Sanding Disk

heathicus · Post by **heathicus** » Wed Apr 18, 2012 12:12 pm

I threw this together in Paint real quick to show what I'm talking about. Maybe it applies to what you're doing, maybe it doesn't.

The big black block is the arbor. The horizontal gray line is the horizontal plane of the arbor. The vertical gray line is the vertical plane of the saw blade/sanding disk. The diagonal blue line is the runout. The red lines, labeled A and B, indicate places where the runout is measured.

Measuring the runout at point B will result in a higher number than measuring at point A. Because the further you get from the arbor (or the center of the blade, or the horizontal plane that is the reference for the runout, or the pivot point, or whatever you want to call it), the further the blade will be from the dial indicator (the vertical plane of reference).

[ATTACH]17166[/ATTACH]

charlese · Post by **charlese** » Wed Apr 18, 2012 1:58 pm

dusty wrote:OKAY. I'll try this differently. I did not get the feedback that I was hoping for. Maybe I'll do better by requestingspecific comment.....

.....At this point, I would say that I have .006" runout (attributable to the blade and arbor). Is that a correct analysis? Or am I incorrectly labeling this test result? I have, by virtue of another test not described here, declared that I have .002" runout on the main drive shaft. Therefore, .004" of this variation I would attribute to the arbor and blade.

Now, having found and quantified that high spot, I set the dial indicator to .000" and rotated the high spot to the other side of the blade (outfeed side). At this point, I measure .008". I would, at this point, say that my table is not parallel to the blade and that it is off by .008". Is that a correct analysis?

BTW The old plywood blade looks to be quite flat for a saw blade. Maybe it should be my alignment tool (reference plane).

SPECIFIC COMMENT:---- of "is that a correct analysis?"
The answer is NO!

Reasons for this answer: ------ There are many variables (some identified) in these measurements that are not accounted for. For example lets first talk about your identified runout on the main drive shaft. Although this runout has been identified, how is the high/low points of the shaft related to the high/low points of the saw blade.

Additionally, how is any of the runout of this arbor related to the positioning of the saw blade.

Are you sure the arbor tightens always in the same circumference location on the saw blade as when tested earlier?

How about the slight rotational location of the arbor on the spindle when tightened. We can not be absolutely positive the set screw sits on the arbor in exactly the same spot (speaking of thousandths here) and the set screw has not wiggled or slanted slightly in the threads of the arbor.

Don't know what you are trying to show here, but if you can ignore many variables (in thousandths of inches) then maybe your table is perfectly aligned and you can't get any better.

Maybe there is some excellent reasoning when saying .005" is close enough for alignment. The largest confusion/concern over the question "is this a correct analysis" is trying to quantify Thousandths " in a system that is not in any way manufactured to those standards.

An analogy:---- It is almost like trying to measure the squareness of a house or a large room in hundredths of inches.

Once again - I will state that just because we are provided the tools to measure to the nearest thousandth of an inch - does not mean that we can, or should, try to apply this preciseness to everything.

wa2crk · Post by **wa2crk** » Wed Apr 18, 2012 1:58 pm

Dusty
Your test sounds like it might be valid but you may be mixing some apples with your oranges. For a test to be valid, comparisons must be made using the same methodology. I don't believe that the shaft runout test and the arbor/blade test should be compared to one another.
The shaft runout and the blade runout test comparison may be different if the arbor is rotated from its original position. IE if you find the high spot on the shaft and blade combination and then demount the blade and rotate the shaft to its low spot and remount the blade then your results may be different provided the blade remains in its original orientation.
Things have to remain in the same orientation to each throughout each test for the tests to be valid.
Whhen you found the high and low spots on the blade were they 90 deg or 180 deg apart? Taking readings at every 45 deg and marking the blade may give more meaningful results.
Most blade manufacturers have a limit for acceptable runout. Check with the manufacturer for runout or flatness tolerance.
They might be able to give some advice on how they take their measurements.
Bill V

dusty · Post by **dusty** » Wed Apr 18, 2012 4:56 pm

charlese wrote:SPECIFIC COMMENT:---- of "is that a correct analysis?"
The answer is NO!

Reasons for this answer: ------ There are many variables (some identified) in these measurements that are not accounted for. For example lets first talk about your identified runout on the main drive shaft. Although this runout has been identified, how is the high/low points of the shaft related to the high/low points of the saw blade. That relationship has NOT been established and since shaft runout is only .002" I don't believe it needs to be. Yes, that runout contributes to all other measurements made on the blade and can not be simply disregarded.

Additionally, how is any of the runout of this arbor related to the positioning of the saw blade. I have experimented and runout measured at the gullet of the blade can be changed by rotating the blade on the arbor. Again, since I had relatively small runout numbers, I did not do that for this test. It is important to note that runout can be improved by fine tuning the blade to arbor relationship.

Are you sure the arbor tightens always in the same circumference location on the saw blade as when tested earlier? It probably does not.

How about the slight rotational location of the arbor on the spindle when tightened. We can not be absolutely positive the set screw sits on the arbor in exactly the same spot (speaking of thousandths here) and the set screw has not wiggled or slanted slightly in the threads of the arbor. I'd be willing to bet a Shopsmith that it is somewhat different every time the arbor is removed from and then reattached to the spindle.

Don't know what you are trying to show here, but if you can ignore many variables (in thousandths of inches) then maybe your table is perfectly aligned and you can't get any better. If one was striving for perfection, no variable can be ignored. As for my table alignment, I already know that it is very, very close and that it need not be tampered with right now. I also know that there are a number of things that can change that.

Maybe there is some excellent reasoning when saying .005" is close enough for alignment. The largest confusion/concern over the question "is this a correct analysis" is trying to quantify Thousandths " in a system that is not in any way manufactured to those standards. I was not attempting to quantify "thousands". I was attempting to provoke discussion that would identify contributing factors to misalignment.

An analogy:---- It is almost like trying to measure the squareness of a house or a large room in hundredths of inches.

Once again - I will state that just because we are provided the tools to measure to the nearest thousandth of an inch - does not mean that we can, or should, try to apply this preciseness to everything. Point made.

Thank you Charlese for all of your very constructive comments.

dusty · Post by **dusty** » Wed Apr 18, 2012 5:10 pm

wa2crk wrote:Dusty
Your test sounds like it might be valid but you may be mixing some apples with your oranges. For a test to be valid, comparisons must be made using the same methodology. I don't believe that the shaft runout test and the arbor/blade test should be compared to one another. I did not mean to co-mingle shaft runout and "blade wobble" in the same discussion. Please note that my comment to shaft runout (.002") made reference to a different test not detailed here.
The shaft runout and the blade runout test comparison may be different if the arbor is rotated from its original position. IE if you find the high spot on the shaft and blade combination and then demount the blade and rotate the shaft to its low spot and remount the blade then your results may be different provided the blade remains in its original orientation. Very good point which was als0o made by Charlese.
Things have to remain in the same orientation to each throughout each test for the tests to be valid.
When you found the high and low spots on the blade were they 90 deg or 180 deg apart? Taking readings at every 45 deg and marking the blade may give more meaningful results. I will do that just to see what the results are. I do have one saw blade that seems to be almost perfectly flat. The reading changes only .002" as I rotate the blade through a full rotation.
Most blade manufacturers have a limit for acceptable runout. Check with the manufacturer for runout or flatness tolerance. All but one of my blades are Shopsmith. Can you tell me what the Shopsmith standards are for these blades.
They might be able to give some advice on how they take their measurements.
Bill V

I hope that these discussions have been helpful to someone and that if anyone has questions regarding these performance characteristics they will bring them to the attention of this forum. Discussions like this can be very informative even for those who do not believe in tight tolerances on a table saw or drill press.

wa2crk · Post by **wa2crk** » Wed Apr 18, 2012 10:12 pm

Dusty;
For the flatness tolerances you will have to check with the manufacturer of the blade or with Jim McCann at SS.
Bill V

wa2crk · Post by **wa2crk** » Wed Apr 18, 2012 10:17 pm

Dusty;
For the flatness tolerances you will have to check with the manufacturer of the blade or with Jim McCann at SS.
Bill V

woodburner · Post by **woodburner** » Fri Apr 20, 2012 2:47 am

I'm not sure what the acceptable runout tolerance for the SS sanding disk is, but I'm sure it will be off something due to weather, humidity, and the like. Same for saw blades. I'm sure it can change from day to day depending on these factors.

I mount my dial indicator in the miter slot of my saw table to check the alignment of my table to the sanding disks and saw blades, and also the alignment of my rip fence, as taught to me by former SS instructor Rick Davis and returning instructor Doug Reid. Yes, you need to know how much runout your blades, etc. have, but I don't think you can actually adjust the runout of these items indidually. You will want to adjust the table to the blade, etc., not the blade to the table (hope that makes sense).

jimthej · Post by **jimthej** » Fri Apr 20, 2012 12:31 pm

I'm going to change my name to Alfred E. Newman.

Then go cut wood!

nuhobby · Post by **nuhobby** » Fri Apr 20, 2012 2:09 pm

Like this ?
[ATTACH]17216[/ATTACH]