Jim, You just mentioned a critical factor. They are not assembled. The proper mesh of the jaws depends on face to face mating at the rivet about as much as warpage. This is probably key to inspecting the part. Make a fixture which clamps the part down ONLY IN THE RIVET ZONE, to a precision flat or short cylinder.
Then, use one or more indicators along the edge to indicate deviation from a straight line. If they are warped, you might need three points to inspect: one close to the rivet area, one halfway out and one toward the end. If you need warp measurement, and you need electronic output, it sounds like these need to be measured on teh fly during an operation. So any fancy hand clamping etc is out. If teh parts are presented in a reliable orientation and location within the given limits.
Then I can think of one possible pretty accurate way. If they are randomly oriented, or the 'misposition noise' is bigger than teh measurement, I'm out of ideas. But, since you need to measure warp, one measurement isn't good enough. You need at least two, and probably three. My idea is air flow, like a fancy bore gage. Three 'flow heads' along the part need to have equal flows within a certain limit when the piece is presented closely to them. It should be possible to do with photosensing of the 'dancer' position in the type of flowmeter that has a small ball or other 'dancer' moving in a diverging tube.
Height position depends on flow. Warps would lead to differences in flow. And, the nice thing is that it should find twist also. The proximity detectors are good, but I don't know that ones in that sensitivity range are very cheap. And, the output probably requires more processing than wiring up a series of photodetectors to see if the shadow is in teh right area. The photodetectors on the flow meter should get you a fairly absolute go-nogo if the part is presented reliably in one position, as on a carousel machine.
Of course a better flowmeter woudl so it as well. The rivet area and the blade must be in the same plane. The handle part doesn't matter much. If you are going to build a fixture to measure and straighten these, then you'd want the measurement device to be integral with the fixture. If you arranged to clamp the blade at the rivet hole area, with the blade part straight up, you could set up a dial indicator to show what straight up should actually be. You'd be sensing near the tip of the blade.
Nov 5, 1999 - indicator will light up to indicate that the motor can be turned on. LOAD INDICATOR. The RDP E725 digital indicator, in combination with two LVDT transducers, can measure the gap between rollers in precision manufacturing processes. The indicator can be set to display individual readings.
A grooved tool could be slipped sideways onto the blade with it's handle on the opposite side from the indicator, and pressured up or down to apply a corrective force in the right direction, with instant readout results on the indicator. I think you need to arrange to fix the indicator to a solid metal block, to which you can quickly and securely clamp a blade for testing and correction. Put a pin on the side of the block (which might be held in a vise) and locate the blade's rivet hole on that pin. Hinge another solid block up to the blade and clamp somehow to sandwich the blade. This clamping should extend to the edge of the metal surrounding the rivet hole. This way you won't be bending the blade at the rivet hole. This must be rigid to resist the forces you're applying to correct a blade.
You won't be hammering on anything, so any indicator in touch with the blade will be safe from damage. A 1 inch dial indicator would do just fine.
I think the hardest part of this is the arrangement you make to clamp the blade against the side of the block with the indicator on it. I imagine this should be fast for the sake of production.
For the actual bending, a slot cut into the side of a square bar near one end should make a suitable tool. To zero the indicator, a piece of known straight material would first be clamped into this jig. After calibration, the fun begins.
How about this? Have a slot that the part is slipped into with the side to be measured facing down.
The bottom of the slot is the flat reference surface. There's a light to the left and a photo-voltaic sensor to the right of the part. A gentle spring on top nudges the part toward the reference surface. Have suitable 'seals' to block the light from going anywhere but through the gap between the part and the reference surface. A microswitch turns the light on when the part is all the way in. The sensor reads the amount of light passing through the gap.
![Manuals Manuals](http://s2.studylib.net/store/data/005419843_1-5566460bb5ee857e7b30c4fd24c13b0e-260x520.png)
If the reading is greater than X, the part is out of spec. The same arrangement could be used for checking the part with air flowing between the reference surface and the part, and measuring the pressure differential from one side to the other. Thanks guys, I'm gonna shower and go to work armed with your ideas. Just a follow up.
The parts should be fairly clean. I would like to test the parts before any expensive grinding or assembly is done.
Identify the warped parts and either fix em or scrap em at that point since they are destine for the dumpster anyway. I'm thinking the grinder operator could place the blanks into a slotted block with the end to be measured sticking out. A known flate piece would be used to qualify how a flat part would measure. I'm familiar with Keyence and have their cataloge out in my shop. Everyone's suggestions are being considered, appreciate the input.
If I come up something, I suppose they will want me to cancel my duck hunting trip for this weekend to build the thing. Geesh.I should have retired 2 months ago. Thanks again, Jim.
Basically you need to make a fixture that locates the parts on a three point mount, aligned along one axis that you can predetermine. So basically it's a fixture with three ball points facing up, and two dowel pins to locate acoss one edge. To measure 'flatness' you need to either check the upper surface along its entire length, or maybe at three spots depending on how the error is happening. I would say, 'operator loads part in fixture, and clamps. Then operator lowers dial gage to one end of part, and hits 'enter' key on the PC. He then sweeps gage (using linear rail or long raduis arm) along the length of the part, and hits enter again on PC.' You now have a full profile of height measured within a half thou for that part, and it can be compared with limits stored to light up the red light, or the green light.
Build a functional grade once and do not use any dial gauges at all if possible.but how to build it?? Toolmakers, any ideas? Am I way off base here? Ummm.could you just build a functional gauge (or set of gauges) that is shaped as the maximum permissible envelope of a part that is acceptable? I realize this might pass the thinnest part with out-of spec flatness.I mean make a very precise slot that the QA guy sticks the part in, and if it is flat enough, it will pass? If you can stick it all the way in, electrical contact is made?
OK, how about have the first functional gauge measure thickness at a single point, then have that information sent to a second gauge that is the maximal envelope for flatness of that part.gee, that's complicated.it would be really cool if you could do it all in one fell swoop. I lilke J Tiers idea of an air gauge. Make a block with air holes each connected to separate lines with ball indicaators. Probably use 4 different ports. Lay the arm on the fixture, hold with minimum pressure The balls would raise up in the tube relative to the air leakage or lack of in their separate paths. Supporting the part at the hinge point is probably right but with just.003 too much clamp force would distort the part enough to give a good reading. I'll email you with a few sketches.
What I'm seeing is an integrated clamp/test fixture. Basically, take two flat plates of steel. Attach a hard plastic sheet that is.003' thick to the 'insides' of them. So it'd go Plate, plastic, plastic, Plate.
Fixture the top and bottom plate-aluminum things to clamp. Add an electrical spring contact behind the clamping area on the handle.
In operation, the piece goes between the two plates. Pneumatic cylinder presses the plates together, so that the handle is clamped by the.003' plastic on both sides.
This way the thickness variation doesn't matter. When the plates come together, they'll be floating on the thickness of the rivet section. Now just check for continuity - if the blade is warped 'upwards'.003 or more, it'll contact the top plate and you'll get continuity.
![Emachine e725 manual Emachine e725 manual](http://www.rdpe.com/images/disp/wp-act.jpg)
If its warped down more than.003 it'll contact the bottom plate and you get continuity in the second circuit. If neither, the blade is good. I'd probably do the fixturing on a rotary indexer. Jim: We are currently building equipment that scans and provides part profiles down to 1 micron (0.00004') We went the laser route - Keyence makes some nice stuff and the Keyence folks in your neck of the woods are pretty good. The laser we are using resolves down to 0.01 microns, but our platform stability and scan rate won't allow us to get meaningful info below 1 micron.
Another way to do this is as a conductivity test. Ground the stamping and set two insulated scribes up 0.003 and 0.000 from the base and hook up the tips to a digital counter. Slide the scribe from stem to stern and any count above zero indicates bow.
A third way is to use an LVDT. RDP Electrosense had a nice control and decent probes. We have used them for years for this sort of thing with great results. A D500 series probe and an E725 readout would probably work well. Sony has a unit as well (J&L or MSC) that would work.
The lasers, counters or the digital gages can be had with PLC compatible outputs. On a side note, I am delighted to see a major US manufacturer paying this sort of attention to detail. On this current job I am working on, we have had about six shipments rejected for various defects in assembly and quality.:mad: Then again, we really like Ideal's wire strippers. (And if that's not the Ideal logo, well, sorry about that but we do like 'em!).
. process (426). temperature (167) ambient temperature freeze. current (120) DC AC current loop. position (105) rudder angle.
voltage (104) DC voltage AC voltage high-voltage. universal (68). status (38). frequency (35). speed (31) wind speed. power (21).
load (15). ground fault (12). humidity (11) relative humidity.
flow (11) mass flow. short-circuit (10). rate (8). torque (7). discharge (6).
pressure (6). force (5).
phase sequence (5). fieldbus (3). tilt (2). noise (1).
not specified (18).