[jvalger]

Hi Brandon,

Yeah, that is a slick technique. I could be wrong, but I think that's how the interior mandrels for commodity carbon fiber tubing are made.

Re the interface between the carbon outer layer and E/G core, the problem is that putting an elastic layer like toughened epoxy between the two means the carbon doesn't stiffen the E/G beyond what the epoxy transfers across the boundary. The carbon layer is a lot of work and brings problems of its own, so it has to make a big difference in stiffness or it isn't worth the effort.

I honestly don't know whether this is making a mountain out of a molehill, only that it isn't obvious what would happen.

Jeff

[turnpoint]

Hi Jeff,
I'm getting out of my experience level here-- this is the realm of a good composite engineer.

--The way I'm seeing it is that the carbon tubing would take the load and the E/G would dampen vibration. So as long as the two materials were bonded --the E/G would dampen vibration. Also, off the top of my head the CTE of carbon/epoxy should be fairly similar to the CTE of carbon/granite and thus there shouldn't be too much problem with the effects of heat expansion at room temperatures-- my gut feeling would be that this would be negligible. This would be fairly easy to calculate.
But I am completely talking out of my a__ on this. So consulting a composite engineer to do the math on this would be a good investment before making a big machine.

[Eric R.]

Hi Brandon,

Sound very interesting. Thanks!

Eric

[the4thseal]

Or, do it yourself. Simpler than it might seem with manually-operated, diy equipment. Pass the fabric around one roller at the bottom of a resin bath to wet it, over a second, higher roller, then down between two rollers (one acting as the drive roller) to press out the excess resin. If you are careful, you can even make the V between the two main rollers double as the bath - the fabric itself prevents the resin flowing out the bottom. If you don't want to build your own I've seen these online.

Jeff
Reply With Quote


do you have a picture of this?

[veteq]

Hello people,

Been reading the last pages and think that we are on the right track:
Most important are:
Quality epoxy, mixture off aggregate and vibration 3G 0,25mm.

Next to this forum i`ve read the complete German forum, was a pain:drowning:

Mr. Jolincks says that 8% epoxy by Volume is the max.
The Germans say 8% by Weight is good and 6% is best.

I would like to use the Silimix aggregate nr-282 and the R-g.de epoxy
L1100 + Harder EPH294, there properties are (mix off the Germans) :
aggregate = 2,03 g/cm^3
epoxy+harder = 1,100154 g/cm^3 ( 100 / 30 )

So if i would like to have 100 kg off EG this is the math:

96,5 kg off aggregate = 47537 cm^3
4,5 kg off epoxy = 4090 cm^3 ( 4,5 % by weight )

Total off Volume is = 51627 cm^3

By volume this would be the mixture off Mr. Jolinck with 8% since,
(4090 / 51627) x 100 = 7,9 %

This is 4,5 % off epoxy by weight, thats a quarter less then the Germans best mix and the max for Mr. Jolinck, could you guys say something about it???

If we take the german mix and compare some more by math:
6 % by weight = 10,5 % by volume
8 % by weight = 13,8 % by Volume

The thing is the Germans say that the last one, 8 % by weight is good.
Here the 8 % by Volume would be the max.
I was thinking at 10 % by weight, thats 17 % by Volume,
what would happen? The machine in the jpeg has used the 17 % by volume.

Then another question,
here vibration is between 10 and 30 Hz with 0,25mm displacement
The Germans use around 70 Hz and ??? displacement.
The vibration is to let the particels settle and fill up the air gaps, if we use to much epoxy there will be a pool on top off the mould. What would happen if you keep vibrating for an extra long time.

If you take a low amount epoxy, lets say 7% by volume and keep on pounding on the EG to compress, wouldn`t the air cas be gone???

The Germans vibrate between 5 and 10 minuts.
Mr. Jolincks says that you should vibrate when epoxy starts to cure, so should we fill the mould and start vibrate, or do we have to do it at the end off the potlife???

Long story, Mr. Jolinck, greybeard, Walter and all other,
Cann you give me some feedback about my findings.
The target for me is a small machine with really low forces, the L-frame in the Jpeg is 500/500/260 mm.


Kind regards and compliment to all, especially the Die hard`s,

Roy B

Ps. are there out there 220Volt vibration motors???

[veteq]

About the vibrator,

Those unit are very expensive, i found this calculation on the net and made a small excel file.
ok, then we know the needed centrifugal force and the wattage off the motor, if i would like to design a vibrater myself, what do i need to do with the data.
The unbalans weight that cann be seen in the picture has a cutout, so the CG has a distance to the axis of revolution.

Is that the formula (sorry metric now)

F = m * w^2 x r

F = centrifugal force (N)
m = mass (kg) (is this of the total unbalans weight?)
w = angular velocity (rad/s)
r = distance to CG from axes off revolution (m) (is this the distance thats the peak peak dimension for the amplitude???)

So compare the Centrifugal force calculated here with the calculated CF from the excel file, then the forces in the design are right???

Was thinking about a DC servo nema 34 with a Geckodrive G320 and a seperate unweight in a housing controlled by mach3.

regards,

roy

[lgalla]

Sorry for not following the thread in detail,but will post a few short comments on carbon fiber and E/G.
Carbon fiber will add to the stifness of E/G,but has very poor vibration damping.C/F is like steel and rings like a bell.The jist of the thread was to use E/G as a vibration damper.Steel tubes filled with E/G would dampen vibrations much better than C/F.On another note,Kevlar has vibration damping qualities.In aircraft parts,Kevlar absorbs impacts and deforms similar to aluinium.If you need to increase stifness,Kevlar is a better choice than C/F
larry

[Eric R.]

Hi Roy,

If you are looking for 230V Vibration motors, check on the German eBay site for Rüttelmotor: (just click the link).

Regards,

Eric

[garfieldsimons]

Episs
Companys fabricating machine bases of epoxy granite are not in the business of selling the raw material.I did a lot of research on the subject and there is little to be found.Sand is basicly granite and probably a good filler.
Idon't remember exactly the figures,but epoxy granite has 10 times the damping factor of cast iron and 50 times the damping factor of steel.
THE COST??? 1gal. epoxy $80 Maybee 10 sq ft/gal/1/4 thick.Figure it out,its
gunna cost mega bucks even with sand fill at 50%.
This is just off the top of my head.If anyone wants more accurate figures I will look them up again.
Any how I figured out a slab of real granite is 1/2 the price of epoxy granite,maybe less.A 4X8'X4"epoxy granite table top could cost 5or6 grand or more.A black granite surface plate 4X8'X8"with stand is $2600 from an E-Bay store.If anyone is still interested I will try to answer any questions or re-research my findings.
Larry

I got a a 2 x 4 block of granit for $100 at an auction. Nobody wants them and they are a pain to move. But they can be had reasonably cheap. Side note my block of granite came with a cmm. And threaded mounting holes.
Richard

[ckelloug]

Hi there veteq Roy,

The mass percentage of epoxy is useful only for measurement purposes. The volume percentage of epoxy is what determines the physical properties of the mixture.

According to the compressible packing model from Francois de Larrard's <u>Concrete Mixture Proportioning: A Scientific Approach</u>, the maximum achievable aggregate percentage is 92% plus or minus 3%. Thus no mixture with more than 11% epoxy by volume could be optimal and no mixture with less than 5% epoxy by volume is possible. The model assumes vibratory compaction. Slightly better compaction is available with huge hydraulic presses but that isn't what we're about here.

This model results from de Larrard are entirely consistent with Mr. Jolincks and the Ph.D. thesis of German Castillio (in French) from the Usinages group, and disagrees with the German group.

There is no question that the German group has done some good machine engineering but I have always felt that their material design was suboptimal. According to their last post in this thread, the Germans are using the Fuller distribution which is demonstrably suboptimal providing at least 6% lower density than the optimal distribution and 5% worse than virtually every other common aggregate distribution. AS for your question about 17% by volume, see the <A href=http://www.cnczone.com/forums/showpost.php?p=444339&postcount=3119>nomograph</A> on pg 260 generated from a formula in A.J. Kinloch's <u>Advances in Resins and Structural Primers</u> and follow the cyan line to 83% aggregate=17% epoxy. You'll find that pine wood with a young's modulus of 2e6psi has a similar stiffness to that proposed E/G mixture. I have tested a few similar materials myself to failure in ASTM D790 flexural tests and gotten modulus numbers that match the nomograph results.

The German group have shown that there is no problem working with a high-tech plastic-wood like formula but if you need stiffness for a large part, there's no substitute for using the optimal aggregate mixture. As long as you can do engineering analysis either on paper or FEM to ensure that you meet deflection and strength criteria you could build a machine out of a suboptimal E/G like the one you propose. The amount of epoxy needed is that amount which fills the voids in the aggregate and perhaps a tiny bit more to aid in processing. The perfect epoxy is a lot less important than an aggregate distribution that achieves a density of at least 92% when compacted.

The vibration numbers I have seen cited in both Castillo and de Larrard suggest that the entire mold must be vibrated at between 2g and 4g at 50 or 60 hz. This is a somewhat difficult proposition for the average contributor here. In general, the longer the mixture is vibrated, the more it compacts. What you can get away with depends on the pot life of the mixture. Most of the epoxy formulations I have considered reasonable have pot lives in the 4-8 hour region and it's almost certain that vibrating this long cannot be good. I have only made small samples with minimal vibration so far so I can't shed any light on the right way to do it. My only certainty is that to make a single production quality E/G machine requires more capital that getting a whole bunch of cast iron and scraping it.

If you are using a suboptimal aggregate mixture, it will segregate and as a result, there is a balance between vibration time and compaction. The de Larrard Minimum-S mixture is ideal as it can be vibrated with limited segregation but I don't have my compressible packing model code implementing it correctly right now so I can't solve the general case for arbitrary aggregate.

There are huge 220V 3 phase vibration motors available but they are expensive. I have a 3 horsepower unit over at my lab that I found surplus for maybe 1/5 the cost of a new one. I think it's 15 Hz by default but with a nice VFD, I should be able to get it up to 30Hz or a bit more. A standard concrete vibrator is too small by all accounts I have read. The 3HP unit I have weighs 400 lbs and the rotors are so heavy I can barely spin them with my foot. The problem is that you need a sturdy table to which you can attach the vibrator and isolators so that you move only the table, vibrator and contents without trying to move the building the vibrator is in. A large steel plate for the table and the vibrator itself will add substantial weight to the weight of the part in your calculation. The deflection will also be based on the weight of the table and the stiffness of the springs on which the table rests. Stiffer springs will provide less deflection and less isolation from the building.



Material Design Conclusion:

If you would be confident in holding the desired deflections on your machine if it were made of softwood like pine, your proposed material design is sufficient. This is a very good approximation as the moduli are almost exactly equal between pine and epoxy-rich E/G. Approximately 2.5 times better stiffness would be available with an optimal material.

On Vibrators and Vibration:

With regards to your second post about building a vibration system, it is my opinion that your 3rd design has the potential to fail catastrophically. The reason that commercial vibrators are expensive is that they have huge bearings and are very heavily built. The forces at the bearings of an off balance rotating item are astronomical. I have reservations about the third drawing because you have the rotating mass rather far from the bearing. This will tend to cause cantilever deflection in the shaft and I would worry about catastrophic failure of either the shaft itself or the bearings. The bearings of a precision motor are not built to handle the abuse of running as a vibrator for long but the setup is probably sufficient for a one-off. If you are okay with the potential of catastrophic failure and loss of the motor, make sure to build a sturdy containment housing to catch the airborne parts.

With regards to the calculations, I see nothing wrong with what you have done but I think your weight estimate of the entire setup might be too low. Pay special attention to the section on isolation in the Deca vibrator page you cited. This static system deflection that they cite is the amount that the vibrator, table etc deflect the springs that it rests on due to their weight. The numbers that they give here are a surrogate for solving the second order differential equation of motion for the vibration table. See <A href=http://www.thevmcgroup.com/springmounts.html> The VMC Group</A> web page and get a copy of the Korfund Dynamics Catalog under downloadable PDF's to get a better feel for what you might need.

Also note that commercial vibrators have adjustable rotors that allow the amount of off-balance to be adjusted by controlling the angular separation of two parts of the rotor. The page you are designing from seems to assume you will use this adjustment to correct the operation of the system into the nominal parameters.

With the type of motor you are proposing, I think I personally would want to belt drive the off balance rotor and house it separately in a housing with very substantial bearings. This would at least preserve the motor should something fail.

At any rate, I wish you the best of luck in your endeavor and think I can speak for everyone when I say we look forward to hearing about your progress.

Regards,
Cameron

[jvalger]

Or, do it yourself. Simpler than it might seem with manually-operated, diy equipment.


do you have a picture of this?

Look at this one for an example: http://www.prosetepoxy.com/process_equipment.html

Jeff

[the4thseal]

that makes sense.....thank you

[MaxxD]

Hi to everyone,

Hope this thread is not dead yet? I just spent the last several evenings reading the whole thing! Amazing with what has been done so far. I have a few ideas to through into the mix (pun intended).

I know the vacuum ideas have been brought up before, but has anyone considered vacuuming while mixing? I'm thinking of a chamber made of pvc, with a lid that has a rotary vacuum joint, that rotates like a cement mixer, and includes vibration as well. That way your mix should be completely de-gassed, all of the particles covered with epoxy, and the mix should be random. Could also use the vibrations to pour the mix out, since it seems to flow while being vibrated.

Why is it that your adding the particles to the epoxy exactly? In the candy making factories, when they're putting a "candy shell" on those little pieces of candy, they pour the "coating" onto a rotating batch of candies. All of 'em rolling around and bumping into and over one another makes sure that they all get a nice fairly even layer. Maybe something to this?

Has anyone used both vibration and vacuum on the mold yet? I'm not talking huge amounts of either, but if your using the slow cure (400 min) epoxy and have a larger mold, wouldn't the vibration and vacuum cause the mix to flow and de-air? Also haven't seen anyone talk of pouring in sections, IE. pour 1/3 vibrate/vacuum , repeat, repeat. Again, not alot of vibrating, just enough to flow out. Seems that the thiner layers would give up the air more easily.

Last idea and I'll leave ya'll alone. I see some mentions of vacuum bagging the mold and mix. That is probably the way I'll go, but have you thought about double baggin'? (reminds me of an old joke). What I mean is you can put the inner bag under vacuum, and the outer bag under pressure. Not alot, maybe 2 bar. That way your mix and mold are effectively under 30 psi instead of 15, Twice the compression (without 2X the vacuum) ought to eliminate those pesky air bubbles and if vibration is included as well, then maximum packing, minimum epoxy here we come!

Of course I'm only tossing around ideas here, and as always your mileage (as well as your ratios) may vary!:idea::idea:

Steve

[Steven.ji]

In terms of mineral casting for machine bed, it's no necessary to mix in vacuum.

For vacuum mixture of mineral casting, I found the exsiting process for lawrence pump

The picture as reference.

[ckelloug]

In response to a PM from Steven Ji,

I've posted a few size distributions on the thread based on the model given in Concrete Mixture Proportioning: A Scientific Approach by Francois de Larrard.

Unfortunately, the current version of the optimizing software I wrote is implementing a solution that only works for the trivial case with individual aggregates where changing the percentage of one size does not change another size and isn't suitable for finding an optimal mixture with real aggregates. The model appears to be predicting density correctly but the optimizer is currently driving the particle distribution to the wrong solution.

Compared to models I have worked on for other things, the model for particle packing provided in de Larrard is brutally difficult to solve for non-trivial cases since there is an n^2 dependency of every particle size on every other size. My original implementation can solve the model optimally in trivial cases but doesn't implement enough of the theory for non-trivial cases.

From de Larrard's work, the simplest mixture is equal parts of everything over 4 orders of magntiude size difference between the biggest and smallest particles. This isn't optimal and will segregate during vibration, it's still better than the fuller's formula mixture used by the german group. The model prediction in de Larrard's book for this formula is for a density around 89% if I remember correctly.

Another formula which appears to be very good is in Hexion's mineral casting epoxy datasheet however the epoxy itself is only available in europe. Google for this reference or look through the thread index and find the reference. This presents Hexion's epoxy and their reference formula for aggregate:
HEX_HCI_556_Mineral_Casting.pdf

I've dragged out the calculation as an intellectual exercise as I would like to create a provably optimal solution using de Larrard's model.

The cost is $250 and the google books reference for de Larrard's work is:

http://books.google.com/books?id=tXH...age&q=&f=false

Of course, one could always work out the formula by trial and error but I doubt too many of us have the resources to do that.

The reference for Kinloch's book is the following. I have misquoted the title a number of times. I checked out the book from the library and don't have a copy as it is almost $300 U.S.

The Precise title is "Structural Adhesives: Developments in Resins and Primers" by A.J. Kinloch found in the following google books link:

http://books.google.com/books?id=2uo...age&q=&f=false

In general, not a day goes by that I don't think about E/G. I know however it is difficult for me to proceed quickly on the research because some expensive and hard to obtain items are needed to test theories and handle materials safely. The theory is a lot easier than the testing and even it requires some difficult work. I'm sitting at my computer today to work on a new model version that can correctly pack the Agsco aggregates.

Commercially available mixed size aggregates with proportions fixed by the production process are much more difficult to utilize than a large number of single sized aggregates with proportions determined by our own mixing.

Regards All,

Cameron

[ckelloug]

Why is it that your adding the particles to the epoxy exactly? In the candy making factories, when they're putting a "candy shell" on those little pieces of candy, they pour the "coating" onto a rotating batch of candies. All of 'em rolling around and bumping into and over one another makes sure that they all get a nice fairly even layer. Maybe something to this?

The particles provide a stiffness to the epoxy making it suitable as a structural material. Most epoxy on it's own is about 20% more flexible than pine. What you need to do is to get the particles to pack together with epoxy filling all of the space between particles. This is almost a 4000 post thread because minimizing the spaces between the particles is easier said than done. I personally don't have enough lab experience to say whether precoating the particles will help or not.

Has anyone used both vibration and vacuum on the mold yet? I'm not talking huge amounts of either, but if your using the slow cure (400 min) epoxy and have a larger mold, wouldn't the vibration and vacuum cause the mix to flow and de-air? Also haven't seen anyone talk of pouring in sections, IE. pour 1/3 vibrate/vacuum , repeat, repeat. Again, not alot of vibrating, just enough to flow out. Seems that the thiner layers would give up the air more easily.

Yove' got the right idea but between German Castillo's Ph.D. thesis and de Larrard's book, I think it's been proven by the big boys that it isn't just a little bit of vacuum and vibration that are required. It's huge amounts of vibration (between 2g and 4g) over the entire part required to get the particles to settle.

The problem with vacuum is that I have found in my lab that you need better than 29.9 inches hg of vacuum for it to effect much of anything, especially with materials higher than 300 cps viscosity. It gets really sticky (pardon the pun) when you realize that the parts most of us are trying to make will be in the 20-100 lbs range. Keep in mind that to hold tolerances in the machine tool range with vacuum, you need to have a mold that will hold .0001 inches tolerance over a large area with 15 psi vacuum pressure which creates stresses in the thousands of lbs of force.

Last idea and I'll leave ya'll alone. I see some mentions of vacuum bagging the mold and mix. That is probably the way I'll go, but have you thought about double baggin'? (reminds me of an old joke). What I mean is you can put the inner bag under vacuum, and the outer bag under pressure. Not alot, maybe 2 bar. That way your mix and mold are effectively under 30 psi instead of 15, Twice the compression (without 2X the vacuum) ought to eliminate those pesky air bubbles and if vibration is included as well, then maximum packing, minimum epoxy here we come!

I think trying to bag the outside with further pressure could have some technical difficulties. An ordinary vacuum bag will almost surely explode under pressure in my opinion. Vacuum bagging works because the substrate handles the mechanical forces while the bag provides a seal. I think for this to work, you'd have to either put your part into a tractor tire or a hard walled pressure chamber. The idea is captivating however and if you refine it further, please share.

Of course I'm only tossing around ideas here, and as always your mileage (as well as your ratios) may vary!:idea::idea:

Steve

Many ideas are tossed around here and some are just tossed but it is the process of doing that this makes the E/G thread an interesting productive and fun place to share work on E/G.

Keep the ideas flowing.:cheers:

Regards all,

Cameron

[MaxxD]

Hi Cameron,

Thanks for the quick reply! I guess I didn't word that quite right. I understand the reason for the particles, and that the strength comes from them. My question was more along the lines of "why are you adding the particles to the epoxy instead of the other way around?" I only ask because in the sweets business they add the coating to the particles as they are slowly spinning in a copper vessel (kinda like a cement mixer). If they put the coating in first and then add the particles, the particles tend to stick to each other instead of separating into individual pieces. I kinda thought with a very dry (low epoxy content) mix that this might be a way to de-air(if done under a vacuum) and ensure a complete coverage of the particles. As to whether the vacuum bag will hold or not, well I'm using 6mm poly and heat sealing my own edges with tape and a soldering iron. Can't afford the "real" vacuum bags. I don't think it will be that much pressure, but again I've been wrong plenty of times before. Also, I understand about the amount of pressure and vibration, and I probably didn't word that part correctly before either. What I meant was that the vibration and mold vacuuming should be done separatorly and in stages of increasing amounts. If your mix is under even a moderate vacuum I don't think the vibrations will be sufficent tocause your mix to flow and allow the smallest of particles to find their places. So I was proposing to do a sequence of vibrating and vacuuming at progressivley higher amounts until you reach the max on your equipment. Also I think I was wrong before about the amount of pressure exerted on the mix. If (and thats a to be seen "if") the outer bag can hold 30 psi and the vacuum bag approaches near -15 psi, then I think it would be like 45 psi, or around 3 tons per square foot! I'm gonna try it and see either way, just hope I don't have an e/g covered shop from the ex-implosion!

Steve

[robp1104]

This may be way too late in this discussion ...

When I was in a microelectronics lab many years ago, we mounted a micro-probe on a tombstone, 20"x30"x4" - since it wasn't inscibed or carved - pretty cheap, and an excellent surface finish. And I'm sure you could get bigger. Stone and synthetic stone Kitchen counter tops also spring to mind, although they aren't as thick (laminate two or three layers?)

Rob

[Steven.ji]

Hello people,

Been reading the last pages and think that we are on the right track:
Most important are:
Quality epoxy, mixture off aggregate and vibration 3G 0,25mm.

-------- -------- ------

Dear Friend

I appriciated your machine

With respect to your material, would you pls advice me

1. what's compressive strength of your epxoy granite?
2. what's bending strengh of your epoxy granite?
3. what's e-modulus of your epoxy granite/
4. is there any creep of your epoxy granite for a long time?
5. is there any air bubble of your epoxy granite?

[bookmarkmaster]

Abstract To maximize the productivity of precision products such as molds and dies, machine tools should be operated at high speeds without vibration. As the operation speeds of machine tools are increased, the vibration problem has become a major constraint of manufacturing of precision products. The two important functional requirements of machine tool bed for precision machine tools are high structural stiffness and high damping, which cannot be satisfied simultaneously if conventional metallic materials are used for bed structure because conventional high stiffness metals have low damping and vice versa. This paper presents the application of hybrid polymer concrete for precision machine tool beds. The hybrid polymer concrete bed composed of welded steel structure faces and polymer concrete core was designed and manufactured for a high-speed gantry type milling machine through static and dynamic analyses using finite element method. The developed hybrid machine tool bed showed good damping characteristics over wide range of frequency (η = 2.93–5.69%) and was stable during high speed machining process when the spindle angular speed and acceleration of slide were 35,000 rpm and 30 m/s2, respectively.

bookmarks