Thread Tools
This thread is privately moderated by tspeer, who may elect to delete unwanted replies.
Dec 26, 2017, 06:05 PM
Registered User
tspeer's Avatar
Thread OP
Build Log

Radian Flight Test Project


I plan to use my eFlite Radian 2m Basic glider as a flight test testbed. Here's the basic plan of attack:

1. Establish test objectives
2. Determine test and analysis methodology
3. Design and install instrumentation
4. Calibrate instruments
5. Calibrate position error
6. Conduct flight tests
7. Analyze data

Some of these require a fair amount of discussion, so I've broken each of them into a separate comment. I'll edit the comments as I go along to document what I'm doing.
Last edited by tspeer; Jan 04, 2018 at 02:07 PM.
Sign up now
to remove ads between posts
Jan 04, 2018, 01:37 PM
Registered User
tspeer's Avatar
Thread OP

1. Establish test objectives


I want to quantitatively measure the performance of the Radian with the stock wings and then use the same fuselage and instrumentation with new wings to determine the performance of the new wing designs.

This is in support of my design of a new 2 m electric glider with rudder, elevator, spoiler control. I want to try out wings with flexible trailing edges that will automatically adjust the camber between low speed and high speed flight. I also want to experiment with devices that may reduce the severity of flow separation and make for a more gentle stall onset. I plan to work through the development issues with the Radian and then incorporate the lessons learned in the new design.
Last edited by tspeer; Jan 04, 2018 at 01:52 PM.
Jan 04, 2018, 01:38 PM
Registered User
tspeer's Avatar
Thread OP

2. Determine test and analysis methodology


The testing will use the turning flight technique described in the Lark test report to estimate the standard day performance and the nondimensional drag polars.

Simplified equations of motion will be used to write a Kalman filter for processing the data. The Kalman filter will estimate the horizontal winds, static pressure error coefficient, true airspeed, turn rate, normal load factor, bank angle and rate of descent. The methods in the Lark report will be used to calculate the drag polars. Standard day performance will be estimated by correcting the test day measurements and compared with predictions from the drag polars.

Pitot-static system position error calibration will be done using GPS measurements of ground speed, based on the assumption that all of the position error in the airspeed and altitude is due to errors in the static pressure measurements. This paper describes a similar approach.
Last edited by tspeer; Jan 04, 2018 at 01:59 PM.
Jan 04, 2018, 01:38 PM
Registered User
tspeer's Avatar
Thread OP

3. Design and install instrumentation


Instrumentation will consist of off-the shelf Spektrum telemetry sensors. These include:
- TM1000 telemetry transmitter (SPM9458)
- Air flight pack voltage sensor (SPMA9556)
- DSMX/DSMR telemetry temperature sensor (SPMA9571)
- Aircraft telemetry GPS sensor (SPMA9587)
- Aircraft telemetry variometer sensor (SPMA9589)
- Aircraft telemetry airspeed indicator (SPMA9574)

Ground atmospheric conditions will be measured using a Blue Maestro Pebble environment monitor and data logger. The Pebble measures barometric pressure, temperature and humidity, logs the data, and transmits the logged data via Bluetooth. The Pebble uses the Bosch Sensortec BME280 sensor. The sensor specifications are:
TEMPERATURE SENSOR Maximum accuracy of +/- 0.5 Celsius between 0 C - 60 C with a full range of -30 C and +85 C Degrees Celsius. Resolution of 0.1 C
HUMIDITY SENSOR Maximum accuracy of +/- 3% Relative Humidity between 20% RH - 100% RH.. Full range 0% - 100% RH, Resolution of 0.1% RH
PRESSURE SENSOR +/- 0.12hPA relative pressure (hPa) between 700hPa - 900hPa with a full range between 300 hPa and 1100 hPa. Resolution of 0.1hPa

The barometric altitude measured by the variometer has two problems - there is no nipple for connecting tubing to a static source, and the altitude reading is set to zero at power-up. The lack of connection to a static source means the variometer sensor must be physically located such that its surrounding pressure is the reference static pressure. This is different from the static pressure measured by the airspeed sensor, which uses static ports on the Prandtl probe. The airspeed sensor and the variometer will be mounted in a sealed container that is connected to the static pressure ports so that both sensors read the same static pressure. The Prandtl tube will be used to provide the static pressure. Alternatively, a trailing static pressure source may be used.

The barometric height above ground reported by the variometer will be added to the pressure altitude calculated from the Pebble to calculate the pressure altitude of the plane and the static pressure measured by the Prandtl tube. Humidity at altitude will be assumed to be the same as the humidity at ground level, since the altitudes to be flown will be under 500 ft agl. The ambient temperature at altitude will be measured using telemetry temperature probe mounted on the outside of the fuselage instead of being used to measure battery or ESC temperature.

The attached picture shows the components fabricated for installing the instrumentation. The pitot-static probe is to be mounted to the fuselage so the tubing connections between the sensors and the probe do not have to be broken and remade when the wings are removed and installed. A fuselage mounted probe also means the same installation can be used for different sets of wings. A blade mount attached to the fuselage above or behind the wing was considered, but rejected so as not to disturb the plastic wing carry-through structure. The canopy was selected as the most convenient location for mounting the pitot-static probe. The chamber for the airspeed and variometer sensors could be located under the canopy and the connections did not need to be disturbed when the canopy was removed or installed.

The photo shows the Prandtl probe mounted to the top of a blade made from 1/8" balsa with 1/64" plywood skins. A bamboo skewer formed the leading edge and forward spar. Another bamboo skewer spar was inlaid between the passages for the static pressure tube and total pressure tube. The blade was sized to provide approximately 2 cm of support for the tube, followed by the length of the total pressure connection, a radius of approximately 1 cm for the total pressure tube to turn from horizontal to vertical within the blade, and a tapered trailing edge. The blade had to be situated far enough forward to allow the total pressure tube to penetrate the canopy ahead of the aft canopy retention magnet. This brought the head of the pitot probe within the arc swept by the folding propeller. The blade needed to be tall enough to avoid interference with the prop as it started up.

The blade mount for the Prandtl tube added vertical area and mass ahead of the center of gravity. A ventral fin was added to the tail to compensate for the blade mount's effect on directional stability.

The canopy has a forward facing inlet that delivers ram air to the cockpit. The inside of the fuselage is vented through the bottom in a region of the fuselage that probably has reduced pressure compared to freestream ambient pressure. The pressure at any point within the fuselage cavity will be a function of airspeed and may change as components are shifted around, such as when removing and replacing the battery. This makes the inside of the fuselage unsuitable for measuring the static pressure.

In order for the airspeed sensor and variometer to measure the same static pressure, the sensors will be located inside a plastic pill bottle. The bottle will be vented to the static pressure source, forming a plenum chamber that provides the same static pressure to the variometer and airspeed sensor. The bottle will be fitted with a brass tube passing through the wall of the chamber so tubing can be attached to carry the total pressure to the airspeed sensor. An X-bus extension cable will also pass through a sealed penetration of the bottle to connect the sensors to the TM1000 telemetry transmitter. The bottle is attached to the underside of the canopy, using a rubber band to hold it to the structure for the pitot-static tube blade mount.

The existing magnets are not sufficient to retain the canopy with the Prandtl tube blade mount attached. Rails, made of 1/8" square balsa, were glued to the inside of the canopy. Matching mounts, consisting of 1/8" balsa with over-hanging 1/64" plywood lips, were glued to the canopy sill. When the canopy is installed the rails on the canopy clip into the grooves formed between the plywood lips and the canopy sill. This retains the canopy despite side loads on the blade. The canopy is sprung outward to disengage the rails when removing the canopy.

The bamboo skewer spars from the blade mount penetrate the canopy and pass through holes in the cross-members. The cross-members are made of 1/8" balsa with 1/64" plywood skins. The bamboo skewers were glued to the cross members, locking the blade in place. The cross members were also glued to the underside of the canopy using a plastic-bonding epoxy.

The ventral fin was constructed of 1/16" balsa with 1/64" plywood skins. Carbon rods inset in the balsa core were inserted into the aft fuselage to strengthen the connection of the fin when it was glued to the aft fuselage.

All the instrumentation mounts were painted orange and the instrumentation wiring marked with orange electrical tape, in accordance with MIL-STD-27733, Modification and Marking Requirements for Test Equipment in Aerospace Vehicles and Related Support Equipment.
Last edited by tspeer; Jan 04, 2018 at 02:15 PM.
Jan 04, 2018, 01:40 PM
Registered User
tspeer's Avatar
Thread OP

4. Calibrate instruments


No transducer actually measures what it is intended to measure. For example, a temperature sensor may actually measure the change in electrical resistance due to a change in temperature. As a result, it is necessary to calibrate the sensors on the ground to determine calibration curves and corrections that need to be added to the measured values to obtain the actual values of the measurands. The chain of calibrations is:
indicated values from the sensors + instrument calibration corrections + corrections for position error = calibrated measurements.

The telemetry altitude, airspeed and temperature sensors will be calibrated by putting the fuselage in an environmental chamber whose pressure and temperature can be varied. A thermocouple temperature sensor with a calibration traceable to the National Institute of Standards will be used as the temperature reference. Static pressure will be measured using a water-filled manometer. The Pebble environment monitor humidity measurement will be calibrated by placing it in a calibration atmosphere of 75% humidity. The Pebble temperature sensor will be calibrated using the environmental chamber at the same time as the telemetry temperature sensor.
Last edited by tspeer; Jan 04, 2018 at 02:05 PM.
Jan 04, 2018, 01:40 PM
Registered User
tspeer's Avatar
Thread OP

5. Calibrate position error


The total pressure measured by the Prandtl tube is the same everywhere in the flowfield except in the boundary layer or the wake of the propeller. Accurate measurements during powered flight are not required because the test objective is to measure the performance in gliding flight. So the Prandtl tube is assumed to correctly measure the total pressure. The static pressure, however, varies as a function of where the measurements are made and this variation is affected by airspeed, air density, and angles of attack and sideslip. These differences can be modeled with a static pressure error coefficient that is a function of airspeed. In steady flight, angle of attack is correlated with airspeed and the sideslip angles are small. The nondimensional coefficient accounts for the effects of air density and the velocity-squared influence on the static pressure error.

The static pressure error can be calibrated by comparing the airspeed with an independent measurement of speed. GPS speed will be used for this purpose. GPS measures ground speed, which is affected by the winds. The wind speed and direction can be separated from the airspeed by flying in circles or cloverleaf patterns. The wind is assumed to be constant during the interval of the maneuver, so the differences between airspeed and GPS speed that are a function of the airplane's course can be attributed to the wind.

The static pressure error coefficient will be determined in a manner similar to AIAA 2010-1350, A GPS-Based Pitot-Static Calibration Method Using Global Output-Error Optimization by John V. Foster and Kevin Cunningham of NASA Langley Research Center, Hampton, VA, 23681-2199. This requires flying at several headings to distinguish true airspeed and winds from ground speed and true course. The total pressure will be assumed to be accurately measured and all errors in instrument-corrected airspeed will be attributed to static pressure error. The static pressure position error corrections determined from the airspeed measurements will be applied to the static pressure measurements to obtain the calibrated airspeed and pressure altitude measurements.
Last edited by tspeer; Jan 04, 2018 at 02:05 PM.
Jan 04, 2018, 01:41 PM
Registered User
tspeer's Avatar
Thread OP

6. Conduct flight tests.


The flight tests will consist of glides in gentle circles, similar to the tests used in the Lark report. The plane will be trimmed for hands-off flight and sticks will be used as little as possible so as to maintain steady conditions. Pitch trim will be used to vary airspeed, with two or more circles flown at the same speed. The size of the circles will be adjusted using rudder trim.

The data will be logged to the SD card in the transmitter and transferred to a computer after the flights.
Last edited by tspeer; Jan 04, 2018 at 02:05 PM.
Jan 04, 2018, 01:42 PM
Registered User
tspeer's Avatar
Thread OP

7. Analyze data


(Reserved)
Last edited by tspeer; Jan 04, 2018 at 02:06 PM.
Mar 11, 2018, 05:07 PM
Registered User
tspeer's Avatar
Thread OP
Here are some preliminary flight test results. These are the first data with the vario and airspeed sensors contained in a plenum chamber that is plumbed to the pitot-static probe. I've not done any instrument calibration, yet, and I also have a lot of work to do in verifying the installation is leak-free and working correctly. That there are issues with the installation is evident from the data!

There are 6 flights shown. For each parameter, I've plotted all 6 flights and then broken them out individually so you can see what's going on.

There's an interesting oscillation in the airspeed. For the most part, I was flying in circles with the glider trimmed for hands-off flight, so I'm not exciting these airspeed oscillations.
Last edited by tspeer; Mar 11, 2018 at 05:35 PM.
Jun 05, 2018, 04:45 AM
JLJ
JLJ
Registered User
JLJ's Avatar
It seems to me that you are on an enjoyable project :-) Thank you for sharing!

"There's an interesting oscillation in the airspeed.". May I ask to expand this point?
Jun 05, 2018, 12:11 PM
Registered User
tspeer's Avatar
Thread OP
The classic airplane dynamic modes include a low frequency oscillation called the phugoid mode. It is an exchange between altitude and airspeed at essentially constant angle of attack. I haven't analyzed the data enough to know if that is what is happening or not.

The glider is grounded for the time being. I broke off the strut holding the pitot probe when recovering it from a tree. I've recently acquired a 3D printer, and I plan on making a new air data system. The variometer will be plumbed to a total pressure probe (so as to eliminate position error) and true airspeed will be measured with a free-spinning propeller. The rotor and support strut will be 3D printed.
Jun 07, 2018, 03:40 AM
JLJ
JLJ
Registered User
JLJ's Avatar
Thank you for the answer. I'm used to air data probes, thank you for sharing, the material that you posted is rare to be found online.

I'm looking forward to reading your next posts!


Quick Reply
Message:
Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Project: Dynamic Load Testing for FPV Racing Batteries slaven Batteries and Chargers 34 Sep 13, 2017 08:35 AM
Cool Ares 1-X Launch (First test flight of Project constellation hardware) FrogChief Life, The Universe, and Politics 60 Feb 02, 2010 05:47 AM