[WhoMe]

I'm wanting to know if there is a better liquid to use in an Archimedes Principle solid density measurement than water or methanol. Water has a tendency to create bubbles on the surface of the solid, whereas methanol does not. However, the effect of temperature on the density measurement when using methanol is twice what it is when using water. Does anyone know of a non-corrosive (to steel) liquid with low surface tension and a low coefficient of thermal expansion at room temperature?

[curiouscat]

Can you use degassed water?

[curiouscat]

Does the temp. coefficent matter so much? Keeping temp. constant to say within 3 C is not that hard.

Methanol has a Volumetric Coefficient of Expansion ~ 0.00149 1/K.

Let's say you had a 1 kg steel sample. i.e. 124 cc. That's a 0.5 cc error in measured volume. i.e. 0.3% error in measured density.

Is that too bad? What accuracy are you shooting for?

[WhoMe]

I can give the degassed water a try. Do you think that will reduce the bubbles on the surface of a part with less than perfect surface finish? I was thinking that the bubbles are forming as the part is immersed and surface tension jumps little voids on the part surface. Degassing will apparently change the density of the water, at least until air dissolves in it again. Any idea what the density of freshly degassed water is?

Yes, you're right about the accuracy. I'm trying to get another digit, preferably two.

[curiouscat]

I can give the degassed water a try. Do you think that will reduce the bubbles on the surface of a part with less than perfect surface finish?

Not sure at all. But worth a shot?

[Arkcon]

Maybe you can add just a touch of solvent, or detergent, to prevent bubbles from clinging to the object.  You'll need to determine the density in any case, so why limit yourself to just one pure substance?

[Enthalpy]

Even if the liquid's temperature is not perfectly controlled, you can measure it to <0.1°C with any ammeter probe and apply a correction.

Hydrocarbons wet steel very well and protect it against corrosion. Shorter ones are thinner but flammable and volatile. A compromise would be around kerosene, Diesel oil, or the more reproducible and healthy cis-Pinane, 2,6-dimethyloctane, mesitylene...

I suppose your part has a complex shape? Because, if you can machine a small cylinder of the unknown steel, measuring the dimensions and weight is much more accurate.