| | |
My primary goal was to increase the rigidity of the column in the x direction.
The joint is pinned and bolted- I probably used too many fasteners though I rather have too many than too few.
The holes in the column casting where reamed and the pins are a slip fit in their holes. I was afraid of creating internal stresses in the column when the plate was applied but I am confidant that I am okay there. I really didn't want to ruin the column!
I'm waiting on some AC bearings and a boring head to arrive in the mail. I plan on installing 7201 bearings on the Z axis. The stock thrust bearings went bad and initiated breaking down the mill in the first place.
I will be machining a bracket that will link the ball nut to the head.
This is VERY nice!Originally Posted by Thorpydo
It is something I have been planning on doing to my X3 when I convert to CNC. The main difference in what I have been planning is that I will fill the column with Epoxy / Granite and use that as the "fasteners". I am hoping the E/G would minimize the internal stresses in adding the back.
VERY NICE!!!
"Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away." Antoine de Saint-Exupery (1900 - 1944)
I've done some FEA on the assembly and the results are interesting.
The bottom surface of the column is 'fixed' and a 500lb load is applied to the flat of the cylinder, representing the head. The material is cast carbon steel.
Here are pictures from the first simulation- without a backplate. max deflection is .0026 in.
And here is a simulation with the backplate.
I am limited to using simulation express with solidworks and so can only simulate using one material. Therefore, the backplate is simulated as cast carbon steel and assumed to be rigidly attached to the column.
The max deflection is actually more, at .0029in but the model appears to twist significantly less, and instead translates where the slot starts.
It will be interesting to see how it cuts!
interesting project!
It appears you are mounting the ballnut to the head itself?
I think this would be a very smart thing to do. A lot of x3 owners complain of when you jog the z axis in one direction the head tilts in another, because it is driven from the rear.
I agree it is smart. But if the screw and nut location don't change (which it looks like is the case here), the torque moments are identical. Therefore there would be no difference in the rocking effect you mention. However, I suspect the rocking is due to a different issue anyway (possibly a poor fit of the Z axis gib).
"Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away." Antoine de Saint-Exupery (1900 - 1944)
The ball nut will mount to the head via a machined part, through the slot in the column. As Hawk points out, the head won't be feeling anything differently.
However, while I had the machine apart, I did regrind the z gib which should help prevent the head from tilting on direction change. The original finish was poor and the fit terrible. The taper was off by about 10 thou. I now have a nice ground surface that matches the z carriage. I had to shim the gib as taking that much material off forced me to do so. I also installed a locking nut on the z gib adjustment screw. The screw itself would move slightly with each direction change and installing the screw prevented this.
I lengthened the lever arm on the head to get a more accurate representation of deflection.
The results look much more promising. Torsion plays a much bigger role in deflection when the lever are is longer.
Without backplate I'm getting a deflection of .0175 in.
With the backplate I'm getting a deflection of .0065 in.
I'm not concerned with the numbers themselves as the load applied is somewhat trivial. In this case, I applied 500lbs but I don't know what a reasonable cutting load would be. Close to a three to one improvement.
oops maybe i don't quite understand properly what your doing. I see you milled a slot in the head-side of your column but there doesn't appear to be any way the ballnut can mount to the arms that reach from the head to the rear of the column, if you get what im saying.
that is a nice machine you have access too!
I'm pretty sure you do understand what he is doing.
What you may be missing is that the force to move the head is being applied from the same location as it was the old way. It doesn't matter if the bracketry takes a circuitous path around the back of the column or if it is directly through a slot in the front. It still applies the force from the same place. Therefore the head won't see any difference.
The biggest issue with a rear bracket is making it rigid (which was done pretty well in this case). Of course the rear bracket makes it impossible to put a back on the column also.
"Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away." Antoine de Saint-Exupery (1900 - 1944)
Thorpydo, you may also want to examine the rigidity in the Y direction. However this may be more dependent on the structure of the base casting. If you were to model the base while loading the column in the Y direction, I suspect you will find a deficiency at least as significant as the column (especially since it is my guess that the stresses are higher due to the leverage of the column).
If you take a look at the bottom of the base, you will notice there is a lack of structure to handle the bending loads applied by the column in the fore/aft direction. It would be interesting to see the "hot spots" in an FEA model. I suspect you will find a need for more structure in the flat area between the column and the Y axis dovetail.
The easiest way I can think of to supplement this, is to cast the base full of Epoxy / Granite.
"Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away." Antoine de Saint-Exupery (1900 - 1944)
that actually not true. the moment arm is moved to the ball screw mount in the middle of the column as opposed to an inch behind it. if the system were magically perfect and rigid, it wouldnt matter, but it isnt. doing what hes done should make a significant difference in reducing head "tilt" as well as binding in the ways allowing smoother motion and higher speeds/accelerations. its the side brackets that cause all the problems, not so much the ball screw position. the brackets are a terrible design, very flimsy mounting. i have a smaller sieg kx1 machine, and it behaves the same way (except even worse because the z saddle is very short).
Posting Permissions
| | |
LinkBack URL
About LinkBacks
