suggestions on GDT strategy

[pdybeck]

I am interested in hearing how some of you might approach a definition using GD&T of a ball screw assembly containing machined ends. I am interested in making sure the ball nut will always be able to be positioned/aligned within the center of rotation of the screw and machined ends/journals by a certain amount, for assembly consideration at the next level up. The ball nut and screw contain some radial lash and this can be utilized to satisfy the centering condition. I would be interested in making sure this condition applies over the whole range of travel of the nut on the threaded portion of the screw. I can clarify the application if needed, but can't post any pictures/drawings.

Pete

 

[CheckerRon]

I worked at a company that made ball screws for aircraft controls and lead screws for nuclear control rod drives, but this was in the earlier days of GD&T (1968 - 1975). Our primary controls were circular and total runout of end journals and screw length, for straightness and circularity using an A-B datum structure. The main concern was control of cululative lead error on the thread to maintain the design centering over the travel. As I recall it was .0002 max inch per inch. Memory fades on the ball nut controls, but I believe runout was also the prime controller.

 

[CheckerRon]

CUMULATIVE lead error, that is.

 

[pdybeck]

I'm more interested in the convention used to define the centering of the ball nut relative to the machined journals. I understand lead error and the like and that is not my aim/focus here. I am considering placing a true position tolerance on the ball nut relative to the machined journal. How do I state that the radial lash of the screw can be utilized to satisfy this condition? I have considered a note saying exactly this, attached to a true postion tolerance on the nut body. It would read something like this... "This tolerance condition must be satisfied for all locations of the ball nut along the ball screw thread. Radial lash between the ball nut and ball screw can be utilized to satisfy this requirement."

Is there a better way to state this? Is there a better way to tolerance this? Is there a GD&T convention that can be used to state clearly what I have just said?

 

[CheckerRon]

This may be a situation where the best solution is using that drawing note flagged to the Position FCF. I would see the radial lash as a clearance that can be added to the position tolerance value, using MMC for feature tolerance variances. How will you measure it however --- by rotational measurement by indicators off the journals as datums? That is why I was thinking about runout callouts, but at the assembly level as well as on the piece parts.

 

[pdybeck]

We could use a functional gage to check and make sure the gage does not bind on the machined journal as we translate/rotate the ball nut and gage together along the full thread length of the screw. The gage would locate on the ball nut. The gage would contain a hole that is sized over the journal diameter by the amount specified in a true position tolerance. Just thinking this through... Does that make sense?

Pete

 

[CheckerRon]

Yes, such a gage set at virtual condition should work well.

 

[pdybeck]

I dont think adding radial lash to the positional tolerance requirement, would give me what I want. I think using radial lash/clearance to satsify or put the nut into a satisfying condition is more what I want. I may be wrong...

Pete

 

[CheckerRon]

In retrospect, I think I agree. This would allow the part variances to use up that necessary claearance.

 

[pdybeck]

CheckerRon, thanks for the help.

Pete