Thread: Gage Block Sstack Algorithm

09252021, 08:08 PM #1
Gage Block Stack Algorithm
reference;
is there an algorithmic description of stacking gage blocks?
Have read the above thread and was curious as to whether anyone came up with a algorithm or program to find the required gage blocks for setting a 5" sine bar for instance. Excel maybe ?
Secondly, I have a really nice 34 piece ScherrTumico set and need to know if this set will have the correct blocks for doing this ?
Thanks,
gatzLast edited by gatz; 09252021 at 08:57 PM. Reason: mispelling in title


09252021, 09:57 PM #2
CNC Machinist Calculator app will do that. Just enter the angle and the length of bar and it will spit out the stack height. ~$11 when I bought it years ago. Use it almost daily at work.
Sent from my SMN960U using Tapatalk

09252021, 11:41 PM #3
cb750chris (used to ride a cl360 myself a long time ago), I think OP is asking about the procedure which Leigh described in post #2 in the linked thread. The question is not "how tall a stack do I need for an angle", but rather "which blocks out of my set of 84 do I choose to make up the desired stack".

09252021, 11:54 PM #4
Nice! my poor cb has been neglected for a few years now in need of new top end seals. It spews oil from the hockey pucks in the head.
I guess I should have read the linked post. I wouldn't be able to describe that well without the person standing next to me and just showing them how I do it.
Sent from my SMN960U using Tapatalk

09262021, 11:49 AM #5
Looking for a program that will calculate as described here.
How to Select Gage Blocks for Use in Combinations
Google came up with this..
How to Select Gage Blocks for Use in Combinations online calculator
Gage Block Calculator  iPhone iPad Apps! Appsuke!


09262021, 12:55 PM #6

SIP6A liked this post

09262021, 01:00 PM #7
I have an Excel tab I use, but it's nothing more than subtraction for each decade value. I tried, and failed, to write a simple routine that automatically chooses the block sizes because the sets are different. A note from my tab says, "There may be several combinations that will get you there. A bit of trial and error will be needed such that subsequent choices exist in the set you're using. An example is that there are no blocks in the typical set between 0.05 and 0.1 inches, making direct elimination of those hundredths digits impossible."

09272021, 01:38 PM #8
This is actually a very complex problem to solve, too complicated for excel I would think.
As Conrad mentioned there are several logic problems that need to be evaluated.
It will vary based on the number of blocks available and the desired resolution/tolerance, best practice would be for the user to supply a list/txt/csv file containing lines for each block size they have on hand. If the blocks have an inspection report you could also solve for stackup error and wring film.
for an 81 Piece set the program might look something like...
Assume Target Value = (X)
Get desired precision from the user (nearest 0.001" or o.0001" ?)
1: (X) less than or equal to 0.100 Must fall directly on a gage block size.
2: (X) between 0.101 and 0.150 Can only have a resolution of 0.001"
3: If desired resolution is 0.0001 and (X) is between 0.101 and 0.201, Rule #1 +0.100x
applies
4: If the desired resolution is 0.0001 and (X) contains an integer ((X) > 1.0) and the
decimal value is less than X.201, we know were are going to want to use either the
0.850/0.900/0.950 + the correct 0.100x block + a block equal to the remaining value.
5: Now we can Iteratively solve for whatever value might remain. first solve for the
decimal value and then solve for the integer value.
While evaluating these things, you also need to keep track of each block used so as to not select a block that has already been used. step number 4 / 5 might take several iterations to get right, as its possible to run out of integer blocks. This requires you to to build up a value equal to the desired Integer from several smaller blocks. If this isn't possible  the program has to "roll back" to step number 4, and retry with a different combo of blocks until it gets it right.
Note that this isn't meant to be a rant, I thought this was kind of and interesting problem I may try to tackle at some point, perhaps a desktop type app that could sell for a few bucks. I think this is the way to solve it, though its tough to say until you really get into it.

BrandonCarriveau liked this post

09272021, 01:54 PM #9
If you can't make up a stack of gage blocks doing the ARITHMETIC in your head you don't have any business working in a machine shop.
Last edited by SIP6A; 09272021 at 05:51 PM.


09272021, 06:58 PM #10
Tony and SIP6A,
The ARITHMETIC is not an issue for me to make a finite stack of available gage blocks.
I've done this many times using the same method as you are referring to and as seen on YouTube. I have been in the machinetool business since the early '70's
Our general machine shop had many customers with varying jobs entailing much math. This was in the days before CAD or CAM. At first I had a small calculator with TRIG on it to assist in solving some of the stickier problems. Later, I got a TI 30 and thought I had the world whipped. I recall solving a TRIG problem where 3 arcs had to be tangent.
The coordinates were then programmed into Gcode for our CNC mill to cut the shape.
The mill was a bit lacking in the way it handled cutter comp, so that was another set of values to have to calculate.
In fact, there was a salesperson demonstrating a new software that they claimed would produce DXF files given the 3 arcs. So I challenged them to solve the problem using new arc values. I got done well before they did with the help of the TI 30.
Using "the gray matter between my ears" became a daily routine.
(Some days, the head just plain hurt)
Making a program or algorithm to solve it IS the challenge! Heck, I don't even have a Sine Bar anymore and seldom have to work in angles other than "standard" as in a angleblock set in my home shop.
My interest was mostly academic; with the hope that someone had done the task and could share the results.
The APP mentioned above may be a good deal at 99¢
It also allows one to pick which set of blocks are to be used.
I may just buy it to see how it works.
gatz

fusker liked this post

09272021, 08:13 PM #11
The APP mentioned above may be a good deal at 99¢
But one has to walk to the computer to use it.
In the good old days, we didn't have calculators or computers. I think I paid $70 for my first calculator..
Oh, but we did have a hand lever adding machine.

camscan liked this post

09272021, 08:35 PM #12

09272021, 08:43 PM #13
Don't underestimate the abacus. I recently read that Nikon's early lens designs were done by a computer a woman with an abacus that worked for the lens designer, doing ray tracing by hand. They did remarkably well.
Now I'll have to try throwing something together with Excel and maybe the Visual Basic that runs with it.

mhajicek liked this post

09272021, 08:59 PM #14
I think Excel can handle this raw. Even easier if allowed to use VBA.
Here is maybe more a challenge, write a Gcode macro to do it.
Set one variable to the set piece size (81,36,112,etc), another to the desired and run it outputting the stack needed and options.
No matter how programmed multiple answers to many final sizes but obviously less blocks win. If working to tenths what to do with the error of putting two or more block together?
If 8 blocks used do you comp for this interfacing and short the added number verses dimension? And by how much?
In the real world one should be able to look at the block set in front of you and do this in your head with no calculator or computer needed.
Bob

10052021, 04:48 PM #15
Block include the wring thickness in their size.
...if clean
...unworn
...properly wrung
If we're really picky there might be a situation where 1+2 blocks are closer to target than a 3 block following the error chart, but I'll bet in the real world getting rid of an interface is better than whatever the listed error is. I agree that this is easily done in one's head, but I have the benefit of an 87 piece (metric) set, so not many compromises need to be made.
If we really want to have fun we should debate the best way to determine 3 gauge pins to check a round hole larger than our largest pin. I've done that by brute forcing in MATLAB, but wouldn't mind a more efficient way.

10082021, 10:28 AM #16
I've setup a spreadsheet for myself that tells me what the block stack should be for the sine bar.
Input Angle/degrees required (decimal)
Input Sine Bar Length
You are provided with the gage block stack size needed and displays the conversion of degrees decimal to Deg, Min, Sec as reference.

10082021, 03:31 PM #17

10082021, 03:57 PM #18
Bookmarks