Added methodology from FAA pub.

1 files changed, 125 insertions, 1 deletions

diff --git a/results.tex b/results.tex
index 9944c1e..4e36c06 100644
--- a/results.tex
+++ b/results.tex
@@ -7,4 +7,128 @@
 
 
 \section{Configuration One}
-\section{Configuration Two}
\ No newline at end of file
+\subsection{Turboprop Engine Data}
+Figure 11 shows the power
+turbine shaft torque output and rotational speed throughout the test
+run. The impact of the power generator can be seen in the mechanical
+system data. The turboprop was taken to full throttle and left there
+for the duration of the electrical system testing. The turboprop
+maintains a specific shaft speed at each throttle setting and so should
+not change throughout the run. The decrease in RPM around the 355-s
+mark was caused by the load of the generator slowing it down. After
+5-s, the engine control unit brought the turboprop back to the intended
+speed. However, it can be seen that the engine was now producing
+more torque at that same RPM, and thus more power.
+The turboprop is rated for a peak shaft speed of 2158-RPM, which
+was used in the design of the generator system. The peak value
+measured during testing was 2136-RPM. So, the generator was
+spinning at 2990-RPM and the design speed for the generator is
+3000-RPM, so it was operating nominally. The peak power
+produced by the turboprop was about 150-kW. The nominal max
+continuous output of the engine is 160-kW, but the engine can
+temporarily boost to 180-kW for takeoff. This means that
+throughout testing, the maximum power rating of the engine was
+never exceeded. The engine performed well, and no otherwise
+unexpected phenomenon happened to the engine throughout testing.
+\begin{figure}[h]
+	\centering
+	\includegraphics[width=\textwidth]{img/test1prop.png}
+	\caption{\label{test1prop}Turboprop Shaft Torque and Speed}
+\end{figure}
+\subsection{Electrical System Performance}
+Figure 12 shows the
+voltage of the central bus and the current flow from the generator and
+battery. The ESC throttle positions are labeled in the figure. The
+voltage data contain some noise caused by the electrical output of
+the generator. The cause of the noise being something else, such as
+the vibration of the instrumentation, is unlikely because it only
+appears when the generator hits the designed speed and exceeds the
+voltage of the battery. The Arduino and probes that measure the
+main bus voltage are also not located on the engine, further
+increasing the notion that the noise is caused by the generator
+creating a noisy DC signal.\\
+Figure 13 shows the total electrical and mechanical power during
+the same period as Fig. 12, allowing for direct comparison. The
+trends for the voltage data are quite clear. The battery begins at
+106.5-V and ends at 105.9-V. Under peak load at around 365-s, the
+minimum operating voltage of 100-V is achieved. From left to right,
+the changes in ESC throttle position can be seen. Starting from the
+left the battery begins charging at 14.4-A. The two ESCs are then
+commanded to start at the expected nearly 15-A current draw, but the
+left ESC failed to engage properly.\\
+Surprisingly, the medium throttle position on the right ESC was
+achieved by drawing ~75-A, but the amount of current from the
+generator saw only a minor increase. Instead, the difference was
+almost entirely picked up by the battery. This can also be seen in the
+power data. There is a small bump in generator power at the ESC
+medium point, but the power decreases slightly and is picked up by
+the battery. The mechanical data show a drop in rotational speed at
+about the same time the medium throttle point was engaged.
+Whenever the load was increased, it slowed down the rotational
+speed (and thus voltage) of the generator preventing it from picking
+up the difference. The generator was slowed down from the near
+constant ~2990-RPM to 2904-RPM, which would have reduced the
+output voltage by 3-V. Because the output voltage of the generator is
+so close to the battery voltage, this decrease in voltage significantly
+impacted the current output. The battery responded to the increased
+load almost instantly but based on the power/RPM data, it took ~5-s
+for the turboprop ECU to adequately react to the increase in
+mechanical load.\\
+Once the high ESC throttle position was reached, a significant
+amount of current ~150-A total was discharged. The voltage then
+dropped by several volts and the amount of current from the
+generator increased significantly to its maximum of 39-A. At this
+point, around 4-kW was being extracted from the turboprop using
+the generator, while about 13-kW of power came from the battery.\\
+Afterward, the second medium throttle position was hit. The total
+current was ~40-A compared to the previous 75-A. The reason for
+the significant difference may have been an error made by the
+operator, who may have missed the target ESC throttle positions.
+The power data, however, provides evidence that the total power
+output of the propeller was continuously decreasing despite the
+operator holding the throttle steady for each position. The simple
+explanation may be that the inertia of the propeller and electric
+motor is quite large. The electric motor did not require as much
+power because it was slowing down instead of speeding up. The final
+observation is that the resting current output of the generator
+recharging the battery is slightly higher than at the beginning of the
+run. This is because the voltage difference between the generator
+and battery increased since the battery discharged some of its power.
+\begin{figure}[h]
+	\centering
+	\includegraphics[width=\textwidth]{img/test1electric.png}
+	\caption{\label{test1electric}Voltage and Current Data}
+\end{figure}
+\begin{figure}[h]
+	\centering
+	\includegraphics[width=\textwidth]{img/test1power.png}
+	\caption{\label{test1power}Mechanical versus Electrical Power}
+\end{figure}
+\subsection{Electronic Speed Controller Failure}
+After the main test was completed, the left ESC that had failed to
+start properly was investigated. The ESC and electric motor had
+previously been successfully tested during an all-electric check-out
+test run. The electric motor was again tested on battery power to
+figure out what was wrong. During this attempt at starting the motor,
+the ESC saw a significant in-rush current as shown in Fig. 14. This
+caused spontaneous failure and the rupturing of the capacitors inside
+the ESC creating a minor fire. The current draw during failure shown
+in Fig. 14 is an underestimate. The sensor used to monitor the battery
+current maxes out slightly over 500-A, so the instantaneous
+maximum values may have been higher than 550-A.\\
+The electrical data during the failure were of particular interest.
+During previous testing, an ESC had desoldered its connection point
+with its main power wire, which had removed the ESC load from the
+main bus. In this case, the ESC shorted, causing a substantial
+unexpected load of 46-kW of electrical power on the central bus.
+This caused significant strain on the battery, whose maximum rated
+current was around 350-A. Had this event occurred during full
+system testing, then it could have damaged the turboprop engine or
+the battery.\\
+The acquisition rate of the current sensor data is slow compared to
+the failure causing steep step changes. However, it appears that as
+the failed system was failing, the minimum of the oscillating current
+expenditure creeps up from around 20-A (at 423-s) to about 50-A
+(428-s). This gave rise to clear recommendations for the handling of
+electrical failure modes.
+\section{Configuration Two}