[dfro]

In polymer concrete, the specific heat is very high, and coefficient of heat conduction is very low. That mean it's not easy to conduct heat to outside, then temperature of headstock will increase quickly, which will result precision lost.

How do you deal with this problem above?

Steven,
You bring up a good point. I have not considered heat dissipation.
The spindle will run at a modest 2,600 maximum rpm. I do not plan on running it for long durations, either. So, I do not see it being an issue.

To me, every choice in design is a trade-off. One material has good heat dissipation, but it is not very rigid - aluminum. Or one material is very strong, but it rings - steel. And cost is always a factor.

I began with a cheap, thin walled headstock casting, with thousands of tiny bubble voids in the casting, and only three holes to bolt it to the lathe bed. With e/g, I turned it into a massive, rigid, vibration damping, solid headstock, with four holes to bolt it to the lathe bed.

I think any loss of precision due to poor heat dissipation will be offset by the added advantages.

However, if heat dissipation proves to be a problem, I could see replacing the cheap, plastic covers that keep dust and grim out of the bearings. I could make water tight covers with a hose nipple on each one that allow me to pump lubricating coolant through the bearings and the headstock.

We will see, once I get it working.

Thanks,
Dave

[lgalla]

Dave,we are not building machines to produce Hubbel mirrors in climate controlled rooms with passport ackess.We are not running 24/7.I must agree with your post 100%.The engineers manage to convince the shop guys or the hands on machinist that their ideas are no good.I have experience where we the shop guys saved the engineers asses.Lets say engineering wanted 10% ep0xy.The shop guys know this was impossible so we tip the scales a little at 15 % and things worked out OK.When invintory asks of the 5 %,We had a spill.The number crunchers can prove our designs are poo and can stop us from persuing new ideas.I am guilty of listening to the number crunchers and stopped thinking about new ideas.Hats off to you for.forging on
Larry

[Steven.ji]

Steven,
You bring up a good point. I have not considered heat dissipation.
The spindle will run at a modest 2,600 maximum rpm. I do not plan on running damping, solid headstock, with four holes to bolt it to the lathe bed.

-----------------------
to pump lubricating coolant through the bearings and the headstock.

We will see, once I get it working.

Thanks,
Dave

Is is possible to cast condenser pipe in side?

In terms of your method, can you present by picture?


Regards

Steven

[BillTodd]

Is is possible to cast condenser pipe in side?

In terms of your method, can you present by picture?


Regards

Steven

That's a good idea Steven

A heat pipe like that would shift the heat to the outside, to a heat-sink of some description, really well.

It might be possible to cast copper rods or blocks into the EG for the same purpose, if the cost of a custom heat-pipe were prohibitive.

Bill

[ckelloug]

Larry,

There is occasionally a reason for engineering. There are also reasons for tried and true shop methods. Totally disregarding engineering is an unwise outlook I do however agree with you that fretting over the mix ratios for E/G is probably not crucial on a small machine.

For quartz, 20% epoxy gives Young's modulus as about 2.5e6 psi which is a touch better than the modulus of yellow pine. The difference between 80% and 90% fill is a factor of two in modulus(stiffness).

For a small machine, the packing density won't likely matter as the forces involved are very small and the distances are small. For a heavy part that is long, it will matter as deflection in a beam is proportional to the length cubed and this term quickly overwhelms the stiffness of the material.

Dave is doing a great job on his machine and producing something. The parts he has shown us are small and as a result, they aren't likely to have any problems with low modulus. They might even get a slight bonus to damping due to the low fill content.

I apologize if I have disuaded anyone from building E/G stuff as my goal is to get the formulas right and design commercial quality material. I don't have a machine to make right now.

If I can be of assistance, I'm glad to help, otherwise, I'm just going to take a backseat on this thread until I have some more results.

Regards all,

Cameron

[dfro]

Steven,
I think you have a good idea with the condenser pipes cast into the headstock. One of the great advantages of e/g is that you can easily cast all of these things into the part. Putting condenser channels like you show into a solid metal part would require a 3D printer that prints metal ($$$$).

You could also directly circulate lubrication/coolant through the bearings and the cavity in the headstock that houses the spindle. You would need to make sure that the dust caps on each side are watertight, though. An input tube could be cast near the front headstock bearing, and an output tube could be cast near the rear headstock bearing. Again, very easy to do with e/g.

Cameron and Larry,
I agree with both of you. I think the abstract/mathematical world of the engineer and the real/practical world of the builder end up finding some equilibrium when designing and building a machine. I am glad Steven brought up heat dissipation. I never considered heat dissipation being a problem. And, in a few days we have come up with a couple very practical and easy solutions, in my opinion.

Cameron,
I think a second line of tests could really boost people's confidence in trying to build with e/g. I am sure you and others are going to continue to optimize the e/g formula, but I think it is ready, as is, to be used as a building material. Yes, my 20% mix may be only as strong as yellow pine as far as the tensile strength, but it is massive and rigid and vibration damping - pine is not.

Making the e/g easier to work with by adding a higher percentage of epoxy causes a loss of tensile strength, but adding embedded steel rod (rebar) can offset this loss. It is very easy to embed steel rod into an e/g casting and thus dramatically increase its tensile strength. A line of break tests on e/g samples with embedded metal rods of different size would be of great value, I think. A set of parameters and guidelines could then be developed to help people design machines - i.e. how many rods to put in a casting of a certain size, how close to place adjacent rods, what diameter rod is necessary, loss of damping as the rod density does up in the casting, etc.

We could end up with something like those great CAD drawn guidelines that the German e/g builder shared on this thread.

Also, I believe in starting slow and small. After my lathe, I think I will build a little gantry mill with a 6" x 6" table, mainly to mill circuit boards. With small, slow machine build projects, I gain big in experience. If it doesn't work out, I have not wasted a lot of epoxy, aggregate, and other materials.

Thanks,
Dave

[stevv05]

This is my first post on this site. Bear with me. I have been following this thread a long time. I have a top tek DM45 Mill. I have already filled the base completly with the E-G mix. I am using West Systems 105, 205, 404. Included some cast iron piecies for good measure. It can't hurt. The column is next and then the lathe. I am in the early stages of this hobby. Hope to learn more

Steve05

[lgalla]

Cameron,
U are the number one E/G guy!Please continue the good work and post.
Larry

[RotarySMP]

I agree that it was great that Steven brought up heat dissipation, as it belongs discussed in this thread.

For your specific application, I would not start redesigning anything to fix a problem until I had run the thing and seen if the problem even exists. Since you already own the parts, and there is little or no opportunity cost involved in down time, there is no reason to start wasting your time fixing theoretical problems at the moment.

Given the cost of components (a pair of very cheap off the shelf bearings), you may even find that the problem exists, but if it does not cause any problem on parts made, you may never fix it. The solution in that case may be simply to accept poor bearing life, and replace the things each year or so.

Since the standard bearings Sieg fits are crappy Chinese ones, a swap to decent western name brand bearings might also be a final solution.

[RotarySMP]

deleted duplicate

[dfro]

For your specific application, I would not start redesigning anything to fix a problem until I had run the thing and seen if the problem even exists. Since you already own the parts, and there is little or no opportunity cost involved in down time, there is no reason to start wasting your time fixing theoretical problems at the moment.

That is definitely my plan.

A condenser pipe would be an easy addition to an e/g headstock design. Adding water tight dust caps and running coolant/lube through the bearings and the spindle cavity would be more involved, but very doable, I think.

Thanks,
Dave

[oxford]

Mike from Precision Epoxy wrote back to me today.

Had some promising things to say about their Santex System epoxy for the bulk of the base:

http://www.precisionepoxy.com/SantexSystem.htm

Said they had an excellent self leveling epoxy to put on the top of that
Offered to give me a call, but my phone lines are down since last storm

It will be at least another ten days before I can call him.

Any thoughts on the material in his link?

[johnohara]

Hi Everyone~

I've been researching E/G for a number of years. Research papers, expired patents, published articles, and of course, this thread.

My test samples are getting better and the Aggregate:Epoxy ratio is getting closer to the ideal (93%:7%) each time.

This thread has grown quite large and contains a tremendous amount of theoretical and practical information on how to make E/G. But actual instances of diy E/G still only make up a small percentage of the overall number of posts.

I'm not complaining nor am I being critical.

I would like to propose to everyone here -- including the most active members and posters -- that an additional thread be created for E/G called "Practical Epoxy-Granite Usage" which contains posts from members who have actually cast something using E/G.

It could be something as simple as casting in a plastic bowl to something as complex as that made by ITW, Epucret, and others. I envision postings that include pictures of attempts, discussions of problems in the process, tricks with moldmaking, actual experiences with materials, vendors used, etc. The more pictures and videos the better. All mask and gloves type stuff.

Please let me know what you think.

~John

[stevv05]

John
I agree. Too much talk and no action. Some of the information is very helpful and some is useless in my real world. I have read most of the posts, I came to a decision and started to physically work on my projects.


My square column mill is at the bottom of the list for anything. So now I have a plan to improve it to some degree.

I thank all who have contribute to this thread.

Steve

[RomanLini]

I'm interested in the possibility of using much larger granite pieces in the mix. That should increase stiffness and allow a higher ratio of epoxy in the mix, making it easier to pour and de-gas.

I'm thinking something like (or actually use) cracked stone roadbase of marble sized pieces or larger.

[RotarySMP]

The recommendation of the the German E/G book (I posted the ISBN number about 2000 posts back) was to have your biggest aggregate 1/5 the thickness of your thinnest feature

[greybeard]

I'm interested in the possibility of using much larger granite pieces in the mix. That should increase stiffness ......

I'm sorry, Romanlini, but that would have completely the opposite effect.

While I can understand the impatience of the hands-on guys who want to turn Walter's original question into a result on the bench, all the necessary details on what would be the best approach are already here.

It's only going to get easier if the manufacturers of the resins come out with an entirely new formulation that pours like water without any change in volume as it polymerises.

Till then, the research on the diy method is aimed at small changes in additives that will give disproportionately large improvements in strength, and modes of handling that give a better cast.
Regards,
John

[brunog]

John,Cameron,
Speaking of low viscosity, I found this site last week, they have 2 products that look interesting.

The low viscosity has a compressive strength of 11,000 psi and viscosity of 175 cps.
The interesting characteristic of this epoxy is that it clings to wet aggregate.

The No Blush product is what they would recommend for EG and has various thicknesses depending on the hardener used.

Please let me know what you think of the specs.

Best regards,

Bruno

[ckelloug]

Bruno,

The no-blush stuff isn't as strong as the other stuff I've tested and I bet it has a lower modulus. I would guess that it is an epoxy reactively diluted with C12-C14 glycidyl ether.

The Low-V stuff looks pretty commensurate in flexural modulus and flexural strength to other stuff I have tested. I would guess that this is reactively diluted with cresyl glycidyl ether.

Of the two, the Low-V stuff is better.

In general, the epoxy is not the sensitive part of the formula to get maximum stiffness aka modulus. High aggregate fill is the sensitive part. If you're not worried about modulus, almost anything will work.

Assuming an unoptimized aggregate mix with a packing density of about 75%-80%, any of the epoxies that have ever been suggested here will give you a flexural modulus of about 2,000,000 psi and a flexural strength around 2000psi. This is similar to yellow pine. For a small machine, none of it will matter.

RomanLini,

Bigger aggregate particles won't increase modulus. An aggregate distribution with larger maximum size is easier to get a high packing density for high modulus but if you just use a bunch of large particles without the smaller sizes, you will get a decrease in modulus. Larger particles also lower the strength as they are more likely to contain a flaw of the Griffith critical size. RotarySMP Mark is right that 1/5 of feature size is a good rule of thumb for max aggregate size.

stevv05,
It might be good to post a practical experience thread. This thread is very long and a new thread will likely get more visibility initially. It will also likely disappear quickly however as there are not a lot of people building E/G stuff yet.

oxford,

The Santex system worries me because no physical data are given for the properties of the system. That doesn't mean it is bad but I like to know what I am getting. Per my comment above to bruno however, for building a small machine, almost any epoxy will work unless you really need the modulus of an optimal mixture.

johnohara,

Can you post anything about the specifics of what you have done? I'm also happy to do flexural strength and flexural modulus tests on samples if you are interested.

dfro Dave,
Kudos for your work on the lathe headstock. You are on the right track. Can you post what mixture you have been using?

To the E/G crew here:

In general, I've been lurking here more than posting lately as I've noticed that my presence tends to drastically cut down on discussion. I don't have much to report anyway. I'm still working on acquiring a vibrating table and getting back to work on E/G but idiotic paperwork problems are taking up my research time.

Please don't let the more technical discussions here dissuade anyone from working on E/G or posting what you have done here. I would especially encourage posting the mixtures you are using and what you did. My goal was to produce optimal material but others have machines they need to build.

Not that I've had any takers before but our thread index needs updating. If anybody wants to index that last 2000 posts I can show you how to make a file that my custom index generator software can process.

Regards all,
Cameron

[johnohara]

Here are some pics of my latest attempt.



1. Sides are shiney from the cast. Nothing modified. The top has been sanded using a belt sander and then polished using 3000 grit (same as a countertop).



2. Close-up of the top showing aggregate distribution.



3. Close-up from the side showing the glossy finish.



4. Coin for comparison. Note the "bug holes" below and to the right. This piece was hand mixed. I vibrated it by tapping the sides and top of a plastic bowl with my fingers.



5. The underside of the piece -- actually the top when it was cast. I used a heat gun to clear the surface bubbles.

This casting weighs 590 grams and is 16% epoxy by weight.