How do you measure rocker arms

[peppy]

I am trying to measure the rocker arms. I am using Wayne's book, and the spec for the shaft is 17.992-18.000 mine measure 19.810.

The spec on the bore is 18.016-18.027 and mine measure 20.060.

Am I measuring this wrong or is this correct for an 88 3.2?

[cgarr]

The bore without the bushings is going to be 20mm did you take the bushings out? Check your mic, looks like your about 2mm off on each.

[peppy]

Originally Posted by cgarr

The bore without the bushings is going to be 20mm did you take the bushings out? Check your mic, looks like your about 2mm off on each.

The bushing is still in.

I am not using a mic, I was using a dial caliper. Is that my problem?

[cgarr]

I mic the shafts and have a small bore gauge for the rockers, a dial caliper will get you withing a thou but your off a lot more, almost 80 thou or 2mm for some reason. Can you feel a lot of play in the shaft and rocker?

[peppy]

Graig, there is no discernible play when the shaft is in the rocker. Maybe I should let the machine shop measure them.

[304065]

Dial calipers do not have the repeatability or resolution necessary for that measurement.

It is a lot cheaper to take it to a machine shop.

However, if you like doing things yourself (I do) then the way to do it is with pin gages in a go/not-go setup. You will need two, one 18.027 (the high end of the tolerance) and another one that represents the wear limit, say 18.05mm. Pin gages come in a variety of types depending on how accurate you want to be-- class X, class Z-- the difference is the tolerance and smoothness in microinches.

For a NOT go gage, which is something too big that you don't want to fit, there is a tolerance that is negative, e.g. the pin can be up to that tolerance smaller than the actual dimension. Putting the tolerance on the negative side assures you that the pin might be too small, but not too big, ensuring that if the gage fits, your rocker bushing is at least below the limit.

(Of COURSE some six sigma fanatics are going to try and take me to town about statistical process controls, how any sample around a limit will exhibit variance, there's no such thing as a one-sided tolerance, etc.-- I know all that, I'm trying not to confuse the gentleman so let me close out this parenthetical and get back to work)

Anyway, McMaster Carr sells these inexpensively. To order you specify the exact size to 0.01mm measurement and whether you want a go (plus tolerance) or not-go (minus tolerance). So an 18.02 for the NOT and an 18.01 for the GO should be good enough, of course you could get the exact three-decimal ones if you wanted, but that requires calling a gage shop.

Good luck!

[MatthewBrum]

I think that your calipers are probably wrong or not zero'ed out. Make sure that the calipers are clean and read zero when closed. However you should be using a bore gauge and a micrometer for this measurement, or as John stated a go/no-go pin gauge. Just remeber you can not measure outside diameters with nothing but a micrometer.

You can buy a decent set of bore gauges and american micrometers (0-4") for about $200 USD. Both the micrometer and bore gauges take some time and training to use properly. If you only need to measure the rocker shafts and rocker bushing bore then you may not want to make the investment.

[peppy]

Thanks for all the info.

Best,

[cgarr]

I never had good luck using gage pins in out of round holes.

[304065]

Craig,

I see what you mean, if the hole is elliptical, the not-go gauge might contact the narrow sides of the bushing and trick you into thinking the bore is in spec, when in fact the other end is too big.

I suppose a bore gage could be used with the appropriate small head and it could be rotated inside the bushing to give you a picture of the dimensions and the ovality.

Using the bore gage with a micrometer on the stand is not easy for the beginner. Some do it with predictable regularity-- I would prefer something like an intrimik or mitutoyo holtest ($,$$$) for the exact size-- bue even those can't tell you if it's oval.