Design for a 'sliding-head' style CNC Lathe

[d_chappell]

Ok, so heres the story.

I am currently doing a project for college in which I hope to design and build a sliding head style CNC lathe. I have done a basic mock up of the frame of the machine in Google Sketchup;



If you can't see the above picture for some reason, I have also posted it at my website http://www.factoryathome.com/xyzgantry.html

The general idea goes as follows. The parts of the frame that are in yellow are to be made from this steel profiling. The grey parts of the frame are aluminium as I do not believe that these parts of the structure are integral to the core strength of the machine (I could be wrong).

The lathe chuck will be located on the platform that slides parallel to the ground, whereas the cutting tools will be attached to the plate that is at right angles to the ground. The cutting tool plate will be able to move left-right, up-down and so multiple tools can be mounted for different functions.

The guide rails and bearings are 16mm ones like this, whilst the leadscrews moving them are 12mm like this.

The stepper motors I think will be NEMA 23's and the motor drivers will be the AVR STMD's like this

I have not decided on the lathe motor and driver yet. I think a DC motor that is controlled with a driver from Homann Designs will be a possibility. Like this one. I hope to be able to control the direction, speed, and position of the spindle so that I can do some milling operations as well as turning.

The workpiece will extend out of the chuck and be supported at the other end by a lathe steady. The difficult part here is that I do not know of any steadies that allow movement in and out of it as well as allowing rotational movement. Does anyone know of a steady that has ball bearings instead of regular rotary bearings?

Please pick apart my ideas and point out any flaws you see. I am pretty certain that I will have missed something obvious.

Best Regards
Dave

[wildwestpat]

Hi Dave

Can you post the graphic showing your proposed machine layout and indicate the size. For some reason all I can see is the little red cross and not the graphic file! OK problem solved as you have added a link to your web site for this build and I have now commented on the proposed machine.

Had a quick look at your selection of frame profile and I did not see the all important moment of inrtia quoted for the sections. I also saw that the sections have a very lax spec on straightness and twist. This will make build difficult. Do you get marks for calculating the deflection in the basic frame work? If so then you need more data on the section materia. Think about and try to reduce the number of joints between the critical parts. Think about the parallel rails - yes they can be adjusted but how ridgid is that frame? Why not use a single piece of plate or even section to hold the rail supports apart on a fit and forget principle.

Is the head stock mounted on the platform to the left of your drawing and the tools mounted arround the square opening to the right.

The normal moving steady uses bronze or brass nose pieces and there are adaptations that use minature ball races to dreduce friction. This link will give you an idea
http://www.aluminium-profile.co.uk/a...m_Profile.html . I guess the steady is to remain adjacent to the cutting tool holding frame and slide on the same rails as the headstock.

Can you list the advantages of your design over that of a conventional capstan lathe?



Good luck regards Pat

[d_chappell]

Hi Pat

Thank you so much for your reply, it's given me lots to think about.

Had a quick look at your selection of frame profile and I did not see the all important moment of inrtia quoted for the sections.

I found some more details specs on another page. It quotes the MOI as being 7.44cm^4 and the resistance moment as 3.72 cm^3.

I do not know details such as straightness and twist but I'm sure I can email the company as request more specs. They have lots more technical details for their aluminium frames, and I suspect that it is because the steel profiles are new to their product range that they have not yet provided the same level of information.

Do you get marks for calculating the deflection in the basic frame work?

The project has quite a wide brief, it is basically a matter of choosing my own project and submitting it for approval before carrying on with the full report. I get marked as much on project management, time management, and technical report writing skills as anything else. Saying that, I do intend to include technical calculations anyway for their own sake.

Think about and try to reduce the number of joints between the critical parts. Think about the parallel rails - yes they can be adjusted but how ridgid is that frame? Why not use a single piece of plate or even section to hold the rail supports apart on a fit and forget principle.

Some good ideas. I've made a modified CAD file of what I think you mean. Does this look something like what you were getting at?

The metal plates are based on a 6mm thick 500x250mm mild steel plate thats available on ebay. There is also a 5mm thick stainless steel plate - more than twice the price though.

Is the head stock mounted on the platform to the left of your drawing and the tools mounted arround the square opening to the right.

Yes, thats correct.

The normal moving steady uses bronze or brass nose pieces and there are adaptations that use minature ball races to dreduce friction. This link will give you an idea

I wasn't able to find anything on the link to do do with steadies. Did you mean to post that particular link?

Can you list the advantages of your design over that of a conventional capstan lathe?

My main thinking was that I could work with quite long pieces of material so that when one job has finished being machined the material would move forward and another job could start automatically. My inspiration was this video on youtube.

Once again, think you so much for the response

Regards
Dave

[wildwestpat]

Hi Dave

Good work on reducing the joints in the frame work. You may like to think of using gusset plates on the main joints that couple the headstock to the tooling plate. These are available and can be fitted after the frame is constructed if required. As far as I am aware this is the cheapest source of aluminimum profiles in the UK and I have to thank another contributer to this site for the link (Hemsworthlad and he has posted some nice pics of a large machine) :-

http://www.aluminium-profile.co.uk/a..._Movement.html

The link I intended to paste is :-
http://littlemachineshop.com/product...2418&category=

The only problem is the support of the stock prior to machining and support of the finished portion of the work. The video you gave the link to is making parts with a 10 to1 length diameter ratio and is I suspect the best they can achieve with out using steadies. The bar stock is limited in length and feeds automatically through the chuck in this type of production machine. The chuck is normally a special type of collet as the same size material is used all the time.

You are thinking obviously about the marking of your project and a few calculations and post build measurements of deflection would be a nice touch to the final project folder.

There is a good principal used in engineering called KISS (Keep It Simple Stupid) which is a good notion to keep in mind all the time you are designing.

Good luck - Regards Pat

Good luck

[Dave]

The design of your tool is driven by the parts you intend to make on it - Swiss-style turning machines such as shown in the video you linked are something of a specialty item, primarily applicable to large production runs or specific parts that can be tricky to make another way. I've done a lot of prototype fabrication (I work mostly in R&D environments, so everything I do is prototypes or very short run production), and none of the shops I regularly use has any type of lathe beyond the typical CNC capstan lathe with automatic tool turret.

If you're committed to the sliding head design, though, I would pay particular attention to rigidity. The plates you added are a good start - but I'd be looking at boxing the section under the headstock to provide better torsional rigidity. The rails for the sliding head could be moved farther apart as well, which will both move them farther apart (greater radius of gyration, and thus more resistance to twisting) and will move them closer to the outer corner structural members (better support, and the plates can tie into this to create your box section). Also, check the deformation of your tool mount plate - it's quite rigid for forces radial to the workpiece, but for axial forces it's liable to have rigidity problems, and axial forces will not be inconsequential.

What about mounting a live center on a second plate that moves in concert with the headstock? That avoids some complexities of using a steady rest, although it also limits your ability to do parting or generate very small pieces.

It looks like you've made a good start to the design, though. Much more fun than the nuclear waste remediation system that I worked on for my senior design project, way back when!