Totally unacceptable. Return it and get a refund. 0.15mm would be just about ok.

Exact and more details are important, terminology is as well. Since you mentioned checking with an indicator I suspect you mean run out and not play as each has completely different meanings. But if you mean play as in the arbor is still loose? As Diogenes asked, what spindle taper is the machine? And 1.5 mm / .060" shows something is very wrong. Unusable in fact. At most it should be around .025 mm / .001" or less for a decently ground arbor. Either you have bought the incorrect arbor taper for it, or on something like a R8 taper machine, possibly the spindles internal key isn't engaged with the side slot in the arbor. If it is R8? The arbors front taper should be pulled up fully within the spindle taper and almost flush with the end of the spindle. About all you should then see is the bottom taper that accepts the drill chuck taper below the spindle nose. That can only be done if the side slot is indexed properly to engage with that internal key.

Earlier posts seem to suggest INT40 in a Beaver Mill

Each small division round the large scale is 0.01 mm as is printed across the dial. So that's 0.15mm TIR which isn't bad though not wonderful.

That is the trouble with thought, it can mislead you. Best not to think but to test the indicator with something of a known dimension (thickness) and see how much the indicator moves. For instance, a search could tell you the thickness of every coin that is legal tender in the UK. Passing one of each of those under the indicator would teach you a lot.

Given how badly thought can divert you from truth, 'look', as you have done with the arbor face, is at least an order of magnitude worse. Now you know what each of the little lines on the indicator mean, poke it up inside the spindle, even if it is at an angle, and actually measure if there is any run out in the spindle itself.

You have paid money for the tool. Why 'look' when you can recoup some of that money?

All you can say for sure is that there is run out on the arbor. You have not yet identified or isolated the source of that runout. You will soon become bored, as will your supplier, if you return a few arbors as faulty and then subsequently discover there is an issue with the milling machine spindle.

The new arbour has probably got centre holes at each end. Is there any way for you to set it up between centres, such as in a lathe, and check if it is bent? I ask this as the usual reason for changing arbors when I was at work is because someone swung a large radial drill around and clouted a vice with a drill chuck. The neck of the arbor was the weak spot that bent as a result. I would assume a new arbor without any obvious bruising is not going to be bent and that the spindle is likely the source of your runout, especially since you have tried different positions and got the same result.

The other thing to do is when you measure the runout on the end of the arbor mark the high spot on the arbor and the spindle Then see if it moves with the arbor when the arbor is tried in a different position or stays in the same place on the spindle.

Martin C

This is what I would do first. Turn a centre in your lathe chuck and then put the new arbor between that and the tailstock centre. Put a dial gauge on the wide part of the taper and turn it by hand between the centres. If there is runout the part is bad, if none then the spindle needs looking at.

I advise against starting with high accuracy measuring. Trouble is it's skilled work that's easy to get wrong. Dozens of booby traps. One is not making sure the instruments are OK before starting, check that dial and learn how to use it!

As unpractised measuring applied randomly with dodgy kit is dangerously misleading, I suggest better to start simply by cutting metal with the machine. If it cuts OK, stop worrying. Measure the work it produces, not the machine. Distinctly ratty tools can still produce good results. Takes longer because the operator has to compensate and double-check, but quite severe backlash and bearing wear can be tolerated.

Only take measurements if the machine doesn't perform adequately, and remember measurement is only one of several ways of pinning down problems.

Best engineering advice ever: if it ain't bust don't fix it.

Dave