[ckelloug]

cnclite1,

ad_bfl was planning on working on the vibratory spectrum although he is a bit quiet of late.

The sizes that you have specified for ceramic spheres are much too large for optimum modulus and compressive strength and the size range is much too small. To keep the aggregate percentage up, one needs 4 orders of magnitude of difference between the biggest and smallest sizes. We have been thinking of using 3M zeeospheres which are tiny ceramic spheres for the smaller sizes.


There are published results in the A.J. Kinloch book that explain that increased aggregate size leads to larger critical flaws vis-a-vis griffith critical flaw size. Also, with the low size range, you'll have so much epoxy that the resultant material will not be very rigid. Epoxy is strong but has an abysmal modulus compared to aggregate.

Would recommend the strong glass fiber, s glass if I remember right. Also, make sure that the fiber length is about 1 inch as rule of thumb shows that this is almost as strong as full length fibers if they are aligned.

Most of the work we've done is with reichhold 37-127/37-606 available from shopmaninc as their low viscosity slow setting formulation. Neither is ideal due to low strength, low modulus, deairing problems and slightly higher viscosity than some other choices. I haven't been able to run to ground the other choices.


For aggregate, see
http://www.cnczone.com/forums/showpo...postcount=3241

For the relationship between epoxy percentage and stiffness aka modulus see:
http://www.cnczone.com/forums/showpo...postcount=3119

Regards all,

Cameron

[CncLite1]

Thanks for the recap on the prior work ... it a LONG thread and there is a lot of info sprinkled through it.

I am going to see what I can find ... something other then spherical aggregate ... given that spheres provide the min density in packing. I will also try to move to something porous to maximize bond strength.

A porous ceramic in Hexagonal form would be ideal I am thinking.

Additionally, I will work towards a method that gives reproducible fiber orientation given the constraints of aggregate and viscosity.

Thinking of beaded strands ... off to the chalkboard ...

[Markus53]

This post probably doesn't really fit well with the rest so not sure if this will be appreciated,... but i'll take my chances

There are some German guys that have started a small company making custom made epoxy granite bases. They started off experimenting a while back and they share what they are doing in the best know German cnc forum.

Here are some pictures they posted of a new base they made for a customer








I don't know these guys, just think the pictures are pretty cool

sorry for the giant image size

Greezt,

Markus

[vishnu]

Thats really good, have been following this thread for a little while its really nice to see good results. any more details regarding the composition and price. Its really encouraging novice engineers into making hitech gadgets. Well my suggestion to this thread would be to consolidate and put up a wiki for people to know the details and also can take a look at whats been finally done after all the technical discussions and hard work and time put into R&D for manufacturing composite castings.

[MaX-MoD]

I agree, a wiki would be neat

Hey, is there anyboby that tried to use traditionnal concrete out there?
I am building a lathe wt a concrete base, I'm interested by feedbacks of ppl that have tried this before

[ckelloug]

Markus,

The German group has been known to post over here now and then if it's the guys I'm thinking of. I can't think of the poster's name at the moment. Knowing what they are up to is always good here.

There are several folks here that have made some test parts but no complete machines yet. greybeard??? Most of the regulars here have day jobs. I'm just finishing 6 weeks at a remote site with no daytime access to phone or internet for my day job which is why I haven't posted any data lately.

The aggregate formula posted here earlier and the us composites low viscosity epoxy formula (Reichhold 37-127 with Reichold 37-606) is the most basic of formulas. I've been studying how to make a mixture that makes stiffer parts as this original mixture in my view is on the low side of the useful range of strength/stiffness. I've been trying to engineer a system that will provide as high or higher stiffness and strength than the commercial formulations.

Pardoning the pun, the number of concrete results is still small although the basic theory and advice from people with industry experience has been invaluable in making progress and learning about the process. I have two promising epoxy samples on my lab bench 2500 miles from my current location and will likely post more data on these epoxy formulations next week when I get back. I have a small materials test machine and accurate laboratory balances for measuring components as well as stocks of epoxy and additive samples.

Vishnu and Maxmod,

Since folks are talking about wikis, the most valuable help right now would be for someone to work on indexing the posts beyond the ones already indexed in the E/G thread index which is a sticky on this forum. I've written software that will take a file of hand created post summaries and link them to their posts for the index thread. This is a feature that would be nice for the jelsoft folks who wrote this bbs software to implement. . .

If anybody has a burning desire to contribute, I can show you how to edit the master index file and you can work on updating it, send it back, and I'll rerun my program on it and create and post the updated index. PM me for details.

While there is a lot of dead air here lately, my work and that of others I keep in touch with hasn't ground to a halt, it's just being slowed by real life. . .

Regards all,

Cameron

[brunog]

Markus,
I am sure the work on the photos psoted are from thomasz who has posted a few time on this thread. It looks like his second generation machine he was planing to build, i will find his web site address and post it.

Best regards

Bruno

[brunog]

Markus,
The photos are the FS2MG machine base from Thomas Zietz, you can see other photos on his website:
www.thomas-zietz.de

I know Thomas has posted a lot of information on the german forum: peters CNC-Ecke at:
www.cncecke.de

Best regards

Bruno

[ticica]

Cameron,

have you managed to get or make a larger fixture for your test machine? I am interested in making some samples for you to test, but the 2mm for largest aggregate size is a little bit limiting.

Regards,
Ivan

[ckelloug]

ticica,

I am finally back from perpetual travel.

I don't have a bigger test fixture but I do have a diamond tile saw with a sample cutting fixture I made. I can saw reasonably shaped samples to fit the 1/2 x 3/8 x 7 1/2 dimension the fixture is rated for. The process isn't perfect but it isn't too bad.

Glad to see folks still have an interest in this. It's definitely a hobby that sounds suspiciously like work

Regards all,
Cameron

[Mel White]

Marcus
Do you have a contact information on the castings that you had on 9/4?

Mel White

[ticica]

ticica,

I don't have a bigger test fixture but I do have a diamond tile saw with a sample cutting fixture I made. I can saw reasonably shaped samples to fit the 1/2 x 3/8 x 7 1/2 dimension the fixture is rated for. The process isn't perfect but it isn't too bad.

Regards all,
Cameron

What size mold should I make to make it convenient for you and allow for having 15mm diameter rocks?
Is it true that cutting such non-uniform material will not disrupt the strength in a way that we are not accounting for? Or, actually, it might be even better a predictor of strength than a sample in which mold walls 'pushed' aggregate away...

Ivan

[ckelloug]

Hi Ivan,

The general rule from de Larrard's book on particle packing is that the container needs to be 5 times the diameter of the largest aggregate pieces in order to avoid the result being dominated by the effects of the edge of the mold. This means a rod about 75 millimeters diameter in your case and about 200mm long. I haven't thought up a strategy yet for cutting test samples from such a piece but if you cast one and send it, I'll figure it out.

I haven't started handling the aggregates yet (though I have tested several other peoples' samples with aggregate) as I'm not satisfied with the epoxy formulations I have been working with. I have been doing my epoxy work with a 1/2 wide by 6 inch deep by 7 1/2 inch long open topped mold made from UHMW polyethylene.

The theory and some published test results suggest that the larger the aggregate pieces are, the lower the overall strength of the piece due to the effects of crack propagation in line with the results predicted by the Griffith Model. The inverse correlation between strength and particle size was published in A.J. Kinloch's book titled IIRC, "Advances in Structural Resins and Primers". Nobody in our merry band has tried aggregate pieces 15 mm in diameter.

The other rule of thumb in getting the maximum rigidity aka stiffness aka modulus is that you want 4 orders of magnitude of difference between your largest and smallest particles to ensure a large aggregate percentage and thus a part that will have a high modulus e.g. be stiff.

Whether sectioning samples on a diamond saw provides a positive or negative effect on the data is not fully known experimentally. Theory and some of my testing shows a negative effect caused by having exposed aggregate on the bottom of the sample. Epoxy generally has a much higher tensile strength than aggregate and even a thin layer of epoxy that is contiguous along the bottom of a sample makes for a large increase in ultimate flexural strength.

The theory above filtered through my best judgment suggests that the flexural strength will be lower for a sawed sample than a cast one while the flexural modulus should be similar.

Regards all,

Cameron

[Calthorp]

Hey you guys got me thinking!

How about using epoxy with aluminium or other metal powder as a cold castable that can be machined as a metal.
Wonder does it have some of the properties of the metal powder or would it be completly different

[probinson]

Sure, You can use any filler you like to create your own composite, it just won't behave like EG. One of the most important properties of EG is dimensional stability. If you substitute fillers that are resilient or have large thermal expansion characteristics you will lose that property but gain others.

Pete

[ckelloug]

Calthorp,

What you are suggesting has been done and is commercially available. PTM&W makes these here in the states. See http://www.ptm-w.com/index.asp?pgid=21

The machinability of these compounds is an advantage over E/G. One based on cast iron would probably be nice. One based on aluminum probably wouldn't be dimensionally stable enough to be that great for high precision tools. I haven't researched these metal epoxy compounds but I know that they are used in low-to medium production run die stamping where they are apparently a big cost savings.

Regards,

Cameron

[Calthorp]

Thanks guys,

They probably got it here in NZ but I am off to the hardware store to get some epoxy & have a play with different powders The local quary has plenty of crusher dust so can give graywacke-epoxy a go too (no granite close by) have a sculptor friend that uses granite but her dust would be probably 10% grinding wheel & too fine.

Cal

[ckelloug]

Hi all,

I've finally run the first remotely successful sample of Hexion Epon 813 /Epikure 3046. The mix ratio is 1.5:1 by mass of 813/3046.

The ASTM D790 result was:
Flexural Strength 9080psi plus or minus 300psi
Flexural Modulus 252ksi plus or minus 31ksi

The Hexion 813/3046 combo is still not as good as the Reichhold 37-127/37-606 we started with and has the added disadvantage of being quite sensitive to mix ratio. The sensitivity is made worse by the fact that Hexion does not publish an amine hydrogen equivalent weight for the 3046 hardener meaning that the final mix ratio has to be determined by testing.


I also tried 1.5:1 +5% of resin mass as Accelerator 3046 but the batch failed to harden fully and it wasn't worth starting the formal test.



I also tried a 1.75 : 1 813/3046 plus 5% of resin weight as Epikure 3253 tertiary amine accelerator. The results were not good enough to finish running the tests but the strength appears around 2000psi and the modulus about 62.5 ksi.

I'm currently waiting for the cure of a 1.6:1 813/3046 plus 5% 3253. I'm hoping this is better. The addition of 3253 at the 5% level seems to induce some homeopolymerization in the epoxy resulting in a change in the required mix ratio. I have tried to compensate by adding a small excess of epoxy to account for it but 15% extra by mass is too much.

I realize that this data is about as fun as ... well as fun as watching slow curing epoxy dry. For the intrepid machine builders, the Low viscosity slow curing stuff from US Composites is still the best stuff I've tested so far.

I suppose it is about time to try hexion resin with reichhold hardener and visa versa to try to determine which part of this mixture sucks and then to get the fume hood and break out the isophorone diamine.

Regards all,

Cameron

[bloefeld]

I'm sorry to jump in so far from the start of the thread, but epoxy is my life :wee:

If you can give me a bit of background on your objectives, I am sure I can give you some good and inexpensive solutions.

Obviously you are a sophisticated user of epoxies, and from the looks of your experiments are using high quality products.

I am a big fan of pure Bis A or Novolac systems. Huntsman makes a line of jeffamine curatives and accelerators that create systems that yield numbers considerably higher than what you are showing here. I can give you some stoichiometry to work from.

It seems to me from the little bits of the string I have read that you are going to need some filler materials to help stabilize the material as it cures and to reduce the chances of stress cracking. You will also probably need to cure in heat and then do a post cure heat to get maximum performance from the materials.

I would really like to help out on this project, and if you can give me a small amount of background, I'm sure that I have some value to add.

Cheers,

Bloefeld

[bloefeld]

Hi,

I am new to this thread, so please bear with me.

It seems that the idea is to make a composite base for machines that has many of the same properties of cast iron but at lower weight and cost.

The primary issue I can see in using polymer materials is the differential thermal expansion properties of the materials. The metal components will expand and contract at least one and perhaps two orders of magnitude less than the polymer components.

A new approach may be to make the machine bases out of vacuum bagged composite materials and allow the tracks, screws etc. to 'float' on them. This would allow the use of honeycomb or balsa core materials that have excellent shear properties and when made into thick enough panels can be incredibly stiff.

Another approach is to make a composite using metal fibre reinforcements; the increased modulus of the metals will constrain the thermal dynamics of the polymer materials.

Fillers, such as wolastanite, that have very high aspect ratio's are inexpensive and improve the stability of the epoxy and increase the strength to the point that many of the resulting properties are similar to those of ceramics.

Ceramic whiskers are expensive, but small amounts increase physical properties in almost unbelievable ways, as do some of the various nano clays and carbons.

Inhance Inc, is a source of functional fillers that can improve mechanical properties and improve damping. They make a Titanium Carbide powder, bonded to functionalised UHMW that makes a very dense, yet harmonically stable composite from a wide array of thermo-set materials.

Technique is also important; making parts under vacuum and properly controlling the exothermic reaction of the thermo-set materials as well as post-curing the resulting part for a long-enough period of time will fully cure the structure. Without these steps, the thermo-set systems never completely cure, and will shrink at a slowly reducing rate for decades.

None of these issues is that problematic and can be easily addressed in relatively simple ways; providing your wife doesn't mind having epoxy in her oven for days on end.

Cheers,

Bloefeld