It could be voltage induced in the towel bar surface (electricians say phantom voltage). But a finite number of milliamperes will flow if you "short" the metal towel bar to the nearby metal faucet. It is not unusual to feel a power line frequency vibration when rubbing your hand on the bar while your bare feet are flat on the floor.
High impedance voltmeters (most digital models) are very apt to show a noticeable voltage reading from an object to ground from induced voltage.
Now if the number of ma that would flow if you shorted the bar to the tap is high, say, more then ten, we have excessive leakage, enough to count as a genuine fault from a live part inside to the metal bar. That is a defect.
Connect a small wattage, say, 7 watt, incandescent lamp between the bar and the tap. If it glows, not necessarily brightly, or trips a ground fault current interrupter then the current leakage is great enough to be a fault.
If the lamp does not glow then you can use the in-line amperes function of a multimeter to measure the number of milliamperes. The subject subcircuit is between the hot feed to the towel bar and the faucet and the load is the dielectric (insulating material between the energized part(s) inside and the exposed bar.
When you do just the meter measurement, you create a subcircuit with two loads in series: the dielectric, and the meter itself. The larger the impedance of the meter compared with the impedance of the dielectric (at power line frequency), the larger the voltage reading on the meter, up to the supply voltage.
(If you did not do the lamp test and there was a real fault, the resulting current could blow out the meter during an in line amperes test.)
More than about 5 ma should trip a GFCI (RCD).
High impedance voltmeters (most digital models) are very apt to show a noticeable voltage reading from an object to ground from induced voltage.
Now if the number of ma that would flow if you shorted the bar to the tap is high, say, more then ten, we have excessive leakage, enough to count as a genuine fault from a live part inside to the metal bar. That is a defect.
Connect a small wattage, say, 7 watt, incandescent lamp between the bar and the tap. If it glows, not necessarily brightly, or trips a ground fault current interrupter then the current leakage is great enough to be a fault.
If the lamp does not glow then you can use the in-line amperes function of a multimeter to measure the number of milliamperes. The subject subcircuit is between the hot feed to the towel bar and the faucet and the load is the dielectric (insulating material between the energized part(s) inside and the exposed bar.
When you do just the meter measurement, you create a subcircuit with two loads in series: the dielectric, and the meter itself. The larger the impedance of the meter compared with the impedance of the dielectric (at power line frequency), the larger the voltage reading on the meter, up to the supply voltage.
(If you did not do the lamp test and there was a real fault, the resulting current could blow out the meter during an in line amperes test.)
More than about 5 ma should trip a GFCI (RCD).