Searle's Apparatus for Halloysite Nanotube Coatings

Sat, 01 Apr 2023 00:00:00 +0000

Thermal regulating paint can cut the energy use of a temperature controlled building by adding an insulating barrier to the walls. The insulating paints on the market are not cost or performance competitive with expanding polyurethane foam, which is what keeps them out of general use. The Advanced Materials Research Lab at Louisiana Tech was working on latex paint doped with halloysite nanotubes to close that gap, and the work needed an instrument that could measure the thermal conductivity of a coating.

Searle’s apparatus is the classic method for measuring the thermal conductivity of a solid rod. It is cheap and simple, which is why it is normally a teaching instrument rather than a lab one. The goal here was not to reinvent it but to modernize it.

Building the instrument

Glass mercury thermometers were replaced with thermocouples feeding a computer logger. The steam jacket became an electric steam generator. Analog balances and timers became digital.

Apparatus layout. The left arrangement measures a solid copper rod as a baseline. The right splits the rod and places the coating at the interface, so the measured drop reflects the coating.

The apparatus open and closed. Cotton batting packs both chambers, and the steam chamber is lined with aluminum tape to reflect heat back in.

The specimen is a 12 in copper bar, 1 in in diameter, cut in half. One half sits on the heated side and one on the chilled side, and the coating under test goes on the interface between them. Measuring the solid rod first gives a baseline to compare against.

The housing is plywood, 3/4 in outside and 1/2 in for the internal supports, with roughly 1 in holes in the internal panels to carry the rods. The steam chamber is lined with aluminum tape to reflect heat back in and cut losses, and copper tubing carries heat from the rod into the chilled water. Both chambers are packed with cotton batting.

Instrumentation and procedure

Four thermocouples run to an eight channel USB logger, with the probes on the rod spaced 10.16 cm apart. A digital balance under the outlet records the mass of chilled water passing through, which is what the conductivity calculation needs.

Steam is held at 100 °C and the chilled water at 13 to 14 °C, with lab ambient between 18 and 22 °C. Each test runs 30 minutes and logs at 5 minute intervals. The apparatus then needs about 24 hours to return to thermal equilibrium before the next run, which is the practical limit on how fast samples can be worked through.

Simulating the instrument

I built five models of the apparatus in SolidWorks and ran transient thermal analysis on each, increasing the complexity step by step to see how much detail the results actually needed. The set covers the solid rod, the split rod with a coating, versions driven by thermal loads and heat flux taken from the first two, and an aluminum cylinder used for heat capacity testing. Copper was modeled at 390 W/m·K and the 3003 aluminum at 170 W/m·K.

Transient thermal analysis of the apparatus. The contour plot shows heat moving from the steam chamber along the rod, and the temperature plot tracks two probe locations over the run.

Limitations

Specific heat was not determined experimentally and was taken from literature, which is a direct source of error in the transient results. The simulated assembly is also simplified against the real apparatus, since the software could not carry every detail of the build.

Credit

This was carried out at the Louisiana Tech Advanced Materials Research Lab with Hawa Hashm, John Kraft, John Matthews, David Hall, Yuri Lvov, and Shaurav Alam. The coating study the apparatus supported was published in Construction and Building Materials, Vol. 392.

Transient Thermal FEA of Ohmic and Convection Curing

Thu, 19 May 2022 00:00:00 +0000

Geopolymer needs heat to cure. An oven does the job but it is slow, and it is no help at all if the goal is curing material as it leaves a print nozzle. Before settling on electrode geometry for the printer, I wanted numbers on how much faster ohmic heating actually is.

I modeled three configurations in SolidWorks. The first is a standard 2 by 4 in Teflon test cylinder in a convection oven. The second is a large ohmic apparatus with steel electrodes on either end of that same cylinder. The third is a smaller pusher plate apparatus with an internal spring that holds pressure on the sample so current is applied evenly.

Modeling ohmic heating in a tool that does not support it

SolidWorks does not simulate ohmic heating natively. Third party add-ons for it exist and were priced well beyond what this project could justify.

Ohmic heating is volumetric, so it can be approximated. Instead of heating a surface, I applied a heat power load to the body of the geopolymer, which puts generation through the volume the way current does. The power values came from lab measurements rather than from the solver.

Checking the uniform heating assumption

Applying power uniformly through the body is only valid if current reaches the whole sample. At 60 Hz, with a measured sample resistance of 20 Ω and a conductivity of 2.497 S/m, the skin depth works out to roughly 41 m. The sample is 2 in across, so penetration is nowhere near a constraint and the geopolymer can be modeled as a single uniformly heated cylinder.

If the skin depth had come out smaller than the sample, the model would have needed two concentric cylinders with power applied only to the outer shell.

Thermal power and results

Thermal power for each setup was calculated from lab data. The convection figure came from heating a 385 g sample in a 70 °C oven for one hour and measuring the rise from 30 °C to 61 °C.

Configuration	Thermal power	Time to reach 70 °C
Convection oven	2.165 W	about 2 hours
Ohmic, 2 by 4 in cylinder	743.6 W	about 20 seconds
Ohmic, pusher plate	433 W	about 40 seconds

Convection reached the target at about two hours, matching the analytical prediction. The large ohmic apparatus reached it in about 20 seconds and peaked at 90.7 °C internally.

Transient thermal contours for the 2 by 4 in ohmic apparatus. External view on the left, section view on the right showing the hot core between the steel electrodes.

Temperature at the sample center against time for the ohmic apparatus. The 70 °C target is crossed at roughly 20 seconds.

The pusher plate apparatus took about 40 seconds. Heating there was more localized, which I put down to the lower applied power together with convective losses through the pusher plate assembly.

The simulation also showed something the hand calculation did not. Even with power applied uniformly through the body, heat propagation started near the center of the sample and spread outward.

What it was for

The comparison gave a defensible basis for electrode geometry and curing parameters on the ohmic curing printer. Three orders of magnitude in thermal power, and the difference between two hours and twenty seconds, is what makes curing during deposition worth attempting.

Limitations

Material properties for geopolymer were taken from literature rather than measured, and the model assumes they hold constant through the cure. Contacts were treated as bonded. The steady state convection result put peak temperature at 51.9 °C, which is likely a thermal stress singularity at the edges of the cylinder lid rather than a real reading, with the more plausible value nearer 42 °C.

SolidWorks | Stephen Timothy Gordon II

Searle's Apparatus for Halloysite Nanotube Coatings

Building the instrument

Instrumentation and procedure

Simulating the instrument

Limitations

Credit

Transient Thermal FEA of Ohmic and Convection Curing

Modeling ohmic heating in a tool that does not support it

Checking the uniform heating assumption

Thermal power and results

What it was for

Limitations