Correct way for bolting a machine tool to their base

[FLG]

Dear forum

I have been wondering for some time what is the correct way for bolting a machine tool to their base, as I have seen people mounting it the "normal way", that is the machine over the base, and others mounting the machine in the bolts with nuts over the base, leaving a gap between the machine and the base, like floating over the base.

Over some time I have seen pictures of both ways of mounting diferents bench top machines ( mills and lathes) like this, I imagine that bolting the machine over nuts make it easier to level, but at the risk of bending the base if the nuts are not at the same height and also there is a reduced area of contact between the base and the machine. The other way will requiere shims under the machine or under the base to level / straight the machine up.

What is the correct way of mounting a machine to the base? asuming that the base is a rough fabrication of welded steel plate as is the norm from import machines.

The picture shows the floating arrangement, over a small import surface grinder, for reference.




Many thanks in advance.

[dieselpilot]

If it's precision machine and the mating surfaces of the base and machine aren't precise, some means of adjustment is necessary to prevent distortion of the machine. No different than leveling on the ground. A three point mount would be less critical, like the grinder in the first photo, but those mounts provide for leveling.

[uuu]

The UK-made Myford lathe is a good example of a popular machine needing adjustment.  The lathe bed is not really rigid enough, and the motor hanging off the back of one end can introduce a twist that, if not corrected, leads to tapered results.

Wilf

[Jo]

Both of those base units are light weight welded cabinets and I find it difficult that they could "warp" the machine tool bolted to it. If you look at the bottom of the base castings of both of those machines you will note it is a big flat surface (probably hollow) which is a much better surface to bolt to than relying on the surface area of some nuts.

I accept that light weight hobby machines may bend if bolted to a non square surface but heat is likely to be a bigger problem. My Prazimat for instance has a prismatic bed but I found that if it sat with the sun was on one end of the bed for a day I could measure a noticeable twist  I over came that by bolting the length of the bed to a 25mm thick piece of black steel plate. It has not tried moving since 

Jo

[vtsteam]

Both ways work if you want to adjust something. You can shim under a solidly bolted base to remove twist. I'd prefer that way as more solid, but depends on the machine, the weight, the size of adjuster bolts, etc.

There are often floor adjusters also for leveling which throws another variable into the equation. Leveling and twist aren't the same thing. And as mentioned, if the base itself has flex, neither way solves that problem. A stiifer base or more reinforcement does.

[steamer]

Tough one to answer.....

If the bottom of the machine casting is well finished flat.....and I mean machined flat,  then gently bolting it down to a mating flat surface is fine.....but that isn't what you describe.

Under the circumstances you describe, I would "float" it on studs as shown taking great care not to twist the base of the machine....and then snug the nuts down to hold it there.

Whatever method you choose, it should leave the machine casting in it's lowest stress possible, and as close to the condition it was originally machined in, to maintain proper alignment.


Dave

[Overbuilt and Overkill]

Fwiw and everything I've read but not experienced with having my own surface grinder (yet) indicates a high degree of precision leveling is required to get the very best accuracy from them. I'd suggest both those pictures were efforts at getting the grinders into that level position using screw thread adjustments. Engineering wise, I'd also suggest there's much better methods to increase the foot print and amount of proper and rigid enough support for those surface grinders. For the most part and fortunately for us, vertical type mills in the sizes we might be using are already stiff enough that a slight out of level condition will show little to no effect. Yes a perfectly level condition might allow setting some parts with a level. Imo there's still better and far more accurate methods using indicators and precision sine bars when the accuracy levels are high enough to require it.

But those level requirements change immediately once you get to something like a long and fairly narrow lathe bed. Yes the manufacturer's normally add cross ribbing between the lathe's longitudinal ways in an effort to stiffen the bed as much as possible. Most of that stiffening has been added so the lathe bed resists the cutting forces better and the ways remain parallel to each other. Those lathe beds are still quite flexible just because of there length to width ratio. Because of that, it really takes little force to twist the beds enough to matter. Simply bolting any bench top or on it's own normally light weight sheet metal cabinet mounted lathe down to either simply isn't anywhere close to being good enough. Wooden constructed benches also suffer with impossible to fully solve dimensional changes just due to temperature and humidity variations. There really is no available paint or sealant that can prevent those from happening on a 100% guaranteed basis.

In general, posts about "lathe leveling" only go so far. That supposedly exact level is still only a static condition. It doesn't factor in the multiple points within a lathe that have variable amounts of different force vectors exerted throughout the whole machine whenever the dynamic cutting loads are introduced. Once the lathes accuracy expectations get high enough, everyone of those deflections do become important. I'd suggest that below .001"/.0254 mm accuracy, machine tool parts really do start to act somewhat like there made of rubber. And for less than any 1 ton lathe I've ever leveled, they always required further but miniscule tweaks on the screw adjustments to get them to actually turn, bore and drill truly parallel to the head stock. And for the most part, even multi ton lathes, bed mills, HBM's etc almost always come from the manufacturer with designed for the purpose screw adjustments to both fully support and level them no matter how large and heavy they are. uuu's mention of the Myford lathe example is a good one. Myford themselves sold what were called raising blocks that were also screw adjustable and in my opinion far superior to the very old recommendations I've seen mentioned in my old Model Engineer magazines about using tapered wooden shingles as a method of leveling the lathe bed. I can't think of a more frustrating method of adjustment that will also change as soon as the temperatures and humidity does. But for bench or cabinet mounted lathes, buying one is only one part of what needs to be considered as a system that needs to work together. Jo's use of that 25mm steel plate is one I've used myself and combined with some method of using adjustment screws for my lathes, does work really well. What anyone chooses to do with anything I've mentioned would of course depend on what the end user considers as being good enough for the work there doing.

Greg

[steamer]

Fwiw and everything I've read but not experienced with having my own surface grinder (yet) indicates a high degree of precision leveling is required to get the very best accuracy from them. I'd suggest both those pictures were efforts at getting the grinders into that level position using screw thread adjustments. Engineering wise, I'd also suggest there's much better methods to increase the foot print and amount of proper and rigid enough support for those surface grinders. For the most part and fortunately for us, vertical type mills in the sizes we might be using are already stiff enough that a slight out of level condition will show little to no effect. Yes a perfectly level condition might allow setting some parts with a level. Imo there's still better and far more accurate methods using indicators and precision sine bars when the accuracy levels are high enough to require it.

But those level requirements change immediately once you get to something like a long and fairly narrow lathe bed. Yes the manufacturer's normally add cross ribbing between the lathe's longitudinal ways in an effort to stiffen the bed as much as possible. Most of that stiffening has been added so the lathe bed resists the cutting forces better and the ways remain parallel to each other. Those lathe beds are still quite flexible just because of there length to width ratio. Because of that, it really takes little force to twist the beds enough to matter. Simply bolting any bench top or on it's own normally light weight sheet metal cabinet mounted lathe down to either simply isn't anywhere close to being good enough. Wooden constructed benches also suffer with impossible to fully solve dimensional changes just due to temperature and humidity variations. There really is no available paint or sealant that can prevent those from happening on a 100% guaranteed basis.

In general, posts about "lathe leveling" only go so far. That supposedly exact level is still only a static condition. It doesn't factor in the multiple points within a lathe that have variable amounts of different force vectors exerted throughout the whole machine whenever the dynamic cutting loads are introduced. Once the lathes accuracy expectations get high enough, everyone of those deflections do become important. I'd suggest that below .001"/.0254 mm accuracy, machine tool parts really do start to act somewhat like there made of rubber. And for less than any 1 ton lathe I've ever leveled, they always required further but miniscule tweaks on the screw adjustments to get them to actually turn, bore and drill truly parallel to the head stock. And for the most part, even multi ton lathes, bed mills, HBM's etc almost always come from the manufacturer with designed for the purpose screw adjustments to both fully support and level them no matter how large and heavy they are. uuu's mention of the Myford lathe example is a good one. Myford themselves sold what were called raising blocks that were also screw adjustable and in my opinion far superior to the very old recommendations I've seen mentioned in my old Model Engineer magazines about using tapered wooden shingles as a method of leveling the lathe bed. I can't think of a more frustrating method of adjustment that will also change as soon as the temperatures and humidity does. But for bench or cabinet mounted lathes, buying one is only one part of what needs to be considered as a system that needs to work together. Jo's use of that 25mm steel plate is one I've used myself and combined with some method of using adjustment screws for my lathes, does work really well. What anyone chooses to do with anything I've mentioned would of course depend on what the end user considers as being good enough for the work there doing.

Greg

Greg   I agree with you...and as you pointed out, every lathe is a bit different.   I have a 4 1/2 foot SB9 that I do most of my work with that has been rebuilt and is straight, but the bench will wander every once in a while and with that long bed it can get a twist going pretty easy.   I have to keep and eye on it.    One thing I have found recently to be handy is the SB pivoting bend stand.   It was designed as a pivoting foot and it allowed the user to put some twist via two set screws to make the lathe turn straight.   As my lathe is straight, I leave the adjusting screws stay slack and let the TS end of the lathe float on that pivot pin, and she turns nice and straight all the time now....    A bit wierd but it's working for me..

Dave

[Overbuilt and Overkill]

I think South Bend put a lot of thought into that pivoting tail stock foot idea Dave. I've never owned one so far, but the idea behind it is supposed to work very well as you've found. There now quite old designs, but some of the British lathe manufactures like the Adept http://www.lathes.co.uk/adept/index.html and what I would guess are lathe designs meant mostly for the hobby market of that time used a much larger head stock foot for bolting the lathe down. The lower lathe bed casting then tapers up from that foot all the way out to the tail stock end. Logically I can only surmise that design choice was made as an effort to remove any influences from whatever bench it was bolted down to. A large number of the watch maker type lathes use the exact same principal. Depending on the bench materials used, then rigidly bolting down a lathe at both ends may not always be advisable. The 1" thick steel plate I used under my smallest lathe only has two bolts attaching it to the bench top at the head stock end. The tail stock end is allowed to float. Then I used screw adjustments to rigidly hold the bed at both ends and do final level adjustments of it's bed to that plate. So it's a bit like your designed to do the same South Bend. Still not the best for holding exact adjustments over very long periods of time if anything moves enough, but it does work far better than just bolting it down to the bench top.

On our lighter weight bench top sized lathes, it's also not only about obtaining a very rigid and stable mounting surface. Apparently during WW II many with home shops in the U.S. were doing light manufacturing piece work for that war effort. At least one version of the story is that interchangeable part accuracy was becoming enough of a problem the department of defense got involved, and since so many were using there machines, they approached South Bend for a solution. They did some experimentation to find a cheap cost effective method. I can't seem to now re-find that war time SB produced bulletin, but what they came up with was to cast a heavy bench top out of steel reinforced concrete with either studs or female threaded inserts cast into those bench tops. Then the lathe got precision leveled with those. Just like my steel plate, all that extra mass is added directly into the lathe bed itself and allows large improvements in overall machine rigidity, accuracy and better surface finishes.

The subject of lathe leveling gets mentioned on every machine tool related forum I can think of. Exactly why its so important and what that twisted bed effect will be is rarely explained well enough for the more entry level members. The best way to visualize what happens is to think of any twist in the bed as either slowly rolling the tool tip into or away from the part in the longitudinal axis as the carriage travels along that bed twist. But its also not a completely linear effect. A .001" /.0254 mm variation of twist along the lathe bed does NOT create the same amount of difference in the parts finished diameter. That twist and because the tool tip is well above the lathe bed itself is magnified to a much larger change in the diameter and amount of taper the part will have. A .0005" / .0127 mm twist in a lathe bed could easily result in a .005" / .127 mm or more part change once the lathe size becomes accurate and large enough. It's also why even the expensive Mitutoyo level I have only gets my lathes very close, and the last final adjustments have to be done by actual measurements on the parts under real world cutting conditions. Requiring that level of accuracy is certainly debatable for most of us and what we might be doing. But knowing it can be done with the equipment you have is I think still worthwhile.

Greg.