Higher end multimeters feature "True RMS" rather than "averaging" voltage measurement. Is this applicable only for AC voltage measurement?
Usually only AC ranges, which are AC coupled so you lose the DC component of the signal. The reason why you would not want the true-RMS to apply to the DC voltage range is that the accuracy of true-RMS measurements is usually not nearly as good as averaging DC measurements.
For example, the Fluke 170 series datasheet here, shows these specifications:
The True-RMS AC measurements are almost two orders of magnitude less accurate than the DC measurements.
A bit higher end is the Agilent 34401A
You can see that the accuracy is more than an order of magnitude worse for the True-RMS AC range.
Practically defining the high end, the Agilent 3458 offers several AC methods of measuring RMS AC. Even the best (and most restricted) True RMS AC measurement technique is more than two orders of magnitude less accuracy than the DC ranges (a few ppm).
Here's a one-chip RMS converter, the LTC1966
It depends on the multimeter. For DC measurements, true RMS is irrelevant since the RMS of a pure DC voltage is equal to the DC value. Most multimeters are AC coupled so if a signal with both DC and AC is applied, the multimeter will only read the RMS value of the AC part of the signal. Even then the measurement is limited by the low frequency response of the multimeter which is usually in the 10 Hz region (it is, of course. also limited by its high frequency response which can range from the low kHz to several MHz for typical multimeters). If the multimeter is DC coupled on its true RMS ranges then it will read the RMS value of the signal including the DC component. However for most applications, this measurement is usually not of interest which is why most meters are AC coupled on their AC ranges.
"True RMS" denotes a meter that really is measuring the RMS value of a signal, so it works on any waveform, AC or DC.
It's the other meters — the ones that measure either peak or average and then scale it to an "RMS" reading — that depend on the waveform being a sinusoid (AC) signal.
At least in theory — many meters block DC when in an AC measurement mode, because this makes it easy to measure both the DC bias of a signal and its AC component separately by just switching modes.