[altaic]

I have been thinking about using self leveling (table top) epoxy to get a flat surface for linear rails. I would like to have the inserts sit in the E/G mix held by a fixture above the surface. However, I don't have a precision surface that would actually come in contact with the epoxy to keep it straight. It would rather be suspended above the EG and just hold the inserts in position. Now, this would allow the epoxy to climb (or drop?) as a result of capillary effect. Does anyone know whether this would be a problem? How does one model this scenario to get the curvature of the liquid against the steel to see if it leaves enough clearance for the mounting holes? I imagine this could also be a problem around edges of DIY surface plates.

Hi ticica,
I wouldn't know how to exactly determine the meniscus, however I have thought about your application before. How about first making a level surface plate using E/G, and then use that to hold the rails. This surface plate would be the start of the mold. IOW, construct the mold upside-down, cut grooves/niches in this surface plate to hold the rails, then add walls. So, put release on the mold, place rails and ensure the gap between the rails and mold are sealed, and pour epoxy in. Also, in my experience the surface touching the mold has a much better finish than the exposed surface, which is a bonus. I hope my description wasn't too confusing.

Will

Edit: I just realized I don't completely understand what you mean by inserts, so I'm unsure of their capillary effect... I don't know if I helped at all. If you're worried about alignment and flatness before/during cure, however, it might be better to just drill holes for the rails in the E/G and mount it the old-fashioned way.

[greybeard]

If it's any help, I've successfully drilled 8mm holes in E/G (sand)using a carbide tipped masonry bit. My largest aggregate size was only 2mm, so may not be so easy with larger aggregate.
John

[brunog]

Will, Ticica,
The whole idea behind inserts is to set mechanical adapters that will permit assembly of different machine frame parts (fix gantry, leveling pads, steel plates, etc) or components such as linear rails, balls screws or fixtures.

The main objective behind the use of inserts is to prevent the use of remachining like drilling and resurfacing, however this cannot be obtained without the use of proper product assembly and mold design planning.The same exercise can be done for inserting piping and conduits for air, lubricant and electrical inside the machine's EG casting

The first photo of post #30 of this thread is a good example of the use of inserts in a EG mold.

Best regards

Bruno

[rowbare]

I have been thinking about using self leveling (table top) epoxy to get a flat surface for linear rails. I would like to have the inserts sit in the E/G mix held by a fixture above the surface. However, I don't have a precision surface that would actually come in contact with the epoxy to keep it straight. It would rather be suspended above the EG and just hold the inserts in position. Now, this would allow the epoxy to climb (or drop?) as a result of capillary effect. Does anyone know whether this would be a problem? How does one model this scenario to get the curvature of the liquid against the steel to see if it leaves enough clearance for the mounting holes? I imagine this could also be a problem around edges of DIY surface plates.

Maybe this is the kind of stuff you should be looking for it self leveled to .003 over a large area...

http://www.cnczone.com/forums/showth...777#post282777

bob

[ticica]

Will, Ticica,
The whole idea behind inserts is to set mechanical adapters that will permit assembly of different machine frame parts (fix gantry, leveling pads, steel plates, etc) or components such as linear rails, balls screws or fixtures.

The main objective behind the use of inserts is to prevent the use of remachining like drilling and resurfacing, however this cannot be obtained without the use of proper product assembly and mold design planning.The same exercise can be done for inserting piping and conduits for air, lubricant and electrical inside the machine's EG casting

The first photo of post #30 of this thread is a good example of the use of inserts in a EG mold.

Best regards

Bruno

Bruno,

I also wanted to avoid the machining of EG block. The first idea that comes to one's mind is to use self leveling epoxy to create a flat surface to which the rails would attach. However, since we do not want to drill the EG, one needs to put inserts in. In the photo you pointed to (The first photo of post #30), there are two metal rails that will be embedded in EG, to which one could screw the rails. Those would have to be precisely machined flat, and also the whole mold would have to rest on a surface plate to keep them aligned. That is what I wanted to avoid. I want to do this without having a flat reference surface at all, if possible. I would embed inserts into EG, but would let the self leveling epoxy provide the flatness. This all seems doable, if the meniscus that will form due to the surface tension between the inserts and the epoxy does not disturb the flatness.
On the other hand, maybe the only way to really find out is to make an experiment...

Ivan

[altaic]

Bruno, that setup in post #30 was pretty much what I was describing, except I was talking about securing the rails in the E/G itself. Using inserts is definitely the way to go, though; way simpler mold design, and it allows for replacing the rails if desired later on.

Ivan, how about pouring a self-leveled epoxy plate and stick the inserts in place using spray adhesive, magnets in the surface plate, or however else... Probably want to mask off the area with the spray adhesive (before attaching the inserts), add the rest of the mold (walls, etc.), then spray release over everything. Then unmask the adhesive, place the inserts, true them up (probably want to do this before adding the walls, but make sure not to get release on the inserts), and pour your E/G.

The problem with suspending the inserts on the surface of the E/G is that if you don't have a flat surface to start with, you're not going to be able to suspend them flatly. Additionally, truing them up would be very difficult while the E/G is setting up; better to have them set in the mold, trued and double checked before pouring the E/G.

[brunog]

[quote=altaic;552694]Bruno, that setup in post #30 was pretty much what I was describing, except I was talking about securing the rails in the E/G itself. Using inserts is definitely the way to go, though; way simpler mold design, and it allows for replacing the rails if desired later on.

Will,
Actually what was secured on the mold were steel plates on which the rails were to be attached. It's definitely not a good idea to embed consumable items in epoxy.

Best regards

Bruno

[brunog]

John,
Is your spincasting project still alive?

Best regards

Bruno

[brunog]

Ivan,
It's still possible to set inserts mask them with something that will be thinneer that the coat of self leveling epoxy so that it covers the masked area, of course you will have to "dig in" the epoxy to unplug the inserts, however there will be no meniscus. As to the countour of the bed, Larry was suggesting to let the coat of epoxy drip over the edges to prevent them.

Best regards

Bruno

[ticica]

If it's any help, I've successfully drilled 8mm holes in E/G (sand)using a carbide tipped masonry bit. My largest aggregate size was only 2mm, so may not be so easy with larger aggregate.
John

Once you had the holes drilled, what did you use to fasten to the EG?

[altaic]

Will,
Actually what was secured on the mold were steel plates on which the rails were to be attached. It's definitely not a good idea to embed consumable items in epoxy.

Best regards

Bruno

I guess I'm still not clear on what you mean by insert. I thought you were talking about the steel plates, which I agree should be the parts embedded in the E/G.

Will

[ticica]

I guess I'm still not clear on what you mean by insert. I thought you were talking about the steel plates, which I agree should be the parts embedded in the E/G.

Will

No. By inserts, I meant threaded inserts used to bolt the rails down.

[ckelloug]

Hi Will,

You are right that you can add a solvent to epoxy (non-reactive dilutant) and evaporate it out but it usually doesn't turn out very well from what I have read. I believe acetone was one that I have seen used in the non-reactive dilutant manner. That's why the low viscosity RTM epoxies use reactive dilutants which can stay in the epoxy. A bit more reactive dilutant can be added than the conventional formulas have in them though it will affect ultimate strength and chemical resistance. Cresyl glycidyl ether is the most effective non-toxic reactive dilutant I've seen. Butyl glycidyl ether works better but is toxic and smelly.

I got the impression talking to bloefeld that the surface energy of epoxy is low enough to wet aggregate and all that is needed is sufficient pressure to force the epoxy to wet the whole mass. The measurement of surface energy is done with a surface contact goniometer which measures the contact angle of a droplet with a plate of known surface energy. I'm going to add some 3M fluorosurfactant to my epoxy to minimize surface energy when I get there.

Using some kind of secondary adhesive to stick the titanium carbide particles to the surface is an excellent idea and one I have been thinking about too. Haven't figured it out yet though.

It sounds like you are making excellent progress and I wish you the best of luck. My network access is minimal this week as I'm on travel in NJ right now. But I'll be checking in periodically.

Two notes to your proposed procedure:

The research guys from concrete all mix their aggregate before doing anything, If you don't, you won't achieve maximum density because the distribution won't be random.

From my experience with Hexion 813, degassing at less than 29.9 inches HG at 20C won't accomplish anything which means that a rotary vane pump is the only way to go.

Regards all,

Cameron

[greybeard]

Once you had the holes drilled, what did you use to fasten to the EG?

I was placing M4 nuts, so I used a just pourable mix of epoxy/G200 zeeospheres.
A drop of candlewax melted in to fill up the threads kept the centre clean, as they were sunk just below the surface.
The nuts were for a low load fixing, and it seems to have worked ok.

John

[brunog]

No. By inserts, I meant threaded inserts used to bolt the rails down.

Will,Ivan,
Maybe the term "insert" is incorrect, the term "anchor" is more appropriate.

This being said, you should anchor down every part necessary, including support plates for rails, motors etc...

best regards

Bruno

[altaic]

Cameron, again thanks for the info and tips. I'll hopefully have some results pretty soon.

I'm just about ready to make some samples... Unfortunately, one of my vacuum fittings has been delayed until Wednesday

Anyhow, I sketched up a basic mold (body, breather/bleeder, spacer, top) and made an animation. It doesn't show the vacuum bag or the vibrator (outside the bag, since it's not rated for high temperatures), but it gives an idea. (sorry for the poor resolution)

<embed width="448" height="361" type="application/x-shockwave-flash" wmode="transparent" src="http://i63.photobucket.com/player.swf?file=http://vid63.photobucket.com/albums/h131/altaic/bar_mold.flv">

Edit: I just found thomasz's posts from last year with example renderings of key E/G design ideas, including inserts/anchors and such. Good stuff! #2610 #2611

[dsquire]

Talking about Walter, does anyone know what happened to him?
He simply disappeared.

Regards

Sandi

Sandi

I Checked and Walter's last post was Jan 09, 2008 post #2622. If you go back and read that post it will it may explain.

Cheers

Don

[ckelloug]

Will,

That looks like an interesting molding setup but I'd worry about the demolding from what I can see. I'd pray to the gods of mold release before testing this as it looks like you don't have positive access to more than one side and as a result you may not be able to get the part out ever even if the mold release is working. My mold has two plates with a solid spacer ring that separates them and is sealed by o-ring seals to keep air out and epoxy in. The 300-500cps epoxy will run through any gap you can see and some you can't so the seals are more important than you might think. I thus end up with the epoxy stuck to the thin spacer ring once I peel the plates off and the o-ring is strong enough compressively to unseat the epoxy from the plates though it needs help coming off the spacer ring. I like to demold pure epoxy hot but this appears to allow the piece of epoxy to distort which means a production part might have to be cooled to room temperature before demolding.

Epoxy is crafty stuff and may get out at the top when the thing is under vacuum. That might be a blessing and it might not. Based on my experience with my first test mold that had a few thosandths of gap at the bottom and no vacuum, epoxy goes only where you don't want it: a tiny stream of epoxy dribbled out of the mold and proceeded to glue the entire damn thing to the bench. Vacuum fittings are a pain. I use 1/4 plastic nylon for my infusion lines with plastic valves and threaded fittings which use o-ring joints with built in retainers. Something I need always seems to be on Backorder: order early and order plenty.

Thomasz is definitely the reigning expert in making useful stuff. He also has an E/G recipe. I'm more interested in the aspect of getting the material optimal for this and perhaps other applications while he is building machines now which as far as I can determine are very good. You can always make a cross section bigger if you're having trouble holding deflections numbers.

His web site is http://www.thomas-zietz.de/

Regards all,

Cameron

[sigma relief]

In regards to self leveling epoxy, I doubt .003 accuracy is enough to use as a datum on any system worth using linear rails on. It could be used for non critical areas that just need to be smooth, but you would be introducing a lot of uncontrollable error.

In stead of casting right side up and floating a surface of thin epoxy, casting upside down with a well prepared mold will give a nicer more predictable finish. Additionally, the surface will be harder because it will not be just epoxy.

Unless you are making a large machine, you can use inexpensive datums accurate to better than .001 as part of the mold. An 18" square B grade granite surface plate can be had for $20 from many discount tool suppliers. As long as you don't get too big, prices are reasonable.

Using such a jig plate covered with a consistent layer of mold release would form a good starting point for mounting linear rails. Placing threaded inserts on the surface plate with a small amount of low strength adhesive will provide for securing the rails. If you need a straight linear datum for alignment of the rails, use precision ground steel stock lightly glued to the surface plate (just make sure the glue is applied evenly thick or to one side of the stock not used as a refrence edge). While they will have no draft, they should peel out of the finished part if you use a good mold release.

If done properly, the remainder of the mold could sit on top of the surface plate and be held down with straps or clamps. Wax can be used to seal the bottom of the mold if needed. A perk of this method is demolding without complete teardown of the mold may be possible by simply removing the bottom and sliding the ring shaped mold off.

Larger parts can be acheived by using the surface plate in a plain mold to replicate itself. Careful mold design would allow a new extension to be cast beside the origional part. As long as the "Real" plate is under both the cured portion and the area being cast, there should be a decent continuity in flatness. Attention must be paid to support while curing so this thing doens't warp, but it could be used to add just a few more inches to a table or other large area. This technique could also be used to make smaller datums for use in other areas of the mold, but at this point, a ground or flycut plate would most likely work

John

[ticica]

In regards to self leveling epoxy, I doubt .003 accuracy is enough to use as a datum on any system worth using linear rails on. It could be used for non critical areas that just need to be smooth, but you would be introducing a lot of uncontrollable error.

I thought it was more on the order of 0.0003"/ft flatness. This post has more on the flatness: http://www.cnczone.com/forums/showth...t=55625&page=2.

Also, I seriously doubt that a thin layer of epoxy will have too much give. Given compressive elastic modulus of 4.5GPa (http://www.efunda.com/Materials/poly...poxy&MinorID=2), assume we pour 3mm thick layer of epoxy (recommended at http://www.uscomposites.com/kk121.html). Also assume that the loading force will be uniformly spread to about 2cm x 10cm area by the linear rail under one bearing. To find out the load on one bearing that would sink it 0.03mm (~1 thou), for example, we would use the following calculation

4.5GPa = stress / strain =>
stress = 4.5GPa * (0.03mm/3mm) =>
force/area = 4.5GPa * (0.03mm/3mm) =>
force = 4.5GPa * (0.03mm/3mm) * (0.02m * 0.1m) = 90,000N or roughly 9,000kgs or 20,000lbs


I am a little biased towards using the TT epoxy as the surface since I will need a rather long X axis.

Ivan