TFE731-5BR  11/22/19

There is another 1000 lbs of thrust available with the TFE731-5BR but likely intake issues that would drive the need to de-rate the output.  The idea of flat rated thrust irrespective of ambient temperatures or altitude for that matter is mighty attractive.  Hot sections are on condition for my maintenance plan so a hot section that ran, say, 100 degC cooler would last a very long time.  There would be a weight penalty (something like 75 lbs - I need to check) which would require more nose ballast.  Still, a hundred more pounds for hundreds more in thrust is mighty attractive.

Ed is busy generating inlet geometry from the L-139 drawings and exit geometry from our fan duct and exit tube design.  Computational fluid dynamics (CFD) is next to get a feel for where problems will arise then its time to instrument up an installation with pressure sensors.  It has been my experience that CFD without real world correlation is helpful but far from definitive.

The whole air tract changes with a -5BR installation.  This rendering gives an indication of what is needed.  Once again, composites provide a distinct advantage in shaping ducts that do not disrupt airflow.

The L39MS uses the DV2 motor (shown below) with around 1000 lbs more thrust.  This would seem to indicate that a -5 with similar thrust could be made to work.  The CFD  modeling and analysis has begun.  There are a couple of pics of the initial models but these are VERY early days in the process and the model requires a lot more work before simulation runs are possible.  As mentioned above, CFD results are nice but to be truly useful they must be correlated to real world pressure data and the models adjusted accordingly.  I'm working on both a -3 and -5 inlet with integral pressure sensors around the fan periphery to collect just such data.  We are at the start of a very long road.

Now that Grey is fully into Phase 1 test flights with a dedicated test pilot (as of mid 12/18), my focus has turned to a couple of sets of -3 conversion parts and full on the -5 installation in UDuck.....  I'm not sure what we will end up with but we are going to try.

Well, I did not see that coming.  We lost access to Grey to do customer demos so UDuck is becoming Mallard but is being built with a -3 instead of a -5.  We have purchased "Mojave" and, as of today (6/11/19), it is being collected in Mojave to be shipped to Florida.  This will now be our -5 project.  You can knock us down but we will get right back up and re-double our efforts.  We are well down the path of completing Mallard as our in house -3 demo plane and are, once again, working on the -5 conversion.  We need to get it done in time for Reno next year :)  Below is a rendering from our gas flow path work looking mostly at the differences between the -3 and -5 post fan gas path.  Here is a more complete look at the model.

The first fan exit duct part is off the mold and has been oven cured.  It is test fit to the motor and we will soon be rolling it into the fuselage to start working on everything from the exit tube to bleed air plumbing.  It looks like the exit tube may need to be a monolithic carbon structure.  We will need to get some -5BR exhaust temperature readings before making that decision.

Many thanks to Paul and his patience in working with us on improving our composite technique.  The -5 duct part is the first to use his improved breather technique and results were fantastic.  Very good resin control and vastly superior material compaction.  Our previous parts were good and these are better.  There is more improvement to come as we incorporate more of Paul's teachings.

Its a work in process.


 

-5BR powered pickup - a fast way to get your motors home

Hans with his current work of art

The Falcone exit Nacelle is a work of art. Pulling a skin to act as a mold was straight forward.

The CnC framework model

The L39MS uses the DV2

The DV2 appears to be a higher bypass fan motor

Ed's Inlet Model

Very early days in the CFD work

Just starting to get the models sorted

Ed got a little curious about overall airframe effects - next thing we know, the airframe is modeled

It is interesting to see how the inlets pull air from the surroundings

and how air comes off the nose

Some renderings from the Parametric Model

-3 -5 Gas Path Differences

Here are some pictures of the -5 Fan Exit Duct Mold Build

Sealing is important for Vacuum Integrity

End plates must be true if the flanges are to fit the motor and seal

Bagging the part with the new breather technique

First part coming off the mold

First part off the mold

Initial Test Fit

Latest part and process using satin weave full layer sheets