RUNOUT Measuring Method for a Driving Shaft

[allenchu]

A Drive Shaft is going to be measured, but there are some considerations that I don’t have the answers yet and don’t know if the measuring method I planed is correct or not. Please help me to provide the answers or your suggestions.

Drive Shaft Description:

Overall Length: 380 mm

Diameter A: 40 +0 -0.2 mm, 300mm in length, Total Runout relative to Datum B-C within 0.005mm (5 Micrometer)
Two diameters, B and C, collectively establish a single datum axis.

Diameter B: 25 +0 -0.1 mm, 45mm in length (one of the end, as Datum B)
Diameter C: 17 +0 -0.1 mm, 35mm in length (the another end, as Datum C)

The adjacent surface of Dia. A and C’s Total Runout to Datum B-C within 0.005mm (5 Micrometer) also

The measuring procedure I plan to do:

Both Datum B and C is set a bearing, the bearing ID should fit the Shaft’s diameter.
Put the Shaft with 2 bearings installed on 2 same type of V-Block
Rotate the Shaft and get Runout reading on the dial gage to see if it is within the spec. (0.005mm)

Question:
Does the measuring method correct?
(I doubt if the CMM can measure the Total Runout, and the CMM’s accuracy is one of the considerations too. I don’t know if the CMM can be used to measure the Total Runout’ here if the CMM’s accuracy is 0.004 mm)

What the Runout/ID tolerance of the bearing should be? Does it have to less than 0.002 mm?

It must have a clearance between the Shaft and Bearing’s ID; does it will affect the measuring result?

How to set the center line of Datum B-C to parallel to the Granite Surface Plate? (Both end’s diameter are different.)

Diameter A, B, C’s diameter tolerance is much bigger than the Runout requirement, does it will affect the measuring result?

The Runout tolerance in this case is very closed; I guess it may have the controversies of measuring result between the customer and my company. Do you think so? (The equipment and tools used have tolerance too. I am worry about the tolerance accumulation issue.)

Thank you

 

[powerhound]

allenchu,
What material is this? This is absolutely the tightest tolerance for the area involved that I have ever seen. From what I've read, you need to hold a total runout tolerance of less than .0002 of an inch over a 300mm length at a diameter of 1.574 inches. Your concern that the accuracy of your measuring equipment doesn't even hold that tolerance is a valid one.
You should not check the runout with bearings mounted unless the drawing states it as such. I think your logic is correct because that will be it's functioning posture but if it's not on the print, you shouldn't do it. If it's possible, you may mention that to the designer though because it makes sense.
Can I ask what this part is for? I'm trying to figure out what in the world could possibly be so critical about this part that it would require tolerance this tight.

Powerhound
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[allenchu]

Thank you for your suggestion,Powerhound.
I don't know what it will be used for. The customer didn't provide the answer and I can not answer your question due to the NDA.

Could you give me some more suggestions about this issue?


Allenchu

 

[dingy2]

The 0.005 micron spec is tough and gauge error could chew up a lot of the tolerance.

You said you placed diameters B & C in a bearing and then placed them in a V-block. Could the bearing ID to OD have circular runout of some value? - probably.

I probably would set it up a little differently as long as you have the equipment and the calibration has been confirmed.

Set datum B in a divider head or chuck (depends how good the chuck is) and then true up the other end with a digital indicator and stand. I would use a 4 way - zeroing out the 12 o'clock - 6 and then 3 - 9 o'clock. In this way, we reduced the roundness effect on the datum.

Now for total runout using a gauge with a resolution of at least 0.001 microns.

Perform a circular runout on one end of the feature zeroing out the indicator first. Note the readings.

Move to the centre or along the feature. Perform a circular runout again but do not zero out. Note the readings again.

Perform this a few times along the feature.

To acquire the total runout, review all your readings together and then the value would be the highest reading combined to the lowest reading combined.



Dave D.

http://www.qmsi.ca

 

[TheTick]

V-blocks are not really appropriate for runout. You need a spindle. V-blocks will cause deflection due to variance in roundness of datums.

 

[powerhound]

Using V-blocks is not the way to do this. Not only because of the obvious reason stated by TheTick, but because you would need V-blocks of different heights because the end shafts are different diameters. Most matched sets of V-blocks or parallels are only certified to within .0002 of each other so they will not be sufficient for this function being as that's the entire tolerance range of your driveshaft. You would need to hold each end in a collet and the collet centerlines would have to be aligned to within 0,0005mm of each other and each would have to rotate as you moved an indicator along the 300mm length of the surface to be checked.
One point you asked about that I didn't address in my first response was whether or not the size tolerance affected the measuring result. The answer is no. Your geometric tolerance is a refinement of the size tolerance and that's what it should be. You can be from +0 to -0,2mm in size and then whatever size it is, the total runout to A-B cannot exceed 0,005mm.
I'm am still completely befuddled by this requirement. If it's a drive shaft then the surface of the 40mm diameter is probably non-functional. Is there any chance that they accidentally put a value in the Feature Control Frame that was in inches on a print that is in millimeters? At my place of employment we do work for a customer that sent us a set of prints for quote that I immediately did a "NO BID" on for exactly that reason. The print was in mm and the tolerance block had a standard 3 place decimal tolerance of .005. When our sales guy came to me, raising hell about why I didn't think we could make the part I pointed out the .005mm standard tolerance and when he went to the customer about it they realized their screw up. They just forgot to change the block to mm. Now we build those parts using prints with a redlined tolerance block.
Anyway, I wish I could be more help but this requirement is the tightest I have ever seen on something that was not a gauge. Please keep us posted on the progress of this project and let me us know how you ultimately reconcile this issue.

Powerhound
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II