[lgalla]

You can get at least .005 or better.You must use TTEpoxy or table top epoxy as it is right for the job.Standard laminating resin may "blush"which is a waxy film on the top surface.The good news is TTepoxy is cheap,$25/gal.
The first coat is the seal coat which can be rolled on.The second coat is poured in the middle and you are aiming for 1/8".1gal1/8"thick is 16sq/ft.
If you dam,there is the possibility of surface tension making a small ridge at the outer edge.Personally,I prefer the epoxy to overflow the edge.Very messy indeed.
Gota go,Anyone want mixing tips and(flame2) ?
larry

[ckelloug]

dfro,

Welcome to our pot of stone soup Thanks for the tip on small quantities of teflon powder.

You will be the first one on this thread to try to cast parts of a precision metalworking machine so I would encourage you to be careful as most of the work here is still rather experimental. Quickly, Epoxy granite is only good to 1000-3000 psi in tension/flexion in AustinT's test so big pieces of it or reinforcement will be required for it to survive high tension or flexion loadings. If you need true metalworking precision stable to .0001, some large cross sections will be needed if you want to forgo reinforcement.

As for casting an epoxy level surface, I will first say that I'm the thread's theoretician and I'm just about to start pouring epoxy. What I say about the leveling epoxy is theory which I believe to be correct but haven't yet tried.

You will want to roughen the surface of the bed as much as is practical before applying the leveling coat. From past post by others on this thread, it's probably also a good idea to cast this while the epoxy in the bed is still achieving a cure which is going to be sometime within hours of the original casting; How much it matters if any is unknown by me.

I don't think that standard epoxy will turn out to be all that hard compared to tools and clamps etc. so I've been thinking about how to reinforce it yet still have it flow well. For the standard epoxy, It would be wise to wait a week, preferably at summer temperatures, before placing any pressure on the leveling coat as these epoxies take a long time to achieve full cure and accuracy will go down if you clamp something against the leveling epoxy before it is as hard as it is going to get. You will also want to level the ways (hopefully with a precision level) before you pour the surface coat so that you have a uniform thickness and hopefully level ways.

The meniscus that forms will depend on the ratio of the surface energy of the epoxy to the surface energy of the mold. If I am understanding the theory correctly the meniscus will be concave towards the edge if the epoxy has a lower surface energy than the mold edge and the meniscus will be convex if the epoxy has a greater surface energy than the mold edge. If they are equal then there should be relatively little or no meniscus. While I haven't tested this, I suspect that if the mold edge were made of hardened epoxy cast earlier that that would be the best you could do other than slopping it over the edge like Larry suggested.

Larry is the thread's expert on actual epoxy so I have to defer to his advice as being better than mine here: I do personally suspect, although I haven't tried it, that the shopmaninc 635 epoxy than many members have used will be okay as a leveling material provided it is mixed accurately.

On the topic of mixing, I think I have identified a bit of a problem with the directions for the 635 epoxy. Larry pointed out blush issues for laminating epoxy but it is my understanding that imprecise mixing is the number one cause of blush. In short, the ideal mix ratio for the 635 epoxy is not 2:1 by volume like the directions state, it's actually 2.06 to 1 by volume for a 3% error.

In detail, the Epoxide Equivalent Weight (EEW) of the A component (Reichhold 37-127) is 197 with a specific gravity of 9.2 lbs per gallon. The Amine Hydrogen Equivalent Weight (AHEW) of the B component (Reichhold 37-606) is 85 with a specific gravity of 8.2 lbs per gallon. 197/85 gives a ratio 2.32:1 by weight between part a and part b which equates to 2.06:1 by volume.

Since the directions specify a 2:1 volume mix ratio, there is a slight excess of hardener in the epoxy mixture produced via the uscomposites directions. In itself, the 3% error might not mean much but given any further error on the part of the user, it might in theory cause batches that do not set properly or are a bit soft.

see http://www.reichhold.com/docs/litera...ochure2007.pdf

For Reichhold's explanation of epoxy mixing.

So, welcome dfro and good luck with your machine. If any of us can help, please make sure to post and ask.

Regards all,

Cameron

[the4thseal]

i understand the portion about surface tension i just did not think that the tension of the epoxy would be the same as the forms when the epoxy is constrained. as for unconstrained wouldn't the epoxy just keep running to the point that there would be a gentle arch going from the center of the edge. very very shallow but still some kind of doming. this might just might be me obsessing but in my world it is never that easy. there is always some unseen detail that makes itself know at an inconvenient time.:-(

[krazatchu]

Hi all...

I am a big fan of this thread, have been following it almost daily since its conception...

Anyhoo... I came across an interesting mixing/de-aerating product while searching for something unrelated...

http://www.thinky.co.jp/english/w-rentaro_e.html

It's a centrifugal mixer... runs at 400G's which really isn't that much.
I didn't look at it too closely but with a bit of math one could figure out the nessesary spin rates of the larger radius and the smaller radius...

I figure the g force from the smaller radius of spin must exceed g force from the larger radius to keep the material from falling back into itself and re-aerating...

Might be possible to quickly mix and de-air the E and the G at the same time.

It seems the smaller radius is tilted to prevent overflow and eliminate sealing of the top as well...

I'm thinking the smaller radius could be belt driven from a stationary pulley in the center that allows for tilt...


Well, there is my minor contribution.

Thanks for the most interesting thread...

Krazatchu

[dfro]

cameron,

I much prefer mixing epoxy and silicone mold rubber by weight. I have an electronic gram scale I use. Usually, the data sheets tell you the exact ratio by weight of the two components. I imagine companies get as close to the 2:1 ratio or the 1:1 ratio so it is very easy for people to mix by volume or to use pumps, like in the West Systems epoxies (5:1 ratio).

I have cast many parts in epoxy and I know that if I under-pour the mold, I will have a concave maniscus. If I over-pour the mold and the fluid begins to bulge over the edge, the maniscus will be convex. I guess very careful initial leveling of the rough cast table or lathe bed is the key. Also a very even outer edge to contain the leveling epoxy is important. Then a very careful pour that leaves a slight bulge would be what I would go for. If it oozed over the sides a little, I would carefully wipe it off the sides, while being very careful not to bump the table.

The idea of casting the outer ridge into the original rough casting of a mill table or lathe bed make sense to me now. I will do that instead of using aluminium strips to box in the leveling epoxy.

I also want to mention that I have had great success in using pressure to cast parts in epoxy. I have two pressure pots and an air compressor as well as a vaccuum chamber and a vaccuum pump. While this method would not work for the large castings, it could work for smaller castings like headstocks or carriages.

This is how I cast bubble-free parts in epoxy. After thoroughly mixing the epoxy resin and hardener together, I put the pot in the vaccuum chamber and remove most of the bubbles. Then, I pour the parts into the silicone molds. This invariably introduces bubbles back into the parts.

I then put the mold into the pressure pot and put as much pressure as I safely can on the parts. All, and I mean all, of the bubbles are squeezed back into solution and disappear. I have read that after curing, the compressed gas will slowly off-gas out of the parts, since plastic is not air tight.

I was very pleased with the consistent results I got using this method. Before, when I only used vaccuum, I had to throw away over half of my epoxy castings due to surface bubbles.

Come to think of it, you might be able to vaccuum bag large castings and use atmospheric pressure to help compress the e/g into the mold and possibly compress some of the bubbles. I have come up with a very cheap and easy way to vaccuum bag that does not require all of the overpriced equipment - except for the vaccuum pump.

I built a frame, like a picture frame, out of wood and sealed it on all sides with epoxy. Coming out of one side, I drilled a hole and fixed a nipple that the vaccum tube attaches to. This tube then goes to the vaccuum pump. On the top of the frame, I affixed clear vinyl sheet, which you can get by the yard from any large fabric store. I used sticky mastic tape to stick it to the frame. This tape is like putty and seals any gaps or holes that air might leak through. I got this from The Gougeon Brothers who make West Systems epoxy. At the corners of the frame, I also buckle up the plastic sheet so that it is loose and can form around the objects I am gluing. To the bottom of the frame, I attached three parallel layers of weather stripping tape. It is the black, foamy stuff that has pressure sensative adhesive on one side. This can be bought at any hardware store.

The frame then sits on top of a large sheet of melamine covered plywood. I make sure to wax the melamine surface, so that any glue that gets on it will release. As I pull a vaccuum on the parts I am gluing, the frame sucks down onto the melamine sheet and makes a very decent vaccuum, at least for my purposes. This might be used to put some pressure on the e/g as it cures. Epoxy tends to squeeze all over the place, so there might be some places where the mold gets glued to the casting. I will be making one time molds, though, so I won't mind chisseling parts of it off the casting in places.

I could post some pictures of the vaccuum bagging frame, if my description is confusing. It is probably more hastle than it is worth for this application, though.

I wonder what is the tensile strength of cast iron or aluminium? The 1000-3000 psi of e/g sounds good to me. I plan to reenforce the castings with threaded rods and rebar.

To me this process seems only slightly more difficult than making the mdf type gantry mills that many people are making. In some ways it seems much easier to me - i.e. leveling and stiffening. The hardest part is knowing what materials to use and how to mix them together, and you guys are rapidly figuring this out. Walter's initial castings look great!

Dave

[martinw]

The meniscus that forms will depend on the ratio of the surface energy of the epoxy to the surface energy of the mold. If I am understanding the theory correctly the meniscus will be concave towards the edge if the epoxy has a lower surface energy than the mold edge and the meniscus will be convex if the epoxy has a greater surface energy than the mold edge. If they are equal then there should be relatively little or no meniscus. While I haven't tested this, I suspect that if the mold edge were made of hardened epoxy cast earlier that that would be the best you could do other than slopping it over the edge like Larry suggested.


Cameron

Dear Cameron and Larry,

This is slightly off-beam, but this meniscus matter is confusing me.

Just suppose you want to build a large reference plate and use low viscosity epoxy contained within a rectangular dam to find its own level. I think that (if epoxy behaves like water), if you do not overfill the dam, the meniscus will "climb up" the dam at the edge. If you overfill on the other hand, the epoxy at the top of the dam will be at a lower level than the level of the main body ( i.e. the meniscus "climbs" down ).

Now, because you will be unable to set the level of the top of the dam absolutely level all round, if you take the overfill approach, at some places you will overfill (climb down meniscus) and at some places you will not overfill (climb up meniscus). Is this correct?

Best wishes,

Martin

[lgalla]

Martin,
Dam or no Dam, that is the question.Negative or positive meniscus can just be filed off.If you dam and have overflow in areas,I would not trust the levelity of the plate.I am comfortable with overflow,no dam,as I have done this on bar tabletops.Daming is OK but you must not overflow.
Any filler is probably a bad idea as it will impead flow.TTepoxy is made for the purpose of flow and leveling.Should not mess with the formula.
Mixing:Ratios are very important and only difficult with small batches as you can be many % out.Larger batches are easy to get accurate.Buy graduated plastic mix containers at Home Depot or a paint store.Pour your A to the desired desired quantity then pour B to the correct ratio and mix.Do not use separate A&B measuring vessels,you would be off ratio for sure.
Too err with slightly more hardener is OK.
Now you have mixed for 3or4 minutes and pour.Resist the temptation to scrape the sides of the mix vessel to get every last drop of epoxy out.It will not be 100% mixed from the side walls.
How do you know your plate is perfect?It will be the smoothest,glossyist surface you have ever seen.Any divits will show and require another pour.
Read;www.precisionepoxy.com/ Navel postgraduate school project.

[HelicalCut]

"I wonder what is the tensile strength of cast iron or aluminium? The 1000-3000 psi of e/g sounds good to me. I plan to reenforce the castings with threaded rods and rebar."

Cast iron around 30000 psi but can be up to 170000psi for some of the nodular types. Aluminium 20000psi to 50000psi. For a machine bed tensile strength is a secondary issue, deflection is the primary concern, that is why a small landis No 1 cylindrical grinder which may only see 40 lbs cutting force weighs 3 tons. Aluminium has far too much expansion (3x cast iron), lacks rigidity and has poor wearing characteristics in a typical machine tool environment which is why no quality machine tool builder uses it for machine bases.

Use Rebar it is cheap and much stronger than most threaded rods. Be careful with using a soft material like epoxy for ways, if it is exposed the chips will burn it and wear it away in no time also I doubt epoxy bears upon itself very well, the filled epoxied like moglice bear very well on steel, note moglice is usually used in situations where it is protected from hot chips.

[Mcgyver]

"Use Rebar it is cheap and much stronger than most threaded rods. Be careful with using a soft material like epoxy for ways, if it is exposed the chips will burn it and wear it away in no time also I doubt epoxy bears upon itself very well, the filled epoxied like moglice bear very well on steel, note moglice is usually used in situations where it is protected from hot chips.

Helical, I agree with a lot of what you say in your post, but moment of inertia has to enter into things to take full design advantage of steel's higher modulus of elasticity, in other words, the cross section matters hugely in how much rigidity is achieved. Consider pound for pound which is more rigid, a tube or bar? or a 2x8 joist laid horizontal or vertical. compared to using tube or better still the whole machine having an exoskeleton of steel, rod won't do much, or least won't do near as much as the steel could if present with a better section.

the objective, is pound for pound to increase rigidity and this is as much a function of section as it is material.

i still like the exoskeleton approach where a normalized steel fabrication gets filled with EG. only 63 more projects to go and I'll get to act with you guys on it rather than talk

[ckelloug]

Hi all,

HelicalCut makes some interesting points as far as moduli and stiffness go but they weren't were I was going with the train of thought. My point about the low flexural strength numbers for E/G was that the skinny sections used in low grade home shop metalworking machines made of metal will never work for making machines out of E/G. A 3/8 thick section 1/2 inch wide and 6 inches long will break at around 35 pounds of force. If your brain is calibrated for designing things from aluminum, be careful as E/G is a lot different in strength and stiffness. . .

There are going to be some serious problems with using leveling epoxy for ways and I hadn't even thought about McGyver's point about burning up the epoxy with hot chips. Unlike a large cast surface plate where edge effects are minimal, machine ways are much smaller and will have much more serious edge effects. To be able to surface ways with cast leveling epoxy, we're going to have to work out how to cast small sections with edge effects that are actually flat. I think I know how to make a material that will be much more wear and heat resistant than generic tabletop epoxy based on my research but I don't yet understand the edge effects problem. I think it will be very much about surface energy of the epoxy vs. the mold. The formula for bearings will be quite different from the ones for machine parts and likely very different from the epoxy for surfacing ways.

Re Martin's commments about water climbing the edge of the vessel, it must be remembered that water wets glass. The same vessel out of teflon with water doesn't have a significant concave meniscus according to the folks that make teflon graduated cylinders for measuring. Very little wets teflon. Remember also that mercury in the same glass has a convex meniscus. In short, the materials necessary to deal with meniscuses and leveling epoxy may be outside of what we can generally intuit without resorting to theory since they aren't present in our day to day lives.

Regards all,

Cameron

[dfro]

I ran into this item on ebay:

http://cgi.ebay.com/ws/eBayISAPI.dll...Name=WDVW&rd=1

It is a table suspended by springs, that has a motor and belt driven vibrator under it. This could work nicely for de-airing and settling many of the molds people will want to make. Add a speed control and you could really settle in on a frequency that works best for each E/G formulation.

[dfro]

Cameron, Walter, HelicaCut, Mcgyver, and all,

Here is how I propose to make a base for a router mill. I would like to start with a modest 18 x 24 table. In the base of the mold (which will become the top of the table), I will route a channel with a 1/8" round endmill bit. When the table casting is turned rightside up, this will produce an even dam all the way around the top of the table, which will contain the leveling epoxy. I will also place several rectangular, solid steel bars lengthwise along the bottom of the mold (will be on top of table when it is turned rightside up). In these I will secure bolts that attach them to crosswise 14 guage, 1.5" x 1.5", steel, hollow tubes that are filled with foam. This will form a steel grid to stiffen the casting.

When the casting is demolded and turned rightside up, the evenly spaced, solid steel strips will be visible. I then plan on leveling the table with a precision level that I have and pouring the leveling epoxy. This will bury these strips 1/8" under the leveling epoxy. I do not plan on predrilling any holes in the steel strips, because I do not see a way to be very accurate in their placement; and also, they would interfere with the leveling step (bubbles, uneven shrinkage, etc.). I will dye the leveling epoxy black, with an oil-based black dye. A very, very little goes a long way. I also figure that the rounded edges of the e/g dam along the perimiter of the table will decrease maniscus issues - there will be a gradual slope up to the top of the dam.

Once the leveling is done and the table has cured for several days, I can then take some linear guide rails and center punch their bolt hole placement along the outer two lengthwise steel strips that are on either side of the table (and under the leveling epoxy surface). Then I will drill and tap through the leveling epoxy and into the steel bars. I will make some kind of drill bushing jig to help me in making my hand drilled holes perpendicular to the table.

I do not intend to use the leveling epoxy as a bearing surface. I want it to be a level surface to attach guide rails to.

Once the gantry is done and the mill is working, I would then cnc a grid of holes in the other steel strips. I could then attach steel or aluminum bars to the table to make T-ways. This surface I could then cut flat with a fly cutter or end mill.

In the sides of the mold, I will embed predrilled and tapped steel plugs for attaching the stepper motor plates. I will also have drilled and tapped steel bolts extending out of the top of the mold, which will accept the table leveling feet, when the casting is turned over.

I am not sure how thick to make the table (how high to make the sides of the mold). Any thoughts?

When I pour and vibrate the main casting, I also plan to push a waxed piece of melamine covered mdf on top of the mold. I will then place my vaccuum bag frame over the entire mold and pull a vaccuum on the mold and casting. If any epoxy ends up glueing the mold together I do not mind because I plan on breaking it apart when I demold

Please, tell me if you think this is a good approach for getting a rigid, flat table with precisely aligned linear guid rails.

If I wanted to do circular linear shafts, I would buy the kind with threaded holes that use evenly spaced threaded rods to attach and align them to the table. I would affix steel tubes on 4" centers along the outer steel bars. The steel tubes would key into holes drilled half way into the steel bars. I would probably "jb weld" (epoxy) them in place. These steel tubes would go all the way through the casting and stick out the bottom, like the leveling bolts that I explained earlier. Once the casting and leveling is finished, I could then drill through the outer steel bars and the leveling epoxy from underneath the table, using the steel tubes as a guide. These holes can now be used to bolt the circular linear shafts in place from the top and the bottom of the cast table.

I am concerned that the circular linear shafts are not stiff enough. Supporting the shafts with a threaded bolt every four inches seems to be 'hanging in the wind'. Any sideways forces (along the y axis) are only being opposed by a few threaded rods sticking up. Any thoughts on the advantages or disadvantages of cicular linear shafts, I would appreciate. I do like the advantage of the large choice of interchangable bearing choices from inline skate bearings, to bronze or teflon bushings, to open-type ball bearing bushings.

I will make some cad sketches, soon and post them.

Thanks,
Dave

[dfro]

Another idea:

I don't need a precision level to level the mill table for the leveling epoxy pour. All I need to do it pour water into the 1/8" depression created by the ridges all along the outer edge of the e/g table. I can adjust the leveling feet on the table until the water is overflowing evenly all around the table and any kind of maniscus looks even. Then I can mop up the water, dry off the table with a heat gun, mix up the leveling epoxy while the table cools, and pour the epoxy. Water acts as a great level. What do you guys think?

Since I do have a precision level, I am sure I will check my results with it, though.

[greybeard]

I'm not sure where this concern for a meniscus at the edge of any flat, be it a way or a large table, is coming from.
Surely any component that is going to make use of the flat can be designed to avoid that part of the surface. It must be constrained from sideways movement by a second surface anyway, so the combination of the two will prevent any edge distortion affecting its use.

Perhaps I'm missing something fundamental, so I'm happy to receive any engineering lessons that I obviously need.

John

[martinw]

Hi all,

.

Re Martin's commments about water climbing the edge of the vessel, it must be remembered that water wets glass. The same vessel out of teflon with water doesn't have a significant concave meniscus according to the folks that make teflon graduated cylinders for measuring. Very little wets teflon. Remember also that mercury in the same glass has a convex meniscus. In short, the materials necessary to deal with meniscuses and leveling epoxy may be outside of what we can generally intuit without resorting to theory since they aren't present in our day to day lives.



Cameron

Dear Cameron.

Thanks for correcting me about menisci (?). Looks like I've been parading my ignorance in public, yet again.

Best wishes,

Martin

[ckelloug]

Martin,

Don't worry.

I think we all need engineering lessons on this leveling epoxy meniscus stuff. In short, until it gets tried, what will actually work is an open question. I know what I said about mercury and glass and water and teflon is correct but I don't know how wet epoxy will behave other than guesses based on my understanding of the theory which may be imperfect. If there were no surface tension, the epoxy would run over the edges of the mold.

The wikipedia article on surface tension is very good:
http://en.wikipedia.org/wiki/Surface_tension

The article on menusci is also good. Thanx to Martin for the correct plural.
http://en.wikipedia.org/wiki/Meniscus

Avoiding the edges where the edge effect occurs would be awesome. My sense of dread indicates that this may be a lot farther from the edge than we would first think.

If anybody has used tabletop epoxy, is the idea to pour it into a constrained space or does it's relatively high surface energy and viscosity cause it to spread to a uniform thickness in a sort of round shaped blob?


Dave,

I think the vacuum bagging frame will not be able to degas a thick mineral filled epoxy slab although nobody has tried it yet on this thread and posted the result. The pressure the vacuum bag frame applies on the melamine board however, along with vibration should go a long way towards making a good casting with few voids. The water idea is good although getting everything to beyond bone dry is highly advisable. Also, I think I'd use carbon black rather than an oil based pigment for the dye as epoxy and oil don't mix well. The steel strips will be covered in epoxy but make sure to grind off all of the scale and clean these with a strong solvent as oil is a rather big problem in getting epoxy to stick.

If you build the spindle first, you might be able to use it as a drill press for the linear guide drilling.

Since you have steel reinforcements, 2 inches is probably enough for the table thickness. Keep in mind that with weight on the table that the part you really need to reinforce is the bottom as it's in tension and E/G is week in tenison. The E/G on the top is in compression where it's half the strength of aluminum.

There's probably some more interesting stuff to say but I can't think of it right now. I'll be back later and in the mean time good luck, Dave.

--Cameron

[dfro]

Cameron,

Thanks for the advice. I have tried to mix carbon black power into epoxy before and it globs very badly. In trying to get the power to whet and to disperse, I end up whipping lots of extra air into the epoxy -still with small clumps everywhere. Any powder clumps would ruin an epoxy surface plate.

The oil-based pigment that I have used in the past I bought from the same plastics company that sold me the epoxy. The oil helps the pigment disperse into the epoxy and it is a very tiny amount - .01% or less. However, I do not want to reduce the strength and rigidity of the epoxy. So, I (we) must carefully consider what to do.

I think a call to the epoxy surface plate company would be a good thing. I will ask them how they are tinting their epoxy. I am also interested in buying their video on how they make their plates:

http://www.precisionepoxy.com/

Some comments form Igalla, the epoxy expert, would be helpful.

I will keep the placement of the steel reenforcement in mind. Your comments on compression and tension are helpful. And like you said, my reason for vaccuum bagging is not to suck air out of the E/G material, it is to clamp the E/G into the mold very tightly and possibly compress some air bubbles back into solution.

It is truly amazing to see how tightly a vaccuum bagged part gets sucked onto the table. And while the air is being sucked out you can hold everything in place through the bag. Once the air is evacuated, there is no moving it! Vaccuum bagging is very elegent - no huge presses and wrenches - just a frame I put together for about $30, a melamine sheet underneath, a vaccuum pump, and miles of atmospheric air collumn pressing down.

Cad drawings are on the way.

Dave

[martinw]

Martin,
Dam or no Dam, that is the question.Negative or positive meniscus can just be filed off.If you dam and have overflow in areas,I would not trust the levelity of the plate.I am comfortable with overflow,no dam,as I have done this on bar tabletops.Daming is OK but you must not overflow.

Dear Larry,

Errr... damned if you do, and damned if you don't..???. The duel continues..

Best wishes,

Martin

[blurrycustoms]

Hey all you epoxy-granite lovers, I've reading this thread, like many others, with excitement, and wanted to get some opinions on the designs discussed in the below thread.

http://cnczone.com/forums/showthread.php?t=43154

My idea is to build a large machine, and I wanted to know what the biggest (as far as weight goes) machine anyone here has cast.

[lgalla]

Gotta agree with Greybeard,I need engineering lessons.Meniscus:Learn something new everyday.I thought MENISCUS was something from your knose or nose.Apparently it refers to knee injuries.Make the top larger than necessary and pick or file off the Meniscus.
Dave,if you want pigment for a surface plate,buy pigment paste as good pigment dispersion is difficult without a ballmill.
In a vacuum,with a caul or sheet of melomine pushing down,any high spots or packed aggregate will localize the pressure.
Cameron,In table top coating,there is no dam.The epoxy simply flows over the edge.What you said is true.Simply a poured puddle in the middle and the epoxy will flow and level.Large tops may need encourangement with a pickle roller.Larry