adding and removing a vacuum fixture (or adding aggregate through an opening) would be very useful. I'm thinking a removable airtight top plate on the mold for vacuuming with a quick-open hole big enough to keep pouring in mixture every couple of minutes between vacuuming.

5. Oven cure if desired leaving the plate and bleeder and optionally the weight in place.

I don't think either are optional. With under 20% epoxy and a thermally conductive aggregate it won't get anywhere near the correct exotherm temperature or enough elevated curing time. An incubator is needed.

As for the weight I am only guessing but it's logical that the mixture should be kept in compression until it's a solid. Removing the weight would likely disturb the top surface and unpack it.

[Kevin_Johnson]

I have been slowly plowing through the front end of thread-maximus. It's a good experience -- I often think, "why doesn't someone bring up X ?" and shortly thereafter someone does. Really wonderful MIT thesis in there -- took a bit of time to read that. Links to explore and so on. I have ordered the text Utz V. Jackisch / Utz V. J;Mineralguss für den Maschinen. Prior to the mention of the ISBN number deeper in the thread I had spent some time trying to discover the title. So it goes.

Returning to the back end at the risk of these very suggestions lying in the middle of the thread...

Much depends on the exact nature of the mold but here is one scenario. I am assuming that the mold top is open. First make a plate that fits the opening but that has a border opening (whilst allowing for shifting "to and fro" -- that's a term of art) that restricts the passage of the smallest aggregate element. If you're using microspheres you may have to afix a straining screen of some sort to the border. The idea is to allow excess epoxy fluid and air bubbles to pass out but not the aggregate.

The interior surface of the plate should be convexly curved such that it is not simply imparting a steady normal (perpendicular) force to the aggregate volume surface (so the exterior surface of the plate will be concave). The superpave gyratory compactor works on a similar principle. The interior surface of the plate should have an epoxy-phobic coating. I suggest a silicone rubber membrane glued to it. It should last a lot longer than multiple waxings. It can be uniformally adhered by using a vacuum forming apparatus.

The concavity on the exterior surface of the plate will accept distributed weights. If you create an even and contiguous surface that still sits proud of the mold borders when the mold is (partially) filled with mixture then you can use a balanced/centered heavy metal plate to load it (relatively inexpensive). I use lead ingots to hold down metal on my plasma cutter.

You can apply vibration to the mold, table and/or plate. The curved interior cover surface will allow migration of entrapped air bubbles and excess epoxy to the borders for removal.

When the material has reached a firm set the cover should be carefully removed. Blowing compressed air into a border gap between the silicone rubber membrane and mixture should help separate it. You might also try running cool water over it to help it contract away from the setting mixture in the mold.

When the cover has been removed the resulting concavity can be filled with a tamped in high aggregate ratio mix and leveled with a screed board/plate.

Hope this is useful to someone.

[ckelloug]

Roman,

I agree with you for the most part of your last message.

I disagree however about the mixture re-aerating itself during vibration without a weight. Once the epoxy gets down to a reasonably low level, the particles are stuck together and tend to densify quite a bit when vibrated without a weight. At least it has in the experiments I've run in the low epoxy regime. The aggregate don't however densify to the desired extent and I agree a weight will be necessary. I was suggesting an intermediate step before adding the weight to avoid the bleeder fabric being overwhelmed in the process.

On the topic of top-up, you mean with aggregate epoxy mix and not with epoxy right?

I agree with you that vacuum during the compaction would be nice but I'm not sure it's practical for non-trivial specimens unless you want to make vacuum tight molds which introduces its own issues. I'm thinking that if the mixture has a deairing agent added and is thoroughly degassed after the aggregate is mixed with excess epoxy that it is unlikely that a lot of air will creep back into it. There will definitely be a bit and I have observed a thin film of air trapped on the mold surfaces. The biggest problem I can see with vacuum compaction is that it ideally requires an autoclave the size of New Zealand for big parts.

Exactly what the cure regimen is quite dependent on the hardener in use. I've been using isophorone diamine which will set fine at room temperature and is slow enough that heat diffuses out of most mixtures about as fast as it is produced. DSC studies show however when the mixture is kept at 85C the optimum mechanical properties are produced. Due to the slow reaction rate, IPDA always needs external heat even taking exotherm into account.

Kevin,

Where did you order Utz V. Jackisch / Utz V. J;Mineralguss für den Maschinen? I can't read German which is probably why I missed noticing this book but it looks like it would be good to have if I can find a local translator.

The superpave gyratory compactor concept looks quite interesting however I'm not sure it's very practical for molds much bigger than test samples. I'm not so sure the curved weight works in the same way as the superpave gyratory compactor because the superpave system uses a cyclic force with a rotating vector as the line of action. A curved weight merely produces a field of (x,y) dependent constant force vectors whose line of action only changes slightly as the weight sinks into the material.

Most of the literature I have been able to find suggests vertical only vibration as the prime method of compaction. I'll draw the dangerous conclusion that most of us agree agree now that weight on top during a vibration step will be most desirable/necessary. I'll need more of a sales pitch to buy the curved weight idea however.

I think the idea of a silicone membrane attached to the weight is priceless however. Is this along the lines of what you were thinking of?

Item # DSP8038CG-032-36, - 80 Durometer Red Solid Silicone Sheeting - Commercial Grade .032 Thk x 36" Wide - Price is per linear yard. Click on Quantity Discount for pricing. on Diversified Silicone Products, Inc.

At any rate, I think we're actually converging on realistic methods of making E/G items. I'm amazed the amount and diversity of the knowledge being shared here.

Regards all and thanks,

Cameron

[Kevin_Johnson]

... Kevin,

Where did you order Utz V. Jackisch / Utz V. J;Mineralguss für den Maschinen? I can't read German which is probably why I missed noticing this book but it looks like it would be good to have if I can find a local translator.

Buchhandlung - Bücher - CDs - DVDs bestellen - bei buch.ch

It was about 47 Francs shipped. I do not know if they have additional copies.

The superpave gyratory compactor concept looks quite interesting however I'm not sure it's very practical for molds much bigger than test samples. I'm not so sure the curved weight works in the same way as the superpave gyratory compactor because the superpave system uses a cyclic force with a rotating vector as the line of action. A curved weight merely produces a field of (x,y) dependent constant force vectors whose line of action only changes slightly as the weight sinks into the material.

When you tamp material into a mold to try to fill the crevices and work out air bubbles this is the same gist. For practical reasons it would be difficult to tilt a large mold and tamp vertically into these areas. However, this technique is done when pouring dental impressions with dental stone (and epoxy). When you prepare impressions for plating you must tamp/stipple metallic powders onto the surface to make it conductive.

The curved weight introduces variable force vectors when the table, mold or plate is vibrated. Think of the system rather than the components individually.

Most of the literature I have been able to find suggests vertical only vibration as the prime method of compaction.

I assure you as a technician that molds are often tilted when filling them and this changes the orientation of the force vectors from the vibrator. Go to any dental office and watch an impression being poured.

For practical reasons it may be difficult to tilt a large mold just as it is difficult to tilt a dynamometer with a running engine attached but it is possible (and instantiated). Slush casting of metals uses a variation of this technique/idea.

Earlier in the thread it is mentioned that vibrators can be attached to the sides of molds so this is actually done in this field.

I'll draw the dangerous conclusion that most of us agree agree now that weight on top during a vibration step will be most desirable/necessary. I'll need more of a sales pitch to buy the curved weight idea however.

I think you have the means at your disposal to investigate it empirically. Simply add a curved surface of known volume to your laboratory device for compacting microspheres.

I think the idea of a silicone membrane attached to the weight is priceless however. Is this along the lines of what you were thinking of?

Yes. I buy sheets of Teflon but have also looked at this type of material. Cosworth makes flapper valves out of it.

Item # DSP8038CG-032-36, - 80 Durometer Red Solid Silicone Sheeting - Commercial Grade .032 Thk x 36" Wide - Price is per linear yard. Click on Quantity Discount for pricing. on Diversified Silicone Products, Inc.

At any rate, I think we're actually converging on realistic methods of making E/G items. I'm amazed the amount and diversity of the knowledge being shared here.

Regards all and thanks,

Cameron

[ckelloug]

Buchhandlung - Bücher - CDs - DVDs bestellen - bei buch.ch

It was about 47 Francs shipped. I do not know if they have additional copies.

Thanks!

When you tamp material into a mold to try to fill the crevices and work out air bubbles this is the same gist. For practical reasons it would be difficult to tilt a large mold and tamp vertically into these areas. However, this technique is done when pouring dental impressions with dental stone (and epoxy). When you prepare impressions for plating you must tamp/stipple metallic powders onto the surface to make it conductive.

The curved weight introduces variable force vectors when the table, mold or plate is vibrated. Think of the system rather than the components individually.


I assure you as a technician that molds are often tilted when filling them and this changes the orientation of the force vectors from the vibrator. Go to any dental office and watch an impression being poured.

For practical reasons it may be difficult to tilt a large mold just as it is difficult to tilt a dynamometer with a running engine attached but it is possible (and instantiated). Slush casting of metals uses a variation of this technique/idea.

Tilting seems to have a lot of advantages but pure tilting isn't what the superpave compactor is really doing. In the superpave compactor, the tilting causes the hydraulic ram compressing the mixture to exert a dynamic kneading force almost like a bread machine. The patent points out that the superpave compactor is used to simulate traffic on a sample of paving material, not to compress paving material for an ordinary road. Because it's designed essentially to cause damage to a sample of road surface, I'm not so sure that it is quite what I would want to use.

I didn't think about the curved weight with the right assumptions when I argued against it the first time:
I originally considered the case of vertical only vibration where the convex curved weight on top of the mixture is in static equilibrium and doesn't appreciably knead the material in vertical only vibration.

Under the assumption of circular vibration (which I wasn't considering) , the weight will obviously rock providing kneading action.

Earlier in the thread it is mentioned that vibrators can be attached to the sides of molds so this is actually done in this field.


I think you have the means at your disposal to investigate it empirically. Simply add a curved surface of known volume to your laboratory device for compacting microspheres.

The microsphere compacting apparatus has a rigid piston with a precision bore. The bore combined with my using it on a vertical only vibration table means that I cannot use this apparatus to simulate any sort of rocking motion.

Yes. I buy sheets of Teflon but have also looked at this type of material. Cosworth makes flapper valves out of it.

Thanks. I've got some teflon sheet in stock somewhere. But. . . Will it be flat and more importantly, Can I find it?

[RomanLini]

...
I disagree however about the mixture re-aerating itself during vibration without a weight. Once the epoxy gets down to a reasonably low level, the particles are stuck together and tend to densify quite a bit when vibrated without a weight.
...

Sorry if I was no clear there, the point I was trying to make re air entrapment is from the pour process. You just can't easily pour something with such a low epoxy content, and this is compounded by your "perfect aggregate ratios" because they are designed to fill all the spaces so particles do not slide easily over each other. It behaves as a thixotropic ie like a solid at the top where the air will get trapped as new mixture is poured on top of that. I'm also anticipating that a machine mold will be a little more complex than a cube or a bucket, and will have thinner sections that will definitely become air traps.

It is going to be very difficult to pour the mixture in the mold without getting air entrapment, and again it's going to be difficult to remove that air if it is more than an inch or so deep. That's why I suggested multiple pour/vac cycles every inch or two in depth.

Making a vacuum tight mold is not that difficult as it has to be epoxy tight as a minimum and a bead of silicone around the mold before casting takes very little effort. Then the top cover just needs a gasket, and we are already talking silicone sheeting...

...
On the topic of top-up, you mean with aggregate epoxy mix and not with epoxy right?
...

Correct. From my experience when you add new mixture in the centre it is a little cone shaped (convex) and the weight of it sinks and displaces the excess epoxy over the outside rim of the mold. Eventually it will be well packed and have no liquid on top.

[Kevin_Johnson]

Tilting seems to have a lot of advantages but pure tilting isn't what the superpave compactor is really doing. In the superpave compactor, the tilting causes the hydraulic ram compressing the mixture to exert a dynamic kneading force almost like a bread machine.

Yes, and if you move the vibration source on the exterior of the mold the movement of the walls will create the same localized kneading action with respect to the suspended curved surface. Because excitatory waveforms travel throughout a structure there would be localized kneading in this sense even with a static vibrator location. The major effect would be normal to the source and in close proximity to it, of course, as you mention below. Variable frequency and amplitude control would be valuable.

The patent points out that the superpave compactor is used to simulate traffic on a sample of paving material, not to compress paving material for an ordinary road. Because it's designed essentially to cause damage to a sample of road surface, I'm not so sure that it is quite what I would want to use.

I suggest that a "steamroller" (archaic term now) -- used to compress pavement during its initial construction -- also uses a localized kneading action with respect to the newly paved surface as a whole.

The "damage" that the superpave compactor inflicts on continually viscous asphalt samples in testing is not quite applicable to cured concrete pavement which is more analogous to cured epoxy granite. I believe the compactor, however, is also used to compact proposed aggregate mixtures sans the binding agents in order to measure maximum theoretical specific gravity. Clearly this would be germane to a discussion of how to compact aggregate in a mold.

I didn't think about the curved weight with the right assumptions when I argued against it the first time:
I originally considered the case of vertical only vibration where the convex curved weight on top of the mixture is in static equilibrium and doesn't appreciably knead the material in vertical only vibration.

Under the assumption of circular vibration (which I wasn't considering) , the weight will obviously rock providing kneading action.


The microsphere compacting apparatus has a rigid piston with a precision bore. The bore combined with my using it on a vertical only vibration table means that I cannot use this apparatus to simulate any sort of rocking motion.

See this research which may suggest that a curved, rather than planar, surface, would be more effective in improving the surface compaction of the curved boundaries of the cylinder mold by bringing the incremental x,z force components into a more normal radial orientation. This also suggests that an analysis of the intended use of the object being made is important along with the possible desireability of having vertical wall surfaces be of similar quality to horizontal wall surfaces.

With the above research cues at hand, adding a curved surface to the piston would allow comparative inspections of the surface walls to be made -- assuming the processed sample can be removed from the cylinder intact.

Thanks. I've got some teflon sheet in stock somewhere. But. . . Will it be flat and more importantly, Can I find it?

I thought about using pure Teflon as the separating medium but it unfortunately would allow aggregate to embed into it and permanently deform it as well -- this would probably create a real mess later on.

A drawback to the silicone membrane is that some of the compaction force would be dissipated into the elastic material. Compromises everywhere.

[Nivea]

Further to the above, I've just found JMicroVision, which might be even more useful at JMicroVision: Image Analysis Toolbox
Same questions, as above.

John

John,


Thanks for posting this. I highly recommend trying it out, it's very easy to use. Take a photo of your aggregate sprinkled on a contrasting colored background, import it, play with a few parameter sliders to best highlight the quartz particles (works very well) and it automatically calculates each partial size. From there it has a huge variety of built in tools to sort, compare, graph and even output to excel.


I think this is going to be a very useful tool, for me at least. I have ordered 18 sieves 30cm diameter in the following sizes (micron) 2360, 1700, 1400, 1180, 1000, 880, 700, 600, 500, 425, 300, 250, 212, 180, 150, 125, 106, 74. That is, 8 to 200 mesh. $US234 from China! So I have been wondering how I might verify the particle sizes and this is the perfect tool for the job. I can even take a photo of the wire mesh and determine the maximum and minimum aperture variation.


It can also calculate particle density against it's background. So I can take an image of a cross section of a cured EG sample and calculate the packing density.


Here are some screen grabs. I sprinkled some fine sand on a sheet of black poly and tried it out.

Shae.

[greybeard]

Great Shae, and thanks for the pics. I'm sure it will give a lot of people encouragement to try it.
As soon as I spotted it, I thought it would be a great tool to get us from a pile of "quarry sand" to one of the optimized mixes.
Once the mixture is analysed, you'll need sieves to extract the components that are either in excess, or add those that are too low. Bit of head scratching maths, but nothing compared with some that have been posted.

I'm amazed at that price for eighteen seives, but given their source, perhaps I shouldn't be

John