Newbie- Endeavour for a Alternative to Spindle Probes

[Ashish B]

Hi.

I am trying to develop a alternative to Spindle Probes used on Machining centre.
A spindle probe finds the centre of hole in Great Accuracy within Minutes. In the Attached Document, i have describe the Alternative & practical limitations to the Technique.

A spindle probe costs around lacs of Rupees. In the Technique method in the attachment, i have suggested a Zero Investment Plan.

Please have patience to be through the attachment & Imagine the Solution to the Problem.

Thanking You.

Ash

[Superman]

There is no problem

The error you talk about, does not exist.

The conact point of the dial indicator ( when the axis is adjusted to read zero ) will be ( when rotated 180° to touch the other side ) the same distace from the spindle centre line. Say the contact was 10° below the centreline on the right side, when touching the left, contact will be 10° below c-line.
In effect, the spindle c-line will be the same distance in radius fron the profile edge ( when setting an ID or an OD )

This is no different from using a blued up dowel in a drill-chuck

Anyway, a good operator will use that method to roughly set the centre,
then when set on centre, set the dial to check the bore on a full spin action checking and comparing his readings at the quadrant points. All this should only take him 2 minutes, tops. This is a skill taught to a 1st year apprentice, like sharpening a drill, this is what can stuff a good part if done incorrectly

Most of our machines have auto-gauging to set zeros when doing production jobs

[wildwestpat]

Hi Ash

I am replying in the thread rather than resonding to your PM as it may spark a debate.

With the use of a DTI the angle of approach is not very important as the aim of the centering exercise is to get zero deflection. The only provisor is that the sensing part of the plunger mechanism or lever has both free movement and is roughly in the middle of its travel. Angles of 45 degrees have little effect on the eventual result so I don't see the problem you are trying to solve.

Yes the encoder approach will refference the work to axis that can be moved. Again the precision with which the adjustment axis is lined up with the 'measurement' axis has only a second order effect as the aim is to get zero deflection. Yes it would be possible to calculate the required correction and there are electronic DTIs that would permit this to be done. However there is no reason why this should not be made an iterative proces that is repeated until no detectable difference exists.

Your problem is that of detecting the difference in offset of the work piece from the machine axis centre of rotation and the proposal to measure this off set is in my opinion going to make the equipment very expensive. The movement of the axis could be used to 'measure' the initial off sets and that measuremnt system would include all the inacuracys - lost motion backlash and resolution of the axis sensor. If the machine was equiped with super quality glass axis encoders then this would be possible. However in my view those using machines that have these high accuracy DROs (Digital Read OUT) in the axis would be likely to resort to refference planes and jigs to solve any centering problem and for super critical work would be able to fall back on the instrumentation of a proficient tool room setter.

For the above reasons it is my opinion that there is some merit in software that would work with a contact sensing probe to successively reduce the centering error. The process should then be repeated several times as the error reduces to the point where the accuracy of the machine is reached. The real life problem will be in getting the software to recognise that the machine imprecise movement limit has been reached. It might be psssible to 'box' the true centre by deliberately off setting the 'found' center in both directions and 'measuring the error - IF - correctly centered the error would be the same but of opposite sign. This is how a skilled operator uses the DTI when extra precision is required.



Hope this helps.

Regards

Pat

[sinha_nsit]

Originally Posted by Superman There is no problem The error you talk about, does not exist. The conact point of the dial indicator ( when the axis is adjusted to read zero ) will be ( when rotated 180° to touch the other side ) the same distace from the spindle centre line. Say the contact was 10° below the centreline on the right side, when touching the left, contact will be 10° below c-line. In effect, the spindle c-line will be the same distance in radius fron the profile edge ( when setting an ID or an OD )

I think your statement is correct only when the X-movement and the Y-movement are along the diameter of the hole. If not, the distance between the spindle centre and the profile edge will change when the spindle is rotated by 180 degree.

Ashish is correct when he says that some error would be involved because of non-zero angle of the lever. But making the angle 0 using T-slots is time consuming as it would involve a lot of trial and error.

Wildwestpat suggestion of an iterative method can be used:
1. Locate the centre by the suggested method.
2. Now place the spindle at the centre calculated in the previous step.
3. Go to step 1.
If the angle is not too different from 0 degree, and the process is started approximately at the centre line of the hole, I believe the calculation would converge in 3-4 iterations.

If there is a dial gauge with exactly vertical arm, like a touch probe, is available, there would be no such error.

Ashish should be congratulated for floating a good idea.

[wildwestpat]

Hi Ash

Can a suggest you look at conventional tooling companies offerings. For example probes for DTIs by the Japanese maker Mitutoyo are available that can measure almost any concievable hole or edge. Pages 197 to 200 list probles for use with their DTIs and similar ones are available from the other manufacturers. Here is the link to the Mitutoyo catalogue.



http://www2.mitutoyo.de/ebooks/engli...dex.html#/201/

The problem you are trying to solve is limited to the one-off shop where a job has to be resetup for some reason. People undertaking this sort of work will be well versed in the use of the DTI method and even if they don't have the relevant probe the cost is relatively small.

Regards

Pat

[sinha_nsit]

Is there a dial indicator which can accurately sense horizontal displacement while being exactly vertical?

[Charlie Gary]

Hi Ash,
As I looked at your pictures I thought of how I find hole centers when the indicator is not big enough to touch all sides. With large parts I find I have to rotate, or "sweep", the indicator at each position to get the highest reading on the dial. This eliminates errors in trying to align the needle to a specific angle.
You are correct when you say the skill of the machinist has a direct impact on set-up time. The only method I have found to correct this is to create a check list of steps needed to find the center of the part, and then have one-on-one training with those who can't do it in a timely manner. If the trainee questions the check list, first ask yourself if the question is valid. If so, the check list needs to be revised. If not, you have to find the best method to communicate with the trainee until he/she understands what needs to be done to complete the task in a profitable manner. Deviating from a proven check list should not be allowed.
Skilled labor cannot be replaced cheaply. You can invest a lot of time trying to create an automatic machine function, and you may succeed, or you can invest time in training the workers. Either way, you will have to invest something into the process.

[Ashish B]

Hi

Pat -
I really Admire your Technique about using a DTI probe.
But I am using a Lever type Dial Indicator & I relate problems in the Thread regarding the Lever Type Mechanism.
I have attached a Pic of Lever Type Dial Indicator.
Also Pat the Tag “Alternative to Spindle Probes”, is only mentioned because I used the Same pattern of Centre Finding. I don’t compare the Cost. What I suggested is a Zero Investment Plan.
I scare that u are not understanding me correctly. Also Your comments about to achieve the True centre of the Hole needs Retakes is what I Disagree.
As per my predication, we should be able to calculate the Centre within 5 to 8 microns & that too in the First Take itself. I refer to a hole very accurate in Ovality dimension. Generally we ream the hole which has an accuracy of 10 to 15 microns, but just for Instance if u try to align a Hole/Boss which is a Masterpiece, I bet we can achieve great accuracy, provided we Filter out the Previously mentioned Error.


Superman –
Your conception is very natural, but it is a Prejudice.
I was of your perpesective when this Idea Striked my Brain. To revalidate it I sketched a Geometry in CAD software. Please Refer attachment for further clarification.
Also to narrate you 1 story, when I used to work in a Public Limited Company. We had 3 machines & they used to make money day & Night. When I joined, it was a shock for me to believe that Operators used to switch off the machines (in night shift) because they were not able to Find the centre of Hole.
As in Night shift, mostly u don’t have a Managerial control so they had such freedom. At the Grass Root Level, operators are unskilled, uneducated & also we cannot expect them to be robotic in nature the whole shift. The clever way is “Filter the Process, so it is Beneficial to them as well as For us.
In ISO Standard & Environment, “Improve Process Capability “is the Correct Word for all these Issues.
To increase the Gravity of my Point, just Imagine that u have pyramid sort of profile mounted on Plate having 2 holes . As we all know, its not only Hole Alignment but also we have Inline Alignment. The worst case in this scenario is that the operator has to tram the Spindle Up-Side-Down on either sides to check the Inline alignment First & after achieving that , u turn to Hole Alignment.
I don’t criticize the Operator Tendency for Delay. Its Very Natural.
If I would have been In his place, I would behave the same way, they do. No owner can expect a employee to be 100 percent efficient the whole day.
“Human is Not A ROBOT”.


sinha nsit –
Hi,
Thanks for you appreciation for the Technique.
I know that Dialing the Dial Tip’s Side to the T slot is very Time Consuming
Can u suggest some other method of Filtering the Angle Error?
As u suggested, we should not take retakes for Computing the centre of Holes. If we endeavour to replace the Old method, we have to achieve Good Accuracy & also Reduce Time.
“Any new Technique runs successfully only when it walks on wheels of Good Accuracy & Reduced Time”.
The Tool Holder in which we clamp the dial (at correct angular position ) can be kept undisturbed or booked. I clarify it that it is not a Investment of Tool Holder but it’s a Booked. No one in the shop floor should remove that Dial, so whenever a new setting is there, just plug the Tool holder & derive centre in fractions of Time.
Also we Consume so much time aligning every Part, so Can’t we spare some time Once for Aligning the Dial to the correct position?.


Charlie Gary –
Hi,
Its nothing Depends on the Hole Size. As attached a Snap of Dial Indicators, u can Pickup Hole/Boss of any size by this Indicator.
Well as I told you before, human is not a robot. Some can find centre within seconds, some takes holes.
Also humans/operator gives up Sometime. Sometimes, operators are tried, they are frustrated because they are consuming more time than regularly required, the employee is shouting at him, he is physically present on the machine but he is thinking of some Personal Problems & the list continues to go on & on……………
Well I don’t want to express the Accuracy of Setup when the operators has so many things to deal. We all are humans & Not Robots.


Hope I clarified the Gravity of My Point & gave branches to Technical Discussion.
Thanks all for your time, Patience & Support.
Ashish B

[Superman]

Hang on,
You PMed me to say how I clock a hole or boss

now you tell me it is incorrect,........by drawings on a piece of paper.

After 30 years of "experience" of clocking holes, bores, blocks etc.

You are right, if operators are not trained in the proper choice and usage of tools and instuments, you will always get junk.
If an operator doesn't know how to use such a basic instrument, why are they on a machine, you may as well replace them with a robot.

How and with what do they measure the parts with ?
Surely understandig a dial indicator's reading is not that much different to a dial caliper let alone trying to understand a vernier scale

Unskilled operators should not have the responsiblities of setting up a part, it is the engineer and/or programmer to design a method that avoids the possiblitiy of making junk. An unskilled operator is the weakest link, not the dial indicator

[Ashish B]

No STEVE,

Don't Misunderstand me.

I just tried to rectify your Misinterpretation.
U said that the Lever Needle will be exact 180 degree apart. No doubt that u are correct, but the Intial Angular Error of Dial Preloaded into it, counts.

It is very much clear that we want the Centre of Hole & we get it with respect to Machine Coordinate Axis. To eanble us to do so, our Lever type dial should have a movement Parallael to the Machine Axis Movement, than & then only U can find the Centre of Hole Accurately & Very Accurately.

Ash.

[wildwestpat]

Hi Ash

If you look at the link I quoted you will see that the DTI is a generic name for any dial / electronic measuring device that uses a plunger. The manufacturers provide and most machine operators are well versed in adapting the instrument they have to measure and find the centre of holes as well as edges. The plunger type of DTI is easily adapted to most measuring situations because the tip of the plunger unscrews permitting various sensors to be attached. For example a wire paper clip can be bent with a loop to go under the ball-ended probe tip and be bent to be at a suitable angle (90 degrees) to let you measure orthogonally the true deflection of the hole wall. The same action being achieved with the lever type of probe.

To return to the pics of the problem you have posed. Whilst you may ream a hole to the accuracy you are quoting locating all the features of the work piece to the same degree of accuracy is a demanding task. The accuracy of all the 'features' on the work piece have both a size and a spacial tolerance one to another. Before you start to hypothesize on re-giging and measuring hole centres you must have a clear drawing that shows all the tolerances as well as nominal sizes and locations of all features. This is a basic design requirement. By dimensioning and working out the acceptable range of tolerances from the theoretical but unrealisable perfect form you will then have a finite objective on what accuracy is actually required. For good production practice the tolerance banding is kept as wide as possible taking into account all contributing sources of error.

It is only when you have done this that you can start to home in on the way the work can be machined. Experience will soon indicate what is achievable with the tools available and what the true cost is of making the tolerances far tighter than is necessary. Again a good text book will cover the theory of production tolerancing and the impact on cost including any scrap items where the tolerances have all gone the wrong way. This is highly statistical and not the subject for a debate!

If you are using the holes to locate the work it would be much better to use three holes. Factors to think about assuming you have worked out what tolerances are necessary would include:-

The circle of confusion in placing the 'reamed' holes you are about to use for re-locating the work to the machine axis.

The accuracy of the holes in size – are they all the same etc.

Use of three 'reamed' holes to mate with a suitable jig (Again jig design is a highly specialised.)

Accuracy of and jig pin fit.

Relocation accuracy of the jig bed plate on the machine table.

The accuracy of any measuring devices used.

The accuracy of the co-ordinates of the machine tool. Including back lash – lost motion of al types including tool flexing in addition to lead screw or DRO scale imperfections.

When you add up all these small and annoying contributions to the degrading of your perfect vision you will have established some views on how the work can be done or on how to simplify the requirement. The use of computer 3D images is no substitute for properly dimensioning and tolerance budgeting . The better CAD 3D packages include the ability to look at tolerances and part interchangeability as well as stresses and strains in the proposed parts. 2D drawings and first cut code are also only a button press or so away alowing either conventional manual manufacture of CNC.

Unfortunately there is no short cut to the process outlined above. Hope this helps as I get the feeling you need to get machining and get a feel of the real world and the necessity for having to work to achievable tolerances.

Regards

Pat

[Ashish B]

Hi Pat,

Can u Attach a sequence with pictures or Videos on how you use a DTI.
(also u can mail me at - ashish.bhosale@yahoo.co.in)

May be i am missing something. As u told it reduces the Time, i believe that it will take the Same Time.

I guess the Dial tip is like a wiggler, so the operator has to jog at the centre.

I apologize if i am wrong.

Also Coming to the Issue - IS HOW WE CAN FILTER THE ANGULAR ERROR ?

Is there any special Instrument like Boring Bar Presetter or CMM Machine. But still the error is that u don't know whether the locks ( provided on the tool holder to clamp in the spindle ) are parallel to the Machine Axis or not. SO EVEN IF WE ALIGN THE DIAL ON CMM MACHINE BY PICKING UP AN REFERENCE OF THE SIDE LOCKS, THAN TOO HOW WE CAN BE ASSURE.

Also i guess we need to talk to the Machine Manufacturer fo the Same.

I have attached a snap to identify the Side LOCKS.

ASH

Thanks for your Time, Support & Patience.

Still Needs a Sound Correction.