[mackeym]

I've been looking for an easy way to turn a non-flat steel surface (error~.020" over 56" length) into something suitable for linear rail mounting. My idea is to use a low viscosity epoxy and pour it into a mode built on the steel plate to a depth of about 1/8". I thought i would post this to see what the general consensus is here.

One of the major factors will be the viscosity of the unhardened epoxy mixture. Several companies sell some low viscosity stuff: MAS epoxy looks to be the most promising. Their standard "Mas resin" with a slow hardener has a viscosity of 550-650cst and a work time around 23 min). They also have lower viscosity stuff around 120cst but the minimum quantity is somehting like 4gal. I'm going to give the Mas standard resin a test (mounted on a granite surface plate to accuractly measure the resulting surface).

Two options (from West systems epoxy) for lower the mix viscosity are:
1. Heating the components before mixing (drawback here is the decreased set time)
2. Adding several percent solvent (several were listed). this has several major drawback like lower epoxy strength and increased shrinkage

[the4thseal]

look at this thread almost everything is discussed hear on the topic. it is long but bountiful.

http://www.cnczone.com/forums/showthread.php?t=30155

[harryn]

Your idea is a great one. The epoxy properties to make it work are slightly more complex than you might think, but life is good. Here is a link to a company that makes exactly what you need. They do this for a living, so it is a tried and tested solution. I have discussed this application with them and the ability of their epoxy to go "flat" is amazing.

http://www.precisionepoxy.com/

Approx $ 50 / gallon for each component.

[audioandy1762]

Hi

Have a read through this thread I know Hayden used epoxy to mount his rails on.

http://www.cnczone.com/forums/showthread.php?t=30751

Andy

[ratrace2]

Moglice.com

[cuemaker1]

look into system3.com.

[svenakela]

I've done it too, works like a charm.

[bodbson]

I f the idea is to fill in small gaps why not use bondo? the key as I see it is to find a reference edge to use to show what needs filling. You might use a length of aluminum angle. Lay it down on the surface and shine a light under it. Use the bondo to fill in the gaps an sand it smooth with a hard flat sanding block. Just a thought.

[mackeym]

ha...thanks for all the replys

[mackeym]

Svenakela, do you remember what grade/type of epoxy you used. Also, what size of area and thickness did you make.

[keebler303]

Your idea is a great one. The epoxy properties to make it work are slightly more complex than you might think, but life is good. Here is a link to a company that makes exactly what you need. They do this for a living, so it is a tried and tested solution. I have discussed this application with them and the ability of their epoxy to go "flat" is amazing.

http://www.precisionepoxy.com/

Approx $ 50 / gallon for each component.

They say its perfectly flat but its actually only as good as the radius of the earth allows. The epoxy will actually be arched at the radius of the earth. Think of it like a smaller version of the "flat" ocean, which wraps itself around the earth.

The error doesn't matter for this application but it is interesting to ponder.

Matt

[Mr.Chips]

They say its perfectly flat but its actually only as good as the radius of the earth allows. The epoxy will actually be arched at the radius of the earth. Think of it like a smaller version of the "flat" ocean, which wraps itself around the earth.


Matt

OK you math wizards If the rail is 4 feet long, how much will it be arched in the center? (Height on center compared to the ends)

[ger21]

OK you math wizards If the rail is 4 feet long, how much will it be arched in the center? (Height on center compared to the ends)

Not sure, but the radius of the arc is about 251,064,000 inches (3962 miles).

AutoCAD won't let me draw it, because it's too straight.

[lgalla]

Mack,
The curvature of the earth is .000126/mile or 8"/mile.
The right epoxy is table top epoxy.Regular laminating epoxy may blush.Rather than go on about it,use the correct material.TT epoxy for $20/2quarts.
http://www.uscomposites.com/kk121.html
Check my avatar.
Larry

[svenakela]

Svenakela, do you remember what grade/type of epoxy you used. Also, what size of area and thickness did you make.

NM 258, Swedish high quality brand. It's not even 1 mm at the thickest place. I layed the gantry horizontal on a polished stone surface, aligned it in 90 degrees, made a small barrier around the beams with tape and the poored the resin. The epoxy is post treated with heat and now it's really hard and very straight.

[mackeym]

ha, that's right; there will be some curvature. Based on the equation of a circle, the amount of error or vertical distance (from the center to either end of the epoxy plate) in inches is:
Vertical Error=(6367000-(6367000^2-EpoxyLength^2/4)^(0.5))

The input "EpoxyLength" must be inputted in units of meter
For a 4ft length epoxy plate, error is 1.17x10-6 in or about 3 nm

Which is essentially 0 for our purposes. The surface tension will also cause additional bowing in the same direction. This might or might not have a much larger affect than the curvature of the earth.

[mackeym]

Wait, there's an error in that equation, i'll post the right one a little later

[mackeym]

Nevermind, that equation was right. The .000126/mile or 8"/mile number is NOT the distance one would be interested in for surface plate flatness. It is right, but not in terms of surface plate error. I attached a picture showing the difference if anyone is interested.

[vger]

I think you will be about 0.000001147" high in the center.
Steve

[lgalla]

OK,
I guess the curvature of the earth is not a concern.Svenakela's method of the tape is the best way.Daming or a mold is difficult as epoxy will seep through the tinyist hole.On a long cure,epoxy may penetrate masking tape.Gaffers tape works better.
.003 is easy to get,possibly.001 with lowering the viscosity by applied heat.If you need better,lapping with a granite surface plate is in order.
Larry