Ni 9213

To verify that the stimulus environment was comparable across contexts, we measured air temperature in the test arenas, task availability and brood dummy temperature in a control experiment, repeating procedures from the non-social context for 10 bees and from the random social context for 10 groups, using bees from two separate colonies. Five bees and five groups of 10 bees were tested from each colony. Air temperature (ALMEMO 2290–8 data logger) and the number of available brood dummies was recorded per minute. Brood dummies temperature was logged at 1 Hz using temperature sensors (Omega thermocouples) running along the longitudinal axis of the dummies, ending at the tip directly under the wax layer and connected to a data logging device and custom programs (NI cRIO-9074 and NI 9213, LabVIEW 2010 Version 10.0.1, National Instruments, Germany). For statistical analysis, fanning threshold temperature was assigned to the mean wax temperature during the one-minute bin in which long fanning was first observed, as extracted from this data set.

The cross‐sectional image of the CWO film was obtained with a scanning electron microscopy (SEM, JSM‐7100, JEOL). The UV–vis–NIR (0.3–2.5 µm) transmittance (T) and absorption (A) spectra of the transparent solar absorber, transparent photovoltaic cell, and their integrated window were measured using a spectrometer (Lambda 950, Perkin Elmer) equipped with a 150 mm integrating sphere. For the temperature response measurement, a solar simulator (Oriel Sol2A, Newport) was used to provide standard and stable 1‐sun power (1 kW m⁻²), and T‐type thermocouples were used to measure the steady–state temperature and connected to a data acquisition device (NI 9213, National Instrument) for data recording. For the TPV electrical characterization, a Keithley 2420 SourceMeter was used to obtain I–V characteristics under simulated AM 1.5G solar illumination. The photovoltaic cell layouts,^[15] in which a matte black background was placed on the back of the TPV device, were used so that illumination from the environment or reflection could be eliminated for both I–V measurements.

Tensile testing was conducted using a tensile testing machine Instron 3369. During the tensile test experiments, four steps were performed: i) The SMA spring was heated by a 5 V power supply and kept at the same length for 1 min. ii) The SMA spring was compressed with a speed of 0.5 mm ⁻¹s until its output force reached zero. iii) The SMA spring was extended to its original length with a speed of 0.5 mm ⁻¹s. iv) The SMA spring was kept at the same length and cooled for 2 min.
A K‐type polyimide flat film thermocouple (TC Direct) was attached to the TWSMA spring to measure its real‐time temperature. A thermocouple module (NI‐9213, National Instruments) carried by a compact‐DAQ chassis (USB‐9162, National Instruments) was used to record the temperature data. The data acquisition program was based on LabVIEW, and the sampling frequency was 20 Hz.