[mackeym]

Thanks for the suggestions. I agree the hole should be about 0.05" larger diameter than the insert.

For epxoy, the Redhead G5 is still my second choice just because the literature on it seems to only consider concrete as the base material. Loctite mentions concrete, stone, brick, and cinderblock. I also like that the bond strength is specified (1900 psi which is high). I think the hole roughness of granite is smoother than concrete and the high bond strength of the loctite will keep the insert from pulling out. The bond strengh of the Redhead G5 could be just as high or higher, but i didn't see it in the literature.

I've updated my website log with my attempts at using masonry bits to drill holes in granite. http://mackey-machining-rocketry.com/CNC%20Lathe.html

[will gilmore]

im following your webpage closely. what are you using for inserts? 316 stainless? are you grooving them on the outside? that starrett link shows their inserts

[mackeym]

I was thinking about stainless 304. This really doesn't need the extra corrosion resistance of 316, and it's not worth the extra money (60% more expensive at speedymetals.com). Yeah, i'm going to groove them. I think i'm going to mimic their grooves by turning it a little off balance in a 4-jaw chuck.

[will gilmore]

316 has much higher yield strength and slightly higher tensile strength but it is more expensive

[nicad]

uh, wow.
As for drilling the granite, I would seriously look at diamond drills, probably core drills. I get some from http://www.ukam.com/. The diamond drills found on Ebay are usually cheap for a reason.
Also, use coolant (water) when drilling. It will make the drills last many times longer, since they are really grinding instead of cutting.

[jalessi]

Maybe you will have some extra space on the slab to make fudge and some pralines too!

[mackeym]

Wow...that looks like a good place to get the core drill i need! Thanks for the link! Very reasonable pricing too. I had already bought one from toolocity and just got it today, but they sent the wrong size 3/8" instead of 1/2". Now i have to wait for them to send another one. I'll most likely get one from UKAM...as the one i'll be getting probably won't last to drill 50 holes to 1 3/16" depth.

Ha, i do have extra space to make some fudge.

[petee]

Petee's 1st post :-)!

My two penneth on diamond core drills.

Not all diamond core drills are the same. Sometimes people refer to 'soft' diamond cutters and 'hard' diamond cutters. I think I'm correct in saying that what they're actually refering to is the makeup of the material which is impregnated with diamonds. Some bonding agents are soft and some are hard. Which you use depends on the material to be cut. The idea of the bonding agent is to wear and continuously present fresh diamonds to the cutting surface. Sometimes when cutting a material with the wrong type of cutter you can 'glaze' the cutting surface, it gets jammed up with debris and stops cutting. It can be cleaned by cutting a soft but abrasive material such as lightweight thermal block.

Unfortunately most cutter sales and even a lot of the hire shops don't state or even know what type they have. I come across cheap cutters cuting hard materials very well and expensive cutters cutting the same material poorly.

People in the UK will know what I mean by cutting Accrington brick. Its a widely used hard glazed brick in older houses especially in the north of England. Get the cutter right and it will slip through it and hardly wear. Get it wrong and it'll go nowhere even after a long time, the cutter rises in temperature easily and then there's danger of unbrazing a tooth. Watch out for that one as if it lets go it might embed part of it in the bottom of the cut making it very hard to further. If its obviously not going well watch out for colour changes in the metal immediately behind the teeth as the first sign of temperatures getting excessive.

So a soft bonded diamond cutter may cut the hard granite better than a hard bonded cutter as it more readily releases diamond to the cutting surface.

Also its important to know if the cutter should be used wet or dry.

Can anybody verify this? Its based on research I did after having a problem cutting hard brick with diamond cutters. Just by chance I got the correct cutter first time round but not the second!

After seeing your efforts of getting that slab in there I really want to see this thing fly. Watching with interest. :-)

[mackeym]

What you say sounds accurate and a lot of places that sell diamond core bits don't give any info. I've just ordered what i think to be a good bit from http://www.ukam.com/ (the green universical brazed core drill 9/16"). They have a lot of good info on their website. Along with the bit, I'm going to get an aluminum oxide dressing stick from mcmaster (150 grit) in case i need it. According to Ukam, the swivel adapter, which is used to supply coolant thru the center of the bit, significantly extends the life of a core drill. I have a rough design and parts list to make a similar device to get coolant thru the center of the bit (rather than buying one for ~$150).

I recently got 1/2" 304 stainless steel bar stock for the inserts. I also got two double disk shaft couplings from Ruland (DCD21) to couple the motor to ballscrews.

This last weekend, i visited my hometown to look at my mom's dryer (not working); to fix it i would have had to go through all the wiring and switches. She opted to buy a new one, so i took the motor and blower. I have a relevant use for it. The 15hp rotary phase converter i have is a little loud for my appartment, so i previously enclosed and soundproffed it (it's about as loud as my refrigerator now). For cooling i was using a Hoffman fan hooked to a temperature controller. I'm going to replace that fan with the dryer blower...which has a LOT more airflow.

[youngjim]

Hi! I just caught on to this build, it seems very interesting. I never even thought of re-tasking a surface plate, even though I spent probably an hour reading about epoxy-granite bases. :idea: Maybe I can talk my buddy out of a corner of his garage...

[petee]

I've been chewing this over in my head as from what I've read so far, I'm heading the same way as you are. Namely using a granite surface plate as start point. One thing that sticks out for me at the moment is twice now I've read that the linear profile rails are NOT straight when they arrive new. Clamping them to the reference edge on the machine base makes them so. If so it'll need a reference edge before using them as a guide to cut the rest of the holes. So I'm wondering has anybody here had 1st hand experience of this? If they're not straight how far out are they?

Speaking to a surface plate supplier today they had a reference measuring machine there which used M10 bolts to bolt down stuff to the plate, the inserts where about 20mm dia (sorry 25/32" :-) ). Maybe the larger insert gives better strength margin when glued in?

If its a wet cutter the coolant sounds sounds good to keep the cut clean. So the cutter is cutting granite and not a granite/diamond slurry.

[youngjim]

Speaking to a surface plate supplier today they had a reference measuring machine there which used M10 bolts to bolt down stuff to the plate, the inserts where about 20mm dia (sorry 25/32" :-) ). Maybe the larger insert gives better strength margin when glued in?

I would agree with that, greater diameter for more surface area. In any other material I'd say fire the holes straight through and bolt it down, but seeing how much trouble it is to work with this stuff makes inserts a compelling option.

[mackeym]

Regarding the "reference edge" mentioned in the two previous posts, is this a horizontal or vertical plane reference? Since the horizontal is provided by the surface plate, i'm guessing it's the vertical reference.

About these rails not being straight, i'll post what i find about my rails when i get to this part. It is, however, easy to mis-interpret the rails not being straight since it takes very little force to bend them a few thousandths. If they are in fact not straight (with respect to a horizontal plane) it doesn't matter much because they have to be bolted to a flat surface (if it's not straight, it will be straight when bolted to the flat surface; if it is straight, it will stay straight when bolted to the flat surface). If bolted to a non-flat surface (with respect to a horizontal plane), the rail WILL conform to the non-flat surface. The only situation that a non-straight linear rail (with respect to horizontal) can have an effect on is if one wanted to initially set the rail on epxoy putty so that the putty conforms to the rail contour (can be tricky to get the right epxoy putty consistancy). Sorry for the wordyness...i'm just trying to convey things clearly.

Another option in addition to clamping the rail against a vertical reference is to choose screws that have .01 or .02" diameter clearance in the rail bolt holes. Drill and tap the inserts using the rail as a guide and tighten the rail down just a little. Then measure the straightness, and push/pull the rail straight (if not straight) and tighten the screws down. This can then be used as a reference for the second rail. There is more information about rail mounting and acccuracy in the NSK website literature. I did not see any info about the straightness before mounting. Just a wild guess is that the straightness (WRT a vertical plane) is not off by more than 0.003".

About the insert diameter, on the Starrett website in a previous post, they found that the granite fractures (~10,000-12,000 lbs) before the inserts pull out (insert diameter was 0.625"). One concern i have for larger diameter inserts is local bending of the rail under the bolt (assuming the insert is below the surface of the surface plate). This is why i want to use 1/2" inserts. I've gone thru all the calcs and found that there is still a decent safety factor.

[digits]

I don't know what rails you guys are using, but my 35mm Hiwin rails are damn stiff - it would take a very large force to bend them at all. As I understand it, they have a reference edge because all measurements are relative - when using a pair of rails, you bolt one down, and it defines the reference points for that axis. When you bolt the other one down, you adjust it relative to the other rail, so that they are both perfectly parallel, regardless of whether that are at some random angle relative to the edges of your machine's base.

It is quite possible that you can distort these rails by bolting them down, but they do seem to start off straight to me - just put a carriage on one, stick an indicator on the carriage and run it up and down, indicating off anything you know to be straight.

[jim_geib]

I have been watching this thread with intrest, because I am doing the same thing. Not on such a grand scale, but a 18x24 surface plate. I just gaot a Hardinge DV59 spindle and I am looking for rails on ebay.

On the comment the rail are not streight, I put 15mm rails on my smithy's Z travle and they were streight as an arrow. I clamped the outer one in place drilled and tapped one end, then allined it to the spindle and drilled and tapped the far end. After checking the basic alinment I drilled and tapped the rest of the holes and bolted it down, then used the front rail to aline the rear rail.
http://i86.photobucket.com/albums/k1...b/DSC00092.jpg

I hope the photo comes thru.

[mackeym]

I have a pair of THK SHS25 rails which are 23.5mm wide. It does not take much force to bend them a few thousandths (also because they are quite long ~72"). After i get them bolted i'll post info on the straightness. The fact that it is standard procedure to bolt the rails against a vertical plane must mean there is some kind of error in the straightness (unless torquing the rail mounting bolts can cause some error...i would guess that it would not). One very nice thing about using 2 rails and 4 blocks is that the error of the overall structure is 1/3 to 1/5 the error of either rail by itself.

Nice picture and setup.

[digits]

I have a pair of THK SHS25 rails which are 23.5mm wide. It does not take much force to bend them a few thousandths (also because they are quite long ~72"). After i get them bolted i'll post info on the straightness. The fact that it is standard procedure to bolt the rails against a vertical plane must mean there is some kind of error in the straightness (unless torquing the rail mounting bolts can cause some error...i would guess that it would not). One very nice thing about using 2 rails and 4 blocks is that the error of the overall structure is 1/3 to 1/5 the error of either rail by itself.

Nice picture and setup.

Just because you can bend them, it doesn't mean that they aren't straight when unloaded - I can measure the effect of standing on one corner of my mill, but when I get off it again, it goes back to being square.

When you bolt your rails down, you are (over) constraining them, so either they'll bend to fit what you're bolting them to, or the base will straighten to match the rail! If you bend them much at all though, I'd imagine that the blocks will simply get stuck.

One question I do need to ask though is, what sort of accuracy are you aiming for - what sort of error over what sort of distance? Nothing is perfect, and I imagine you're not going to be running this in a temperature compensated environment, so there's no point in obsessing over accuracy you can't achieve or don't need.

[mackeym]

I'm hoping to stay within 0.001" over the working distance of the lathe.

Just because you can bend them, it doesn't mean that they aren't straight when unloaded - I can measure the effect of standing on one corner of my mill, but when I get off it again, it goes back to being square.

I agree.

[jalessi]

With the mass of the surface plate being what it is, the temperature should remain very constant.

Its like a giant heat sink.


Jeff...

[jim_geib]

I'm hoping to stay within 0.001" over the working distance of the lathe.



I agree.

mackeym,

When I setup my smithy I checked it over about 10-11 inches of travle and it was within 0.0005 and it has held for over a year, and like I said I used only 15mm rails. Your larger rails should work like a dream.

Jim Geib
Mansfield, Oh