I've attached a diagram of what I'm talking about.
I have a large (600x300x10 mm) aluminum block in which I place thermistors to monitor the temperature distribution in the plate (in a standard voltage divider circuit.)
When I have done this before, I have simply wired two leads to each thermistor (one ground, one voltage line) and passed them out of the plate to my PCB. I am now looking at a lot more sesnors (maybe up to 100) and that makes for a lot of wire management, not to mention expensive components like 200-way ribbon cable.
I thought I could half the soldering and number of wires if I ground the aluminum (connect it to GRD of my Arduino) and solder the thermistor ground lead to the aluminum body.
However, this plate will change temperature quite a bit (say, in the range of 60 to -10 C) and I don't know if this makes for a good GND. I'm worried that a plate this large, that changes temperature over an extensive range, would somehow change the GND and make for unreliable and innacurate sensor readings. I certainly don't want the sensors accuracy or reliability to be dependent on plate temperature because the level of my ground is changing.
Am I overthinking this or, to get good and accurate sensor readings, will this plan not work?