I am looking into custom building a CNC lathe. The size would be somewhere in the 10 to 12" swing range. I don't want to convert an existing lathe because 60% of it would be scrap as I want to use precision linear ways and a variable speed spindle. Also, I want to make it as compact as possible.

For the bed I was considering welding up a webbed box section using a combination of 1/2" and 1" steel walls with 1/4" webbing. I would then get this normalized, milled and surface ground. To reduce resonance I was then considering filling it with something like concrete (any other suggestions would be welcome).

The big problem is the headstock. I don't want to make this as it is outside my capabilities. So I was thinking of either just buying a new headstock casting, backgear and spindle or getting one from a parted out machine. Does anyone have any suggestions? Within reason, I am not too concerned about the budget, I care more about ending up with a precision machine when I am finished.

If you buy the right machine, used, you could care less about the parts you have to scrap to make it into cnc. Perhaps even, you could sell off what you scrap, if it is one of the more popular machines.

The seperate headstock of a real lathe is not something that you are going to find all that easily, most likely, its going to come on a bed, etc.

I suppose you could even cabbage a set of linear ways on the sides of an old lathe bed, so that you could make use of the existing spindle and tailstock.

I think you would be far ahead to start with a used cnc lathe, and retrofit a new control on it. Consider the cost of two sets of linear ways, two ballscrews, high precision spindle bearings, a full cnc retrofit kit, building a carriage, tool turret, and a thousand and one other knick knacks. There is just no payback in starting from scratch.

Check out the machines at www.machinetools.com, if you are having a struggle finding something.

I agree 100%, however there are a couple of reasons I don't want to go this route. The first and biggest problem is I don't have space to put an industrial CNC lathe.

Secondly, I am just as interested in this as a project as I am in the end result. I have always wanted to make a lathe.

The lathe would mostly be for model engineering so accuracy, versitility and swing is important. Of less importance is production throughput.

I was planning on getting as much of the linear components from eBay as I can. I have seen some very good deals on THK stuff.

The way I see it, you are going to need a lathe to build one. Else you will be buying parts and basically assembling a lathe. You might consider getting a lathe that isn't quite to your liking and use that to make parts for your lathe.
..A mill would also be an asset .

This alone will be a "project"

...just my two cents...

John (Bloy)

P.S. I am just finishing a router I made and I couldn't have done it without my trusty Shoptask....which is a lathe and a mill combined (yeep!) :)
Shoptask lathe/mill:
http://www.cnczone.com/forums/showthread.php?t=2139
Router:
http://www.cnczone.com/forums/showthread.php?t=5049

Just to clarify I have a lathe (Myford super 7) and CNC milling machine (Tormach) already.

Thinking about the bed, how about using continously cast iron as a basis. Obviously it would require machining and grinding. Places like versa-bar provide all kinds of shapes including hollow.

http://www.versa-bar.com/

I guess you could also get a custom die made but I bet this would be costly.

Oh! Here in CNCzone is STEVIE who has built a lathe and is in the process of building a more massive one. You may gain some ideas from his threads:

http://www.cnczone.com/forums/showthread.php?t=8928.... the smaller one
and

http://www.cnczone.com/forums/showthread.php?t=13350 ....the bigger one

Well, I you are looking for an accurate, stable flat base, you could always use a granite plate. They are already qualified, relitively cheap, and damp vibration well. The only problem is you have to epoxy in threaded inserts, and use a diamond grit coring drill to make the holes. Granite is used as a base for some of the most accurate lathes in the world. Also, for a slightly different configuration, you could mount both axes to the granite (as long as you don't need a tailstock), put the spindle on one carriage, and the cutting tool on the other.


Keith

In wanting to get into a CNC lathe as well I ended up going the used non CNC lathe route and scavenging all the usable parts. Got this Hardinge DV-59 second op lathe for $710 delivered (eBay of course). It has a very high precision head, 5C collet closer, HD cast iron bed, coolant pump and a incredibly HD stand, the doors are 3/16" thick, I just love old stuff :D It will be easily enclosable. These also do not have any of the extra stuff normally on a conventional lathe so the lines are clean. This seemed to be, for me at least a very easy and quick way of jump starting this project, spindles are easy to make, high quality spindle's are not.

The original V-way bed was removed and a 1" x 7" x 36" steel plate will replace it as the base for mounting the THK rails (eBay of course). It will have about a 8.5" swing but could easily be increased with a spacer as the head is a separate unit from the bed. I esentially have all the necessary mechanical and electrical components and at this point expect to end up spending about $2500 or so total.

I like the granite idea. I did a little research and found

http://www.standridgegranite.com

They will custom make machine bases including slots and bonded in inserts. Thinking about it, I really need a slot down the middle to run the leadscrew in. I could use a flat plate and add risers for the rails to run on but from a rigidty point of view I don't really like this. I may mock up a drawing and get a ballpark quote from them to see if this is a feasible approach.

Ken, your deal is the kind of thing I am looking for. However, I wouldn't use anything but the headstock and maybe the bed. The tailstock and carriage and cross-slide would be custom.

Yes, the granite idea might be a good one. Perhaps if you could get a piece with a squared edge, you could fasten the ballscrew mounts along the perpendicular edge, which would create a similar torque position to the feed rack or leadscrew of a conventional lathe with the toolpost mounted on that same side.

Even my large slantbed lathe has the Z ballscrew back behind the bed, and the tool turret is back there, too. It may not be the best design to put the ballscrew down the center between the rails, but I do not know that for a fact.

It may not be the best design to put the ballscrew down the center between the rails, but I do not know that for a fact.

works for my designs; the best part is the forces are inline with the axis center; a feedscrew off this axis centerline creates tangental force that jags the main slide

That was my thinking as well steve. I have to believe most lathes with the leadscrew at the side do this purely to line up with the output from the screwcutting gearbox. They could build the output of the box to run down the center but I guess it would really complicate the headstock and encroch on space used for the spindle.

When considered in the scope of forces that a metal cutting machine will see when working, that tangential force trying to jam up the slide should be neglible. If your slide system is sloppy enough for this to be an issue, then you will have other problems.

You will have a simpler, more rigid, cheaper bed if you make it solid, and drive from an offset leadscrew.

I agree the loads for cutting are quite light. However, I wanted to mount a turret on the crossslide and some operations may involve drilling which requires considerably more force.

I have a couple of questions:

1/ Is it feasible to perform drilling using a turret mounted to the saddle? Do they do this on commercial CNC lathes or do they have a power feed on the tailstock? I am worried about the tool load being offset from the driving load of the ballscrew on the saddle. This distance could be 5 or 6"

2/ Does anyone have pictures of how the saddle is actuated on commercial CNC machines. Most pictures of CNC lathes has all the interesting stuff covered up.

Thanks

After much thought, I have decided to go the retrofit route. The numbers just weren't adding up with the approach I planned.

So I have bought a Hardinge TFB lathe from eBay that isn't in great shape but will be perfect for my needs. I will create a new thread to cover the conversion.


This thread is now closed.

pstockley,
you had wanted to see a comercial cnc lathe uncovered, as did I. I have found a small cnc lathe called omniturn and they have a good picture on there site with everything uncovered. It looks quite similar to Steves lathe that he is building.
http://www.omni-turn.com/Media/GTSLIDE.jpg

Thanks. Thats almost the exact design I came up with. I got a quote for $3500 for a granite base (35" between centres) with bonded in inserts and a centre groove cut out. The big problem I had was finding a suitable spindle. In the end, the only economical way to get a decent spindle was to buy a used lathe. I found the Hardinge complete for under $2500. basically with a bed regrind/saddle fit and new head stock bearings totalling around $1200 I would have a virtually good as new lathe given that most of it would get discarded in the conversion.

I got a quote for $3500 for a granite base (35" between centres)

Please tell me you are not entertaining the idea of spending $3500 on a granite base (chair) :D

No do you think I am nuts! I still think it would make a great bed especially for a smaller lathe. The problem is by the time you get a good spindle your looking north of $7000. You can buy a good old cnc machine for that.

pstockly,
as for the spindle/headstock have you ever checked out dunham tool?
http://www.dunhamtool.com

Well I did not think you were nuts, but I was beginning to wonder :D
Just kidding of course.

My plans are to use 1" x 7" x 36" 1018 at a cost of $65, there have been a couple of anti-backside drive and a couple of pro-backside drive comments, my choice is backside, it offers easier fabrication and protection of the ball screw. I just am unable to imagine any machining forces significant enough to cause any cocking of the slide, but we will see. What I have are the THK HSR25 series, they will carry a substantial load in all directions.

Ken

pstockley.....check the formus....someone here was building a homemade lathe....I think it was listed as something like...."My new lathe; bigger,better" brought to you by Stevie....or something like that....his lathe (1st version) was spectacular....and the 2nd version is....well even better...

Toolalew: Yes I did look at dunham. Their cheap 5c spindle cartridge is $2500 and it only has 2 bearings and is only really good for a small lathe. Plus I really wanted the ability to mount a 4 jaw chuck. For this capability you are looking at $4500.

ViperTx: Yes I have seen stevies lathe and it is very nice. However, its designed for relatively small diameter high speed work (gun barrels). I wanted something more substantial. I've used a lot of different types of lathes and I wouldn't want to go back to something as lightweight as that.

Ken: Are you going to build a box bed with that or just use 1" plate? For a very light weight lathe it might be ok but not if you want to do some serious turning. Just to give you an idea, the bed I spec'ed was 14" wide by 10" deep by 52".

For a smaller lathe, I thought about buying a cheap surface plate and getting a local granite countertop maker to cut it down to size and mill some register slots for the linear ways and a slot down the middle. I still may build a smaller lathe just for fun in the future.

I will be using the existing DV-59 lathe body and spindle, removing the original V-way and replacing it with this plate. It should end up with about 29" max travel and a max 9" swing. I have a 14" x 40" travel manual lathe and have never had need for anything larger then 6" diam or 20" in length so this should fit my needs nicely.

I see why that chunk of granite was so expensive now, I was under the impression it was just a plate and not the entire (rather Large) lathe body.

Ken

Ken, that makes sense to me now. I thought the whole bed was just 1" thick :o Why not just stick with the dovetail bed, is it worn out?

Actually the original bed is in very nice shape but is only 3 3/8" wide, consequently no room for the Rails. You can see this in the pictures a few post back.

Ken

pstockley/Ken Shea-
I love the timing of this thread. I too have just recieved a second operation hardinge lathe. Mine is a bit older, but virtually identical. It has a chamferred way with a single T-slot down the center. It should make it easy to mount a plate, and THK rails for the carriage. What do you think about mounting the THK rails inverted, under the plate? So the bolts go through the rail, and up into the bottom of the plate. One in front of the existing bed, and one behind? I would help keep them clean, and also help keep as large a diamater for work as possible. Also, there is no chance of dropping anything on the rails when changing chucks, collets or workpieces. The carrige would then wrap around this plate, and the ballscrew could be attached to either the front or rear THK bearing blocks. I'm going to try to get a model of this concept started soon.

NEATman

This is the thought that I had for locating the THK rails to help keep them clean, and protected. Also keep the maximum swing possible. (see jpg)

For the cross axis, the rails move and the modules are stationary.

I have some real wild ideas on how to use ballscrews and make it both manual and CNC. I'll try to work on those real soon.

Keith

NEATman,
Since you and I seem to be on the same path I started a new thread on CNC'ing the Hardinge 2nd OP lathe here
http://www.cnczone.com/forums/showthread.php?p=119429#post119429

Did not want to take over Pstockley's thread, which I could see happening.

what advantage do you see to replacing a Hardinge bed with rails? If there's a better lathe made than hardinge i don't know what it is - a lot of that value is in that bed with its broad bearing surface. Wouldn't you be better off to take advantage of that and use a bigger servor/stepper because of the increase friction?

I think you've done very well to get this as a starting point, as you pointed out the bearings alone in a lathe like that are a huge expense if bought seperately. Unlike my geared head lathe, the Hardinge is also belt driven for super smooth operation.

with regards to granite and concrete, there is lots of content here, do some searching for polymer concrete etc.

Hi McGyver,
From my perspective that "broad Bearing Surface" on the DV-59 is only 3 3/8", I also did not get any cross slide or tail stock. This method will also not trash the original bed.

Ken

good point, didn't realize it was that small, different model than I’m thinking of and no tailstock/carriage makes it an easier decision....still feels a bit like seeing boyd cottington having a go at a '38 Bentley :D

I'd like to try polymer contrete and mold some beams and evaluate/ compare stiffness. Do you have a good source for concrete polymer, I'm in Bay area/San Jose

thks
michel

Well, Polymer Concrete sounds great in theory, but from my experiences with it, it tends to be more trouble than it's worth. One large machine that I designed using this polymer concrete with internal steel "I" beam reinforcements kept developing cracks over time. It was an absolute nightmare, as most of the cracks did not develop until after we had shipped the machine to our customer. We literally sued the polymer concrete vendor, and I had to compile all my e-mails with that vendor for evidence.

Rule #1, if you are using steel reinforcements, round and smooth all corners before casting the polymer around it. As the epoxy cures, it is an exothermic reaction, heating as it cures. That heat expands the steel, and when the whole thing is cured, and returns back to room temperature, the steel is stressed. If there are any sharp points on the steel, they become stress risers, and cracks will emminate from them.

Rule #2, Leave plenty of room from the threaded inserts to the edge. An inert too close to the edge can crack out a inverted cone shaped chunk. The tensile strength is only as strong as the glue holding it together. Also, the shape of the inserts should be considered. Make sure that they also don't have sharp points that can concentrate stress.

Rule #3, Plates "Cast in Place" may require scraping after casting, due to the stress created in them from the curing cycle, same as above. A better idea may be to "Grout" the plate in place after casting the base. This insures that there is minimal heating when this small section cures. With the mold design, usually the surfaces that you care about will be in the bottom of the mold when you cast. If anything moves due to the hydraulic pressure, start over. If you are interested in this route, I have a few tricks that I developed for aligning the plates and grouting them in place.

Rule #4, make more than enough "rocks and glue" to fill your mold. I recall seeing one very large machine which used this process, and they miscalculated how much material was required. The prototype had a huge parting line where they had to cast the remaining section onto the original base.

Rule #5, I assume that you will be using wax as a mold release agent to free the mold. Remember to thoruoughly remove all traces of wax, or else you will wind up with a paint job that develops "leopard spots", and peeling paint. This is also quite embarssing when it develops over time, at the customer's facility. Also, a PITA to try to remedy, as everything is mounted to it...

If you plan on going this route, read as much information as you can prior to purchasing material. If everything is carefully considered, and you have a good mold, and a knowledgeable vendor, great things can be made with this material. But there is a reason that there is really only one vendor for this in the US. Even Hardinge is outsourcing this process, and they developed much of the technology in this field.

Sorry for the rant, I just wanted you to know what you might be getting into...

NEATman

Here what I have in mind.

A 6x4x30 rectangular shape beam with 3 1/2" coper tube inside to be able to run temperature coolant to control temperature and blow air during curing to control the stress dur to temperature.
Two rows of steel plate 1"x1"x0.375 to act as insert to be able to bolt a steel plate to receive thk/IKO rail.
Use this setup to meaure stiffness and understand most of the issues.

Thank you for the tips Neatman.
Do you have any pointers on epoxy concrete polymer vendors, website, applications notes ?

The paint and finishing as to be considered but it is not the main concern now.

michel

Michel-
One vendor that I have worked with in the past is Philedelphia Resins. http://www.philadelphiaresins.com/
Also, contact Hardinge, and possibly a european vendor.
http://www.hardingeus.com/
One more is Anocast, a division of Anorad.
http://www.rockwellautomation.com/anorad/products/anocastpolymercomposite/

It is doubtful that they will share thier recipe, or sell the raw materials involved, but check into it.

Good Luck-
NEATman

Well, Polymer Concrete sounds great in theory, but from my experiences with it, it tends to be more trouble than it's worth. One large machine that I designed using this polymer concrete with internal steel "I" beam reinforcements kept developing cracks over time. It was an absolute nightmare, as most of the cracks did not develop until after we had shipped the machine to our customer. We literally sued the polymer concrete vendor, and I had to compile all my e-mails with that vendor for evidence.

Rule #1, if you are using steel reinforcements, round and smooth all corners before casting the polymer around it. As the epoxy cures, it is an exothermic reaction, heating as it cures. That heat expands the steel, and when the whole thing is cured, and returns back to room temperature, the steel is stressed. If there are any sharp points on the steel, they become stress risers, and cracks will emminate from them.

Rule #2, Leave plenty of room from the threaded inserts to the edge. An inert too close to the edge can crack out a inverted cone shaped chunk. The tensile strength is only as strong as the glue holding it together. Also, the shape of the inserts should be considered. Make sure that they also don't have sharp points that can concentrate stress.

Rule #3, Plates "Cast in Place" may require scraping after casting, due to the stress created in them from the curing cycle, same as above. A better idea may be to "Grout" the plate in place after casting the base. This insures that there is minimal heating when this small section cures. With the mold design, usually the surfaces that you care about will be in the bottom of the mold when you cast. If anything moves due to the hydraulic pressure, start over. If you are interested in this route, I have a few tricks that I developed for aligning the plates and grouting them in place.

Rule #4, make more than enough "rocks and glue" to fill your mold. I recall seeing one very large machine which used this process, and they miscalculated how much material was required. The prototype had a huge parting line where they had to cast the remaining section onto the original base.

Rule #5, I assume that you will be using wax as a mold release agent to free the mold. Remember to thoruoughly remove all traces of wax, or else you will wind up with a paint job that develops "leopard spots", and peeling paint. This is also quite embarssing when it develops over time, at the customer's facility. Also, a PITA to try to remedy, as everything is mounted to it...

If you plan on going this route, read as much information as you can prior to purchasing material. If everything is carefully considered, and you have a good mold, and a knowledgeable vendor, great things can be made with this material. But there is a reason that there is really only one vendor for this in the US. Even Hardinge is outsourcing this process, and they developed much of the technology in this field.

Sorry for the rant, I just wanted you to know what you might be getting into...

NEATman


For polymer concrete, it's wrong to cast steel reinforcnce inside.
Nobody can avoid inner stress.

Steven.ji-
First, this is a really old thread.
I should have clarified: the vendor reccomended that I install steel reinforcements in the polymer concrete. They also handled the fabrication of the steel structures. They did not round any of the sharp corners on the steel as they had recommended, and I did not see the steel before the polymer was cast, as the vendor was extremely late in delivering them, and they were rushing to complete them.

The material works well (without steel inside) once you get a good mold made, and have a good process, but I think it's really not feasable for DIY'ers to make one or two machines with this process. They would be better off starting with a granite surface plate & adding inserts.

NEATman

The polymer concrete is introduced in China this year. We have made some seminars in China university. In Chinese image, the polymer concrete is suitable for machine base withstanding compressive streghth, but it is not suitable for portal because of weak tensile strength.
They have a idea that steel fabrication is cast in polymer concrete, which reinforce the tensile strength.
Do you have proved experince for steel fabrication cast in polymer concrete?

Steven.ji
I can understand approval for compressive applications, it works well for that. The machine I had designed was supported at the corners by pneumatic air isolation, so there was a potential for the material to sag under it's own weight. The vendor recommended steel "I" beam inserts. The problem described above is what happened when 5 of these 10,000 lb. castings were made.

I believe that it would be different if Rebar (concrete reinforcing rod) were used, especially in shorter overlapping lengths, so the length change of each rod due to the heating that occurs when the epoxy cures has little effect on stressing the material. Also, keeping the rebar well below the surface should also help prevent cracking.

This material has been used on some of the most accurate diamond turning machines in the world, so I know it works well. That'w why I designed the above mentioned machine with this material. It just needs to be carefully designed and manufactured by a high quality vendor who knows what they are doing, not a place that is behind schedule and cuts corners.

NEATman

NEATman

Do your plant operate in China, if so, we can do cooperate for polymer concrete in China.

We are a German company specializing in polymer concrete, named Rampf, did you heard about the company?

Steven-
This design was many years ago, and I have now moved away from the precision machinery design and into designing medical products. The only machines that I design and make now are for my own use.

Had I known of your company, we may have tried you for the machine that I had designed several years ago.

Thanks for the offer!

NEATman
Keith

Keith

Thank you for your offer
Our company is a German basis company in China, we specialize in machine base of polymer concrete.
In Europe, we have more than 30 years for machine base of polymer concrete.
We're the largest company specializing in machine base of polymer concrete in the world.
You may refer to our website: http://www.epucret.de/en/
What's more, our polymer concrete is also suitable for medical machine, becase of antibacteria property.

Regards

Steven

Keith

I still waiting for your reply.
My e-mail address is steven.ji@rampf-group.com.cn. Would you please communicatie by e-mail?

Steven