Manual inspection is almost becoming a lost art with the rise of CMM equipment available to the shop floor. The aerospace company I worked for had Maxichecks, Zeiss and a monster L&K CMM in our layout inspection area. We also had an old guy named Bud who could inspect a part with a height stand and indicator, a tilt rotary table and a calculator faster than a CMM could. We compared his readings with the CMM and he was spot on every time. The man was an artist at his craft.
Travis is the best quality guy I have worked with, especially considering he's an incredible Machinist as well, you don't find that very often. Great video man!
Excellent video. We need more QC content on here. It is especially important for the machine operators making parts to know how a part is to be checked for various requirements. This only benefits everyone.
The quickest way to do this is to have the surface you want to check facing the granite. Set it on 3 blocks the same size( Gauge blocks are fine). Then run an indicator on the bottom side. This eliminates the need to indicate in flat.
This is probably the quickest and most accurate way for larger faces, like the one he just showed. His technique has much more uncertainty. I use 3ea. 1-2-3 blocks that are lapped and matched with an indicator on a base that slides under the part, no adjustments needed.
While I often did that in gauging gauge plates I made, this way does allow access to the entire surface. Indicating the bottom does not. I would normally check three points on the bottom so I could measure the parallelism of the part, but as this was already known and he just wanted to check flatness, this was best as it gave easy access to the whole surface.
@@dj2740The risk with doing that is that if they're part is a cup shape where you would miss a low spot in the center because the edges are touching the granite not allowing the feeler gauge to see the high spot A common alternative is a flat plate with a plunger indicator poking up the center and you can move the part over it. Indirectly measure the deviation from the surface of the flat plate to the lowest part of the feature which will give you the flatness. I'm going to call the flatness indicator
I am sill learning, but a few questions come to mind: 1. Aren't you measuring parallelism when you set your reference on the bottom and then measure off the top? Taller stands would let you sweep the underside where your plane is established. 2. Adjustable jack stands seem unreliable for precision work with call outs in the .005 range, just setting the part on them could cause more movement then that on the threads, adjustments aside wouldn't fixed stands that have been run through the surface grinder together be more accurate as a base? You guys are awesome and I like how everything is broken into manageable sections for the widest audience, this is a trade that will always have a place in the world and it's great to see enthusiastic teachers of useful skills.
I do not use the bottom for the reference plane but rather use the adjustable jack stands to create the reference plane on the top surface. Takes a little longer but give better access in my opinion.
You are creating a parallel measurement, but you’re doing is by referencing your plane to a known flat plane. So you assume the table is “dead flat” and then the peaks and valleys of the top surface can be referenced to a perfect plane. So yes, but only as a byproduct. The real goal is referencing the measured plane to a “perfect” one. Weirdly enough the jack stands work okay. For super precise stuff they do make high precision jack stands and such. Great questions man
I agree with the previous two replies. If you plan to use the jacks in the fashion when you are sweeping your indicator on the face of the part that the jacks are resting on you must use a point style tip (on the jack). Using the ball mount socket creates a pivot point mini datum pad. This style of feature could cause unintentional interactions with the part if cupped and would not act as reference points. The whole idea is three points make a plane.
One is a point…two is a line.. three are a plane. Simple and effective…. Most machinist will have this inbred but for the rest of us who watch to be educated and amazed its always good to have things to latch onto for our hobbies and work. You must be a super high performing genius not to learn something from these videos….. consistent learning..magic.
Man why can't I like this video twice, an amazingly valuable information please Travis everyone does machining videos on youtube but no one does inspection, please more inspection videos especially old school inspection methods.
All you checked was parallelism, not flatness. Need to set on 3 fixed points, run indicator on the surface you are locating on, (bottom ), & read your TIR. That's flatness. You can't use adjustable points to check flatness unless they are all level within .0001. Aerospace machinist for over 36 years.
@@Frank_inSA That is parallelism to surface plate, not flatness of given feature. Set surface to be checked on tooling balls, sweep locating surface, TIR is flatness. Ran a lot of parts that callout was flatness .0018, parallelism was .005 and we held that for hundreds of parts over 6 months. By the way, that was on titainium.
@@johnhines3591 the surface plate is the reference, you need something to measure against. If the top is parallel to the plate and has no high or low spots it's 100% flat Btw of course you do the same with the 3balls, checking parallelism against the plate, only upside down
@@Frank_inSA Flattness is measured to itself. Must locate on itself to check. Preferably tooling balls so you are locating on points, not another plane.
@@johnhines3591 why can't you guys not take some time to think about? Of course your tooling balls make sure that the part sits 100% parallel to the plate. Or can you take different sized balls? While moving your indicator stand you measure ALWAYS against the plate, which is the reference. Have a good night
@@ILmachiner you can check it both ways accurately so long as you create a perfect plane to measure against on the top surface. This is why we use the adjustable jacks. In a technical sense we are checking against the uncertainty of the surface plate but that is true no matter how you do it.
@@travisjarrett2355 this method is a wast of time. Put the part on 3 points you know are equal hight. That is the plain. The smaller the point on the 3 points the better, or use precision balls.
@@travisjarrett2355 I see the advantage of having the surface facing up where you have full access. It's an extra step but not much work. Setting up is half the fun.
Be careful using sharpies on Stainless steels, they can lead to long term corrosion. Lots of applications forbid anything other than low chloride/low halogen marking implements on a finished surface.
This is a good content, back to basic to things that most people forgot or simply cheaper methods that can be done immediately, especially for a machinist....
Just because these boys have everything, doesn't mean they forgot everything that got them there. Always good to rely on accurate, proven conventional methods to check the accuracy of your parts. Great Video Travis!!
The way you are doing this we’re you are looking for parallel and surface planter accuracy.. you are relying on your step ups as perfectly level after adjustments you have done. What you should do is either set your surface parallel with a know parallel surface gauge, or do as you have done but take your measurements three times by moving the part in rotation 120 degrees and recheck your intended surface. This is the only two ways you can accurately measure parallel surface for planier accuracy. One you have determined planier accuracy is parallel, then you can check your intended surface accuracy, however.. this is also determined by the bottom surface be planier accurate as well. By rotation 120 degrees you can immediately determine your bottom surface accuracy. Mil Spec T5400 quality assurance standard.
Thanks! It amazed me to watch u fellas doing a tool.setup a while back. Yall started with joblocks and the surface plate as your root reference. Johanson was smiling down from heaven ;)
Some old dude I use to work with would do both CMM and this way just to make sure to bad he never really like sharing the knowledge. Unlike myself I enjoy teaching those who are interested.
Excellent video! Simple, clear, concise. Well done. Could have mentioned temperature stabilization for such a heavy part. That could take a full day if you want to get down to tenths. Also you could mention on thinner/lighter parts the need to wear gloves to keep body heat out of the part.
All good points my friend. There are so many subtle changes that can take place when dealing with thousands (or ten thousandths) of an inch. Thanks for watching.
Am nearing retirement.. the reason I mention is! I have been watching these videos for some weeks now.. I've had a good working career, but it's never felt like I really was in the right profession. Anyway I think I should have become a machining engineer 🤔 I have really enjoyed the provisional side of this videos..
The other way to measure flatness is to have 3 gauge blocks. Flip the part over with the surface needing verification supported with the guage blocks. Sweep the surface with a dial test indicator. Flatness is the difference between min and max indicator reading. This works fairly well.
I've seen the indicator point angled almost straight down to the part. Way back learned to keep at least the point as parallel to the part as possible. Seems to be so different with newer machinist
This is indeed correct William and maybe I should have mentioned it. I did try to make the angle as small as possible and still access the whole surface of the part. Compensating for the roughly 20 degree angle (estimated) would put the actual deviation around .0031 (a difference of .0002). For this particular application I was ok with this level of uncertainty. Thanks for watching.
I would have thought it better to check the 3 points on the bottom plane and then vs the top plane, that would establish flatness and parallelism. Because I would think they are very relative to one another, especially with that part being part of a hub of some sort. Correct me if I'm wrong. 🤷♂️
He’s only measuring flatness, the parallelism isn’t relevant. If it was parallel he could just set it on the table as one plane and then measure the top, no need for stands. You would still have to check flatness and parallelism separately though. They may call out a 0.001” flatness with a 0.005” parallelism for example, where you could easily meet parallel but not flat. That’s where this comes in
@@blacklabel6223 yes I understand your point here, but my thinking is, if you have the part raised to level it out, you could then check both, flatness and parallelism with that same setup vs doing 2 as you mentioned. Kinda kill 2 birds, one stone type of thing. Or, while you're there, unless the part has an angled face. But I get ya. 👍
"One is a point, two are a line, three are a plane." These are my thoughts about everything. It's been years that I want to write things down. 😅 Things only become real with a 3-point framework. And it's just everywhere.
What would be the difference? Genuine question. I’m not a metrologist, but setting a 3 point plane parallel to a calibrated reference surface then measuring the plane seems ok to me as long as the probe and reference surface tols are ok. What am I missing?
@@joelawton123 parallelism is the relation of one surface to the other in the same plane as the name "parallelism" suggests. in this case the surface on the jack stands is compared to the surface that is being swept by the indicator.
Thanks Travis for another great video. One thing I don't understand is the reason for the three jacks. Is it because we cannot assume the bottom surface of the part to be flat? Ciao, Marco.
yes that is part of it, the bottom isn't a perfect surface and has it's own tolerences that could add or cancel with the top surface (kinda like a trapazoid vs parallelogram). Using the jacks isolates the top surface.
Thanks Marco. As wictor kempel mentioned that is partly correct. We need to isolate the surface we are checking from any imperfections on the bottom surface (not flat) or maybe more importantly any relations between the two surfaces (not parallel). The jack stands allow us to isolate the surface from these possible imperfections that would distort the readings. Thanks for watching Marco.
Jacks are not needed really. This inspection method is 100% appropriate if the goal is to quantify flatness. Assuming the reference surface is not in a "rocking condition", if the measured surface is parallel to the reference surface (the one with jacks) then one can determine that it is also within the flatness spec BUT if the measured surface it is not parallel within the flatness specification it MAY still meet the flatness spec. There's an overlap here. Parallelism also controls flatness but flatness does not control parallelism. The parallelism and flatness tolerance zones are the same. The main difference is that orientation (parallelism) error may add to the "flatness measurement". Sometimes speed is needed, sometimes accuracy. It all comes down to what you're trying to achieve. Nice video.
As long as your table has zero variation, this is great! I would have my doubts about a table being perfectly flat in such a large area, but I'm sure you guys are very aware of the tables flatness and levelness. As usual, great job!
@@blacklabel6223 No where in this video was this mentioned. It is more important than the whole video, as if the granite is not flat, neither is the part.
You can check your granite with the indicator prior to use. If it reads zero everywhere you are probably OK, just need to have base in multiple directions for a grid, really only necessary for weekly check if you are using it often and maybe wearing it while sliding the base around, or if it's new to you. Next video on how to fix this issue on the machine please 😁, so that we can get a perfect part off.
Measure flatness with an optical flat or auto collimator. I made several up to 12"(305mm) diameter in that size flatness, was 1/20 of wave length, monochrome green.
We have a leveling plates various sizes alot safer than using jacks. They are about 1/4 inch steel plates drilled and threaded holes with bolts with rounded ends that sit on the table. I think they date from the 1960's
I’m sorry @ Travis Jarrett, but you are incorrect or ignore a lot of things in this video. I noticed you already replied to someone else’s comment that you are ignoring cosine error of the indicator, an important thing to state, although in this video your indicator was at a shallow angle so not a big deal in this video. However the biggest issue is that the method of checking flatness is incorrect. Using the ‘put the part on 3 jacks method’ , just zeroing out 3 points on the top surface and then sweeping the indicator over the surface and looking for the total variation doesn’t check the actual flatness. What you need to do is move the indicator back and forth all across the surface, constantly adjusting the jacks up and down to bring the indicator variation to a minimum. This can be a very difficult and time intensive process, which is why using a CMM is usually the best method. And why using a scanning CMM that can measure 100s or even 1000s of points would be best, rather than using a touch trigger CMM that only takes individual point hits. Your method will most likely give you a good idea of what the general flatness of the surface is, but at a fundamental level it is not technically correct. FARO has an okay article on their website that talks about this issue somewhat: www.faro.com/en/Resource-Library/Article/how-to-evaluate-flatness-in-gd-t under the section ‘using a height gage to evaluate flatness.
Nice demo. It would have been good to see the cmm output to compare. I can’t help watching anything like this in imperial units without thinking how much easier this stuff is in metric though 😂
Yup that's how we did it before CMM became more common. About the only thing I would ad in the video is cleaning the table and working surfaces before you start. I know you checked it but for any newbys it would be good to show it.
Educational video for the most part. But you are not measuring flatness the way this is set up, you are measuring deviation in a plane, referenced from another plane, you are actually measuring the parallelism of the feature. In order to properly measure flatness, you must be locating on the plane you are measuring. Flatness is the deviation in a plane referenced to itself. More or less, flip the part over so that the face you are measuring flatness of, is sitting on the 3 jack stands, then sweep the indication across the surface, the total deviation is your overall flatness. 😁
Exactly correct. If this guy is the QC manager, they must be making more money from YT videos than making parts. On any vaguely functioning lathe it would be impossible to get a part not flat better than .0033. I guess if you made it concave/convex by programming a taper face, or the "X" axis wasn't anywhere near perpendicular with the spindle centerline. I spent 42 years designing and building "Super Precision" lathes (that ought to tell you where). .005" flat was hacksaw territory. It's sad seeing all these people eating this up and thanking him for the "lesson". At HBI, this guy would have been working in heat treat...
@@berniepragle948 you're repeating yourself. Maybe you wasn't as good as you think if you cannot imagine that there are always many different ways to skin a cat, besides that what you once have learned
@@Frank_inSA And you would have been sweeping the floor in Heat Treat. I personally designed and ground thousands of parts with this exact back face and hole pattern. They were the back face of the tooling that fits on the spindle noses of lathes, collet chucks, etc. We also ground the mounting face and nose of our lathes after assembly to within 10 millionths of an inch true. Don't let the gloves, "three points...", and fancy video fool you. This guy is clueless. A little life lesson: You don't have to know everything. You just have to know where to find someone who does.
His method is technically fine since he is establishing a level plane using the jacks. Its not how I would do it, but if he were to flip the part and rest on that face and indicate underneath I bet he would get the same answer, faster.
Some times old ways are still the best . Don't forget optical flatness gauges as well . Then optical comparator to ensure concentsity and parallelism between surfaces are Absolute
We have some Renishaw Equator CMMs at work. They take longer to inspect the part machined then it does to actually machine it. Pretty crap when the cnc machine producing them needs to run constantly to hold size..
Looks like it would indeed show you flatness but would not tell you perpendicularness from that face and the face against the jacks correct ? Kyle F. Lead Machinist from Tulsa ok thanks
The problem I have with this method of checking flatness is that if one of the 3 points you set the initial plane is low (if the flatness happened to be way out in that one spot), then your whole plane is now bad and you don’t know about it. When I manually check flatness, I set the part on three points on the surface I’m checking flatness on, then sweep my indicator under the part. The points that it sits on give more surface area, compared to the indicator ball which is has an infinitely small contact point given its spherical shape, for the part to locate on. In this instance it wasn’t an issue due to the part being dished and the flatness essentially got worse as you got farther from the center in every direction, but it could be an issue on other parts. All this being said depending on your tolerance this method is probably always fine. Might be worth mentioning the reason for using 3 points and not 4 too. Just thoughts. Source: my name
That's part of the reason why you're checking the rest of the surface though right, as if you choose one point, and move, and its off as soon as you move away, you immediately know that point is off. If it stays the same, as you move away, and is the same as the surrounding points, you don't have a point that's out.
By the look of it. U guys machine the side with the hole first. Than held it , to do the taper. The part collapse n bent inward.that’s why the high spot is on the outter area.
Ease of access is mainly why I choose the top surface method and the 123 blocks may hide the "highest" or "lowest" spot thereby hiding the real value. Thanks for watching Michael.
or JUST USE feeler gauges and flat granite block DONE without wasting time . Indicator will not give a TRUE flatness reading it will give perpendicularity , feeler gauges and granite block never ever fails.
I see you used a sharpie on your part. Honorable mention if you are making parts for Boeing big no no. We had a major finding where I used to work because we’d write with sharpie marker “MRR” (Material review request) for non conformance parts, but every now and then the parts would be looked over by engineering and deemed conforming. With the sharpie though I guess it can do something to the metal later on down the line which etches it or something. So Boeing actually makes sharpie markers that are to be used with their parts to meet their spec which we found out during an audit. Just thought I’d mention maybe you guys are more familiar with it than I am. I forget the full reason or the actual spec it was under it was years ago now.
For all of us regulars to the channel we need to do better for The TITANS of CNC Machining. They put out great content and educational videos daily for new and seasoned machinist. Please take a moment and support them by hitting the LIKE button. When I see over 12000 views and just over 550 likes I feel like we can do better. Thanks, Just a fan of the channel.
If the face your inspecting has more of a cupped or dish shape, would you then have too indicate from the tallest points of the piece? And would you also have to indicate the "roundness" of that sort of shape?
Manual inspection is almost becoming a lost art with the rise of CMM equipment available to the shop floor. The aerospace company I worked for had Maxichecks, Zeiss and a monster L&K CMM in our layout inspection area. We also had an old guy named Bud who could inspect a part with a height stand and indicator, a tilt rotary table and a calculator faster than a CMM could. We compared his readings with the CMM and he was spot on every time. The man was an artist at his craft.
I have never had either art. One nice 6” starrett dial caliper and I can build you a jet engine 🤣
Sounds like me. Lol not trying to be cocky but I am faster with everyone that uses the cmm vs me using Cadillac, a height stand and indicator.
@@ZacharyTelesca hmm must be really nice calipers.
@@dominic6634 make sure they are zeroed, measure 10 times, remove the outliers and average the rest 🤣
I agree. My company just invested in ViciVision M306 Techno. I'm not going to lie, though; it's pretty bad a**.
Travis is the best quality guy I have worked with, especially considering he's an incredible Machinist as well, you don't find that very often. Great video man!
Excellent video. We need more QC content on here. It is especially important for the machine operators making parts to know how a part is to be checked for various requirements. This only benefits everyone.
Ha ha. You get no argument from me. Thanks for watching.
like your handle (Trumpwon,bigly)
The quickest way to do this is to have the surface you want to check facing the granite. Set it on 3 blocks the same size( Gauge blocks are fine). Then run an indicator on the bottom side.
This eliminates the need to indicate in flat.
This is probably the quickest and most accurate way for larger faces, like the one he just showed. His technique has much more uncertainty. I use 3ea. 1-2-3 blocks that are lapped and matched with an indicator on a base that slides under the part, no adjustments needed.
While I often did that in gauging gauge plates I made, this way does allow access to the entire surface. Indicating the bottom does not. I would normally check three points on the bottom so I could measure the parallelism of the part, but as this was already known and he just wanted to check flatness, this was best as it gave easy access to the whole surface.
feeler gauges and granitew block done and it will give true flatness not perpendicularity like indicator will.
Exactly!...I def have no use (future reference) for this video!
@@dj2740The risk with doing that is that if they're part is a cup shape where you would miss a low spot in the center because the edges are touching the granite not allowing the feeler gauge to see the high spot
A common alternative is a flat plate with a plunger indicator poking up the center and you can move the part over it. Indirectly measure the deviation from the surface of the flat plate to the lowest part of the feature which will give you the flatness. I'm going to call the flatness indicator
I am sill learning, but a few questions come to mind:
1. Aren't you measuring parallelism when you set your reference on the bottom and then measure off the top? Taller stands would let you sweep the underside where your plane is established.
2. Adjustable jack stands seem unreliable for precision work with call outs in the .005 range, just setting the part on them could cause more movement then that on the threads, adjustments aside wouldn't fixed stands that have been run through the surface grinder together be more accurate as a base?
You guys are awesome and I like how everything is broken into manageable sections for the widest audience, this is a trade that will always have a place in the world and it's great to see enthusiastic teachers of useful skills.
I do not use the bottom for the reference plane but rather use the adjustable jack stands to create the reference plane on the top surface. Takes a little longer but give better access in my opinion.
You are creating a parallel measurement, but you’re doing is by referencing your plane to a known flat plane. So you assume the table is “dead flat” and then the peaks and valleys of the top surface can be referenced to a perfect plane. So yes, but only as a byproduct. The real goal is referencing the measured plane to a “perfect” one.
Weirdly enough the jack stands work okay. For super precise stuff they do make high precision jack stands and such.
Great questions man
I agree with the previous two replies. If you plan to use the jacks in the fashion when you are sweeping your indicator on the face of the part that the jacks are resting on you must use a point style tip (on the jack). Using the ball mount socket creates a pivot point mini datum pad. This style of feature could cause unintentional interactions with the part if cupped and would not act as reference points. The whole idea is three points make a plane.
Thanks man, you're a good teacher with clear explanation Travis.
My pleasure. Thanks for watching.
As an inspector in aerospace, I would agree with everything said in this video. That's how you do it :)
One is a point…two is a line.. three are a plane. Simple and effective…. Most machinist will have this inbred but for the rest of us who watch to be educated and amazed its always good to have things to latch onto for our hobbies and work. You must be a super high performing genius not to learn something from these videos….. consistent learning..magic.
Yeah I get that their flatness wasn't at all a tight call out but I'm surprised that you guys tried that?
Please do more of these. Excellent demonstration.
More videos like that!!! Pure educational gold!!! :D
Man why can't I like this video twice, an amazingly valuable information please Travis everyone does machining videos on youtube but no one does inspection, please more inspection videos especially old school inspection methods.
Keep on giving!!! Watching before work starts
Reminds me of when I worked. I was In aerospace/medical. I spent 35 + yrs as machinist, T&D maker and in metrology.
Thanks
as a mechanic support machinist we don't always get the best of inspection equipment (cmms) this can help alot of people
Keep up the tips!
All you checked was parallelism, not flatness.
Need to set on 3 fixed points, run indicator on the surface you are locating on, (bottom ), & read your TIR. That's flatness. You can't use adjustable points to check flatness unless they are all level within .0001.
Aerospace machinist for over 36 years.
No, by leveling on three points he compares to the plate and not the bottom.
@@Frank_inSA
That is parallelism to surface plate, not flatness of given feature.
Set surface to be checked on tooling balls, sweep locating surface, TIR is flatness.
Ran a lot of parts that callout was flatness .0018, parallelism was .005 and we held that for hundreds of parts over 6 months. By the way, that was on titainium.
@@johnhines3591 the surface plate is the reference, you need something to measure against.
If the top is parallel to the plate and has no high or low spots it's 100% flat
Btw of course you do the same with the 3balls, checking parallelism against the plate, only upside down
@@Frank_inSA
Flattness is measured to itself.
Must locate on itself to check. Preferably tooling balls so you are locating on points, not another plane.
@@johnhines3591 why can't you guys not take some time to think about? Of course your tooling balls make sure that the part sits 100% parallel to the plate. Or can you take different sized balls?
While moving your indicator stand you measure ALWAYS against the plate, which is the reference.
Have a good night
I like this method. I usually set up the surface to be inspected upside down on three points and sweep under with a long stem indicator.
Don't change you are checking flatness correctly. This video is checking parallel.
@@ILmachiner you can check it both ways accurately so long as you create a perfect plane to measure against on the top surface. This is why we use the adjustable jacks. In a technical sense we are checking against the uncertainty of the surface plate but that is true no matter how you do it.
@@travisjarrett2355 this method is a wast of time. Put the part on 3 points you know are equal hight. That is the plain. The smaller the point on the 3 points the better, or use precision balls.
@@travisjarrett2355 I see the advantage of having the surface facing up where you have full access. It's an extra step but not much work. Setting up is half the fun.
Be careful using sharpies on Stainless steels, they can lead to long term corrosion. Lots of applications forbid anything other than low chloride/low halogen marking implements on a finished surface.
That is true Andrew. Thank you for adding that.
Ya, that made me anxious 😂
hmmm didn't know that. Thanks.
You not wrong but its situational
You not wrong but its situational
This is a good content, back to basic to things that most people forgot or simply cheaper methods that can be done immediately, especially for a machinist....
very useful, thanks team
Very good work... This knowledge will help many...
Just because these boys have everything, doesn't mean they forgot everything that got them there. Always good to rely on accurate, proven conventional methods to check the accuracy of your parts. Great Video Travis!!
The way you are doing this we’re you are looking for parallel and surface planter accuracy.. you are relying on your step ups as perfectly level after adjustments you have done. What you should do is either set your surface parallel with a know parallel surface gauge, or do as you have done but take your measurements three times by moving the part in rotation 120 degrees and recheck your intended surface. This is the only two ways you can accurately measure parallel surface for planier accuracy. One you have determined planier accuracy is parallel, then you can check your intended surface accuracy, however.. this is also determined by the bottom surface be planier accurate as well. By rotation 120 degrees you can immediately determine your bottom surface accuracy. Mil Spec T5400 quality assurance standard.
Thanks!
It amazed me to watch u fellas doing a tool.setup a while back. Yall started with joblocks and the surface plate as your root reference.
Johanson was smiling down from heaven ;)
Great explanation espacially when adding the comparison of CMM
Appreciate the QC & Inspection videos. Please do more of these. Thanks a lot!
Super helpful!
That is badass. Love the way you clearly explain things. Very easy to follow your teachings. Much love and gratitude
Thank you my friend.
Hello sir, could you provide more videos about other geometric tolerances ?
this is exactly how I would do it, great demonstration
Thank you very much.
Thanks great explanation.
Thanks Rodney.
Really enjoying you videos.
Some old dude I use to work with would do both CMM and this way just to make sure to bad he never really like sharing the knowledge. Unlike myself I enjoy teaching those who are interested.
Excellent video! Simple, clear, concise. Well done. Could have mentioned temperature stabilization for such a heavy part. That could take a full day if you want to get down to tenths. Also you could mention on thinner/lighter parts the need to wear gloves to keep body heat out of the part.
All good points my friend. There are so many subtle changes that can take place when dealing with thousands (or ten thousandths) of an inch. Thanks for watching.
@@travisjarrett2355 Exactly so. Just been bitten by heat from handling making scrap parts a few too many times. 😀😀😀
Excellent content! Everything that I have is manual....but these principles still apply. Great explanation. Thanks!
He does such an excellent job in explaining the Procedure !!
Great video, but you could of just used three 1-2-3 blocks (or 3 gage blocks) to do a rough check like that. Just a suggestion.
seen it done it, show us more complicated QC practices. This is TITAN!!
Awesome man
I love metrology 💖
great video more metrology videos needed to our comunity great job
Am nearing retirement.. the reason I mention is! I have been watching these videos for some weeks now.. I've had a good working career, but it's never felt like I really was in the right profession. Anyway I think I should have become a machining engineer 🤔 I have really enjoyed the provisional side of this videos..
The other way to measure flatness is to have 3 gauge blocks. Flip the part over with the surface needing verification supported with the guage blocks. Sweep the surface with a dial test indicator. Flatness is the difference between min and max indicator reading. This works fairly well.
I've seen the indicator point angled almost straight down to the part. Way back learned to keep at least the point as parallel to the part as possible. Seems to be so different with newer machinist
This is indeed correct William and maybe I should have mentioned it. I did try to make the angle as small as possible and still access the whole surface of the part. Compensating for the roughly 20 degree angle (estimated) would put the actual deviation around .0031 (a difference of .0002). For this particular application I was ok with this level of uncertainty. Thanks for watching.
I would have thought it better to check the 3 points on the bottom plane and then vs the top plane, that would establish flatness and parallelism. Because I would think they are very relative to one another, especially with that part being part of a hub of some sort. Correct me if I'm wrong. 🤷♂️
He’s only measuring flatness, the parallelism isn’t relevant. If it was parallel he could just set it on the table as one plane and then measure the top, no need for stands. You would still have to check flatness and parallelism separately though. They may call out a 0.001” flatness with a 0.005” parallelism for example, where you could easily meet parallel but not flat. That’s where this comes in
@@blacklabel6223 yes I understand your point here, but my thinking is, if you have the part raised to level it out, you could then check both, flatness and parallelism with that same setup vs doing 2 as you mentioned. Kinda kill 2 birds, one stone type of thing. Or, while you're there, unless the part has an angled face. But I get ya. 👍
This guy seems to be on the level.
"Isolating a plane" is the name of the technique.
excellent
Kindly also make video on how to check perpendicularity in gear
I would love to see you do projection, symmetry, and true position without CMM.
Symmetry can be inspected quite easily, rest of those in the otherhand will take some fiddling to do.
I really enjoyed doing true up position!! THE OLD FASHIONED WAY. AND I ALWAYS GO WITH GAGES OVER CMM
"One is a point, two are a line, three are a plane."
These are my thoughts about everything. It's been years that I want to write things down. 😅
Things only become real with a 3-point framework. And it's just everywhere.
Would have been interesting to know what flatness the cmm gave yall but anyway nice and easy way to check parts!
I took 3 circles with the CMM. Inner, middle, and outer on the face. CMM came back at .0031 Thanks for watching.
Love the video 💕
Excellent video thank you!
This is measuring parallelisim. Flatness is measured from the surface facing the granite. At least that’s the way I have always checked flatness.
What would be the difference?
Genuine question.
I’m not a metrologist, but setting a 3 point plane parallel to a calibrated reference surface then measuring the plane seems ok to me as long as the probe and reference surface tols are ok.
What am I missing?
@@joelawton123 this video is measuring parallelism, however, you are also measuring flatness too. @freeCali 650 is correct in what he described.
@@jamonmang6099 ok but what is the difference?
@@joelawton123 parallelism is the relation of one surface to the other in the same plane as the name "parallelism" suggests. in this case the surface on the jack stands is compared to the surface that is being swept by the indicator.
@@jamonmang6099 but if the reference granite is calibrated flat then what is the difference?
Very nicely done, great job explaining things. Thank you.
Thanks Jeff! I appreciate that my friend.
Cool! Now we wait for next level, how to chech parallelism :) I have never seen how it's done.
99% of the time. Use a mic
Thanks Travis for another great video. One thing I don't understand is the reason for the three jacks. Is it because we cannot assume the bottom surface of the part to be flat? Ciao, Marco.
yes that is part of it, the bottom isn't a perfect surface and has it's own tolerences that could add or cancel with the top surface (kinda like a trapazoid vs parallelogram). Using the jacks isolates the top surface.
Thanks Marco. As wictor kempel mentioned that is partly correct. We need to isolate the surface we are checking from any imperfections on the bottom surface (not flat) or maybe more importantly any relations between the two surfaces (not parallel). The jack stands allow us to isolate the surface from these possible imperfections that would distort the readings. Thanks for watching Marco.
@@travisjarrett2355 Thanks for the explanation Travis (and Wictor). You do a great job on your videos, I enjoy watching them a lot.
Jacks are not needed really. This inspection method is 100% appropriate if the goal is to quantify flatness. Assuming the reference surface is not in a "rocking condition", if the measured surface is parallel to the reference surface (the one with jacks) then one can determine that it is also within the flatness spec BUT if the measured surface it is not parallel within the flatness specification it MAY still meet the flatness spec. There's an overlap here. Parallelism also controls flatness but flatness does not control parallelism. The parallelism and flatness tolerance zones are the same. The main difference is that orientation (parallelism) error may add to the "flatness measurement". Sometimes speed is needed, sometimes accuracy. It all comes down to what you're trying to achieve. Nice video.
As long as your table has zero variation, this is great! I would have my doubts about a table being perfectly flat in such a large area, but I'm sure you guys are very aware of the tables flatness and levelness. As usual, great job!
It’s a surface plate, they’re ground dead flat and also have a report to say how flat it really is. Usually sub tenth accuracy
@@blacklabel6223 👍
My concern as well!!
@@blacklabel6223 No where in this video was this mentioned. It is more important than the whole video, as if the granite is not flat, neither is the part.
You can check your granite with the indicator prior to use. If it reads zero everywhere you are probably OK, just need to have base in multiple directions for a grid, really only necessary for weekly check if you are using it often and maybe wearing it while sliding the base around, or if it's new to you.
Next video on how to fix this issue on the machine please 😁, so that we can get a perfect part off.
Measure flatness with an optical flat or auto collimator. I made several up to 12"(305mm) diameter in that size flatness, was 1/20 of wave length, monochrome green.
Very cool.
We have a leveling plates various sizes alot safer than using jacks. They are about 1/4 inch steel plates drilled and threaded holes with bolts with rounded ends that sit on the table. I think they date from the 1960's
I’m sorry @ Travis Jarrett, but you are incorrect or ignore a lot of things in this video. I noticed you already replied to someone else’s comment that you are ignoring cosine error of the indicator, an important thing to state, although in this video your indicator was at a shallow angle so not a big deal in this video. However the biggest issue is that the method of checking flatness is incorrect. Using the ‘put the part on 3 jacks method’ , just zeroing out 3 points on the top surface and then sweeping the indicator over the surface and looking for the total variation doesn’t check the actual flatness. What you need to do is move the indicator back and forth all across the surface, constantly adjusting the jacks up and down to bring the indicator variation to a minimum. This can be a very difficult and time intensive process, which is why using a CMM is usually the best method. And why using a scanning CMM that can measure 100s or even 1000s of points would be best, rather than using a touch trigger CMM that only takes individual point hits. Your method will most likely give you a good idea of what the general flatness of the surface is, but at a fundamental level it is not technically correct. FARO has an okay article on their website that talks about this issue somewhat: www.faro.com/en/Resource-Library/Article/how-to-evaluate-flatness-in-gd-t under the section ‘using a height gage to evaluate flatness.
JUST A FRINDLEY BIT OF ADVICE . NEVER PUT SOME THING DOWN ONTO THE GRANITE . THE BEST WAY IS TO SLIDE IT ON AND SLIDE IT OF . NICE WORK . ⚘⚘⚘⚘
I have 6 of those same jacks
Nice demo.
It would have been good to see the cmm output to compare.
I can’t help watching anything like this in imperial units without thinking how much easier this stuff is in metric though 😂
Yeah, I know the struggle 🤣
Yup that's how we did it before CMM became more common.
About the only thing I would ad in the video is cleaning the table and working surfaces before you start.
I know you checked it but for any newbys it would be good to show it.
Indeed my friend. Thanks for watching.
Thanks 🙏 great video
5 thou flatness, I wish. I could check that with a flat edge. I can see that with my eye. 😂
This is basically how 3d printer beds are calibrated.
Thanks for sharing
Educational video for the most part. But you are not measuring flatness the way this is set up, you are measuring deviation in a plane, referenced from another plane, you are actually measuring the parallelism of the feature. In order to properly measure flatness, you must be locating on the plane you are measuring. Flatness is the deviation in a plane referenced to itself. More or less, flip the part over so that the face you are measuring flatness of, is sitting on the 3 jack stands, then sweep the indication across the surface, the total deviation is your overall flatness. 😁
Exactly correct. If this guy is the QC manager, they must be making more money from YT videos than making parts. On any vaguely functioning lathe it would be impossible to get a part not flat better than .0033. I guess if you made it concave/convex by programming a taper face, or the "X" axis wasn't anywhere near perpendicular with the spindle centerline. I spent 42 years designing and building "Super Precision" lathes (that ought to tell you where). .005" flat was hacksaw territory. It's sad seeing all these people eating this up and thanking him for the "lesson". At HBI, this guy would have been working in heat treat...
Sorry guys, you are wrong.
By leveling it at 3 points it is exactly parallel to the plate, the bottom has no influence in this case
@@berniepragle948 you're repeating yourself. Maybe you wasn't as good as you think if you cannot imagine that there are always many different ways to skin a cat, besides that what you once have learned
@@Frank_inSA
And you would have been sweeping the floor in Heat Treat. I personally designed and ground thousands of parts with this exact back face and hole pattern. They were the back face of the tooling that fits on the spindle noses of lathes, collet chucks, etc. We also ground the mounting face and nose of our lathes after assembly to within 10 millionths of an inch true. Don't let the gloves, "three points...", and fancy video fool you. This guy is clueless. A little life lesson: You don't have to know everything. You just have to know where to find someone who does.
His method is technically fine since he is establishing a level plane using the jacks. Its not how I would do it, but if he were to flip the part and rest on that face and indicate underneath I bet he would get the same answer, faster.
Some times old ways are still the best . Don't forget optical flatness gauges as well . Then optical comparator to ensure concentsity and parallelism between surfaces are Absolute
We have some Renishaw Equator CMMs at work. They take longer to inspect the part machined then it does to actually machine it. Pretty crap when the cnc machine producing them needs to run constantly to hold size..
Very good tip!
This is extremely useful!
Thank you! :)
Great video!
Give us more manual please!
Looks like it would indeed show you flatness but would not tell you perpendicularness from that face and the face against the jacks correct ? Kyle F. Lead Machinist from Tulsa ok thanks
The problem I have with this method of checking flatness is that if one of the 3 points you set the initial plane is low (if the flatness happened to be way out in that one spot), then your whole plane is now bad and you don’t know about it. When I manually check flatness, I set the part on three points on the surface I’m checking flatness on, then sweep my indicator under the part. The points that it sits on give more surface area, compared to the indicator ball which is has an infinitely small contact point given its spherical shape, for the part to locate on. In this instance it wasn’t an issue due to the part being dished and the flatness essentially got worse as you got farther from the center in every direction, but it could be an issue on other parts. All this being said depending on your tolerance this method is probably always fine. Might be worth mentioning the reason for using 3 points and not 4 too. Just thoughts.
Source: my name
That's part of the reason why you're checking the rest of the surface though right, as if you choose one point, and move, and its off as soon as you move away, you immediately know that point is off. If it stays the same, as you move away, and is the same as the surrounding points, you don't have a point that's out.
So satisfying to watch as a machinist lol
oh, i get it, because it references against that three point plane
Dear sir, what should be tripod position angle ??
Do you have any GD&T explanation video?
Make a video on how to check parallelism with the indicator.
Thanks
it's pretty much the same but for parallelism you lean the part directly on the table
So the only difference is the three points?
I always thought you would indicate the bottom surface for flatnesses and the top for parallelism.
Thanks!
Pretty level guy !!😅😅😅
Love it!
Amazing! Thank you Travis. Will you be doing more old school gd&t measurements.
By the look of it. U guys machine the side with the hole first. Than held it , to do the taper. The part collapse n bent inward.that’s why the high spot is on the outter area.
Why wouldn't you set the the part on 3 1x2x3 blocks resting on the surface to be measured and come underneath the part and indicate?
Ease of access is mainly why I choose the top surface method and the 123 blocks may hide the "highest" or "lowest" spot thereby hiding the real value. Thanks for watching Michael.
or JUST USE feeler gauges and flat granite block DONE without wasting time . Indicator will not give a TRUE flatness reading it will give perpendicularity , feeler gauges and granite block never ever fails.
I see you used a sharpie on your part. Honorable mention if you are making parts for Boeing big no no. We had a major finding where I used to work because we’d write with sharpie marker “MRR” (Material review request) for non conformance parts, but every now and then the parts would be looked over by engineering and deemed conforming. With the sharpie though I guess it can do something to the metal later on down the line which etches it or something. So Boeing actually makes sharpie markers that are to be used with their parts to meet their spec which we found out during an audit. Just thought I’d mention maybe you guys are more familiar with it than I am. I forget the full reason or the actual spec it was under it was years ago now.
Halogens in the sharpie can degrade the oxide layer on stainless and lead to corrosion. Lots of people make special markers.
Thanks for raising a low spot in my metrology knowledge.
I wonder if you'd get less than .0033 flatness reading if you zero the high and low points...
For all of us regulars to the channel we need to do better for The TITANS of CNC Machining. They put out great content and educational videos daily for new and seasoned machinist. Please take a moment and support them by hitting the LIKE button. When I see over 12000 views and just over 550 likes I feel like we can do better. Thanks, Just a fan of the channel.
How do you calculate the Measurement errors? I am afraid there are many of them.
If the face your inspecting has more of a cupped or dish shape, would you then have too indicate from the tallest points of the piece? And would you also have to indicate the "roundness" of that sort of shape?
‘We’re going to check flatness the old school way’ As if everyone has a CMM in the back room.
Machinist stone and granite table have one thing in common - I have only seen these on TH-cam. XD
"Everything is crooked! Reality is poison!" This isn't true level.
why put the part on jacks
when the surface plate is already flat? doesn’t make sense to me since the height gauge is on the surface plate anyway
Get an IQ test and then comment. Read other comments ("isolating a plane" method explained).
@@GrandePunto8V stand by my comment