[llilrex]

Anyone know the approximate stiffness values for mgn12 block and rail or similar? I am going to do some fea simulations but I would appreciate corroboration if anyone has any.

thanks,

[peteeng]

Hi LLilrex- The rail is steel and I usually model the car as aluminium. This makes it close to the published stiffness of the car. If you do a search here you will find stuff on how to model cars in FE. You also need to learn about car preload. Depending on what you are using the cars for you will need a preloaded car. cars bought with no preload spec will be zero preload and these will develop clearance very quickly when used on a CNC machine. This will lead to backlash very quickly... Use a medium or heavy preload car to maximise stiffness and minimise wear....Peter

[llilrex]

sounds good to me. for my current project I have crap lying around that I am going to play with. I have over looked them many times in favor of other better rails. I think I originally got these for a 3d printer. anyway I am interested in what I can make with these and some sheet metal as a personal challenge. these cars are preloaded, although I don't know by how much. I am more curios about the stiffness as a matter of professional interest then making an impact on my current design. I expect my backlash through the kinematic chain to exceed the deflection by quite a bit. thanks for the aluminum suggestion (6061?). I will do a search for FE with the cars.

[peteeng]

Hi Ray - all aluminums have similar stiffness 68-71GPa... depends on what book you look into. Ensure when you model the contact area its on the bearing tracks and does not include the top of the track to the bottom of the car... Peter

[llilrex]

yeah 68 for 6061 and 71 for 7071 (according to google), not much of a difference.

[llilrex]

This is what I am getting ready to run FEA on. I am still working on it but I decided to make life hard and weld up a frame out of 16gauge sheet steel. The machine stiffness should be way higher then it needs to be with regard to the frame. but there it is... with cheap rails and a 400W spindle. a side from time and my junk box it should not cost me anything to build. ( I am lying to myself)

regards,

[peteeng]

Hi Ray - I think all machine builders have some self delusion about $$$ in at least one of their cortex lobes. I know I do....Peter

[elfrench]

Will you be gluing the carriages to the center plate?

[peteeng]

Hi Ray - Nature would never build something with absolute regular spacing. If you make one of those thin plates by accident the correct size that vibrates you will have a huge problem. They need to be made odd sizes to stop this sort of harmonic. Run a modal analysis and see what happens... Peter

[llilrex]

[QUOTE=peteeng;2446054]Hi Ray - Nature would never build something with absolute regular spacing. If you make one of those thin plates by accident the correct size that vibrates you will have a huge problem. They need to be made odd sizes to stop this sort of harmonic. Run a modal analysis and see what happens... Peter[/QUOTE
That’s interesting, I will have to play with that.

[peteeng]

Or fill the entire structure with foaming PU. This will stiffen and dampen the structure. 40kg to 400kg/m3 foam is available. Peter

[llilrex]

peter, so obvious of a solution, and simple with a lot of advantages. I am a little mift that I didn't think of that.

elfrench, I will most probably not be gluing the cars to the X carriage. however I will likely use some metal filled epoxy to cast a flat surface for them to attach to. by impression the X carriage against a flat plate. I will likely do the same to the rail seats. I have used this method many times to great effect. or if I feel froggy I will hand scrape them in after milling.

[peteeng]

Hi Ray - PU Good solution - make sure your local stiffness at hard points eg under rails is thick enough. Its sometimes tricky to integrate a thin structure to the hard points. The local deflections can be quite large even though the global stiffness is very big

One thing standard FE does not do is "large deflection" analysis. Or non linear geometric analysis... Thin structures can behave badly and linear won't pick this up. If you have a large deflection solver (or NL geometric) its worthwhile running it to check no warping, oil canning etc is happening. But if you fill with high density PU this issue will be minimised...

A linear solver assumes the deflected shape is the same as the starting shape. A NL geometry solution applies a small amount of load calculates the deflection, then restarts from this deflection and solves again etc etc. In this way if a section warps it is included in the solution. Another way to get insight to this is to run a linear buckling solver. This will tell you the probable buckling shapes of the structure and the load that it will occur at... If your buckling loads are close to 1x applied loads or less then the structure is not stable at that load...

A structure such as this with thin (16g 1.6mm) steel which is welded will have heaps of internal stress trying to pull it into various shapes. Interesting issue that one... soldering or brazing is a better process for this sort of thing...or if you have bender you can flange the parts and rivet or screw together or even epoxy together. Once you fill with PU it will be bullet proof...Peter