IceSense Proof of Concept: Calibrating an Instrumented Figure Skating Blade to Measure On-Ice Forces

Competitive figure skaters often suffer from overuse injuries, which may be due to the high impact forces endured during jump repetitions performed in practice and competition. However, to date, forces during on-ice figure skating have not been quantified due to technological limitations. The purpose of this study was to determine the optimal calibration procedure for a previously developed instrumented figure skating blade (IceSense). Initial calibration was performed by collecting data from the blade while 11 skaters performed off-ice jumps, landing on a force plate in the lab. However, mean peak force measurements from the blade were greater than the desired error threshold of ±10%. Therefore, we designed a series of controlled experiments which included measuring forces from a load cell rigidly attached to the top of the blade concurrently with strain data from the strain gauges on the blade. Forces were applied to the blade by adding weight to a drop tower or by manually applying force in a quasi-static manner. Both methods showed similar accuracy, though using the drop tower allowed precise standardization. Therefore, calibration was performed using the weighted drop method. This calibration was applied to strain gauge data from out-of-sample drop trials, resulting in acceptable estimates of peak force (less than 10% error). Using this calibration, we collected data on one figure skater and present results from an exemplar on-ice double flip jump. Using the IceSense device to quantify on-ice forces in a research setting may help inform training, technique, and equipment design.


Design Improvements
Since publishing the initial design of the IceSense on-ice force sensing system [13], important design improvements have been made to increase the signal quality and the robustness of the measurements. These improvements included re-routing the wiring from the power supply, modifying the design of the Wheatstone bridge board and circuitry, raising the gain of the signal conditioning board, and improving the protection of the electronics. Each improvement is described in detail below.
Wiring ( Figure 1): In the initial wiring scheme, only the data logger was directly connected to the battery. Power was then fed from the data logger to the rest of the boards. While this wiring scheme was simple and in principle ensured that there was a common ground and power provided to each board, the additional power running through the data logger damaged its components; it burned the tracing that carried power through the data logger and distributed it to other boards. To remedy this problem, each board was connected directly to the battery. Although the new system was more cumbersome, it protected the electronics since no board was being used to distribute power to other boards. In addition, all boards were connected to ground at a single point (star grounding) to isolate the analog signal from noise in the digital components, reducing an artifactual periodic spike in the output signal.
Wheatstone Bridge Boards ( Figure 1c): In the initial data acquisition system, the bridge load regulation from the voltage regulator was noisier than expected. This impacted the accuracy of the data acquisition circuitry at the micro-volt level. The simplest solution was to substitute a high-precision voltage reference source capable of driving the bridge circuit. A REF195 low-dropout voltage regulator (Analog Devices, Norwood, MA) accurate to 4 ppm/mA and 5 ppm/C was used. Further, a vibration-stable trimming resistor was incorporated into the circuit to zero the output of the bridge, thus maximizing the signal available to the data acquisition circuitry. Bridge trimming was performed after temperature stability was achieved prior to each collection session.
Signal Conditioning Board (under data logger shown in Figure 1d): To increase the sensitivity of the system, the gain of the signal conditioning board was raised. This gain can be set separately for each channel and depends on the gain resistor of each channel: G = (49.9 kΩ/Rg) + 1, where G is the gain and Rg is the resistance of the gain resistor. The initial system had gain resistors of 604 Ω, resulting in a gain of 83.6. By testing multiple gain resistors while skaters performed jumps on ice, it was determined that a gain resistor of 50 Ω and the resulting gain of 999 maximized the sensitivity of the system while still avoiding saturation. (Figures 1f-g): Using the system off ice and on ice, vibrations (as occur during a landing, for example) often caused channels to shift and re-zero. These vibrations affected the electronic boards and wiring. To reduce the movement of the wires and electronics relative to the boot, the Wheatstone bridge boards were secured to their respective stanchions using tape. Second, the bracket was redesigned to orient the signal conditioning board downward, allowing the wires to be bundled in wire wrap, connected to the signal conditioning board, switch and LED, and taped to the boot and blade to avoid movement against the blade. Care was taken not to secure the wires too tightly, as this was found to cause damage at the connections between boards.

Protecting the Electronics
The electronics are sensitive to water and ice. A boot cover was designed from 3 mm thick neoprene, which is both waterproof and can be stretched to fit tightly around the boot. The bottom portion of the cover was glued to the blade below the stanchions using Hi-Strength 90 Spray Adhesive (3M, St. Paul, MN, USA), which bonded the neoprene to the portion of the aluminum blade holder immediately above the steel blade, protecting all the electronics. The top and back portions of the neoprene panels were joined using Velcro to enable easy access to all electronics.

Temperature Sensitivity
The stanchions and strain gauges (Figure 1b) also respond to changes in temperature. Two different tests were performed to analyze the system's sensitivity to these changes. In the first test, a thermocouple was attached to the middle stanchion (next to one of the strain gauges) and measurements were recorded under three different conditions: room temperature (approximately 23 °C), cold (next to the ice, approximately 7-10 °C), and in contact with the ice (approximately 3-6 °C). Under each of these conditions, voltage was recorded as a static, vertical force of 100 lb was applied to the blade after it had reached thermal equilibrium. The results showed that the different conditions caused an offset in voltage but did not otherwise affect the response.
In the second test, the effect of temperature on calibration accuracy was determined during the force plate validation (see Force Plate Validation above). One calibration matrix was generated from 10 landings with the blade at ice temperature and another calibration matrix from 10 landings with the blade at room temperature. Both calibrations were then applied to 118 landings performed with the blade at ice temperature. To get the blade to ice temperature, the boot-blade system was placed on a 20 cm by 25 cm sheet of ice in a Styrofoam cooler for 25 minutes prior to testing. Between landings, the skater stood on the ice sheet for 20 to 30 seconds, which proved sufficient to maintain the blade at ice temperature. In analyzing the results, we found that temperature did not affect calibration accuracy (results not shown for brevity).
The results of both tests together led to the conclusion that blade temperature changed the baseline values of each channel, but do not affect the strain gage channel responses. To avoid sensor saturation, this change in baseline (caused by the change from room temperature to near-ice temperature) was counteracted by trimming the Wheatstone bridges once the system had equilibrated to the near-ice temperature.