[ckelloug]

Attached are plots of filler modulus vs. filler percentage using average iron at 15e6 psi and average steel at 30e6psi modulus.

For reference the moduli for the commonly discussed materials are approximately:
Quartz 10e6psi
Iron 15e6psi
Steel 30e6psi
Alumina 40e6psi
Silicon Carbide 60e6psi

What the graphs tells us is that to get a material that is as stiff as the 4.9e6psi modulus posted by Accures it needs to have 87% quartz, 86% cast iron, 84% steel, or about 83% silicon carbide by volume. For stiffness, the material chosen doesn't tend to make a big difference until the fill rate gets into the 90% plus range.

I would suspect that steel and iron compositions have a higher tensile strength than the mineral compositions and I also suspect that they will fail in the bond rather than in the filler itself.

In the improvised machine case, I would suspect silicon carbide to be nice because it has a very high modulus and is thus the most likely to have acceptable stiffness performance in the face of questionable gradings and inadequate vibration.

Until the gradings are so good that one achieves a density in the >90% range, the ultimate performance of the fillers as far as stiffness is similar for all the listed materials. Since I don't believe strength will be an issue since a part stiff enough for a machine tool should be strong, I'm going to say that price seems like the chief discriminant to me and that metal particles appear to impart a slight disadvantage when compared to silicon carbide or alumina.

Regards all,
Cameron

[lazlo]

The original mixture of using 1/5th or each aggregate is prone to segregation when vibrated.

I don't have the post number at hand but the post I recently made about the improved mixture is likely to behave better under sufficient vibration.

Yeah, but you're suggesting that you need to vibrate the mold at over 2g for the improved mixture, right? I don't have 400 lb vibrator...

Wow, that can be quoted out of context

[ckelloug]

Lazlo,

I think the operative thing is that you want to vibrate any mixture at 2G according to the French paper. Either the old or new mixture will obtain a higher density when vibrated.

It doesn't really matter what you do if all you want to do is fill something. The problem comes if you are making a structural part that needs a modulus greater than that of pine.

Regards,

Cameron

[jhudler]

Lazlo,

I think the operative thing is that you want to vibrate any mixture at 2G according to the French paper.

Remember this nugget I found.
Note: They have taken the page down... wonder why?

There maximum acceleration works out to 3.2 G's

[ckelloug]

Jack,

I was talking about the French PhD thesis posted by sardyne at www.Usinages.com rather than the one you just referred to. MaxMod brought it to our attention. I read it and summarized it very briefly a few dozen posts back. I'm attaching it here for reference but it is in French.

My understanding of the thesis was that he said that 2g is the minimum that worked well in his experiment. I suspect that at least under some circumstances, more might be better.

So we have the following acceleration specs from different literature:

German Castillo's thesis: 2g
Epucret paper on Knauer CNC vibrator: 3.2g
Francois De Larrard's Book: 4g

All I can say from experience is that my little tiny 1/2amp 120V bin vibrator didn't do much in experiments and I haven't fired up the 7000lb force model so I don't know how samples with higher vibration will do.

Regards all,
Cameron

[lazlo]

All I can say from experience is that my little tiny 1/2amp 120V bin vibrator didn't do much in experiments and I haven't fired up the 7000lb force model so I don't know how samples with higher vibration will do.

I'm guessing you're talking about the 1/2A Harbor Freight vibrator. Is there any intermediate solution between that and the 400lb "Tool Time" vibrator?

[ckelloug]

Lazlo,

The little vibrator I was talking about was a Syntron V6B surplus off of a vacuum chamber I got on e-bay. The tooltime version is an INVICTA BK45-3578. The former is too small but the latter was available surplus at the right price. The right vibrator is undoubtedly in between the two sizes but us scavengers take what we can get.

Off to catch a plane!
Regards all,

Cameron

[jhudler]

Here's what I worked out for vibrating 100,250, and 500 lbs (remember to add the weight of your table and vibrator).

Using .01" displacement as the maximum.

100 LBS Mass
80Hz (4800 rpm) - 840 ft/lbs or 1 1/2 horsepower.
40Hz (2400 rpm) - 200 ft/lbs or 1/3 hp.
20Hz (1200 rpm) - 50 ft/lbs or 1/10 hp.

250 LBS Mass
80Hz (4800 rpm) - 2100 ft/lbs or 4 hp.
40Hz (2400 rpm) - 530 ft/lbs or 1 hp.
20Hz (1200 rpm) - 130 ft/lbs or 1/4 hp.

500 LBS Mass
80Hz (4800 rpm) - 4200 ft/lbs or 8 hp.
40Hz (2400 rpm) - 1100 ft/lbs or 2 hp.
20Hz (1200 rpm) - 261 ft/lbs or 1/2 hp.

Personally I think 80Hz for a HSM is ambitious.

Jack

[J13]

I guess I'll use e/g as fill, not structure. thanks for that graph.

[jhudler]

If you're creating something structural, then fill just won't do.

Steel fill has it's own inherent problems, namely getting it clean of oils and preparing its surface to bond with epoxy. Something minerals does quite nicely.

On the subject of steel, make sure you isolate axes from each other. Or in other words don't have internal steel reinforcement that would transmit vibration from part of the machine to another. This would defeat the damping that EG provides.

Jack

[Mbk]

This thread captured my attention.

Usually for my customers I sell big bases for machine tools in grey cast iron or in welded steel.
I didn't know about this material.

Can be competitive as price in big weights around Kg 30000 (30 Tons)?
Base plates in cast iron and in welded steel need also thermal treatment and an expensive job of CNC machining.

From I have read in this thread the casting of this material can be very precise.
The problem would be usually on the guideways of the machine bases need an induction tempering and you can't make this treatment on bases in epoxy granite.

Any suggestion?

[jhudler]

From I have read in this thread the casting of this material can be very precise.
The problem would be usually on the guideways of the machine bases need an induction tempering and you can't make this treatment on bases in epoxy granite.

Any suggestion?

Welcome the to group!

You can embed the guideways in the casting, or embed mounting blocks that are threaded for the guideways. In the later case, you can place solid mounting blocks and thread after casting.

Search the web for examples of commercial work.

[Mbk]

I'm looking at ZANITE faq and they claim the costs for finished cast parts are between $0.75/lb. to $100.00/lb.
I don't get that $100.00/lb.

Haven't understood very well also the cost of machining ZANITE, because bases usually I deal need tolerances up to 0.01/1000 mm.

[jhudler]

Post machining of composite casting is always expensive.
Generally you design the mold and the parts to be embedded to the tolerances you need.
Or you can leave a hole or groove and grout the part in after casting.

The pricing is based on what you're building, the type of mold needed, precision, and quantity. If you were making a one of a kind machine tool with high precision, then I can see the price approaching or exceeding $100/lb.

Same issues apply to DIY casting as well.

Jack

[Mbk]

Quantities can be interesting: usually I need 30-40 machine tool bases/year of weight between 20000-70000 lbs.
These big pieces in cast iron usually require a part cleaning and sandblasting, thermal treatment, an induction hardening on the guideways, a rough machining, a drilling job and a precise CNC machining on length up to 9-10000 mm with tolerances to keep between 0,05-0,01/1000 mm. So a grinding job sometime is necessary.

I haven't understood very well if with this epoxy granite or similar material I can save money.
Now as now I pay around Euro/Kg 1,27 for the fusion, thermal treatment and sandblasting plus around Euro/Kg 0.9 for the fully machining on bigger pieces.
So the total EXW is around Euro/Kg 2,17.
Need to pay less with this alternative material, but don't know if it can be.

[Mbk]

We are using machines with Polymere-Concrete frames for years.
These are the best machines we ever had. Very stable and high accuracy.

Epucret is a frame manufacturer for Mikron/Hermle etc.

http://www.epucret.de/files/EPU_News_NR_26.pdf

Guideways of our Mikrons seems to be fixed with Epoxy grout after alignment.

Do you know some cost about the casting made by Epucret?

[snookered1ca]

Regarding Cameron's post # 3597, I posted a response from PTM&W quite some time ago on the concept of using steel shot as a filler for the resin but maybe it got lost in the discussion at that time. Their recommendation to me was very simple. Place a quantity of small steel shot in the mold and then pour in the resin until the shot was covered, then repeat until the mold was filled. They didn't mention anything about degassing but I imagine that would be a necessary step. The product they recommended was PT4450.

Bruce

[ckelloug]

Hi all,

I just got back from the UK after a nice meeting with greybeard John. I also managed to order a copy of B.W. Staynes 1972 Ph.D. thesis on Epoxy Granite from the British Library's Ethos service. I'll share after they get done digitizing it.


Bruce,

Thanks for reiterating your post from PTMW. My efforts at thread indexing are about 1600 posts behind right now so I don't always remember everything posted.

Mbk,

Those are some serious parts that you are interested in making. E/G requires some different equipment and thinking than metal parts. I know that huge pieces are done regularly by epucret and Studer in Germany, also by the company that cnczone poster roach works for in the UK and Accures and Zanite here in the U.S. Studer's patent has expired so one can duplicate their material for large machine bases from their patent.

To make the kind of machinery you are interested in will require a capital outlay of vibrators and mixing equipment on a grand scale. I'd estimate that the costs would be about $50,000 to $100,000US for the mixing and vibrating equipment to handle materials on the scale you are talking about. I'd also estimate that there would could be substantial costs to make the mold for the parts. See http://www.accurescasting.com/polymer.html

Rather than surface grinding the parts themselves in E/G, it is preferable to grind the mold so that the mold accurately represents the part you wish to cast. One also would normally drill the mold and accurately place dowel pins so that the holes were cast in place. Precision attachment points are usually best achieved by embedding metallic hardware into the castings. One way of making a sliding bearing would be to embed precision ground metal parts into the casting.

Professor Slocum from MIT recommends that bearing surfaces directly subjected to rolling loads not be made from epoxy due to wear problems. As a result, it's my understanding that most commercial E/G frames use metallic precision linear bearings. There is moglice however which is used to make objects such as acme nuts and in the repair of sliding machine ways.

I think that some of this problem might be overcome by innovative use of ultrahard materials but whether this will work is an open question. I'd suspect that the rolling load issue are more of a problem in the huge machines you want to build as opposed to the multi-hundred pound stuff in my interest range.

As for the $100.00 a pound cost estimate for the upper limit on Zanite's site, I'd suspect that it either was meant to be $1.00 or it includes intricate hard steel molds for a specific unique part with only one unit produced. Based on the raw materials costs, I would guess that around $3.00 a pound is probably the absolute upper limit on the cost of the materials themselves and that would indeed be a strange part.

I would assume that excluding capital costs, a normal industrial scale part could be made for less than $0.25 a pound. Most E/G parts require no post casting operations so once the epoxy has set over the course of a day or two and possible low temperature post casting heat treatments, the part is complete.

One of the problems you will not face is that of making sure to maximize the material properties. When you are making a piece that is 70000lbs, the cross section of the part alone will almost certainly ensure sufficient strength and rigidity. I am interested in making small utility machines that weight only hundreds of pounds and the small cross sections involved mean that slight variations in material have huge consequences in terms of stiffness. For tiny machines, maximum stiffness also minimizes shipping costs which are probably one of the killers in the market for small machines.

At any rate, please feel free to continue posting here about what you are interested in or to PM any of the posters. Thanks for sharing your aspirations.

Regards all,
Cameron

[greybeard]

Glad to hear the BL came up with the Staynes info, and that they hadn't closed the borders before you got back
John

[Mbk]

Need also to understand if I can use this E/G for T slotted surface plates.
Usually these plates are made in gray cast iron EN-GJL-250 where the levelers in in gray cast iron EN-GJL-350.
The requested loading capacity can vary between 15000 to 35000 Kg/square meter.

Considering the great difference compressive strength of the gray cast iron and of the E/G, don't know in this case if the project is workable.