Lever type dial gauge - Usage of a dimensional 'test indicator' for measurement

[mister - 2004]

I have a question regarding the usage of test indicator for measurment.I already know the answer,but would like some feed back from other metrologists.I work for a small company in canada,we make mold parts mostly for Husky.the usage of test indicators is rampant here at this facilty.tolerances are .005 microns total or .0002 inch.remember I said test indicator not dial indicator.Most employees use these the wrong way ,but the owners of the company do not realize how small 5 microns is(they are also managing the place)cosine error is rampant but i can't seem to get thru to them.Angle correction factor for the tip is never implemented.Most are measuring the underside of a shoulder 10mm +.005 with the tip almost vertical.Any articles or info that anyone could share would be great.Our
PPM for the month of June is 240,000 ,rolling average 75,000 over 3 month.
thankx for any input or links.Mister

 

[Marc]

Please define 'Test Indicator'. Are you referring to a Go-NoGo gage?

 

[Cari Spears]

...the usage of test indicators is rampant here at this facilty.tolerances are .005 microns total or .0002 inch.remember I said test indicator not dial indicator...

...Angle correction factor for the tip is never implemented.Most are measuring the underside of a shoulder 10mm +.005 with the tip almost vertical.

Welcome to the cove!

I'm not sure what the exact problem is that you're describing. The test indicators we use for measurement have resolutions of .001", .0005", and .0001". The usage of them is "rampant" here too, as with most places I've worked, so I guess I don't understand how that's a problem.

The last thing I'm confused about is that, as with dial indicators, test indicators are comparison measurement. If the angle of the tip stays the same when you touch on the shoulder as it was when you zero'd on the reference surface - there shouldn't be a need for angle correction factor - should there? Although I agree that they should not be nearly vertical because I would worry about preload and the movement of the tip on the surface being impeded, I don't recall ever seeing an exact angle called out anywhere - I'll dig out my "Metrology 101" text book when I get home tonight.

In my experience - the most common mistake made by users is failing to zero the indicator with enough pre-load.

Mind you - I wouldn't consider myself a metrologist, but I was a floor inspector for years prior to becoming QA Mgr and did a lot of surface plate layout work - mostly with height gages with test indicators and scribers.

 

[Govind]

What Mister is referring is "lever type dial gauge". The position of lever nearly vertical is unacceptable.In this position, one could damage the precision bearing where the lever is mounted.

This can cause significant Cosine error if the lever is at approx 60 deg to the surface.Some Lever dial gauge manufacturers provide a sheet with Cosine Error for various angles along with the product.We use to recieve for TESATEST lever type dial gauges. Keep in mind, this holds good for that particular length of the lever only.

The error will be in the measurement, only when you read on the dial. If the gauge is fitted to the holder of a height gauge and the with the indicator preloaded, Zero at the start (both dial gauge and height gauge) and Zero at the end (only dial gauge) thereby reading the parallelism error in the Height gauge. This way Cosine Error can be eliminated.

I should acknowledge the other member ( Cari Spears) who replied prior to me also meant the same approach.

Another common error using lever dial gauge measurement is not tightening the lever properly. The pointer can losely wobble causing additional errors. Also look out for the wearout of the ball pointer in a Profile projector with good magnification.

Note:
Mister, You are providing data like shoulder dimension..It will be difficult to imagine a part without looking at it. You may want to attach a jpeg picture explaining where exactly the surface being dialed, tolerance,etc.

Also just mentioning ppm, we will not be able to assume that all your defectives are due the parallelism error or another dimension.I guess we have clarified your question on Cosine Error. But if you need more help, you have to provide us with adequate information. Iam attaching a picture of lever dial gauge to make sure that we both talking about the same subject.

Govind.

 

[copternadley]

I have a question regarding the usage of test indicator for measurment.I already know the answer,but would like some feed back from other metrologists.I work for a small company in canada,we make mold parts mostly for Husky.the usage of test indicators is rampant here at this facilty.tolerances are .005 microns total or .0002 inch.remember I said test indicator not dial indicator.Most employees use these the wrong way ,but the owners of the company do not realize how small 5 microns is(they are also managing the place)cosine error is rampant but i can't seem to get thru to them.Angle correction factor for the tip is never implemented.Most are measuring the underside of a shoulder 10mm +.005 with the tip almost vertical.Any articles or info that anyone could share would be great.Our
PPM for the month of June is 240,000 ,rolling average 75,000 over 3 month.
thankx for any input or links.Mister

I should work very well. Maybe there's something wrong with the mechanism.

 

[Umang Vidyarthi]

I have a question regarding the usage of test indicator for measurment.I already know the answer,but would like some feed back from other metrologists.I work for a small company in canada,we make mold parts mostly for Husky.the usage of test indicators is rampant here at this facilty.tolerances are .005 microns total or .0002 inch.remember I said test indicator not dial indicator.Most employees use these the wrong way ,but the owners of the company do not realize how small 5 microns is(they are also managing the place)cosine error is rampant but i can't seem to get thru to them.Angle correction factor for the tip is never implemented.Most are measuring the underside of a shoulder 10mm +.005 with the tip almost vertical.Any articles or info that anyone could share would be great.Our
PPM for the month of June is 240,000 ,rolling average 75,000 over 3 month.
thankx for any input or links.Mister

Hello Mister, Welcome to the Cove

A very good first post.

DTIs (Dial Test Indicator) probe swings in an arc and touches the workpiece at an angle. If the probe is not parrallel nor at tangent to the examined piece, then there will be 'Cosine error'. (Contrary to popular misconception, the cosine error is also present, when the button of a normal 'Dial Indicator' is not perpendicular to the examined surface). Suppose DTI's probe is at 50* and the reading is, say 0.005 then the correct reading after correction would be 0.005X0.64 = 0.0032!! The difference between right and wrong reading gets wider with the angle.

I agree with you that he problem of cosine error is rampant at majority of facilities (yours is not an exception), despite providing enough training to the workers. Those in the know, ignore it considering the tedious correction procedure, while a majority is ignorant of the problem.:mg:

The solution lies in using DTIs with built in cosine correction. There are DTIs having the probe at an angle. Another type comes with pear/flame shape probe in stead of the ball. The cosine error gets corrected automatically for the given degrees of angle.

Hope this helps

Umang

 

[Nyioh]

Hi Umang and Govind,

Both of your explanation is really helpful
take this opportunity to ask some similar question

We have been used dial test indicator to measure polyester embossed height.and there are few challenge regarding the measuring result:

1. we found that the result will fluctuate every time we re-measure it, we believe that it is due to the angle of the level. but it is not possible to have the angle exactly same every measurement.

2. for Umang reply on above discussion regarding DTI with build in cosine correction can you share the product link? thousand thanks

3. Another issue is regarding the parts to be measured, because we are measured on embossed polyester film so do you think the result may be affected by the force of the level which press on the polyester film. Do you think is there any alternative way to measure the embossed height?

Up front Thanks for all precious and valuable idea.

Regards
Nyioh

 

[kgriff]

You might also think about looking at a company called Dorsey Metrology. They make a dial test indicator that, due to its' design, eliminates cosine error. At the company where I work, we use Dorsey indicators in critical applications where any cosine error is unacceptable.
Be aware, however, that it has been my experience that Dorsey indicators are not as robust as traditional test indicators.

 

[Nyioh]

Hi All User of Dial Test Indicator,

kindly look at google for mitutoyo dial gauge indicator manual (sorry can do attachment here)


kindly refer to page 23 for the correction factor for every measurement angle.

But really do not feel this method is so practical because:
1. the measurement angle is not easy to measure and control
2. the measurement angle is not only effect by the level but also the magnetic flexible arm stand.

Hope some one can prove that all above idea was wrong... Please...

Deeply appreciated

Nyioh

 

[Shankara]

Thank you all for sharing valuable information. What is the root cause for cosine error?
and can any one please tell me about physical and mathematical analysis of cosine error?

Warmest regards and thanks in advance
shankara