G
Gordon Clarke
Here goes with an attachment attempt. It's in
jpg as that should allow for different document formats.
jpg as that should allow for different document formats.
S
stefanhg
Hi,
If anyone needs information about dimensions of threads or thread gauges, I can help. Please write here or send a private message.
If anyone needs information about dimensions of threads or thread gauges, I can help. Please write here or send a private message.
Wayne
Gage Crib Worldwide
I am going to come off very critical here, so put on your thick skin.I think I understand where you are going here. Basically you sell a thread measurement system that uses balls attached to a standard caliper. This system is not accurate enough to measure to the full thread tolerance as specified in the codified standards. To help your customers feel comfortable with working to a rounded-off version of the official standards, and to remove from them the complicated task of rounding, you have re-created the thread data by applying a reduced/rounded tolerance. To make the data appear more official you have published charts displaying the thread data in its modified form, with a disclaimer attached.
I can see a problem with this. I sell thread gages for a living, and every day I have to educate customers about the requirements of the standards. I get thread specifications with mis-matched data that needs to be corrected. For example a customer may ask for a 1/4"-20 UNC-2B with pitch diameters of 0.414"/0.419". Both you and I know that according to the standard the 1/4"-20 UNC-2B class-of-fit has specified pitch diameters of 0.4143"/0.4189". What has happened here is some “engineer” pulled the pitch diameter data from some source other than the official standard, like your charts or he had his CAD software set to round to 3-places, and codified it on the drawing for this specific part.
The problems then begin. To what should the gage pitch diameters be made, the print specified class-of-fit or the print specified pitch diameters.
It is an oxymoron; it can not be both ways. The question originates with me, the gage seller because I am not going to sell a gage which is marked -2B if the pitch diameters do not concur with the ANSI standard, but the customer is always right; the print is the final determinant. My customer then has to go back to his customer, or his customer’s customer and explain that some “engineer” used modified data. The engineer gets all defensive because as an engineer he is never wrong; or it gets too expensive to process a print change; or the drafting department is too busy to make the change, and an exception is made. The reality is that the problem just does not go away, it escalates. The next time the customer changes suppliers and the new supplier questions the print it will be said that it has been this way for years, so why change it now; or a change is not going to happen because they want future parts to mate with existing product in the field; or when a third-party inspector picks up the gage marked 1/4-20 UNC with pitch diameters of 0.414"/0.419" and asks why the 2B has not been marked on the gage like the print demands.I can agree with your assertion that there is plenty tolerance in a thread, and using calipers and rounding, the thread tolerance can be reduced. The parts may mate with their intended component and may even pass inspection by a gage, assuming no errors in lead; pitch; flank angle; or minor diameter and no fat-thread or too severe drunken-lead. But, I have talked to the authors of the ANSI thread standards, and much discussion has ensued in seemingly endless meetings over which points in the math are rounded at 5-places or 6-places, just so results are consistent across the industry. If they could get their hands on you I am sure that they would wring your neck for the possible confusion you may cause with the publication of erroneous data. I know that your disclaimer is printed on your chart, but as is usually the case, who reads the fine print. I am sure that they would say that if data is published it should be most accurate, and deviations from the data to be made and justified only at the most local level.
I know that I am not always the most tactful and I hope I did not sound too harsh.
G
Gordon Clarke
Wayne,
Working where you do, and with the products you have, may make you slightly biased
Maybe even a wee bit one-eyed.
I hope you take this the same way as you hoped I’d take what you wrote and keep an open mind. I have become very thick skinned as I’ve heard your “arguments” more than once. Anyway, What would be the point of a discussion forum if all were in agreement? J
First I suggest you look at my profile information or send me a PM and I’ll give you details and specific information. This will help you understand where I’m coming from, as your “assumptions” are both right and wrong.
If you find rounding off dimensions a difficult task, then serious problems must be almost insurmountable obstacles. When I'm figuring out when and what to round up or down, I always take the tolerance size into consideration. I'm often going from inches to metric and vice versa.
As you write, 1/4"-20 UNC-2B has specified tolerances as per ANSI. Anything deviating from that is simply not to standard and must not be specified as such. If 1/4"-20 UNC requires a different pitch diameter than 2B then it must be specified and marked on the gauge and is certainly not 2B. I’ll bet these special tolerances add quite a bit to the standard gauge price. That’s the advantage of measuring pitch diameter – you know where you are. As I also remarked, if a thread gauge is used, then the pitch and thread profile should also be OK.
One company I know in Denmark, has for years made parts with thread OD in mm and pitches in 12 TPI. It lets them take an extra high price on spare parts! Now if this company had spare parts made in the USA, I’d guess that the thread would be made with a 60O flank angle. However it was actually drawn originally as a Whitworth thread (55O) but only specified as 95 x 12 TPI. As long as they made these parts themselves there wasn’t a problem, but when they started to send them to a sub-supplier you can imagine the problems. Some supplied with 60O and others with a 55O flank angle – and no one enquired as to what the tolerance was! If you tell me this couldn’t happen in the USA I won’t believe you J as I’ve been there several times. The company was/is ISO 9001 certified, so they had their gauges sent out to an authorized lab for calibration – apparently without problems! It was only when I was brought into the picture (due to massive customer complaints) that I suggested a solution – which they still use to this day. Write if you want me to tell you my solution.
If only gauges are used to inspect a thread then a possible inspection has no way of knowing how much of the tolerance the machine actually requires and how reliable the process is – probably one of the major factors in quality control and assurance. Most machine operators invariably make a thread too tight as they’d rather squeeze (force?) the thread gauge on or in the thread. How would they know where the middle of the tolerance was without measuring?
Basically I’m advocating to do a quick, safe measurement that operators and inspectors can use to make a qualified decision. I’ll leave “nano”-philosophy to the guys in the white coats and their cotton gloves in their temperature controlled environments.
Some will find what I’ve written approaching blasphemy but a coin has three sides – heads, tails and standing on edge.
Gordon
P.S. I honestly hope this provokes a discussion J As my dad said “I’m not always right, but I’m never wrong”. He did say it with a smile though.
Working where you do, and with the products you have, may make you slightly biased
Maybe even a wee bit one-eyed.I hope you take this the same way as you hoped I’d take what you wrote and keep an open mind. I have become very thick skinned as I’ve heard your “arguments” more than once. Anyway, What would be the point of a discussion forum if all were in agreement? J
First I suggest you look at my profile information or send me a PM and I’ll give you details and specific information. This will help you understand where I’m coming from, as your “assumptions” are both right and wrong.
If you find rounding off dimensions a difficult task, then serious problems must be almost insurmountable obstacles. When I'm figuring out when and what to round up or down, I always take the tolerance size into consideration. I'm often going from inches to metric and vice versa.
As you write, 1/4"-20 UNC-2B has specified tolerances as per ANSI. Anything deviating from that is simply not to standard and must not be specified as such. If 1/4"-20 UNC requires a different pitch diameter than 2B then it must be specified and marked on the gauge and is certainly not 2B. I’ll bet these special tolerances add quite a bit to the standard gauge price. That’s the advantage of measuring pitch diameter – you know where you are. As I also remarked, if a thread gauge is used, then the pitch and thread profile should also be OK.
One company I know in Denmark, has for years made parts with thread OD in mm and pitches in 12 TPI. It lets them take an extra high price on spare parts! Now if this company had spare parts made in the USA, I’d guess that the thread would be made with a 60O flank angle. However it was actually drawn originally as a Whitworth thread (55O) but only specified as 95 x 12 TPI. As long as they made these parts themselves there wasn’t a problem, but when they started to send them to a sub-supplier you can imagine the problems. Some supplied with 60O and others with a 55O flank angle – and no one enquired as to what the tolerance was! If you tell me this couldn’t happen in the USA I won’t believe you J as I’ve been there several times. The company was/is ISO 9001 certified, so they had their gauges sent out to an authorized lab for calibration – apparently without problems! It was only when I was brought into the picture (due to massive customer complaints) that I suggested a solution – which they still use to this day. Write if you want me to tell you my solution.
If only gauges are used to inspect a thread then a possible inspection has no way of knowing how much of the tolerance the machine actually requires and how reliable the process is – probably one of the major factors in quality control and assurance. Most machine operators invariably make a thread too tight as they’d rather squeeze (force?) the thread gauge on or in the thread. How would they know where the middle of the tolerance was without measuring?
Basically I’m advocating to do a quick, safe measurement that operators and inspectors can use to make a qualified decision. I’ll leave “nano”-philosophy to the guys in the white coats and their cotton gloves in their temperature controlled environments.
Some will find what I’ve written approaching blasphemy but a coin has three sides – heads, tails and standing on edge.
Gordon
P.S. I honestly hope this provokes a discussion J As my dad said “I’m not always right, but I’m never wrong”. He did say it with a smile though.
G
Gordon Clarke
Wayne,
I've just thought of another thing regarding when sizes should/can be rounded up or down. In the countries using the metric system (which is the majority) a standard 6 inch caliper is referred to as a 150 mm caliper.
You must admit it would be daft to call a 6 inch caliper a 152.4 mm caliper or a 150 mm caliper for a 5.9055 inch caliper.
To me it's a question of knowing when to use common sense.Not that I'm implying you don't have common sense of course. It's just an example.
We both know that a digital caliper is always a wee bit more than both 6 inches and 150 mm anyway.
I notice in ASME B1.1 a typical dimension is given as 0.003969" - when are dimensions to 6 decimals actually needed? I'm guessing it's only given that accurately so that it can be used for calculating purposes. Try going on to a shop floor and talk about dimensions and tolerances given to 6 decimal points. You'd get some very funny looks in most places
Gordon
P.S. I'm assuming you are as thick-skinned as you hoped I'd be. I love a good discussion. I have been known to change my opinion - but always by logic and common sense - plus of course good humour.
I've just thought of another thing regarding when sizes should/can be rounded up or down. In the countries using the metric system (which is the majority) a standard 6 inch caliper is referred to as a 150 mm caliper.
You must admit it would be daft to call a 6 inch caliper a 152.4 mm caliper or a 150 mm caliper for a 5.9055 inch caliper.
To me it's a question of knowing when to use common sense.Not that I'm implying you don't have common sense of course. It's just an example.
We both know that a digital caliper is always a wee bit more than both 6 inches and 150 mm anyway.
I notice in ASME B1.1 a typical dimension is given as 0.003969" - when are dimensions to 6 decimals actually needed? I'm guessing it's only given that accurately so that it can be used for calculating purposes. Try going on to a shop floor and talk about dimensions and tolerances given to 6 decimal points. You'd get some very funny looks
in most placesGordon
P.S. I'm assuming you are as thick-skinned as you hoped I'd be. I love a good discussion. I have been known to change my opinion - but always by logic and common sense - plus of course good humour.
Wayne
Gage Crib Worldwide
I agree, I also like the discussion, it keeps our minds sharp.
Sorry if I did not make myself clear here. I should have attached
to the statement. As for rounding direction, the resultant should never be beyond the thread tolerance which would always leave a finished thread which would gage correctly.
My original point here was that the different pitch diameter required in many cases is due to engineer error. The extra cost of special tooling could/should be avoided if the thread was correctly designated on the part drawing. The battle I am fighting is that too many engineers, or maybe their detailers, do not have the knowledge of the correct thread specifications nor do they go back to the standard to get the correct information. They just guess, or copy another drawing, or look in some inaccurate non-codified source. This lazy/uneducated approach causes problems for everyone using the drawing for the life of the drawing, which could be a hundred years.
I agree that measuring pitch diameter is a nice thing, but can not be the only check made of a thread. I did find your statement referenced above in your 21st August 2008, 03:44 AM post in the thread titled: Measuring minor diameter of internal threads - How To?
I can agree with this statement as a good way to measure a thread; however; it does require some skill on the part of an operator. This is why GO/NOGO thread gages are popular. Using GO/NOGO gages for checking of a manufactured thread is simple.
Again I agree. If you want to control the thread manufacturing process, additional measurements beyond using GO/NOGO gages are needed. With the tool you have designed, or tri-roll gages, measurements can be documented and SPC can be implemented. This process still does not require the rounding of the parameter values.
The people reading this forum for the most part are the white lab coat guys. What you are promoting, I believe, is still best implemented at the local level. I agree that it can be very useful in the machining process; however; if some one chooses to round, it needs to be their decision and their calculation so that they can defend it.Working where you do, and with the products you have, may make you slightly biased
I’ll leave “nano”-philosophy to the guys in the white coats and their cotton gloves in their temperature controlled environments.
In a "Calibration" related [forum], I'm probably with people who work in temperature controlled environments - my target "audience" is machinists and production technicians in general who measure machine produced components.
Publication of thread data should always be fully accurate. Any competent machinist looking at the accurate data would be able to do any rounding in his head. I still contend that publishing erroneous data only exacerbates the spread of bad information, and that creates problems which are far reaching.
Sorry, this example does not hold. It is apples to oranges. Thread specifications are published on drawings, which are then taken as the final word, regardless of how erroneous they appear. No one would specify a 6" caliper on a drawing. And that is the point. I can foresee discussing a drawing with an erroneous thread specification and hearing the engineer’s defense as: “but I have this chart that I found on the internet.”
I agree with you so many times. This is another time.
What you have said is correct. That is the precision that the B1 committee has gone to assure receptive accurate data whenever someone calculates the data.
The person on the shop floor should not have to worry about the 6 decimal places. They have to look at the thread chart, or the drawing, or the thread engineering software, or the pitch diameters marked on the gage and work from there.Try going on to a shop floor and talk about dimensions and tolerances given to 6 decimal points. You'd get some very funny looks
In summary, the tool you sell is useful if used in conjunction with, at a minimum, a GO gage, but the chart you have published is ripe for unintended abuse. If a chart is published that proposes to assist in calculating thread pitch diameters, regardless of the disclaimers printed on the chart, the data should give the most accurate data. Publishing thread pitch diameters that are “close enough” is just assisting the spread of error.
Whew! This much typing and thinking all in one day is too much like WORK.
G
Gordon Clarke
Wayne,
I like the way you replied but it would I feel, become really complicated if I continued in the same mode.
I’ll try a slightly different approach this time, even although I’m sure we’ll never completely agree about rounding tolerances up or down. A “Rule of Thumb” says that when measuring a tolerance, the tool used should preferably be at least 10 times more accurate than the tolerance itself. I’ve seen this modified in “emergencies” to 5 to 10 times as accurate.
If I was measuring something that had a tolerance of 0.1” I’d be comfortable measuring with something I could read in 0.01”. Using something that could measure in 0.001” would I feel, be “overkill”. If I only had a tool that measured in 0.001” and was making a report on my measurement results, I would round up or down to the nearest 0.01” This would be enough for reliable SPC data. I’m not referring to “life or death” situations or products.
In my basic tables I start off by writing:
Important:
These tables are made as a guideline for calculating pitch diameter tolerances. As the primary purpose is to measure with a digital caliper, most dimensions are calculated to the nearest 0.01 mm.
When a tolerance in a table is used it should be remembered that the tolerance will often be slightly larger for a finer pitch than standard. i.e. the pitch diameter tolerance for M60x2 is larger than for M36x2, which again is larger than that for M16x2 – which is standard.
For exact tolerances use the relevant, approved standard.
Pitch diameter tolerances
Thread
Bolt (6g)
M16x2
M36x2
M60x2
M120x2
+0.212 / -0
+0.224 / -0
+0.236 / -0
+0.250 / -0
-0.038 / -0.198 (0.16)
-0.038 / -0.208 (0.17)
-0.038 / -0.218 (0.18)
-0.038 / -0.228 (0.19)
Sorry about the mm but I am in Europe. As you can see, I explain why I do as I do in the tables, and also show the actual sizes before rounding up or down. If the dimension ends in 0.005 mm then I let the user decide which way to go. I also always give the standard I have used as a reference for the tolerances.
I was at one German company where the technician had been measuring thread pitch diameter with wires. The micrometer was a standard 0 – 25 mm micrometer and in 0.01 m. When he subtracted the wire constant from the measurement result (wire constant sizes are given in 0.001 mm) he arrived at a number that was in 0.001 mm. I find this very misleading and felt he should have rounded up or down. If he had used a 0.001 mm micrometer, now that would have been a different story.
I’ve also found that many machine operators that only use thread gauges have little or no knowledge of the actual tolerance size and when they approach the point when the gauge can almost be used they start to fiddle around and take very little off until the gauge can be “forced” in or on. At least by looking at my tolerances, they can see if the thread gauge can’t go in or on, then they can increase material removal by at least half the tolerance.
In an ideal world every company (no matter how small) would have all necessary relevant standards but it just doesn’t happen. I feel that a little knowledge or insight is better than no knowledge and many producing threaded parts have virtually no knowledge of thread tolerances. To me a thread is just two opposing triangles equally distant from a common centreline.
When I first started with QC I was at a measurement seminar and when we got to the part about threads and thread measurement the first 5 pages were filled with formulae complicated enough to think we were dealing with rocket technology. I could see that the majority of the participants turned their thoughts elsewhere and waited for the next part of the seminar. In my opinion a few simple facts and drawings could have kept people focussed. The formulae could have been saved for last.
Gordon
P.S. You're right - this is hard work
I like the way you replied but it would I feel, become really complicated if I continued in the same mode.
I’ll try a slightly different approach this time, even although I’m sure we’ll never completely agree about rounding tolerances up or down. A “Rule of Thumb” says that when measuring a tolerance, the tool used should preferably be at least 10 times more accurate than the tolerance itself. I’ve seen this modified in “emergencies” to 5 to 10 times as accurate.
If I was measuring something that had a tolerance of 0.1” I’d be comfortable measuring with something I could read in 0.01”. Using something that could measure in 0.001” would I feel, be “overkill”. If I only had a tool that measured in 0.001” and was making a report on my measurement results, I would round up or down to the nearest 0.01” This would be enough for reliable SPC data. I’m not referring to “life or death” situations or products.
In my basic tables I start off by writing:
Important:
These tables are made as a guideline for calculating pitch diameter tolerances. As the primary purpose is to measure with a digital caliper, most dimensions are calculated to the nearest 0.01 mm.
When a tolerance in a table is used it should be remembered that the tolerance will often be slightly larger for a finer pitch than standard. i.e. the pitch diameter tolerance for M60x2 is larger than for M36x2, which again is larger than that for M16x2 – which is standard.
For exact tolerances use the relevant, approved standard.
Pitch diameter tolerances
Thread
Nut (6H)
M16x2
M36x2
M60x2
M120x2
+0.212 / -0
+0.224 / -0
+0.236 / -0
+0.250 / -0
-0.038 / -0.198 (0.16)
-0.038 / -0.208 (0.17)
-0.038 / -0.218 (0.18)
-0.038 / -0.228 (0.19)
Sorry about the mm but I am in Europe. As you can see, I explain why I do as I do in the tables, and also show the actual sizes before rounding up or down. If the dimension ends in 0.005 mm then I let the user decide which way to go. I also always give the standard I have used as a reference for the tolerances.
I was at one German company where the technician had been measuring thread pitch diameter with wires. The micrometer was a standard 0 – 25 mm micrometer and in 0.01 m. When he subtracted the wire constant from the measurement result (wire constant sizes are given in 0.001 mm) he arrived at a number that was in 0.001 mm. I find this very misleading and felt he should have rounded up or down. If he had used a 0.001 mm micrometer, now that would have been a different story.
I’ve also found that many machine operators that only use thread gauges have little or no knowledge of the actual tolerance size and when they approach the point when the gauge can almost be used they start to fiddle around and take very little off until the gauge can be “forced” in or on. At least by looking at my tolerances, they can see if the thread gauge can’t go in or on, then they can increase material removal by at least half the tolerance.
In an ideal world every company (no matter how small) would have all necessary relevant standards but it just doesn’t happen. I feel that a little knowledge or insight is better than no knowledge and many producing threaded parts have virtually no knowledge of thread tolerances. To me a thread is just two opposing triangles equally distant from a common centreline.
When I first started with QC I was at a measurement seminar and when we got to the part about threads and thread measurement the first 5 pages were filled with formulae complicated enough to think we were dealing with rocket technology. I could see that the majority of the participants turned their thoughts elsewhere and waited for the next part of the seminar. In my opinion a few simple facts and drawings could have kept people focussed. The formulae could have been saved for last.
Gordon
P.S. You're right - this is hard work
Q
qcdirect07
Hope someone can help me. I'm not familiar with metric threads. We have parts that have a threaded hole M6x 1.0-6H. These parts will not take the go gage that we purchased but will accept the mating hardware which are M6 bolts. Any idea why they bolts work but the gage does not?
S
stefanhg
Hi qcdirect07,
It is possible that the GO Plug not enter, if your internal thread is on the lower limit, but the threaded hole and the GO Plug gauge are OK.
The other possibility – threaded hole is below the lower limit, workpiece is rejected, but it still can fit the bolt. Normally the bolt is M6x1 6g and the min. allowance is 0,026mm.
To know where you are, you must measure the diameters of the threaded hole.
See the calculations for ISO and ANSI GO plug gauges in the attachment.
Calculations are made whit software QMSys Threads and Gauges 4.8
It is possible that the GO Plug not enter, if your internal thread is on the lower limit, but the threaded hole and the GO Plug gauge are OK.
The other possibility – threaded hole is below the lower limit, workpiece is rejected, but it still can fit the bolt. Normally the bolt is M6x1 6g and the min. allowance is 0,026mm.
To know where you are, you must measure the diameters of the threaded hole.
See the calculations for ISO and ANSI GO plug gauges in the attachment.
Calculations are made whit software QMSys Threads and Gauges 4.8
Last edited: