[Kevin_Johnson]

Hi Cameron.

I've picked up on this particular point as it is exactly what my current "project" entails.
Rather than go hopelessly OT, I'm going to start a new thread devoted to the possibilities of creating small gears from some variation on E/G, using spin casting.
For the project, I need to be able to produce batches of 30+ worm drives components, and needless to say it will start with some of my more off-the-wall ideas.

John

If you have them correctly oriented in the mold with respect to thread handedness and machine spin you can use the Archimedean screw effect to get the particles to pack themselves under multiple Gs of pressure as well as force out less dense air and excess epoxy. A radial or parallel setup could be used. Parallel would probably be better overall given the ready access to large diameter round tubing. CNC cut and indexed endplates would be straight forward and fairly inexpensive to have made. You could direct drive the outer mold tube/case with a simple belt, surplus automotive belt tensioner and variable speed electric motor. A threaded torsion rod along the central axis of each gear with an appropriate mold structure will allow you to place the cured gears under consistent preloaded longitudinal compression and greatly increase their strength as well as provide convenient means of mounting them. A larger longitudinal central rod could be fitted into mounted bearings and a frame to hold the mold position constant with respect to the drive motor.

Balance is going to be a safety issue with the mold itself and so you might want to consider correcting this before use. The frame and drive motor could be used to spin cast other items using a similar mold structure but perhaps different length belts.

[greybeard]

Kevin, I'm going to have to read that several more times....and do some drawings.
There are several ideas in there packed very tightly together
Regards
John

[Kevin_Johnson]

The more you increase the angle (towards the center) the greater the net downward compacting force in the worm gear helix in addition to gravity. Make sure the mold casing is spinning the correct way with respect to the handedness of the helix/spiral. Test with water.

Edit: also clock the top plate in the same direction of the spin with respect to the bottom plate for more compaction. You have the choice for infusion of epoxy into the dry compacted skeleton through the gate. Huge number of ways to approach it.

[greybeard]

Kevin,
As I said, I don't want to go OT as this thread is complex enough as it is.
But I will just say that, if I interpret your sketch correctly, the radial force on each of the worm molds will throw the filler into one side of the worm rather than uniformly around the surface.
My thread dedicated to the possibilities is http://www.cnczone.com/forums/epoxy_...ar_making.html .
I have started on the worm gear, that's the pinion type part, as it might be of more general interest, rather than the worm shaft which presents different problems.

[Kevin_Johnson]

Kevin,
As I said, I don't want to go OT as this thread is complex enough as it is.
But I will just say that, if I interpret your sketch correctly, the radial force on each of the worm molds will throw the filler into one side of the worm rather than uniformly around the surface.
My thread dedicated to the possibilities is http://www.cnczone.com/forums/epoxy_...ar_making.html .
I have started on the worm gear, that's the pinion type part, as it might be of more general interest, rather than the worm shaft which presents different problems.

There is a net downward packing force distributed around the gear when its orientation in the mold is adjusted for the direction of spin. The outer half of the worm gear has greater centripetal acceleration than the inner half: given that the inclined plane opposes itself equally from inner to outer half because of the handedness of rotation if moving in a straight line under constant acceleration then the higher centripetal acceleration of moving in an incrementally larger radius arc will upset this balance. A sort of mental FEA.

You could orient the worm gears radially, as I mentioned, but the economics of sourcing premade tubing of radii consistent with typically useful worm gear lengths is discouraging (say a meter/yard long). Letting the length of a smaller diameter tube carry this burden seems more practical for the hobbyist and safer as well.

You could spin one worm gear along its central axis in a mold (or perhaps a couple more?) but you mentioned that you wanted to run batches of 30 and that doesn't scale very well using a single -- albeit complex -- mold and single investment procedure.

Edit: Ah, I see you are only trying to make very tiny gears of the hose clamp scale. Sorry -- I thought you were wanting something of the order of a gear drive for a cnc machine rather than a guitar. My question would be: why don't you cast them directly out of metal? Dental labs in England could easily and economically make them from bronze, stainless steel, or any number of superalloys under controlled atmosphere and melt conditions. You would simply need to provide plastic or wax patterns. I say this because my grandfather was a master technician in England and people in the labs regularly played around with this sort of thing as did I when I worked with him. He would have been about your father's age, I would wager. If you are looking to make very precise replicas for repair, it would be a distinct advantage to use just the epoxy in order to have a more stable pattern but still one that should completely and cleanly combust in the investment. You can use the silicone molding material for hobbyists to make the mold for the epoxy. That could yield a mold small enough that you could simply spin it around with a rope like a lasso in order to encourage the epoxy to fill the mold. I am not sure you can get any cheaper than that.

Maybe the tonal qualities of epoxy granite?

Seriously, though, I suspect that a local lab owner might be interested in helping you out since the idea is to donate instruments to the local school. If you choose to use an alloy that is not dental related an expense would be buying a dedicated crucible to avoid cross contamination. Some of the metals used in coinage might work okay.

If you want to try something even lower tech then obtain 30 small socket head cap screws or maybe button head cap screws. Some steel angle iron can be used to mount them in. Drill through the angle iron and either thread the hole or tack weld a nut on the other side. The screws can be cross drilled near the head to accept the wire -- just like safety wiring aviation fasteners. A jam nut next to the threaded hole or tack welded nut could be used to lock/hold the desired tension. You don't have to use the hexagonal keyed screws -- they just look a bit more elegant. Hex headed screws would work fine as well. Good luck.

[Kevin_Johnson]

Hi Cameron,

I think the Studer patent issues with meal grain sizes has probably more to do with the general shape of the particles. The money shot there would be to economically produce tiny geometric particles conforming to his specs.

I think he would have already evaluated spherical particles as they were available at that time. It would be interesting to hear or read his thoughts on that general issue of creep with respect to these shapes.

[Nivea]

Cameron,

Do you think it's possible to achieve higher packing densities than with the agsco grades? If so, do you have any suggestions on a mix I might try with my grades?

I have woven wire mesh sizes in the following US mesh: 8 10 12 14 16 18 24 28 35 48 60 100 115 150 200
I have also made round hole sieves in millimeter sizes: 3 4 5 6 7.5 9 11 13 16 19 22 30 50. It's not hard to make any size above 1mm if you think it would help.

The image shows some of my sieved sizes. They are much more uniform than samples I have from local suppliers. I intend to try these mixes with vacuum, vibration and compaction.

[ckelloug]

Nivea,

It is definitely possible to do better than the agsco grades. I used the agsco grades because they are well documented as to what they contain for discussion purposes here.

When I have some time this weekend, I'll study what can be done with your aggregates avoiding some of the smallest particles which are currently giving me fits since they don't behave as we would want.

Regards all,

Cameron

[JeffMN]

There is lots of DIY info about making countertops from concrete. Some of it seems like it would apply: forms, additives, finishing, pouring eg How to Make Concrete Countertops - wikiHow

[Nivea]

Nivea,

It is definitely possible to do better than the agsco grades. I used the agsco grades because they are well documented as to what they contain for discussion purposes here.

When I have some time this weekend, I'll study what can be done with your aggregates avoiding some of the smallest particles which are currently giving me fits since they don't behave as we would want.

Regards all,

Cameron

Feel free to use sizes up to about 45mm if you think it helps. Particularly if you avoid the smaller particles it might be useful to be able to extend the upper range. I can easily make sieves any size you like above 1mm. The ones I listed are the ones I have made already but if it helps to use a particular size not listed, please use it.

I'd like to experiment with dry and water wetted mixes first. A method to accurately measure the densities of the compacted mix would be useful. If the largest aggregate is say 30mm then the container needs to be a good cube of about 4 liters volume. To measure the density one needs to measure the volume of the compacted mix. Not easy to do with good accuracy. Might be best to just use epoxy and simply measure the dimensions of the cured solid block...

Ideas anyone?

Regards,
Shae

[Kevin_Johnson]

Feel free to use sizes up to about 45mm if you think it helps. Particularly if you avoid the smaller particles it might be useful to be able to extend the upper range. I can easily make sieves any size you like above 1mm. The ones I listed are the ones I have made already but if it helps to use a particular size not listed, please use it.

I'd like to experiment with dry and water wetted mixes first. A method to accurately measure the densities of the compacted mix would be useful. If the largest aggregate is say 30mm then the container needs to be a good cube of about 4 liters volume. To measure the density one needs to measure the volume of the compacted mix. Not easy to do with good accuracy. Might be best to just use epoxy and simply measure the dimensions of the cured solid block...

Ideas anyone?

Regards,
Shae

For smaller volumes, an inexpensive but sturdy measuring device you could make would involve an automotive piston and mounted wet sleeve. Old Renault engines used them (Dauphine, R-8, R-16). I think this could yield a device similar to what Cameron uses in the lab.

Larger volumes I think would involve a trip to the hardware store and examining large diameter plastic pipes. You would have to use creativity in making a suitable large diameter but close fitting "piston".

[Kevin_Johnson]

For smaller volumes, an inexpensive but sturdy measuring device you could make would involve an automotive piston and mounted wet sleeve. Old Renault engines used them (Dauphine, R-8, R-16). I think this could yield a device similar to what Cameron uses in the lab.

Larger volumes I think would involve a trip to the hardware store and examining large diameter plastic pipes. You would have to use creativity in making a suitable large diameter but close fitting "piston".

Laugh! I forgot that I was going to make a compression device out of the "slug" of eg that I previously cast in the cylinder!

There you go.

[ckelloug]

Hi nivea (Shae),

I've run the model assuming a virtual packing density Beta of .62 for all the fractions. This isn't a bad assumption but it isn't a great one either. I didn't convert the sizes to sieve sizes. The values in the second column are percentages by volume (rounded so may not quite add to 1). Multiply mass times density to get volume.

22mm 0.36
12mm 0.09
6.6mm 0.08
3.6mm 0.07
2.0mm 0.07
1.1mm 0.06
.60mm 0.05
.33mm 0.04
.18mm 0.04
.10mm 0.12

Predicted Actual Packing Density Phi 0.825

Since this design cuts out at 100 micron, the model results should be reasonable. To get a result more likely to model the material you have, it would be necessary to have packing measurements at 4g vibration with 10kPa of pressure from a weight on top to be consistent with the model.

I'm not a big believer in sizes much above 22mm so I used that as the maximum size.


I also ran another model using the same assumptions but 6 components instead of 10. The packing results are basically equivalent but at least in theory, the 6 component mixture should be somewhat less likely to segregate.

22mm 0.30
7.5mm 0.16,
2.5mm 0.13,
.86mm 0.11,
.29mm 0.08,
.10mm 0.13,

Predicted Actual Packing Density 0.822

Hope this helps.

--Cameron

[ClockMakerDK]

After reading a lot about Epoxy Granite i deside to build my own CNC Router. The Y axis is almost finish and i am going to post all detail about the production. I have pictures of each step. It is based on Mach3 with the latest update. Servomotors with integrated drivers. Se you soon.

[brunog]

Welcome aboard ClockMakerDK,
I can't wait to see the pictures!!

Best regards

Bruno

[brunog]

Cameron,
Have you tested your samples, I'm curious to see what the results will be on them.

Best regards,

Bruno

[ClockMakerDK]

Right now i am adjusting the rails. It seems that i am able to do it within one to two thousand of an millimeter. I am going to make the description in PDF format since it is easy to attach and distribute. Her is a picture where the Y axis is in opright position on my Granite plan. Here you see the meter for lining up the Master Rail.

[ckelloug]

Hi Bruno,

I want to test the samples but have not done so yet. My tile saw setup for sample cutting isn't working too well and I haven't yet managed to depalletize and test the antique surface grinder I got for that purpose. I still need a VFD and a bunch of wiring for the grinder as well as to make wrenches to change the wheel etc. Money is a bit tighter than usual here so I'm trying very hard to avoid buying anything until I am absolutely certain I know what I need.

Regards,

Cameron

[mitip]

A few years ago I owned a Defiance VTX high speed CNC machine with a granite structure. I never got inside to see it in detail but it was there.

So recently I design a benchtop machine to sit on a standard granite surface table base. It works out fine in terms of flatness, but drilling holes to fix to it is a nightmare and any shaping to suit servos and ballscrews 3etc is a challenge.

So I now have a revised design in 3D CAD for a concrete version. The mould is an L shape with recesses in the front faces for ballscrews and servos with aluminium plate moulded in that can be machines afterwards to fit rails. The surface will be smooth and colored concrete with a steel fibre and granite beads. The damping properties have been discussed here, the issue I am after is ease of manufacture, all the other machine parts can be done on a basic bench mill.

Getting rigidity and strength at the L joint is the issue. I considered pre stressing threaded rods along the lengths but getting the strength 'around the bend' where the concrete is in tension not compression due to z cutting forces is the biggest challenge. I have considered moulding two slabs about 6-8" thick and about 14" wide and 24" long with steel threaded rods tensioned externally then using the rods to bolt through a 6" square section steel tube. The other alternative is to mould th complete L shape and set two L shapes sections of welded steel structure in the back of the mould and use these to sit the machine on.

The advantage of this approach is the entire structure does not need to be that accurate because material is left on the aluminium plates for one machine pass to get them flat and square to sit the THK slides afterwards.

I came across this post when researching epoxy concrete mixes. I think anyone who cracks a succesful way to mould a machine carcass is on to a winner.

[ClockMakerDK]

Hello again.

To follow up i send a picture showing the way i am doing it. As mentioned early i am not finish with the documentation so far, but it want be long. I build my machine around aluminum profiles. Each tube is filled with Epoxy Granite for rigidity. The button and the top is also filled. Even on each side of the aluminum for the Rails. All it needs is to be flattened for the rails.

Regards
Jørgen