Track/Guide help

[Ritmo2k]

I have a device which I was positioning with a THK Linear Guide and caged roller bearing. Now, the path is no longer linear, but has varying changes (The path may resemble an exponential curve possibly when done). It was suggested to remove the thk components and just mill a slot and develop an assembly of bearings that can counteract all directions of possible implied thrust.

The device being positioned by the slot requires the slot to orient it with respect to the plates normal.

Since using only two bearings bearing's would cause one side to always rub, I am lost at developing a concept that allows a possibly rectangular array of 4 bearings on the device being positioned. This would require the slot change width as the device rotates and the axis between each opposed bearing contacting opposite sides of the slot moves away from being normal/perpendicular to both slot walls.

Any guidance would be greatly appreciated.
Thanks!

 

[hendersdc]

I would say think about how a railroad car is setup but turn the wheels on their sides. They use 4 in the front and 4 in the rear but you could use 2. If we say your path sits in the X-Y plane you could have two bearings whose centers are orriented symetrically about the direction of motion but still in the X-Y plane that together can rotate about an axis normal to both the direction of motion and the line between bearing centers (the + or - Z direction). Do this in two sets and you have your railroad car. I would think with this approach a "rail" to follow would work better than a slot because as the bearings rotate off of straigt the overall dimension can decrease which would cause you to loose contact with the slot walls if this occurs in a straight away. With a rail the decrease in spacing would not occur because the bearings are already in contact with the sides.

Hope this helps and that I understood what you were looking for correctly.

 

[IRstuff]

You can use either a slot or a rail. If a rail, then you put bearins on either side of the rail. With a slot, you'd do the opposite, the bearings would roll along the sides of the slot, with the bearing support in the middle.

TTFN

FAQ731-376

 

[Ritmo2k]

Hey guys,
I may be missing something which lead me to dismiss the use of a rail using bearings whose centerline axis is normal to the plane the rail deviates within.

Lets visualize a possible assembly as a railcar axle whose wheels have 90 degrees of positive camber. CMIIW, but don't railcar's have that assembly free to swivel about the center point (Sort of like half a trunnion)? My thoughts are that if both axle sets are locked parallel to each other, they will succumb to the same problem but in an opposite fashion. They will become tight on the rail as it curves?

In my assembly if I used this mechanism, I would have to place maybe 4 bearings above and below the plane making contact with it for stability. This could be done, but I am hoping for a simpler solution to ease manufacturing.

Thanks for all the time and thought guys, its helping a great deal!

 

[hendersdc]

Ritmo2k,

I have attached a quick picture of what I am thinking. Lets assume the direction of motion is to the left as viewed in the picture. Also assume that the rail it is following is taking a turn to the right (concave upwards as viewed in the picture). The two bearing pairs would rotate about their pivots on the main body and orient themselves as shown by the blue lines. The distance between the contact surfaces of the two bearings within a pair does not change so a uniform rail thickness is possible. Two more bearings that support the assemblies weight on top of the rail might be desirable as well. If when you say stability you mean that this assembly can not be held on the rail by gravity a dove tail style rail with the side bearings tilted so the contact surface is pointing slightly up would prevent the assembly from coming off and save you a couple of bearings.

Once again I hope this helps.

 

[MintJulep]

How about a linkage?

 

[IRstuff]

I think that we're assuming that the rail does not participate in the traction, i.e., there are road wheels that actually drive and support the vehicle, and that the rail is strictly for steering and is not needed to keep the vehicle either on the plane or to prevent it from leaving the plane.

Nonetheless, even if there is that added requirement, the center rail concept can be adapted to handle that. You'd make the center rail a cylinder supported by a thin rail on the bottom. The wheels would still be on either side, except that the rims are hollowed out to match the cylindrical center rail. This type of rail is used on a LARGE number of roller coasters.

TTFN

FAQ731-376

 

[Ritmo2k]

Appreciate all the insight guys. So I am thinking based on the size of the mechanism and ease of manufacturing, I would mill a slot with chamfered edges (top and bottom) in a plate. I would then use a carriage with two v-grooved concentric studs on one side, and a third in between those two on the other side that is very heavily spring loaded. This should eliminate the need to run more bearings to hold the assembly in the slot, as the weight and forces are very low. Given the replaceable plate is milled accurately then polished, it should provide for a good surface for the v-grooved stud's.

The weight of the assembly should not exceed 10kg and forces applied should not exceed ~150N. The v grooves will see ~12mm on each side of contact with the steel plate. Anyone see anything wrong with this?

Thanks!

 

[hendersdc]

Your solution can work but you might want to watch out for a few things.

-Your v-groove studs should have the groove cut with generous lead in radii (assuming the spring force is axial on the stud) or cut around the entire perimeter (assuming the force is radial) not just a straight groove on one side. This will eliminate any alignment problems as well as making it not bind up in the turns.

-You will need to know the tightest turn the slot will make so you can calculate the travel necessary in your spring loaded studs.

-Also, depending which way you have the springs, the carriage will resist going into turns or coming out of them because this will require compresion of the springs. Depending on your driving mechanism you could have problems keeping the carriage where you want it rather then where the springs want to force it.

 

[Ritmo2k]

Good points...
Thanks everyone!