What's the radius of your blades?

[AgnesNitt]

The glide portion of your blades seems to be the 'rocker size' that the blades are referred to in ads. It's possible to actually measure your blades rockers by hand and calculate the rocker using an online tool

Now there are two options. The first is that the total blade (glide rocker AND the spin rockers) are used to sort of 'average out' the rocker number reported in the ads.
The second option is that only the glide rocker is used.

Following the image below here's what to do:



1. Total 'average' blade radius. Lay a straight edge at the heel of the blade to the front of the blade at the end of the spin rocker. Where does the spin rocker end? For the moment let's assume it ends where the hollow ends. So mark the line on the straight edge. Using a ruler find the length of the line and the highest point on the blade between the marked line and the edge. Feed the data into the calculator.

2. Total glide rocker measurement. Lay a straight edge at the heel of the blade to the point where the spin rocker starts. So mark the line on the straight edge.
Now using a ruler find the height of the highest point between the liine and the edge. You have two measurements, insert them into the calculator.

Report your results here: Blade model, 'full average' rocker, and glide rocker.

Everybody ready!? Grab your blades, a ruler and a straightedge (marker is helpful). GO!

[riley876]

I worked out this way to measure a bit more accurately:

http://www.skatingforum.co.uk/index.php?topic=6727

[Query]

Or just do it by printing out profiles, and sighting the blade against them.

I have an image of blade profiles here.

[AgnesNitt]

Where'd you get that?

[AgnesNitt]

The paramount video is here on facebook. https://www.facebook.com/Paramount-Skates-149449021742190/?fref=ts

[Bill_S]

I'm working on it Agnes... 

[tstop4me]

I believe you need an accurate measurement setup such as Bill's.  I did some calculations.  Assuming a width of 10 inches, then

height = 0.149 inches for a rocker radius of 7 feet, and

height = 0.130 inches for a rocker radius of 8 feet.

(By the way, I fed the height and width values into the calculator and got back the expected radius values.)

The difference in height would only be .019 inches.

Hand tracings and measurements with a ruler won't yield sufficient accuracy to distinguish (for comparison, 1/64 inches = .0156 inches).  Neither will sighting against a template.

I have a couple of dial indicators that I was planning to ditch on eBay.  I think I'll keep them now.  :-)

[Bill_S]

OK, here are the stats on some new 10-1/2" Jackson Ultima Synchro blades that I never got around to mounting.

First the process (click to enlarge in some browsers)...



The miter slot in my table guides a miter bar on the underside of the jig to prevent side-to-side motion. Tolerances are adjustable to provide a tight fit.

I took rocker depth measurements each 1/2" of blade length, and compiled them into a graphic/regression program. After the points were plotted and the scale changed to visually exaggerate the rocker depth, I ran a curve-fitting routine on it. It's a very good fit with an R^2 value of 1.0013. The resulting formula in case someone wants to duplicate this on a CNC machine is...

Y(x)= -(1.1252445*10^(-6))*x^7+(4.2795519*10^(-5))*x^6-(6.249962*10^(-4))*x^5+0.0044961*x^4-0.0169112*x^3+0.0380313*x^2-0.0996406*x+0.23539

You can visually see the nice fit between the two curves...



Oh yeah, baby! Standard deviation is only 0.001" - better than most manufacturing tolerances can be held.

Then I used my cartesian coordinate numbers shown in the table and an online calculator to find the rocker between the 1" from tail to the 5.5" from tail section of the blade. The web calculator is found at

http://mlmsolutions.biz/threepointcircle.php

The rocker measured 7.5'

Advertised is 8'

Can't wait to do another one! ( )

[Query]

Actually, .019" is pretty easy to see, if you have good vision, or reading glasses - e.g., I have a quite readable ruler marked in 1/64" = .015625" increments.

I did not say to cut along the curves. I cut in between, in the white space, then place the blade directly on the paper around a given curve radius, directly next to the curve. Quite easy to see differences that large. You can also see small variations in curvature that appear on some blades.

Though I should note that the back .5 - 1" of most MK and Wilson blades is a bit rounded off, at a different rocker radius, which could mess up measurements.

AgnesNitt, if the question

Where'd you get that?

Was that meant for me? You can probably draw constant radius curves with any good graphics design program. I wrote a very junky short program to draw what I wanted in IDL (you can get a free trial version, which I used), output in postscript, then used PaintShop Pro to convert it to image form.

The problem with the gauges I've seen (including one that used to be available at http://iceskateology.com) is that they don't measure the rocker radius along the edge, nor along the center-line of the blade. They only use 3 points to measure the radius, and offset a distance from the edge. That means that if the offset isn't consistent for those 3 points (e.g., if there is a scratch on the side of the blade, or if it is parabolic, tapered, or any other form of side-honed), you won't get the right radius. You might not even get the same radius if you place the gauge on one side or the other of the blade. There might be fancier gauges that do something resembling a quadrature (least squares) fit, or that are properly designed to track all the measurements into the center - or one could write a fancy program that estimates the radius of curvature of the edge of a scanned image of the blade, and analyzes it, but I haven't seen such.

On the other hand, the gauges are way faster to use. It's a lot of fun to run one along a nominally new, factory sharpened blade, and see just how much rocker variation there is along the length of the blade. In fact, Paramount has a video of doing exactly that on another company's blades.

[AgnesNitt]

Bill YOU ARE DA MAN!



So is that the glide rocker you're reporting? What's the spin rocker?

[tstop4me]

Fantastic job, Bill.  Can't beat proper measurement technique.

[Bill_S]

So is that the glide rocker you're reporting? What's the spin rocker?

Spin rocker is 17.67 inches between the 8" and 10" (from rear) location.

The blade ends at the root of the toe picks at the 10.9" mark, so this starts almost an inch back from the picks. I just picked the location based upon my own spinning technique. It arcs upward fairly quickly to the toe picks beyond that, and we never spin there.

Click to enlarge...

[tstop4me]

Hey, Bill, if you've got the time, it would be great if you could do a sliding window 3-point fit and map the rocker radius as a function of position along the main rocker.  Say, even at 4 positions would be useful.

[Bill_S]

Had a little time before grading the final exams. This is a mental health break. Any teacher will recognize this avoidance tactic. 

Here's what I found using a 2" "sliding window" along the blade. Measured from the tail of the blade...

2"-    7.58' rocker

3"-    7.59'

4"-    7.58'

5"-    7.25'

6"-    9.25' (the blade flattens out a little here)

7"-    9.81' (still a little flat here too)

8"-   3.08'  (rocker shortening toward toe pick as expected)

9"-   1.48'  (17.67  inches)

I didn't have the data here at work for the rest of it, but by now we are well into the spin rocker and closing in on the toe picks.

Recognize that with a smaller window comes less measurement precision. However I believe the flatness evident around 6-7 inches could be real. If that's the case, it appears that the transition between glide rocker and spin rocker is not a perfect tangent between the two arcs. It was probably put there when the blade was factory sharpened. Of course the next sharpening will change everything again anyway, so I wouldn't read too much into it.

[tstop4me]

Thanks, Bill.  Exactly the data we need.  Several skate techs I know all claim that the newer blade manufacturers produce more consistent product than Wilson and MK.  Ostensibly because the newer manufacturers use more CNC, while Wilson and MK still use a lot of manual labor.  But I've never seen any data supporting this.

[AgnesNitt]

Bill, IM me w/ a mailing address and I'll send you 3 sets of blades coro ace, mirage, MK pro (Plus return fees) and you can test them.

I'm buying a dial micrometer to see if I can duplicate this set up on a circular saw at the base woodshop, but I won't be able to do that for weeks.

I wish there was a way to make this portable, a smart skate tech could use it to show skaters they need new blades .

[Query]

Bill, IM me w/ a mailing address and I'll send you 3 sets of blades coro ace, mirage, MK pro (Plus return fees) and you can test them.

I'm buying a dial micrometer to see if I can duplicate this set up on a circular saw at the base woodshop, but I won't be able to do that for weeks.

I wish there was a way to make this portable, a smart skate tech could use it to show skaters they need new blades .

Dial rocker radius gauges are quite portable. But on many blades, the very uneven rocker radii you see may be awful even on new blades.

So is this "wellness gauge":

http://iceskateology.com/Skateology/Blade_Wellness_Gauge.html

It measures the distance from the back toe pick tooth to the point whose tangent goes through the tip of that tooth. A ruler can do the same thing.

But I think it tells you that most Ultima blades are bad when new, because of the different spin rocker shape. People simply don't agree what the shape blades should be.

[tstop4me]

Bill, IM me w/ a mailing address and I'll send you 3 sets of blades coro ace, mirage, MK pro (Plus return fees) and you can test them.

I'm buying a dial micrometer to see if I can duplicate this set up on a circular saw at the base woodshop, but I won't be able to do that for weeks.

I wish there was a way to make this portable, a smart skate tech could use it to show skaters they need new blades .

If you have access to a machine shop instead of a woodshop, I'd go there first.  Dial indicator gauges are common in machine shops that do precision work, and if there is a milling machine with a carriage having a long enough travel to accommodate the length of your blades, that would provide an easier setup.  By the way, if you plan to do a comprehensive survey of a lot of blades, there are digital gauges that output the readings to a computer.  And if you are buds with a skate tech who has a spare blade holder (for sharpeners) that he's willing to lend you, you'll have even less setup work to do.

[Query]

Blade smoothness and skate rocker radius gauges"

IceSkateology's rocker radius gauge measures curvature from 3 points, with a total spacing of 1 3/8". That means the difference in post height of the center post to the outer posts is only about 9 microns different between 7 foot radius and 8 foot radius.

Derivation: If R is the radius in inches, and d=1.375 inches is the distance between the first and 3rd posts, the Pythagorean theorem shows that the middle post height should differ from the outer post heights by R-sqrt(R^2-(d/2)^2).
For 7' rocker, we get 7*12 - sqrt((7*12)^2-(1.375/2)^2) = .00281347718 inches post height difference.
For 8' rocker, we get 8*12 - sqrt((8*12)^2-(1.375/2)^2) = .00246178287 inches post height difference.
The difference between those two values is .000351694313 inches = 8.93 microns.

Does anyone believe that the blades are actually smooth to that size? For an 80 grit wheel, if I found this right, the grit size is around 200 microns, though grit size isn't quite the same thing as surface roughness.

That's why I like the idea that a gauge should ideally do a quadrature (least square) fit, rather than a 3 point fit.

I planned to estimate radius from a scanned image of the blade, using quadrature fits - but I'm not sure that scanners are accurate to that level either.

Maybe surface roughness is part of why radius gauges show so much variation on skate blades. I know the edge definitely isn't that smooth - under a strong magnifying glass, you can often see edge roughness. I'm not sure how to judge sharpened hollow center line roughness, which is what you probably want the gauge to measure.

A legit question is whether 9 microns or so of surface roughness affects blade performance much. Maybe it doesn't matter, and we should ignore detailed radius gauge measurements? As well as blade companies that try to claim other company's blades are awful because rocker radii vary - maybe they just initially polish the hollows of their blades a bit better, which won't last.

Of course, 9 microns might affect blade performance. Diffraction measurements suggest that the lubricating water layer on ice at skating rink temperatures is only about .04 - .05 microns thick. Only a precise empirical test that would be hard to do could answer performance questions.

BTW, I tried to do a cheap version of Bill_S's measurement. I taped a blade to a ruler, and used the depth gauge of a calipers to measure the distance of the edge to the ruler at various points along the ruler. Alas, tape isn't very stable - I got far too much inconsistency in measurement, even for one point. (I also tried use a micrometer, but the variation due to tape movement was still too large.) Sigh. This is harder to do than it seemed.