Measuring Cone

I think I would shadow graph it and take a high def picture,with a reference piece print it then measure that.

Use an Optical comparator - Wikipedia, the free encyclopedia

Now can you please explain me how to use an optical comparator to measure the cone?

I thought the Wikipedia article called for little further comment.

If you have to ask, I doubt that I can.

graduations on the screen, being superimposed over the silhouette, allow the viewer to measure, as if a clear ruler were laid over the image.

1. You need to design a custom gauge, with one probe tip (or two at 90°) of the ideal dimensions, air discharge type.
2. Mount that on a rigid bracket that clamps around the shaft (or whatever the section A-A is) to ensure repeatable positioning.
3. You make two measurings, one with the smallest and one with the biggest acceptable internal diameter/depth, to establish your tolerance range.
4. Scale the analog signal from your gauge to convert it to engineering units.
5. Investigate what the hell an air discharge gauge is, I don't know the correct term.

Regards

How can he make sure the gauge is correct?

It would be another drawing with the same problem.

How much effort or cost do you need to put into this? The positional tolerance is pretty loose at Ø0.25 (mm I guess) & can be relaxed at MMC so you don't need to measure this too accurately.

I have used both of the methods mentioned above. IMHO the optical compartor or shadow graph would probably be the simplest. However this would require you to make a mold of the hole and then measure it on a shadow graph. This is good if you are only required to measure a few parts or only running these parts once. Reason being it is cheap and most machine shops have a shadow graph laying around. However if you are measuring a large number of parts or you will be running them repeatedly the checking fixture is the way to go, but they are usually pretty expensive to make and it will need to be calibrated on a regular basis. Hope this helps.

First, a few comments on this drawing

1. The .250 dia, the 118° conical angle, and the 9.070 distance are conflicting. Either the angle and diameter, the angle and depth, or the depth and diameter are sufficient to define the feature. one of these features should be a REFerence dimension.
2. Based on the feature proportions in the drawing there are mixed units. It appears that the diameter of the cylindrical section is 20 mm and the diameter of the conical hole is .250 inches
3. A standard drill bit will not produce the sharp point shown in this drawing. There will be a small flat truncating the cone. So does the 9.070 dimension reference the actual depth, or the theoretical sharp point?

Before I would waste a lot of time and money on inspecting this, I would go back to the designer and find out what the intent of this feature was, and what dimensions are critical to its intended function.
I am just guessing here, but from the full part drawing, this is some kind of indexing fixture. The features in question are detentes for a ball plunger to seat in. If this is the case, I would use a standard size ball (such as a bearing ball). to establish the angular relation to the datum features and depth. I would treat the 118° angle as a nominal dimension, that is whatever a standard 118° drill cuts.

The 0.25 is in MM the position tolerance of the cone with respect to its theoretical - perfect radial- position. It means that the cone center line can be in a cylinder of 0.25 MM centered on the perfect radial position. It is thus no conflict. If some dimensions are in MM it is VERY unprobable that the tolerances could be in inches.

I stand corrected. The .250 mm is a location tolerance specification.

You have issues with your supplier and their designers. This appears to be an example of designers getting too happy trying to dimension everything they can think of, in multiple ways, and not being concise, selective and smart about what they are trying to say. It's "bubblehead" dimensioning.

I work with one of those. Very hard to pin him down on what he actually wants. I have to work it out for myself by asking what the assembly does etc. I like the idea of putting a steel ball in the hole and then measuring with a micrometer across the shaft and ball. This method could lend itself to a 'go - no go' gauge.

Jim

Also makes the part harder to make, harder to inspect and more expensive.

Optical comparator.

Becherc caught what you and I did not.

The original drawing was a cross section of a solid hub with two "dimples". So there's no way either a shadow graph or an optical comparator will measure this.

Look at the second, complete drawing and you will see that.

Without more information, I'm done.

I'm wondering if a CMM would do it?

Since I believe that the whole point of GD&T is to allow the mating of parts that might fail an inspection dased on coordinate dimensioning, just build a go/no-go gauge.

Could you describe the gauge more in details ?

A device with an ID at the maximum material condition and one at the minimum material condition.

Another with sliding pins to measure the position depth and taper of the "cones" at the two conditions.

How do you measure the taper? In such a way you obtain a combined error value but you do not know what is what. The radial displacement is the sum of radial position deviation, center line eccentricity and angular error. Quite difficult to know how big is every one in order to correct manufacturing procedure and fixtures. If you make only one measurement then to those adds the cone angle error as effect.

I very much doubt that it could be a good solution.

We shall not forget that control is in fact NOT to say "this is scrap" but to obtain informations in order to correct the process.

The solution I suggested does accept a combined error as well since it does not quantify the cone angle error and does not measure the taper.

Depending on the criticality of the "cone" feature, the angle may not be that important.

If it is indeed a set screw seating feature, which we don't know, it MAY not matter.

We just don't have adequate information!!

I agree with you but you brought the taper into discussion and did not anwer my question. What if the taper is important ?

I think a gauge will still work. One for minimum angle; one for maximum angle.

Both gauges fit, good taper.

Could use the mike across both shaft and ball technique with two different sized balls. One ball that fits close to the bottom and one that fits close to the top. Simple geometry or trig will give the angle, knowing the diameter of the balls.

Jim

Yep, i missed it (senior moment). Yes a CMM would do it. The newer motorized CMM is just a simple program, allows rapid inspection of the same part on a production basis.

A CMM will easily measure the location of the drill point however the size and angle may have to be measured with a different method. You may also need a fixture or at least a reference mark in the piece to ensure the orientaion is at least close to the same every time. The reason I bring this up is the CMM is going to need manual hits taken off both the O.D. of the piece and one of the drill points. Even with all that it may be hard to get consistancy in your measurements.

I agree needed 3 contact of the outer diameter (if shaft center line parallel to one of the reference axis) + 2 contact in the cones.

First 3 points give shaft center line and diameter. The lines between computed center and the two cone contacts give the angles. The cone position can be determined by the length between cone contacts and computed center.

If the shaft is not set parallel to an axis then it should be measured in 2 planes to get the center line axial direction knowing it cones axial positions can also be computed.

Results were not consistent because the measuring program was not thought the right way.

For a prototype or a small series it is a good solution, for series it is too time consuming and thus too expensive. If only a statistical control is made it could be OK.

This is ONLY a suggestion how the control can be done provided the series are consistent and the fixture can be made.The principle is to check the CONE position using a ball. For the given dimensions the maximal diameter can be 5.49 mm but it has an edge contact. It is better to have flank contact so that a 1/8" ball would be more acceptable (it is also a standard dimensions for bearing balls).The fixture has a guide as shown and 2 "sensors". The balls are on supports sliding radially (their position is directly related to the cone position in radial direction ) and on a 2 degrees compliant beam which by means to be chosen (strain gauges or micro distance sensors) measure ball transverse displacements in 2 directions (axially and in the plane normal to axis). The device has to be calibrated but has the advantage to give in ONE operation all position errors with respect to theoretical ideal position.If my sketch is not clear enough I can give further explanations. On my opinion it is one of the safe solutions and apparently the simplest to put to work.

Since it is not a "modern" concept using optics or other advanced technologies I expect a lot of negative comments. I am ready to read them with great attention.

Nick

I'll give you a GA for your elaboration of my initial idea of using a bearing ball seated in the cone Note that with a little geometric calculation (done by a cadd program) the depth of the cone can be measured by "miking" over the cylinder plus the ball.

Also, if this is a spindle for an indexing fixture, as I guessed, the mating part of the fixture could be used to check the angular relationship of the detentes to the rectangular feature on the spindle.