Centreing Design

[drawoh]

I am trying to centre a gear on a co-axial drive system. The gear fits onto a shaft Ø52mm outside and a Ø48mm bore. Right now, I am screwing the gear onto an M52X0.8 thread. I am wondering how accurate this is.

How accurately can a pair of machine shops, one of whom is a gear manufacturer, cut threads concentric with the reference diameter? How accurate will the final assembly be?

Has anybody inspected something like this?

JHG

 

[BillPSU]

The accuracy depends on how the parts were made. Were the gear teeth cut locating off the thread or off the minor diameter of the thread or some process diameter? Was the part heat treated after hobbing and or shaving? Was the gear ground after heat treat? Tolerance on gear manufacturing depends on a lot of factors.

My question to you is why are you depending on a thread to drive a gear. Wouldn't a spline or a keyed shaft be easier to produce accurately? What type of a tolerance is allowed on this gear? A gear can be produced to almost any accuracy you desire but the cost will go up.

 

[drawoh]

BillPSU,

Thanks for the response.

The gear is fabricated to AGMA[ ]8. My gear attachment has to work with a 2mm thick wall. It is an optical system with requirements to be small, but to have a large aperture. I am considering a press-fit, but I want to be able to take everything apart. I have designed a heavy[ ]duty wrench to assemble it. Right now, the gear is bronze, but I am considering stainless steel. There is no heat treatment.

The actual problem is that we are considering the use of a resolver instead of the optical encoder we have now. The resolver has very tight centreing requirements.

JHG

 

[btrueblood]

If you must have a thread, can you at least have a partial-depth shoulder on the part, which engages as a near-interference fit to the smooth part of the shaft?