Shaft deflection

[CNCTerry]

Greeting to all,
I have been concerned about shaft deflection and support. Typical solution is to go with a larger dia. shaft and/or shaft support. This would require a rail mount system of some type. Thinking outside of the box, the support is only needed when deflection is present, and then only where the load is at that particular point in time and distance.

So, why not make the shaft support travel with the load. I attached a file of concept that might resolve some deflection issues. The idea is that the load is carried by the shaft and sleeve bearing, then under downward force the load is transfered to a ball bearing running directly under the shaft and the sleeve bearing. This should prevent shaft deflection ???

Please let me know your thoughts.

Thanks

[Lee Lemon]

I considered something like that for a project but since you are buying a round rail that is essentially perfectly straight, the plate that the secondary ball bearings ride on have to be equally straight as well as parallel to the round rail. This was not practical for the scale I needed so mounted shafts were the most viable solution.

[CNCTerry]

Thanks for the reply. I am using 1/2 dia. shaft with bronze sleeve bearings. I wanted to build a new bearing block and thought why not add the ball bearing as a carrier bearing. It would not carry any significant load in the static condition, but under load the external down force would translate to the lower ball bearing. The running plate could be thin alum, steel or perhaps acrylic. As long as there is a contact area underneath. It would not have to be exactly parallel but should be close, as long as the ball bearing does not run off the plate. I was thinking about adding an adjustment so that it could be lowed to just barely making contact with the running plate. This way the bearing would just skim the plate unless under load.

I appreciate all comments, I am here to learn and share.
Thanks
CNCTerry

[Zumba]

That only accounts for gravitational force, straight downwards. What about deflection caused by typical X-Y cutting, Z-axis plunging, and acceleration in all 3-axes?

P.S. Don't use the bronze bushings. The frictional coefficient of those things are horrendous when there's even a trace of misalignment.

[CNCTerry]

You are right, good point. I did not account for the forces in other directions. I will have to make some more design changes. Thanks for the info and the up shot on the bronze bearings.

All comments are welcome, I am here to learn and share
CNCTerry

[dz1]

i actually just had to do something similar to compensate for deflection in my x and y rails

for the Y i did this

http://www.cnczone.com/forums/attach...2&d=1180148216

and for the x i had to do this

http://www.cnczone.com/forums/attach...9&d=1182100947

they operate smooth, will see if my parts show any weirdness in a couple days when i finish the tool mount and begin cutting

[CNCTerry]

DZ1, thanks for the pics and please give me an update when you get ready to cut. It looks like you have a cam adjustment on the back of the bearing mount? I took my machine down and pulled off everything except for the electronics and the base. I am wanting to improve the machine for less deflection, more accuracy and easier to adjust out.

I am going to rebuild her and make her better.. stronger... faster...

Thanks for the reply
CNCTerry

[dz1]

yup all the add on bearings are mounted with eccentric bushings so that I can adjust teh fit. on the Y the bearing stud goes through a bushing with an off center hole, for the two X followers, the bearing is mounted to the al spacer piece, which has a stud offset into its base to allow adjust there.

I have cut a couple signs and circles so far testing and tuning the machine, the follower bearings have compensated for the deflection quite well, there is maybe .050" play in the torsional twist of the x under full load (was about .25") and no noticable sag on the y when tramming the table with a dial indicator

the x play should not be noticable for making signs and stuff, and will not even come into play if i get around to usingit as a plasma table too..

end of the day i knew i was pushing it with a light weight (cough cheap) design but will work out for intended usage. granted i probably cant take a 1/4 cut at 40ipm but my little laminate trimmer cant do that either so its an even balance i guess.