.png)
Tweet
I am enjoying this thread, now if I could see the top of my bench to lay some tools out, I would try some of your combined wisdoms. I believe the first test for a used wrist pin should be, to see if it is still round and then not hour glass shaped from wear. So I will be thinking about this for a while as I am to busy to clean up now. Thanks for the input, Jay
-
"Just build it and be done" -
A few things to think about
OK.
All the previous posts seem to assume/presume that the wrist pins are new, in pristine condition, are round and straight and neither tapered nor worn.
They also seem to assume that the ends are flat and that if tapered that only the outer diameter is touching the test table. If the taper is such that the "high" part is near or at the centre (ie near the central hole) the "square" will be more likely to "rock" (on its end/s).
I would put those wrist pins on a good vee-block and put a good dial indicator on the wrist pin.
First, to check for "round", I'd spin the pin and see that there is no deviation of the dial indicator at several points along its length. The next test would be for straightness. I'd move the pin along the stationary vee-block and under the stationary indicator (without rotating the pin in the process). I'd do that at least at 4 points/lines 90 degree apart and check that there was no deviation on the indicator.
Now, no matter what method is used this far, the main problem is to get the ends machined such that they are "square" to the cylinder face and axis, and preferably with only a narrow (say 1/8") "land" on the outer edges of the end faces with the "relief" to minimise picking up and grit etc. Whether or not this machining is done on a lathe or surface or cylindrical grinder is of no consequence so long as the required outcome is required.
The best or optimum result is achieved with a maximum "end diameter" to "length of cylinder" ratio and with the minimum practical width of the "land" on the end face/s.
I would never use a machine slide to check for this order of accuracy unless I was assured that the machines slides were very accurately "squared" to each other. Further, even on a perfectly squared machine, the clamping can have quite an effect which may cause an indicator to show an error on the "pin" when in fact it is in the machine. (This applies to mills and lathes).
This is the reason I bought my Machinist's Frame Level which is accurate (well the certificate says it is and so far I've had no reason to doubt it) to 0.02mm/meter = 1:50,000 for flat and square.
When I put my "new" (cylindrical) squares on both the mill table - and as a check on the magnetic chuck - on my surface grinder - and put a "Class 1" machinists square on them and there was no discernible change in the (very narrow) "gap"/"line of light" with a strong lamp in the back-ground. The same result applied when I put any two "squares" against each other, then rotated first one and the the other 90 degree against each other and then "end for ended" each in turn and tried one against the other. Same result - every time.
The human eye has an extra-ordinary ability to see very narrow gaps using light and even better to see any change in light "width" variation which may indicate taper or lack of straightness. This is in the order of a "tenth" or so.
Some may recall my thread on these tubular squares that I made and used to "true-up" my angle plates. After milling those angle plates I placed my "Class 1" test square on them as well as my (tubular) machine squares and there was no discernible variation of gaps. As a check of an against my Machinists Frame Square, I repeated the tests using it and it was as good as I could wish for:
Comment
-
Tiffie, I didn't make up the method I described -- it's textbook (Moltrecht).Originally posted by oldtiffie
The machine ways don't come into play because they're not moving while you're measuring the cylinder. And the spindle runout doesn't matter because you're measuring on opposite sides, and adding the errors.
But if you're really worried about it, then use Forrest's spin on the technique, where you make the same measurements with the DTI on a surface gage on a surface plate.
In either case, you'll get much more accurate values than eyeballing it. More specifically, you get discrete values for the cylinder's taper and lean. Once you know those values, you can subtract out the error when you're using the cylinder, for super-precise work (like machining an angle plate)..."Twenty years from now you will be more disappointed by the things that you didn't do than by the ones you did."Comment
-
A couple more comments
A couple more comments are in order here.
First of all, it is not best practice to be moving the indicator/surface guage etc. relative to the work under test as errors can creep in. It is best to move the work under or relative to the indicator and any aids or fixtures, tools etc. that may be used in the process.
If the base of the surface guage is (very) flat that will be better. But an error in either or both the base of the surface guage and the surface plate will be magnified at the highest point of the indicator on the surface guage mast and will be (falsely) indicated (as) less at the lower point of measurement - the relevant ratio is the length of the base of the surface guage in contact with the surface plate to the height of the indicator above the surface plate. If there are any errors - no matter how small - as regards "flatness" of the surface plate between any two or more positions of the surface guage on the surface plate it will be manifested at the indicator position/s and will or may give a "false positive/negative".
"Tapers" as such should be pretty well a non-issue in most cases if the cylinder is sufficiently round, parallel and straight.
It is quite possible and probable that even if the two ends of the cylinder are "flat" that they may not be parallel to each other in which case different "readings/results" may apply to each end of the "square" and must be marked and recorded as such unless only one end is designated for and used as a "square" reference/tool for further work.
My objective was to achieve "squares" that are accurate to my requirements such that they can be used "as is" without further "adjustment" or "allowance/s" (for) as a (pre)condition for use.
I deliberately used methods that would compound my errors with the intent of finding and identifying those errors. I am satisfied with the outcomes. Others may not be as is their indisputable and undisputed and respected right.
I am not much fussed either by "Co-Sine errors" as an indicator is just that - an indicator - and for accurate indicating where a specific quantity is required, should be used in conjunction with a "reference" such as a known size or "slip guages" etc. Ideally the axis of the indicator arm should be parallel to the (in this case) axis of the cylindrical "square" under test.
The potential or real "Co-Sine errors" need to be put into perspective. I suspect that there is a lot of "(over??) hype" with this as while it is certainly applicable in some circumstances, it may not be applicable in all of them. I seem to hear it as a "mantra" when it may not need to be (perhaps its just my hearing).
It is the difference between the "base" and the "opposite side" in triangular geometry and trigonometry. Some examples for "unit" ie "1" length of "base" with respect to opposite side/s" are:
- 5 arc degree: 0.9962 ~ 3.8 per 1,000;
- 10 arc degree: 0.9848 ~ 15.2 per 1,000; and
- 15 arc degree: 0.9659 ~ 34.1 per 1,000
I accept that it will be of concern to others and that they may act as they like - as can I.
As I've said often enough, I have no dispute with "Reference Texts" to the extent that what they say is correct. But I am not going to take them as "Gospel" or as "the one true Bible" etc. I want to know not just what they say but the logic and end-results required of and achievable by them.
That there are often so many held in such high esteem but use different methods to achieve generally similar end-results supports my opinion. I do not like "prescriptive" methods which lean toward the "You must do it this/my way - there is no other" way of "leaning" on the reader.
I have no use for "Machining"-type/related "Evangelists", "Hot Gospelers" or "Catechisms" or the like in my shop - or in my life either.
I have absolutely no problem with anyone (me included at times) using any or all of those prescribed methods conditional on my being able to do it differently to get a result that suits me.Comment
-
LOL Tiffie! That the whole point of this thread! The OP found some wrist pins, and he's asking how to measure their accuracy.Originally posted by oldtiffie
The measurement techniques that Paul, Forrest and I posted are ways to measure whether the cylinder is round and straight -- also known as the Taper and Lean
Then get back in the shop and Get 'R Done MickI have no use for "Machining"-type/related "Evangelists", "Hot Gospelers" or "Catechisms" or the like in my shop - or in my life either.
"Twenty years from now you will be more disappointed by the things that you didn't do than by the ones you did."Comment
-
Squaring up
Well, lazlo as I am such a nice chap, I will give you some "wriggle room".Originally posted by oldtiffieOriginally posted by lazloLOL Tiffie! That the whole point of this thread! The OP found some wrist pins, and he's asking how to measure their accuracy.
The measurement techniques that Paul, Forrest and I posted are ways to measure whether the cylinder is round and straight -- also known as the Taper and Lean
Then get back in the shop and Get 'R Done MickOriginally posted by oldtiffie
Now having been back to my shop as ordered for Penance and Enlightenment (but no Absolution), and "got 'er done" and I return bringing news and glad tidings - and more "info".
It seemed to me that all of you just jumped in assuming the wrist-pins were accurately round, straight and parallel - or perhaps you forgot to mention it.
I checked it first.
Just shown me what I missed and why my method doesn't work as I think it does and within the resources available to me in what is an average HSM shop - it works.
I checked every square against another to ensure that I compounded any errors to make sure that I found them (which I didn't) so that I could correct them if needs be (which I didn't have to).
I thought I'd chase up some more info on this matter of "Master Squares" and checking try-squares.
I know you will be interested and grateful - so here goes.
First of all, I will deal with the testing of a try-square.
There are are two methods which I learned both in the Main Work-shops and the Tool-Room, the first of which is a self-centering method using Tool-makers "buttons" - which is what I thought Forrest was talking about:
which in turn discusses the second and possibly more precise method using a "Master Square" (which might look familiar):
Now, I hope that I have dealt with that comprehensively enough - once and for all.
Next, as the Original Poster (OP) referred to squares and possibly at least some of the methods of making/milling them, I thought I'd cover that as well so that the OP had his questions answered:
Comment
-
No.Originally posted by oldtiffie
Jay (TipTop) posted that he found some wrist pins, and asked how he can measure if they were round and straight. Paul, Forrest and I showed him how to do that.
Case closed.
Not really Mick. We answered Jay's question, and you're going off on tangents about try-squares and toolmaker's buttons...Now, I hope that I have dealt with that comprehensively enough - once and for all.
"Twenty years from now you will be more disappointed by the things that you didn't do than by the ones you did."Comment
-
I dunno. I think Tiffie got there in the end. The very last page of his link references show exactly what Forrest was hinting at in the beginning. You use an indicator and front edge of the base to zero the indictor, then rotate the wrist pin to see any out of square error (double error).
Rather like the Texas rancher trying to impress a visitor with the size of his spread. "If I drive out the gate and along the fenceline all around my ranch, it takes me two days to get back." "Yeah," the visitor said, "I know what you mean. I had an old pickup like that once myself.".
"People will occasionally stumble over the truth, but most of the time they will pick themselves up and carry on" : Winston ChurchillComment
-
You're right Jan -- I didn't bother clicking through all his linksi, but Tiffie posted the entire "Be Square" chapter from James Harvey's "Machine Shop Trade Secrets." I don't know if he realizes that on the last page Jim Harvey describes the exact procedure that I outlined on page 2 of this thread.Originally posted by TGTool
The method I described (from Karl Moltrecht's Machine Shop Practice) adds a third measurement 90° from the first two, so you can calculate lean. The first two measurements, 180° apart, as described by myself, Paul, Forrest and Harvey's book, only give you taper:
Originally posted by lazlo
So using a ordinary surface gage, like Forrest described, or a special-purpose surface gage like Jim Harvey shows, is basically a dedicated fixture to do the 2- or 3 point taper and lean measurements. Like I posted earlier in the thread, you can do the same measurement in the mill or lathe, because the measurements average out the runount of the spindle bearings, and the slide-ways aren't moving.
LOL! Quite aproposRather like the Texas rancher trying to impress a visitor with the size of his spread.
Last edited by lazlo; 06-16-2008, 12:05 PM."Twenty years from now you will be more disappointed by the things that you didn't do than by the ones you did."Comment
-
I'm gettin there?
So, I believe I am starting to understand how, I may (more accurately) check these used wrist pins for taper, concave, or convex and being square. If a person were to take a surface gauge as shown in oldtif's last link, but it had a "V" on the face that contacted the "cylindrical square" and then proceeded to do the tests that have been mentioned by several of you in different ways, he would then have accurate numbers to show the "case" as it may be. I think I have the tools and knowledge to perform these tests. Now I guess I should do it before class, as I am sure the teacher will want to grade my work. Boy did I get myself in a pickle this time, maybe I ought to keep my mouth shut when running into what I thought was just a good score. Oh well. Thank you everyone, for all the input, I really am enjoying the banter and learning at the same time. Of course if it looks as though my take on the information is not quite proper or true I expect I will here about it. Jay"Just build it and be done"Comment
-
Usual precautions apply. That is, "Never ask a cowgirl the size of her spread."Originally posted by lazloLOL! Quite apropos.
"People will occasionally stumble over the truth, but most of the time they will pick themselves up and carry on" : Winston ChurchillComment
-
You don't need the special-purpose surface gage from Jim Harvey's book (that's in Tiff's last link). The general idea is that you run a DTI from the bottom to the top of the cylinder on opposite sides (and between the two, if you want to measure lean).Originally posted by tiptop
So you can either set the cylinder up under a mill spindle, and spin the DTI around with a Zero-Set type indicator holder (like you do when you tram the head), or you can set it up on a lathe faceplate like Paul described, or you can use the crotch of a traditional surface gage, as Forrest described.
By the way, Forrest's method is one of the old machinist's trivia questions: what's the mirror for in a Gerstener chest, and what's the ball for on the end of a surface gage scriber, ...
"Twenty years from now you will be more disappointed by the things that you didn't do than by the ones you did."Comment
-
Getting it square
First of all lazlo, you are wrong.Originally posted by lazloYou don't need the special-purpose surface gage from Jim Harvey's book (that's in Tiff's last link). The general idea is that you run a DTI from the bottom to the top of the cylinder on opposite sides (and between the two, if you want to measure lean).
So you can either set the cylinder up under a mill spindle, and spin the DTI around with a Zero-Set type indicator holder (like you do when you tram the head), or you can set it up on a lathe faceplate like Paul described, or you can use the crotch of a traditional surface gage, as Forrest described.
By the way, Forrest's method is one of the old machinist's trivia questions: what's the mirror for in a Gerstener chest, and what's the ball for on the end of a surface gage scriber, ...
Your method that you espoused at page 2 post #13
.png)
You used a machine slide - whether it be a quill feeding down or a knee feeding up where-as Harvey used a known static square with the the indicator static on the mast and at the same height from the reference base. This method certainly found the "out of square" which manifested itself as a "tilt" indicated by changes in the indicator. It did not and could not have detected any "out of round" or out of straightness in any item similar to the OP's wrist pin/s.
Forrest's method was similar in many respects, but neither he, Harvey nor I moved the indicator up/down - it stayed at the same height through-out.
Using the "Master Square" would have faired no better in that regard.
Your method at post #13 was/is:
First of all you did not say the mill was correctly trammed and if there was a tram error what its direction and magnitude was. Any such tram error if not accounted for would indicate this error as a "false positive/negative" on an otherwise perfect/acceptable "square" (wrist pin?).Originally posted by lazlo
Second, you only sought the error in one axis (either "X" or "Y"). It is quite possible that you got no indicated error where you indicated but that the error was in the other ("X" or "Y" axis as the case may be) - or anywhere in between and you would have missed it. Further, I'd have carried out that procedure at 45 degree intervals - or finer.
Third, you used a machine slide/quill to run the indicator up and down. If the slide or quill was not a very good fit in very good condition there may have been "lateral float" that might also not be detected but will show up as (another) "false positive/negative".
Fourth, you method is not the same as Harvey's or Forrest's either at Post #14 as he did not move the indicator up/down either - only you did that out of all of us.
http://bbs.homeshopmachinist.net/sho...0&postcount=14.
Fifth, the "master square" in the other reference, is as said a perfect square if it is accurately round, straight and with machined ends (lathe or "side-wheeling" on a cylindrical grinder).
Sixth, my methods do not resemble Harvey's at all.
Seventh, I said that I would not use machine slides.
Eighth, the Tool-makers Buttons while perhaps OT as far as the Original Post(er) (OP) is concerned were right on the mark in response to Paul Alciatore at Post #11
Paul's proposition was very close indeed to test the try-square (as suggested by Forrest) as the solution "using 2 lines" was very close to the "Tool-makers Buttons" link at:Originally posted by Paul Alciatore
Ninth, my posts links to Harvey's methods of making a square in a mill are very relevant to the subject of machine/d squares and as an alternative way of making a quite useful and serviceable "machine square".
Paul A at Post# 12 at:
suggested using a lathe face-plate. I would agree with that if the face-plate was known to be very flat and neither concave not convex as the "out of flat" will or may introduce an error or (((180) - (cone included angle))/2) which will be a contributing error ("false positive/negative").
I have explained this previously but the sketch on this link should make it obvious.
Comment
-
Machine Shop Trade Secrets:Originally posted by oldtiffie
"set the indicator to "zero" with the indicator base up against the block you're inspecting. Then check the opposite side of the block and compare the readings".
Lazlo:
- Preload the DTI, and wind the spindle/DTI from the bottom to the top. Measure the difference.
- Without touching the cylindrical square, spin the indicator holder to the opposite site (180° from where you were), find the high spot again and wind the spindle/DTI from the bottom to the top. Measure the difference.
I guess Machine Shop Trade Secrets, and Karl Moltrecht are wrong too
No:Originally posted by OldTiffie
Jim Harvey and Forrest's method, as described, only checks the base! In order to measure the taper and lean of a cylindrical square, you have to run the DTI from the base up the cylinder.Originally posted by Paul Alciatore
Machine Shop Trade Secrets:
"set the indicator to "zero" with the indicator base up against the block you're inspecting. Then check the opposite side of the block and compare the readings. If the arrow does not return to zero, then your block is out of square by half the amount of that reading at that height up the block".
The quill error doesn't matter because you're taking measurements on opposite sides and adding them (same reason Rollie Dad's Method doesn't need a straight bar to measure spindle alignment). So if the quill sticks out a thou along the X-axis, the first measurement will be: (Cylinder Taper + 0.001), when you measure the cylinder on the opposite side the second measurement will be (Cylinder Taper - 0.001). Add the two measurements and divide by 2, and the quill error is gone.Third, you used a machine slide/quill to run the indicator up and down. If the slide or quill was not a very good fit in very good condition there may have been "lateral float" that might also not be detected but will show up as (another) "false positive/negative".
Then why did you post Harvey's method?Sixth, my methods do not resemble Harvey's at all.
Seriously Mick, you do this all the time -- someone asks a question, a bunch of people answer it, and then you jump in 20 posts later with a very long-winded post that wanders all over the place, and is rife with unrelated links. I know you're trying to help, but you end up hijacking the thread...
So now TipTop (Jay) is probably more confused than if he had never posted the question. Great.
Last edited by lazlo; 06-17-2008, 10:16 AM."Twenty years from now you will be more disappointed by the things that you didn't do than by the ones you did."Comment
-
Tedium
Lazlo.
I will respond - again - to your post #29 at:
This is getting tedious. I don't see how I can reduce it to two-syllable words nor do I want to draw too many word pictures or any (more) graphical ones.
First of all, I don't have or have access to Karl Moltrecht's books (as I suspect that neither do most others). I made a point of scanning and posting any reference that I used - Harvey et al.Originally posted by lazloMachine Shop Trade Secrets:
"set the indicator to "zero" with the indicator base up against the block you're inspecting. Then check the opposite side of the block and compare the readings".
Lazlo:
- Preload the DTI, and wind the spindle/DTI from the bottom to the top. Measure the difference.
- Without touching the cylindrical square, spin the indicator holder to the opposite site (180° from where you were), find the high spot again and wind the spindle/DTI from the bottom to the top. Measure the difference.
I guess Machine Shop Trade Secrets, and Karl Moltrecht are wrong too
Moltrecht (and you) were the only ones who moved the indicator vertically in a machine "slide". I did not, nor did Forrest nor Harvey. Forrest and Harvey both referenced the bottom of the known square and adjusted a dial indicator at the top - and left it there.
As you quote and by inference agree with both Forrest and Harvey and that you differ from them in that you moved your indicator vertically over the test piece - and they did not - then it seems that you are wrong - not they/them.
Nope - wrong again. In both the case/s of Forrest and Harvey they both used known/proven squares, set their bases (in Forrest's case, the ball of the mast of his surface guage) to the base of the reference square and zeroed the dial indicator. This is what Forrest said and meant when he said that a guage had been set. The "guage"-line is in fact between the "touch point" of both Forrest's "ball" and Harveys round base and the "zero" on the dial indicator. Both - in all subsequent positions after setting the reference guage line - used the base as a reference in any other position and measured any variation from the vertical/reference/guage line as a deviation on the dial indicator - without moving the dial indicator at all in the vertical plane as you recommended and quoted Moltrecht as a reference source.Jim Harvey and Forrest's method, as described, only checks the base! In order to measure the taper and lean of a cylindrical square, you have to run the DTI from the base up the cylinder.
Forrest and Harvey did - and only did - as per your quote:
Further, the cylindrical "Master Square", if as I said previously, is sufficiently accurately round, straight and parallel and the ends "squared/faced-off" on a lathe or grinder, it is a very accurate and "natural" square that can be made with adequate accuracy by most HSM-ers without any additional tooling (reference square blocks - 1-2-3 and 4-5-6 blocks etc.) or expense and can be repaired or replicated very easily.Machine Shop Trade Secrets:
"set the indicator to "zero" with the indicator base up against the block you're inspecting. Then check the opposite side of the block and compare the readings. If the arrow does not return to zero, then your block is out of square by half the amount of that reading at that height up the block".
That is wrong as well to the extent that any errors in "tram" or in the mechanical alignments and "fit/s" of and in the machine mechanisms - if not identified - will manifest themselves as "false positives/negatives" and show either as errors which do not exist in the item under test or are additive (+ve or -ve) and superimposed on the real state of the test piece. If I could not identify whether any such error existed I would assume that they did unless or until I could prove otherwise.The quill error doesn't matter because you're taking measurements on opposite sides and adding them (same reason Rollie Dad's Method doesn't need a straight bar to measure spindle alignment). So if the quill sticks out a thou along the X-axis, the first measurement will be: (Cylinder Taper + 0.001), when you measure the cylinder on the opposite side the second measurement will be (Cylinder Taper - 0.001). Add the two measurements and divide by 2, and the quill error is gone.
Given the status that Moltrecht has, I would assume that he would have been using a machine of the proven quality of a Jig-Borer or a Metrology class/level machine that was verified/certified and still "In test" and used under appropriately controlled conditions. Anything less would have to be "suspect" at least.
Forrest's and Harvey's methods quite elegantly virtually eliminates that variable and should "stand up" in just about any environment.
Now I will deal with your comment on Paul's method in the lathe.
I dealt with that in terms of "flatness" (and implicit and inferred "square to the spindle"). "V"-ways are by their nature, self-aligning with minimum or no practical errors if straight and not worn laterally with respect to the lathe spindle. They they are so accurate and reliable is indicated - and proven - by their known accuracy in lathes and grinders - and do not need or use gibs or any other adjustments. The ball-races in my surface grinder are a variation and a case in point in that there are two "V"-ways with balls between them - a "straight ball-race". So, I'd accept the V-way set-up in preference to any or most others.Originally Posted by Paul Alciatore
Place the faceplate on the lathe. Place the square against the faceplate. Run the DI along the edge of the square using the longitudinal feed on the lathe.
I suspect that your use of the example of Paul's was a "red herring" as it had nothing to do with Forrest or Harvey's methods. But having said that, I'd support the reliability of the accuracy of Paul's movement a V-way above just about any other on balance - for the reasons given.
Perhaps so - I plead guilty to that. But I must say that you are not innocent in that regard either - a case of "the pot calling the kettle black" perhaps?Seriously Mick, you do this all the time -- someone asks a question, a bunch of people answer it, and then you jump in 20 posts later with a very long-winded post that wanders all over the place, and is rife with unrelated links. I know you're trying to help, but you end up hijacking the thread...
Now, I would not (want to) be so unkind as to think or suggest this was a sort of covert "strong-arm" or intimidation attempt. But in the off chance that it is - it is very overt - so forget it - you've got no chance...............I know you're trying to help, but you end up hijacking the thread..
Perhaps so, but TipTop is quite able to speak for himself without anyone else assuming his role and doing it for him - as it the case for anyone and everyone on the this forum.So now TipTop (Jay) is probably more confused than if he had never posted the question. Great.
We are all equal here.Comment
Comment