[greenene]

I've been thinking about replacing my linear shafts with supported rails. Currently , the gantry sits on two unsupported shafts, which flex when I am drilling. I suspect that the flexing of two y axis( the long axis) results from the bit hitting the workpiece and, during that brief time when the endmill is trying to drill, the force at work pushes back the router, which sits on the x axis. Based on observation, when the router moves slightly back, the y axis are pulled up. This is the flex I am talking about.

Now, I see many design with supported rails having the pillow block and the open face of the bearing facing the ground and the gantry mounted on top. It would appear, therefore, that with the open face of the bearing part facing the ground, the linear system is operating at a fraction of its load bearing capacity.

I am talking specifically about vertical force at work during the drilling operation, and which way the y shafts flex.

If my observation is correct, shouldn't the pillow block with the open face of the bearing be facing up instead. I am referring to the open part of the bearing....

One last question, would it make sense for me to replace 31"long 20mm linear shafts with supported rails, but with the open face of the bearing on its side? If I keep the shaft diameter the same, the mod would be plug and play, with hardly any machining.
I cut mostly mdf and notice some flex, even when I use a pecking method. Again, based on observation, the flex is between .015-.050, depending on the feed rate and the bit diameter.
thanks
tony

[wizard]

I've been thinking about replacing my linear shafts with supported rails.

Unsupported rails over long distance are a bad idea unless you are talking seriously large rails. However even the so called supported rails need a solid base to be fastened to.

Currently , the gantry sits on two unsupported shafts, which flex when I am drilling. I suspect that the flexing of two y axis( the long axis) results from the bit hitting the workpiece and, during that brief time when the endmill is trying to drill, the force at work pushes back the router, which sits on the x axis. Based on observation, when the router moves slightly back, the y axis are pulled up. This is the flex I am talking about.

You may have your X & Y destinations reversed from convention, but this is a continuing area of discussion.

In any event it is a certainty that those rails are deflecting due to cutting loads. All machines do this to some extent or another.

Now, I see many design with supported rails having the pillow block and the open face of the bearing facing the ground and the gantry mounted on top. It would appear, therefore, that with the open face of the bearing part facing the ground, the linear system is operating at a fraction of its load bearing capacity.

Possibly. The thing is you need to analyze all the forces involved here for your specific design to suggest that this is a problem one way or another. I highly doubt that over normal operation the upward thrust is a serious percentage of run time. Most end mills will have a tendency to pull down on the spindle.

I am talking specifically about vertical force at work during the drilling operation, and which way the y shafts flex.

This certainly can be an issue with weak machine designs.

If my observation is correct, shouldn't the pillow block with the open face of the bearing be facing up instead. I am referring to the open part of the bearing....

That would depend upon the bearings load capacity.

One last question, would it make sense for me to replace 31"long 20mm linear shafts with supported rails, but with the open face of the bearing on its side? If I keep the shaft diameter the same, the mod would be plug and play, with hardly any machining.

That would depend upon the design of the machine. However you would be applying force to the rail in what is likely a less than optimal way. Note though that no matter how you orient the rail some machining forces will be applied laterally.

From a practical matter orienting the rails as described may be a bit difficult as far as establishing the slave rail parallel to the master rail.

I cut mostly mdf and notice some flex, even when I use a pecking method. Again, based on observation, the flex is between .015-.050, depending on the feed rate and the bit diameter.
thanks
tony

That is terrible deflection even for a router dedicated to wood machining.

[greenene]

Thank you for your thorough analysis of the issue at hand. You just made the case for not modding, but to look for an upgrade in a machine.
tony

[louieatienza]

I read somewhere that for supported round shafting, the load rating is 100% in the standing (opening facing down) direction, 70% on its side (opening facing sideward), and 30% upside down (opening facing up). So I would say mounting the rails upside down would work, since the only force pushing the gantry down would come from the weight of the gantry itself and gravity. But since the bearing blocks are rated at 300lbs a piece (and you'll have 4) unless your gantry weighs a half ton you should be fine.

[wizard]

Well if you are looking for a reason to upgrade I'm glad I could help! After all a new machine is like Christmas day for many.

However that being said I don't know what sort of design you are working with here. To be perfectly honest some are't worth upgrading in my opinion. On the other hand building a completely new machine is not cheap so it is up to you which way you go.

Another way to look at this is that the table and the associated axis is the foundation of your machine. If it isn't built well you will compromise the whole machine. This is why I recommend that people build a robust table first if funds are limited. It is far easier to improve the gantry and other parts of the machine later. Of course you need to know up front and understand that future upgrades will be required.

Thank you for your thorough analysis of the issue at hand. You just made the case for not modding, but to look for an upgrade in a machine.
tony

[Michaelofburien]

I used 16mm fully supported rails like you describe for all 3 axis of my MDF build. They are only as supported as what they are fastened to. If you slap them onto a 1/4" sheet of MDF they will flex under minimal loads. Build a torsion box underneath them and suddenly you need a micrometer to detect the deflection.

As for the open side of the bearing, its has not really been an issue. They are "strong enough" to keep the table and gantry aligned for me to cut aluminum block. Unless of course you are hanging several hundred pounds off of them. All 3 axis travel smoothly and don't "hang" anywhere.

I did not even consider "unsupported" rails as I wanted as stiff a machine as I could afford to make. Swapping out for supported rails that are mounted solidly I believe would improve your machines stiffness. Regardless of their orientation. And to be clear I have one set of 4 mounted "opening down" for Y, 4 facing rearward for X horizontal and 4 facing rearward at 90 degrees to horizontal for Z. And all 3 axis are strong enough for peck down to the table through a 2" piece of aluminum without visible deflection.

I think replacing your rails sounds like a good idea.

[greenene]

Interesting. Are you saying that you have 4 rails for each axis, meaning that each side of the gantry has two rails in parallel. I thought this was not practical, given the potential for slight misalignment between rails, and the additional inertia on the motors.
tony

[Michaelofburien]

No, to clarify. I have 2 rails per axis with 2 bearings each. 4 bearings for axis.

[dmalicky]

Good comments above.

The orientation of SBR blocks is important, but unfortunately there is no ideal orientation for all axes and loads. Still, for each axis, some orientations are better choices than others.

Louie's post on the tension capacity being 30% of the compression capacity is in the ballpark of experimental stiffness measurements:
- http://www.cnczone.com/forums/linear...ml#post1412778
- Stiffness varies with load, but at a load of 100 lb, the stiffness is about 70k lb/in for tension, 190k lb/in for compression (27%). "Capacity" describes how much load the bearing can take before the balls fail from fatigue, while "stiffness" describes how much the bearing deflects under a certain load. Still, both of them are related to the ball loading, so it's reasonable that they are both around 30%.
- The lateral (side) stiffness is probably in between tension and compression. The balls are oriented fairly good for that, but the rail is only bolted every 6" and it will flex.

The long-axis blocks can be oriented with their open face either down or lateral--both have pros and cons for this axis.
When looking in the side view (along the gantry tube), for drilling especially, the cutter needs to be *in-between* the bearings. If the cutter is hanging off the front, drilling will easily cause the front bearings to switch their load direction, because the cutter has more leverage than the bearings do. This is especially important if the opening is face down. If the cutter is ~half-way in-between the bearings, it would have to lift the entire gantry off the rails before the bearings see tension.

With the unsupported rails of your current machine, the cutter deflection will be worst when the axes are in the middle of their travel. And best when at the ends of the travel.

I'd suggest posting a picture of your machine; as Wizard mentioned, there may be other weak spots that make it difficult to effectively upgrade.