[brunog]

YOU ALWAYS NEED A VIBRATION TABLE, WITH VERTICAL AMPLITUDE OFF 0,25mm
AND 2 to 4 G ACCELERATION.

This is the bottleneck for DIY, because you need to buy a table with 2 motors, 1 drive and good isolation rubbers (at least 1000 dollar total)

Roy,
Ok for vibration, directionnal? I guess Thomas Zietz would not agree with you!
None of his setups were/are directional.
Vibrating motor and isolation rubber costs, come on, with a little creativity you can bring that down to less than $200 CDN. Have you forgotten Steven Donner's (observer) vibrating motor??

Don't forget, I am sold to EG.
I was reacting to Romanlini's comment that making EG is a fussy process, and I still maintain it doesn't have to be.
People have proven that it can easily be done easily and in many ways, it just depends what you want to make with it, and what will be the EG's purpose in the machine construction.

Romanlini,
I understand where you want to go with Perlite or styrofoam pellets, the problem is how can you make sure that there is no accumulation of pellets in some areas that will create weak spots without complicating yourself with a lamination process using fiberglass,carbon fiber or some kind of cloth like certain existing concrete panels in the construction industry, now that would be fussy! :stickpoke

Best regards

Bruno

[veteq]

Hey Bruno,

You are right about the direction off Thomas his vibrator table, i made that one bigger then it is. Let i say it like this, i would use 2 motors.
And that vibrator that Observer made maybe only pleases his woman i think.....:banana:

Think a real motor is neccesary !

[nebbiasound]

Hi, I want to add another small contribution to those living in Europe.
To find components such as resins and fillers, there is also an interesting guide on assembling techniques of composites. Address is: http://www.r-g.de/
Are 2 PDF files, one is the catalog, the other is the guide to the techniques of composites.

Greetings to all and good job.

[RomanLini]

...
Romanlini,
I understand where you want to go with Perlite or styrofoam pellets, the problem is how can you make sure that there is no accumulation of pellets in some areas that will create weak spots without complicating yourself with a lamination process using fiberglass,carbon fiber or some kind of cloth like certain existing concrete panels in the construction industry, now that would be fussy! :stickpoke
...

Some good points, and I agree making the mixture lightweight AND very strong may have its own difficulties. I'm just questioning exactly how strong and how rigid it needs to be, and that depends on what it will be cutting...

Veteg said; "none will mill steel or aluminum properly."

If I was building an EG (let's say "epoxy composite" as it might not qualify as EG) machine it would probably be to cut softer materials like plastics, wood, soft metals so the construction technique would not be the same as a machine designed just to mill steel. The design would be more optimised for what it NEEDS to be able to do.

I wish I had more time at the moment to experiment with some lightweight, easy to work with, rigid composite materials.

[ckelloug]

I'll weigh in on Romanlini's discussion with Bruno and Roy.

The stiffness (modulus) of a quartz E/G has a thoeretical limit of about 5,200,000psi with the epoxy osievrt is using. Using the very best regular epoxies, 5,500,000 psi is the theoretical limit. The reference formula I published with the aluminum oxide should have a higher modulus than I have predicted for it because the models don't have a way of accounting for multiple aggregate materials and I modeled it as quartz.

The only things stopping us from achieving fairly close to the theoretical are the fact that I haven't made a testable sample of the current material in the lab to validate the models and the fact that custom sieved aggregates are going to be necessary to get the modulus up. I've avoided custom aggregates to this point because that makes it harder for others to duplicate the research. I've also stopped mentioning additives since not everybody has access. I'll also point out that the optimal aggregate design is also a function of how much its vibrated. A material vibrated with low energy needs a different design for maximal packing than one with high energy. My reference formula is for the recommended vibration of 4G in the literature.

So, romanlini, in essence you are right about picking the material to match the machine. Raw epoxy with any sort of mineral filler is better than MDF in terms of stiffness and if that's what you're trying to beat, go for it. My own interest is in precision metalworking machines and eventually large ones. The cost of the extra epoxy from going with a suboptimal aggregate design will quickly kill big projects: suboptimal aggregate design also greatly increases thermal expansion and other undesirable second order properties.

One other note, lighter parts tends to raise resonant frequencies which if your luck is bad enough could raise the first resonance to something easily excited by your servos or spindle. So, in my opinion weight is an asset.

[ckelloug]

Hi Nebiasound,

Thanks for the two posts you have given us about vibrators and composite methods. I found the one on vibrators especially interesting.

Regards,

Cameron

[nebbiasound]

Thanks Cameron for the reply, I live in Verona, where does this fair international http://www.marmomacc.it/it/index.asp
I have seen live the process for fabrication of aglomerate Breton, but they use polymer resins Biolene owners. In that area, the waste material from processing poses no problem, so even if there is withdrawal after hardening of the resin do not care about this.
The epoxy resin for this type of processing block is too expensive, polymer resins Biolene exhibits more efficiency in weight / volume, it is also biocompatible with food. I am convinced that the epoxy is the only solution to build at a hobby cnc machine without spending a fortune.
I started to design a 5-axis trunnion and I want to use granite epoxy frame. I follow your discussion with enthusiasm, I read all the posts and I learned many things I did not know. Thanks for writing this, I think you are all on the right path because all this is true.
Dario

[RomanLini]

...
So, romanlini, in essence you are right about picking the material to match the machine. Raw epoxy with any sort of mineral filler is better than MDF in terms of stiffness and if that's what you're trying to beat, go for it. My own interest is in precision metalworking machines and eventually large ones. The cost of the extra epoxy from going with a suboptimal aggregate design will quickly kill big projects: suboptimal aggregate design also greatly increases thermal expansion and other undesirable second order properties.

One other note, lighter parts tends to raise resonant frequencies which if your luck is bad enough could raise the first resonance to something easily excited by your servos or spindle. So, in my opinion weight is an asset.

I agree totally. I've got to confess ignorance, I didn't realise the thread was focussed so much on EG to be used in large steel milling machines.

Sorry if I have derailed it in any way, from the start my posts have been more oriented towards the type of machines I use, ie; smallish router type machines to machine wood/plastics/aluminium etc. In that case the cost of the epoxy itself might only be a couple of hundred dollars even in mix ratios where the epoxy is as high as 25% of the machine weight.

It's a shame there's not many smallish epoxy composite machines like this on the market, I think they would offer a great alternative to the aluminium or MDF framed small machines and could have many of the features "cast" in place as such to improve construction.

Maybe if other people are interested in this idea there could be room for another thread on "small experimental epoxy composite machines" or the like, seeing as this thread is already so large.

[ckelloug]

Hi Romanlini,

I am interested in metalworking machines personally but there's no bias against routers etc here. The design principles are similar and a quartz material of optimum stiffness will suit all machine applications pretty well. You can certainly get by with less than optimal materials if you don't need to keep deflections small or hold .0001 inch tolerances etc. But, by designing a good material here, I'm hoping we can save everyone a lot of trouble in the long run.

I've been focusing on this thread on the material itself rather than building machines out of an obsession with "Getting it Right". I promised myself when I started on this that I would participate here in the development a provably optimal material since there's no reason to let the big monopolies with proprietary formulas have all the fun. It just happens that high performance E/G is cheaper than low performance stuff and has fewer problems with segregation under vibration etc.

Unlike many here, I'm not here to build a single machine. I'd like to develop a process to mass produce excellent metalworking machines knowing enough about the process to avoid a CNCBridges style fiasco.

Everybody who has posted to this thread has contributed to making E/G more practical. This gargantuan discussion is, in my view, the enabler for small E/G machines coming onto the market because it takes the technology out of the domain of proprietary black magic and into the public domain.

I wouldn't be here without Walter's contributions and Walter wouldn't have stayed without Larry's contributions etc. Even in cases where we all disagree, I know I have learned more personally from constructing my counter-arguments than I would have if folks with differing opinions hadn't posted. So, I don't think you've derailed anything at all. Glad to have you aboard and I think all the regulars here are happy to help if you've got a machine design you're thinking about.

In keeping with your comments, one approach to making the structure lighter would be to replace some of the aggregate with something like 3M ceramic microspheres http://solutions.3m.com/wps/portal/3...beWHB23F5LMRgl The H50/10000 EPX are silane treated and thus will stick well to the epoxy.

I think that these glass bubbles will lower the strength of the composite a fair bit and the modulus some but I don't really know how it would behave. Definitely food for thought however.



Regards all,
Cameron

[nebbiasound]

Hello Cameron, I agree fully on the subject to replace that portion of the microspheres, my point is: most materials that make up the aggregate have a predictable form and the more predictable the behavior of the compound itself.
If it were possible to use other types of balls also make more, so totally replace the sand and stones. I found this: http://www.rgpballs.com/eng/prodotti...sion_balls.pdf. There are more balls in materials and alloys. I do not know if this is wrong, certainly more expensive, perhaps find a middle way. You can enter data for the model to obtain might be a new way?
Even a mix of the two techniques can give a new boost.
Thank you for your attention.
Best regards
Dario

[the4thseal]

i have been working on a substitution of granite with steel bearings of various sizes but have not had results worth posting...but i am making progress and will post soon. the rabbit hole is deep.(nuts)

[greybeard]

Hi all.
Just a few observations, to prove I'm still alive, if not kicking too much !

On the subject of the diy approach.
We're all hard pressed for cash, so it seemed to me that having to acquire/build a huge vibrator for what may be a one-off mold pour needs to be avoided. The cost of the epoxy being a large enough component, it doesn't need to be doubled by adding in the capital cost of the equipment.

Consider a mold needing 10 jugs full of EG(note the new international unit of volume, where 1 jug = 10 cups).
Given a working time of say 60 mins for your own epoxy mix, and a mix/vibrate/pour time of say 6 mins per jug, then a vibrator that can do the business for 1 jug is fine. The last pour will go into the mold about the time the first one is starting to build up the exotherm reaction, and the whole mass will become a single solid.
This is obviously an improvement on needing a 10 jug vibrator, but it seems to me that there is no need to stop there. A 1 cup size should also work, and this leads me on to consider if it might be possible to design/build a system that continuously vibrated the mixture as it was being transferred from the mixing vessel to the mold.
This would have the added advantage of needing only a very short distance for the bubbles to escape from the mix.

Regarding air entrapment during the pour. This can be avoided by careful pouring and starting at the lowest corner of the mold, and only adding a steady stream of material to that already in the mold.
If the mix is so lumpy that it doesn't pour, then continuous ramming will be needed. This might be the "killer" to the whole idea, of course.

Another problem might be edges, where there will be a tendency for a lumpy mix to bridge over. I wondered if a separate mix, possible losing the largest aggregate component, might be placed along these difficult areas, without jeopardising the strength of the whole.

Of course, those who have followed my ramblings, know that my own preferred method of de-airing is to spin the whole lot, with all the danger to life and limb that might entail. Much more exciting.

This leads me to my other thoughts on reducing total weight.
Rather than casting a large single block, along with low density cores, why not consider building the frame up from small components - a space frame of EG girders.
They could be simple plain ended struts, with separately molded joints, or molded ends, only requiring epoxy adhesive to build up the frame.
Any variation on this approach would also have the advantage of drastically reducing the size of any equipment needed, both for weighing, vibrating, and handling in general.
Is there any fundamental disadvantage, ie from an engineering point of view, in this approach ? I acknowledge that the placement of insert/fixing points would now be a different ball game, but who said it would be easy.
For most diy-ers, the extra time in the build would be more than worth it for the reduction in equipment costs.

Regards
John

[brunog]

Hey John, nice to hear from you!

I thought that a jug was an old imperial unit of measure!

Your right about edges , Zanite uses a mix with smaller aggregates and line the whole interior of the mould to get smooth surfaces and edges in a similar manner applying gelcoat.

I like the idea of making smaller parts, the only downside of it is the more there are components, the more molds are needed to be made and fitted, but it could be done with carefull planning beforehand.

Best regards,

Bruno

[greybeard]

Hi Bruno,
"More thumbs than expected" - do you have to share a keyboard with someone else ?


John

[brunog]

Hi Bruno,
"More thumbs than expected" - do you have to share a keyboard with someone else ?


John

My wife and I were playing 4 hands on the keyboard!

Seriously, for once I didn't want to write mispelled

Best regards

Bruno

[ger21]

Given a working time of say 60 mins for your own epoxy mix, and a mix/vibrate/pour time of say 6 mins per jug, then a vibrator that can do the business for 1 jug is fine. The last pour will go into the mold about the time the first one is starting to build up the exotherm reaction.

I've made a cup of epoxy-sand with slow setting US Composites epoxy, and didn't see any exotherm. Cure time was extremely slow. Using your method, you'd have at least 12 hours per jug.

[ckelloug]

Hi nebiasound,

On the topic of precision ceramic balls, I'd say they are too expensive. Additionally, the process of making E/G relies on a random distribution of the materials to get a high packing density since it's near-impossible to get an ordered distribution and even if you could, it would have shear planes. The packing density of randomly oriented spheres is .71which is a little better than aggregates but precision spheres may be several orders of magnitude more expensive than approximately round aggregate.

Hi the4thseal,
I did some quick research on steel shot for E/G a few years ago and my opinion from the model is that the modulus is too low to make a material much stiffer than quartz E/G. You won't get a strength improvement either most likely as you will fail in the epoxy between 3000psi and 6000psi tensile.

Greybeard John,

Glad to see you're still about. Making small parts is a good suggestion I think, and if you do it right, the assembly can be made by bolting and grouting. The downside is that the more pieces you have, the more difficult it is to hold critical alignments between them. But, I think careful design could make this very workable.

Best of luck. I hope to get home soon and actually get something made.

Regards all,
Cameron

[the4thseal]

i was looking control....bearing size and control of wetting of balls....also using fibers steel and aramid. i agree with you on calculations though. I am enjoying the trip down the rabbit whole.

[brunog]

Gerry,
The reason you did not get any exotherm is that all the heat was absorbed by the sand.

Best regards

Bruno

[RomanLini]

Correct, with very low % epoxy levels it will never exotherm and you need to endotherm by applying heat from outside. Then post-cure for max strength at a different temperature.

Which is one of the reasons I developed cheap accurate open-source temperature control technology;

SH1-Temp cheap $12 microprocessor temperature controller kit

...
In keeping with your comments, one approach to making the structure lighter would be to replace some of the aggregate with something like 3M ceramic microspheres http://solutions.3m.com/wps/portal/3...beWHB23F5LMRgl The H50/10000 EPX are silane treated and thus will stick well to the epoxy.
...

Yep, I stock the glass microspheres in the workshop here and did mention them earlier, having used them in some experimental composites in the past. I just wanted to add a warning they are VERY nasty for inhalation, they "puff" in the air much worse than something like flour and need special breathing precautions. They are also very LOW strength and need to be supported with chopped glass fibre or other strength components. Microspheres are best thought of as "air" and have all the strength you would expect from it.

On the topic of other balls, shiny/polished balls are one of the very worst adhering surfaces for epoxy as it mainly adheres mechanically. So it will result in very low strength cured composite.